[{"id":"b-resistance","kind":"bottleneck","name":"Acquired resistance to every therapy","aka":[],"tldr":"Nearly every targeted therapy stops working within months to a few years as the tumour adapts.","summary":"Every class of anticancer drug is followed by a catalogue of resistance mechanisms: on-target mutations (EGFR C797S, ESR1, BTK C481S, androgen receptor variants), bypass signalling (MET amplification, PI3K activation), lineage plasticity (adenocarcinoma to small-cell or neuroendocrine transformation), drug efflux, antigen loss after CAR-T or ADC therapy, and payload-specific resistance to topoisomerase inhibitors. Even the best first-line targeted therapies in lung cancer produce median progression-free survival of under two years. Because resistance biopsies are rare and mechanisms are heterogeneous between lesions, the choice of next-line therapy is often guesswork, and combination or sequencing strategies designed to pre-empt resistance are seldom tested prospectively. Degraders that remove the target, dual-payload conjugates, ctDNA-triggered switching, and upfront combinations are the emerging countermeasures.","asOf":"2026-09-08","links":[{"label":"Soria et al., Osimertinib in untreated EGFR-mutated NSCLC (FLAURA, NEJM 2018)","url":"https://doi.org/10.1056/NEJMoa1713137"},{"label":"Vasan, Baselga & Hyman, A view on drug resistance in cancer (Nature 2019)","url":"https://doi.org/10.1038/s41586-019-1730-1"}],"tags":[],"related":["idea-dual-payload-first","idea-payload-switching","idea-btk-degrader-frontline","idea-ctdna-switch-generalised","genie","idea-bio1-mandatory-progression-biopsy","idea-bio1-open-resistance-atlas","idea-bio1-upfront-bypass-combination","idea-bio1-pre-randomised-sequencing-trials","idea-bio1-antigen-mapping-at-progression","idea-bio1-dual-antigen-adc-escape","idea-bio1-adaptive-car-antigen-switch","idea-bio1-drug-holiday-resensitisation","idea-bio1-persister-ferroptosis","idea-bio1-epigenetic-persister-blockade","idea-bio1-mutagenesis-blockade-rev1","idea-bio1-stress-response-blockade","idea-bio1-senolytics-after-therapy","idea-bio1-persister-metabolic-vulnerability","idea-bio1-molecular-progression-add-on","idea-bio1-alternating-schedules","idea-bio1-resistance-mutation-vaccine","idea-bio1-mechanism-matched-access","idea-bio1-resistance-platform-trial","idea-bio1-slfn11-payload-guidance","idea-bio1-stromal-resistance-blockade","idea-bio1-real-world-resistance-surveillance","idea-bio1-approval-linked-progression-sampling","idea-bio1-drug-exposure-sanctuary-mapping","idea-bio1-immunopeptidome-timing","idea-moon-closed-loop-adaptive-therapy"],"cancers":["nsclc","breast-hr-positive","cll","prostate","melanoma","dlbcl"],"sections":[],"technologies":["protac-degrader","dual-payload-adc","bispecific-adc","liquid-biopsy","kinase-inhibitors","adc-payload-neutralizer"],"targets":[],"drugs":["osimertinib","camizestrant","pirtobrutinib","bgb-16673","amivantamab"],"companies":["astrazeneca","beone","nurix","sutro-biopharma","mersana","systimmune"],"institutions":["moffitt","mskcc"],"pathways":[],"terms":["resistance","c797s","esr1-mutation","castration-resistance","efflux-pump","met-amplification","histologic-transformation","endocrine-resistance","bcg-unresponsive","adc-sequencing"],"trials":["serena-6","mariposa","flaura2","bruin-cll-321","cadance-304"],"people":[],"bottlenecks":[],"keyPapers":["paper-vasan-nature"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Median progression-free survival with first-line osimertinib in EGFR-mutant NSCLC (FLAURA), the benchmark single-agent targeted therapy","value":"18.9 months","source":"Soria et al., NEJM 2018","url":"https://doi.org/10.1056/NEJMoa1713137"},{"label":"Median progression-free survival with sotorasib vs docetaxel in previously treated KRAS G12C NSCLC (CodeBreaK 200)","value":"5.6 vs 4.5 months","source":"de Langen et al., Lancet 2023","url":"https://doi.org/10.1016/S0140-6736(23)00221-0"}],"causes":["Tumours are heterogeneous, so a resistant subclone almost always pre-exists the drug.","Single-agent targeting of one node leaves parallel pathways available for bypass.","Resistance biopsies are uncommon outside academic centres, so mechanisms are inferred rather than measured.","Sequential single-agent development is commercially simpler than testing rational upfront combinations.","Lineage plasticity and epigenetic state changes are not captured by DNA sequencing and have few druggable handles."],"currentEfforts":["SERENA-6 established ctDNA-triggered switching to camizestrant on emergence of ESR1 mutation, before progression, as a viable strategy.","Combinations such as amivantamab plus lazertinib (MARIPOSA) and osimertinib plus chemotherapy (FLAURA2) attack EGFR resistance up front.","BTK degraders (BGB-16673) and non-covalent BTK inhibitors (pirtobrutinib) are designed to work after covalent-inhibitor resistance and to pre-empt it in first line.","Dual-payload and bispecific ADCs (Sutro Biopharma, Mersana, SystImmune's BL-B01D1) aim to prevent payload- or antigen-loss resistance.","AACR Project GENIE and cBioPortal make paired pre- and post-resistance sequencing data available for pooled analysis.","Evolutionary and adaptive therapy trials at Moffitt test dosing schedules that maintain drug-sensitive clones to suppress resistant ones."],"successLooksLike":"Resistance mechanisms are measured prospectively (tissue or blood) in every patient at progression, and first-line regimens are chosen to pre-empt the dominant resistance routes so that median progression-free survival on targeted therapy in lung and breast cancer more than doubles."},{"id":"b-ai-validation","kind":"bottleneck","name":"AI that is built but not validated or deployed","aka":[],"tldr":"Thousands of cancer AI models are published; a handful are in clinical use, and fewer have shown they help patients.","summary":"Machine learning models for cancer detection, pathology, prognosis and treatment selection are published by the thousand, but almost all are evaluated retrospectively on data from the institution that built them. Of AI-enabled devices cleared by the FDA up to 2020, nearly all were evaluated only retrospectively and most on a single site, and among deep-learning studies comparing AI with clinicians, only a handful were prospective and two were randomised. Retrospective accuracy is not clinical benefit: models drift as scanners, populations and practice change, integration into workflow is costly, liability is unresolved, and reimbursement rarely exists. The MASAI trial of AI-supported mammography screening is one of the first randomised demonstrations that an AI tool can safely change a cancer pathway. Prospective and randomised evaluation, reporting standards, post-market monitoring for drift, and regulatory pathways for models that keep learning are the requirements for AI to move from papers into care.","asOf":"2026-09-08","links":[{"label":"Wu et al., How medical AI devices are evaluated: limitations and recommendations from an analysis of FDA approvals (Nature Medicine 2021)","url":"https://doi.org/10.1038/s41591-021-01312-x"},{"label":"Nagendran et al., Artificial intelligence versus clinicians: systematic review (BMJ 2020)","url":"https://doi.org/10.1136/bmj.m689"},{"label":"Lång et al., AI-supported mammography screening (MASAI, Lancet Oncology 2023)","url":"https://doi.org/10.1016/S1470-2045(23)00298-X"},{"label":"FDA, Artificial intelligence-enabled medical devices","url":"https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-enabled-medical-devices"}],"tags":[],"related":["idea-multimodal-foundation-model","idea-ai-her2-low-scoring","fda-approvals","idea-data-neutral-ai-evaluator","idea-data-regulatory-sandbox-adaptive-ai","idea-data-ai-liability-safe-harbour","idea-data-ai-reimbursement-tied-to-outcomes","idea-data-silent-trial-before-deployment","idea-data-subgroup-performance-reporting-mandate","idea-data-precompetitive-cancer-foundation-model","idea-data-open-weights-for-public-funded-ai","idea-data-in-silico-trials-calibrated","idea-data-ai-pathology-evaluation-standard","idea-data-multisite-validation-precondition","idea-data-patient-facing-model-cards","idea-data-ai-red-team-programme","idea-data-ai-audit-trail-in-ehr","idea-data-ai-decommissioning-rules","idea-data-llm-documentation-rct-oncology","idea-data-ai-vs-tumour-board-rct","idea-moon-validated-digital-twins","idea-moon-continuous-ai-validation-registry"],"cancers":["breast-hr-positive","prostate","nsclc","melanoma","colorectal"],"sections":["ai-computation"],"technologies":["radiology-ai-screening","digital-pathology-ai","pathology-foundation-model","dermoscopy-ai","ai-trial-matching","ai-drug-design","mammography"],"targets":[],"drugs":["artera-ai-prostate","artera-ai-breast"],"companies":["paige","pathai","lunit","aidoc","artera","owkin","tempus"],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-wu-nat-med","paper-nagendran-bmj"],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"major","metrics":[{"label":"FDA-approved AI medical devices (to 2020) evaluated only retrospectively","value":"126 of 130","source":"Wu et al., Nature Medicine 2021","url":"https://doi.org/10.1038/s41591-021-01312-x"},{"label":"Deep-learning studies comparing AI with clinicians that were randomised trials (systematic review, 2010-2019)","value":"2 of 83 (with 9 prospective non-randomised)","source":"Nagendran et al., BMJ 2020","url":"https://doi.org/10.1136/bmj.m689"},{"label":"Screen-reading workload reduction with AI-supported mammography screening at equal or higher cancer detection (MASAI, randomised, 80,033 women)","value":"44% fewer reads","source":"Lång et al., Lancet Oncology 2023","url":"https://doi.org/10.1016/S1470-2045(23)00298-X"}],"causes":["Retrospective single-site accuracy is cheap to produce and enough to publish, so prospective validation is rarely done.","Models trained on one population and scanner degrade on others (dataset shift) and are not monitored after deployment.","Regulatory clearance has not required evidence of clinical benefit or multi-site validation.","Workflow integration, IT security review and liability allocation cost more than the model.","No reimbursement code exists for most AI outputs, so hospitals have no business case.","Locked models cannot legally be updated without re-clearance, so they age in place."],"currentEfforts":["The MASAI randomised trial in Sweden showed AI-supported mammography screening increased cancer detection while nearly halving reading workload.","Paige Prostate (FDA de novo 2021) and Lunit, Aidoc and PathAI products have cleared regulatory pathways with prospective or multi-site evidence.","CONSORT-AI and SPIRIT-AI extend trial reporting standards to AI interventions, and DECIDE-AI covers early-stage clinical evaluation.","The FDA finalised guidance on Predetermined Change Control Plans (2024) allowing AI devices to update within pre-specified bounds without new submissions.","The EU AI Act (Regulation (EU) 2024/1689) classifies medical AI as high-risk and requires post-market monitoring.","ArteraAI Prostate, a multimodal prognostic model, was included in NCCN prostate guidelines on the basis of validation in phase 3 trial cohorts."],"successLooksLike":"Every AI tool in clinical oncology use has prospective multi-site validation and post-deployment drift monitoring, at least ten AI tools have randomised evidence of improved patient outcomes or equal outcomes at lower cost, and models can be updated safely under regulatory oversight."},{"id":"b-biomarker-validation","kind":"bottleneck","name":"Biomarkers are not validated or standardised","aka":[],"tldr":"Tests that decide who gets a drug are often not validated prospectively and are measured differently in every lab.","summary":"Modern oncology allocates drugs by biomarker, but the biomarkers are held to a lower standard than the drugs. PD-L1 is scored with several non-interchangeable antibodies and cut-offs; HER2-low, which now defines eligibility for trastuzumab deruxtecan, depends on distinguishing IHC 0 from 1+, where pathologist agreement is poor; TMB varies by panel, bioinformatic pipeline and tumour type; HRD assays disagree with each other; and TROP2 and other ADC targets are given without any validated assay at all. Most biomarkers are validated retrospectively within the pivotal trial of one drug, on one assay, then used with different assays in practice. The consequences are patients wrongly included or excluded, irreproducible subgroup results, and combination biomarkers that never reach the clinic. Analytical standardisation, external quality assurance, prospective biomarker-stratified trials, and AI quantification that removes inter-observer variability are the remedies.","asOf":"2026-09-08","links":[{"label":"Fernandez et al., Examination of low ERBB2 protein expression in breast cancer tissue (JAMA Oncology 2022)","url":"https://doi.org/10.1001/jamaoncol.2021.7239"},{"label":"Hirsch et al., Blueprint PD-L1 IHC assay comparison project (JTO 2017)","url":"https://doi.org/10.1016/j.jtho.2016.11.2228"},{"label":"Merino et al., Establishing guidelines to harmonize tumor mutational burden (JITC 2020)","url":"https://doi.org/10.1136/jitc-2019-000147"}],"tags":[],"related":["idea-ai-her2-low-scoring","idea-trop2-pet-selection","adcdb","idea-tr2-cdx-mutual-recognition","idea-tr2-psma-volume-qualification","idea-tr2-bicr-discordance-public","idea-tr2-biomarker-evidence-grading","idea-tr2-tmb-calibration-standard","idea-tr2-hrd-functional-standard","idea-tr2-preanalytics-in-report","idea-tr2-cutpoint-lock","idea-tr2-spatial-biomarker-standards","idea-tr2-radiomics-ibsi-mandate","idea-tr2-biomarker-cwe","idea-tr2-label-assay-concordance","idea-tr2-positivity-rate-surveillance","idea-moon-biomarker-validation-utility"],"cancers":["breast-her2-positive","breast-hr-positive","nsclc","ovarian","gastric","urothelial"],"sections":[],"technologies":["histopathology-ihc","companion-diagnostic","digital-pathology-ai","hrd-testing","cgp","trop2-pet","her2-pet"],"targets":["her2","pdl1","trop2","cldn18-2"],"drugs":["foundationone-cdx","trastuzumab-deruxtecan"],"companies":["foundation-medicine","paige","pathai","roche-genentech"],"institutions":[],"pathways":[],"terms":["her2-low","tps","cps","tmb","hrd","msi","ihc","companion-diagnostic-term"],"trials":["destiny-breast04","destiny-breast06"],"people":[],"bottlenecks":[],"keyPapers":["paper-fernandez-jama-oncol","paper-hirsch-j-thorac-oncol","paper-merino-j-immunother-cancer"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Pathologist concordance in classifying HER2 IHC 0 vs 1+ (the HER2-low boundary) across 18 pathologists","value":"26% agreement","source":"Fernandez et al., JAMA Oncology 2022","url":"https://doi.org/10.1001/jamaoncol.2021.7239"},{"label":"PD-L1 assays in the Blueprint comparison project whose tumour-cell staining was interchangeable (22C3, 28-8, SP263) vs discordant (SP142)","value":"3 concordant, 1 discordant","source":"Hirsch et al., Journal of Thoracic Oncology 2017","url":"https://doi.org/10.1016/j.jtho.2016.11.2228"}],"causes":["Each drug sponsor develops its own companion assay and cut-off, with no obligation to harmonise.","Immunohistochemistry is semi-quantitative and subject to pre-analytical variation and inter-observer disagreement.","Retrospective validation within a single trial is accepted by regulators as sufficient.","External quality assurance schemes are voluntary in many countries.","Reimbursement for testing lags approval, so laboratories adopt cheaper, non-validated substitutes."],"currentEfforts":["The Blueprint PD-L1 IHC Assay Comparison Project and Friends of Cancer Research TMB Harmonization Project align assays across vendors.","The ASCO-CAP HER2 guideline update (2023) added HER2-low reporting standards, and AI-assisted HER2 scoring is being validated to reduce inter-observer variability.","NordiQC and UK NEQAS run external quality assessment for immunohistochemistry and molecular pathology.","The EU In Vitro Diagnostic Regulation (IVDR) raises evidence requirements for companion diagnostics.","FoundationOne CDx and other pan-tumour companion diagnostics consolidate multiple biomarkers on one validated platform.","TROP2 PET and other quantitative imaging biomarkers are being developed as alternatives to tissue IHC for ADC target selection."],"successLooksLike":"Every biomarker used to allocate a drug has a harmonised assay with published analytical validation, participates in mandatory external quality assurance, and produces the same treatment decision for the same sample in any accredited laboratory."},{"id":"b-cachexia-supportive","kind":"bottleneck","name":"Cachexia, toxicity and the limits of the patient","aka":[],"tldr":"Cancer cachexia, the muscle and fat wasting driven by tumour and host inflammatory signals, affects most patients with advanced pancreatic, gastric and lung cancer, and treatment-limiting toxicities decide what dose a patient can receive. Only Japan has an approved cachexia drug, and supportive care research gets a small share of funding relative to its effect.","summary":"Cancer cachexia, a syndrome of muscle and fat loss driven by tumour-derived and host inflammatory signals (IL-6, GDF15, activin), affects most patients with advanced pancreatic, gastric and lung cancer and is implicated in a large share of cancer deaths, yet anamorelin in Japan is the only approved drug anywhere and there is none in the US or Europe. Beyond cachexia, treatment-limiting toxicities determine what dose a patient can receive: neuropathy, cardiotoxicity, cytopenias, interstitial lung disease from ADCs, cytokine release and neurotoxicity from cell therapies, and fatigue. Supportive-care research receives a small share of funding relative to its effect on survival and quality of life, and effective interventions such as structured exercise and geriatric assessment are rarely prescribed. Treating the host is a therapeutic target in its own right.","asOf":"2026-09-08","links":[{"label":"Fearon et al., Definition and classification of cancer cachexia (Lancet Oncology 2011)","url":"https://doi.org/10.1016/S1470-2045(10)70218-7"},{"label":"Groarke et al., Ponsegromab for cancer cachexia (NEJM 2024)","url":"https://doi.org/10.1056/NEJMoa2409515"},{"label":"Roeland et al., ASCO guideline on management of cancer cachexia (JCO 2020)","url":"https://doi.org/10.1200/JCO.20.00611"}],"tags":[],"related":["idea-exercise-as-adjuvant","idea-bio2-protein-plus-training-during-io","idea-bio2-remote-weight-step-monitoring"],"cancers":["pancreatic","gastric","nsclc","esophageal","head-and-neck"],"sections":["supportive-care"],"technologies":["cardio-oncology","exercise-oncology","scalp-cooling","geriatric-assessment"],"targets":[],"drugs":[],"companies":["pfizer"],"institutions":["asco","cruk"],"pathways":[],"terms":["trastuzumab-cardiotoxicity","ild","crs","icans","irae"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-fearon-lancet-oncol","paper-roeland-j-clin-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"major","metrics":[{"label":"Share of cancer deaths in which cachexia is the immediate cause (estimate)","value":"Up to 20%","source":"Argilés et al., Nature Reviews Cancer 2014","url":"https://doi.org/10.1038/nrc3829"},{"label":"Weight gain with the GDF-15 antibody ponsegromab vs placebo at 12 weeks in cancer cachexia (highest dose)","value":"+5.6% body weight difference","source":"Groarke et al., NEJM 2024","url":"https://doi.org/10.1056/NEJMoa2409515"},{"label":"Physician under-reporting of symptomatic toxicities compared with patient self-report across three randomised trials","value":"Substantial under-reporting for all six symptoms studied","source":"Di Maio et al., JCO 2015","url":"https://doi.org/10.1200/JCO.2014.57.9334"}],"causes":["Cachexia is driven by systemic inflammation and neuroendocrine signalling that tumour-directed therapy does not address.","Supportive-care trials have no patent-protected asset to fund them and attract little industry investment.","Toxicity is graded by clinicians on the CTCAE scale, which under-captures patient-experienced symptoms.","Dose-finding uses maximum tolerated dose, so many regimens start above what most patients can sustain.","Nutrition, exercise and psycho-oncology are delivered inconsistently and rarely reimbursed as oncology care."],"currentEfforts":["Pfizer's ponsegromab (anti-GDF-15) produced weight gain and improved activity in a phase 2 trial and is in phase 3.","Anamorelin (Ono) is approved in Japan for cachexia in lung, gastric, pancreatic and colorectal cancer.","The Cancer Grand Challenges CANCAN team is dissecting the neural and metabolic circuits of cachexia.","ASCO published its first cachexia management guideline in 2020, and PRO-CTCAE (NCI) provides a validated patient-reported toxicity instrument.","Cardio-oncology and exercise-oncology services (including the CHALLENGE trial of structured exercise in colon cancer) are formalising host-directed care.","Scalp cooling, biosimilar growth factors and ILD-monitoring algorithms for ADCs are reducing treatment-limiting toxicity."],"successLooksLike":"An anti-cachexia drug improves survival or function in a phase 3 trial and is approved in the US and EU; patient-reported toxicity is collected in every trial and used for dose decisions; supportive-care interventions with survival benefit are reimbursed and delivered to most eligible patients."},{"id":"b-tme-immunosuppression","kind":"bottleneck","name":"Cold tumours and the immunosuppressive microenvironment","aka":[],"tldr":"Most tumours keep the immune system out or asleep, so immunotherapy helps only a minority.","summary":"Checkpoint blockade is the one therapeutic advance of the last two decades for which more than two fifths of US cancer patients are eligible, yet it works durably in a minority of patients and barely at all in several of the cancers that kill the most people. Pancreatic, prostate, most breast and microsatellite-stable colorectal cancers and glioblastoma are 'cold': few infiltrating T cells, low neoantigen load, dense desmoplastic stroma, abundant myeloid suppressor cells and regulatory T cells, hypoxia, and metabolic competition that starves effector cells. Even in 'hot' tumours, exhaustion and antigen loss limit durability. The field has many candidate mechanisms and many single-agent attempts (IDO, TIGIT, STING agonists, oncolytic viruses, CD47) that have failed in phase 3 because the biology of each tumour's exclusion is different and largely unmeasured. Converting cold tumours, or bypassing the microenvironment with engineered cells, bispecifics and radioligands, is the central strategic problem in immuno-oncology.","asOf":"2026-09-08","links":[{"label":"Haslam & Prasad, Estimation of the percentage of US patients eligible for and responding to checkpoint inhibitors (JAMA Netw Open 2019)","url":"https://doi.org/10.1001/jamanetworkopen.2019.2535"},{"label":"Chen & Mellman, Elements of cancer immunity and the cancer-immunity set point (Nature 2017)","url":"https://doi.org/10.1038/nature21349"},{"label":"Binnewies et al., Understanding the tumor immune microenvironment (Nature Medicine 2018)","url":"https://doi.org/10.1038/s41591-018-0014-x"}],"tags":[],"related":["idea-immunotherapy-mss-crc","idea-fap-theranostics-pancancer","idea-mesothelin-car-t-regional","idea-photoimmunotherapy-plus-pd1","immunotherapy-roadmap","cellxgene-hca","idea-bio2-trem2-myeloid-reprogramming","idea-bio2-in-situ-vaccination-solid","idea-bio2-oncolytic-regulated-il12","idea-bio2-engineered-bacteria-payloads","idea-bio2-tumour-anchored-tgfbeta-trap","idea-bio2-lactate-acid-axis","idea-bio2-radiotherapy-sting-fractionation","idea-bio2-tertiary-lymphoid-induction","idea-bio2-cxcr2-neutrophil-blockade","idea-bio2-caf-subtype-assignment","idea-bio2-implantable-microdevice-screen","idea-bio2-vascular-normalisation-window","idea-bio2-complement-c5ar-blockade","idea-bio2-intracavitary-immunotherapy","idea-bio2-myeloid-engager-bispecific","idea-bio2-hypoxia-guided-adenosine","idea-bio2-histotripsy-immune-priming","idea-bio2-trained-immunity-priming","idea-moon-cold-to-hot-programme"],"cancers":["pancreatic","prostate","colorectal","glioblastoma","breast-hr-positive","ovarian"],"sections":[],"technologies":["checkpoint-inhibitor","t-cell-engager","sting-agonist","oncolytic-virus","neoantigen-mrna-vaccine","car-t","photoimmunotherapy","single-cell-spatial","cytokine-therapy"],"targets":["pd1","pdl1","ctla4","lag3","tigit","fap","cd47","dll3"],"drugs":["tarlatamab","fap-2286","intismeran-autogene","autogene-cevumeran","vusolimogene-oderparepvec","tiragolumab","magrolimab","epacadostat","adu-s100"],"companies":["moderna","biontech","replimune","10x-genomics"],"institutions":[],"pathways":["pd1-checkpoint","cgas-sting"],"terms":["cold-vs-hot","tils","tmb","msi","desmoplasia","neoantigen","immunogenic-cell-death","abscopal-effect"],"trials":["keynote-177","checkmate-8hw","niche-2","napoli-3","interpath-001"],"people":[],"bottlenecks":[],"keyPapers":["paper-haslam-jama-netw-open","paper-chen-nature"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"US patients with cancer eligible for a checkpoint inhibitor (2018)","value":"43.6%","source":"Haslam & Prasad, JAMA Network Open 2019","url":"https://doi.org/10.1001/jamanetworkopen.2019.2535"},{"label":"US patients with cancer estimated to respond to a checkpoint inhibitor (2018)","value":"12.5%","source":"Haslam & Prasad, JAMA Network Open 2019","url":"https://doi.org/10.1001/jamanetworkopen.2019.2535"},{"label":"Median progression-free survival with pembrolizumab vs chemotherapy in MSI-high colorectal cancer (KEYNOTE-177), showing the gain when a cold cancer is immunogenic","value":"16.5 vs 8.2 months","source":"André et al., NEJM 2020","url":"https://doi.org/10.1056/NEJMoa2017699"}],"causes":["Low mutational burden yields few neoantigens for T cells to recognise.","Desmoplastic stroma and abnormal vasculature physically exclude lymphocytes.","Myeloid-derived suppressor cells, tumour-associated macrophages and regulatory T cells actively switch off effector cells.","Hypoxia, lactate, adenosine and nutrient depletion in the tumour disable T-cell metabolism.","Loss of MHC class I and antigen-presentation machinery makes tumours invisible even when T cells are present.","Combination trials have mostly added agents empirically without a biomarker for the specific exclusion mechanism."],"currentEfforts":["T-cell engagers and CAR-T bypass the need for endogenous priming; tarlatamab (DLL3) and mesothelin or GD2 CAR-T show that engineered T cells can act in solid tumours.","STING agonists, oncolytic viruses (RP1, T-VEC) and photoimmunotherapy are being tested as in situ vaccines to turn cold lesions hot.","FAP-targeted radioligands and FAP theranostics attack the fibroblast stroma directly.","Neoantigen mRNA vaccines (Moderna/Merck intismeran, BioNTech autogene cevumeran) aim to generate T-cell responses where none exist, including in pancreatic cancer.","Trials in MSS colorectal cancer combine checkpoint blockade with VEGF, MEK or bispecific agents, and NICHE-2 showed dramatic neoadjuvant responses in dMMR disease.","Spatial transcriptomics and multiplex imaging (10x Genomics, Human Cell Atlas) are being used to classify exclusion mechanisms per tumour."],"successLooksLike":"A biomarker-defined strategy produces durable responses in a majority of patients with pancreatic, prostate, MSS colorectal or glioblastoma, and the fraction of all cancer patients who respond to immunotherapy rises from roughly one in eight to a majority."},{"id":"b-data-silos","kind":"bottleneck","name":"Data silos","aka":[],"tldr":"Records, scans, genomes and outcomes sit in separate systems that cannot talk. Every patient's experience is lost to the next.","summary":"The overwhelming majority of patients with cancer are treated outside trials, and what happens to them, their genomics, imaging, pathology, treatment, toxicity and outcome, is recorded in electronic records, laboratory systems, PACS archives and registries that are not linked and cannot be queried together. Trial data are shared rarely: after journals required data-sharing statements, individual patient data could actually be obtained for about 1% of the trials studied. Privacy law, consent models, vendor lock-in, unstructured free text, absent common data standards and the lack of any incentive to share mean that the largest source of evidence, routine care, teaches the system almost nothing, and that each institution's AI is trained on its own slice. Interoperability standards (FHIR, mCODE), federated learning, national health data spaces, and consented patient-controlled data are the technical and legal answers; the missing piece is an obligation to contribute.","asOf":"2026-09-08","links":[{"label":"AACR Project GENIE","url":"https://www.aacr.org/professionals/research/aacr-project-genie/"},{"label":"mCODE: Minimal Common Oncology Data Elements","url":"https://mcodeinitiative.org/"},{"label":"European Health Data Space","url":"https://health.ec.europa.eu/ehealth-digital-health-and-care/european-health-data-space_en"},{"label":"Danchev et al., Evaluation of data sharing after implementation of the ICMJE requirement (JAMA Netw Open 2021)","url":"https://doi.org/10.1001/jamanetworkopen.2020.33972"}],"tags":[],"related":["genie","cbioportal","tcga-gdc","cptac","seer","clinicaltrials-gov","idea-multimodal-foundation-model","idea-data-mcode-payment-condition","idea-data-oncology-api-certification","idea-data-national-cancer-data-space","idea-data-donor-card","idea-data-dynamic-consent-app","idea-data-consent-in-the-pathway","idea-data-accredited-tre-network","idea-data-synthetic-companion-datasets","idea-data-open-deidentification-pipeline","idea-data-national-slide-archive","idea-data-automatic-outcome-linkage","idea-data-open-line-of-therapy-algorithms","idea-data-point-of-care-randomisation","idea-data-procurement-open-api-clauses","idea-data-privacy-preserving-linkage-tokens","idea-data-atlas-to-outcome-linkage","idea-data-radiotherapy-dose-repository","idea-data-surgical-video-commons","idea-data-ipd-deposit-enforcement","idea-data-common-biobank-consent-mta","idea-data-cross-border-rare-cancer-pool","idea-data-quality-scorecards","idea-data-structured-radiology-response","idea-moon-universal-sequencing-learning-system"],"cancers":[],"sections":["ai-computation"],"technologies":["cgp","digital-pathology-ai","pathology-foundation-model"],"targets":[],"drugs":[],"companies":["tempus","owkin","bostongene","patient-data-vault","cancer-commons","foundation-medicine","guardant-health"],"institutions":["aacr","nci","broad-institute","mskcc","dana-farber"],"pathways":[],"terms":["real-world-evidence","ngs"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-danchev-jama-netw-open"],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"critical","metrics":[{"label":"Adults with cancer participating in treatment trials, and hence the share of patients whose outcomes systematically feed evidence","value":"~8%","source":"Unger et al., JNCI 2019","url":"https://doi.org/10.1093/jnci/djy221"},{"label":"Trials with ICMJE data-sharing statements whose individual participant data could actually be obtained by requesters","value":"6 of 487 (about 1%)","source":"Danchev et al., JAMA Network Open 2021","url":"https://doi.org/10.1001/jamanetworkopen.2020.33972"}],"causes":["Electronic health records are optimised for billing and documentation, not for structured clinical data capture.","Vendors and institutions treat data as a proprietary asset.","Privacy law and ethics review make linkage across institutions slow and expensive.","Key variables (stage, response, progression, toxicity) are recorded in free text or not at all.","There is no reward, and often a penalty, for sharing data."],"currentEfforts":["AACR Project GENIE pools clinical-grade sequencing and outcomes from more than a dozen cancer centres for open use.","mCODE (Minimal Common Oncology Data Elements) and HL7 FHIR define a common structured oncology data model, now adopted in US regulation for interoperability.","The European Health Data Space Regulation (2025) creates a legal basis for secondary use of health data across the EU, and Health Data Research UK links NHS datasets for research.","The NCI Genomic Data Commons, cBioPortal and CPTAC make research-grade genomic and proteomic data openly available.","Owkin, Tempus, BostonGene and Flatiron apply federated learning or curated real-world datasets across institutions.","Patient-controlled data initiatives (Cancer Commons, Patient Data Vault, Count Me In at the Broad Institute) let patients share their own records for research."],"successLooksLike":"Structured treatment and outcome data are captured for every patient with cancer and linkable across institutions under a standard consent, and a researcher or regulator can answer a comparative effectiveness question in weeks from routine data with results that match randomised trials."},{"id":"b-dormancy-mrd","kind":"bottleneck","name":"Dormant cells and minimal residual disease","aka":[],"tldr":"After a 'successful' treatment, cells can sleep for years then relapse. We can barely detect them and cannot target them.","summary":"Disseminated tumour cells can persist for years in bone marrow, liver, lung or brain in a non-proliferative state, invisible to imaging and untouched by cytotoxic agents that require cell division. In oestrogen receptor-positive breast cancer, distant recurrence continues at a steady rate for at least twenty years after diagnosis; in colorectal, bladder and lung cancer, detectable circulating tumour DNA after curative surgery predicts relapse with high specificity months before scans. Detection is now possible, but the therapeutic side lags: there are few drugs designed to kill quiescent cells or to keep them asleep, no validated biomarkers of dormancy, and until recently no trials that acted on a molecular residual disease signal. The first ctDNA-guided adjuvant trials show that the signal can be used to de-escalate safely; whether escalation on a positive test improves survival is the open question.","asOf":"2026-09-08","links":[{"label":"Tie et al., ctDNA analysis guiding adjuvant therapy in stage II colon cancer (NEJM 2022)","url":"https://doi.org/10.1056/NEJMoa2200075"},{"label":"Pan et al., 20-year risks of breast-cancer recurrence (NEJM 2017)","url":"https://doi.org/10.1056/NEJMoa1701830"},{"label":"Phan & Croucher, The dormant cancer cell life cycle (Nature Reviews Cancer 2020)","url":"https://doi.org/10.1038/s41568-020-0263-0"}],"tags":[],"related":["idea-late-recurrence-interception","idea-ctdna-guided-adjuvant-crc","idea-ctdna-guided-adjuvant-melanoma","idea-mrd-guided-stop-myeloma","idea-ctdna-escalation-tnbc","idea-bio2-dormancy-selective-screen","idea-bio2-autophagy-block-mrd","idea-bio2-pro-dormancy-therapy","idea-bio2-il1-awakening-blockade","idea-bio2-national-mrd-platform","idea-bio2-mrd-reference-standards","idea-bio2-mrd-triggered-neoantigen-vaccine","idea-bio2-nk-cells-for-mrd","idea-bio2-whole-genome-mrd-depth","idea-bio2-marrow-niche-on-chip","idea-bio2-localise-the-mrd","idea-bio2-circadian-mrd-sampling","idea-bio2-mrd-guided-endocrine-duration","idea-bio2-mrd-clinic-service","idea-bio2-survivor-plasma-biobank","idea-moon-engineered-immune-surveillance","idea-moon-mrd-weather-service","idea-moon-dormancy-eradication-programme"],"cancers":["breast-hr-positive","colorectal","urothelial","nsclc","multiple-myeloma","aml","all-leukemia"],"sections":[],"technologies":["mrd-testing","ngs-mrd-clonoseq","liquid-biopsy","flow-cytometry-mrd","ctdna-lymphoma-monitoring"],"targets":[],"drugs":["signatera"],"companies":["natera","foresight-diagnostics","adaptive-biotechnologies","guardant-health"],"institutions":[],"pathways":[],"terms":["mrd","ctdna","late-recurrence","mrd-negativity-myeloma","mrd-negative-cr","neoadjuvant-adjuvant"],"trials":["dynamic","circulate-japan","imvigor011","cambria","natalee"],"people":[],"bottlenecks":[],"keyPapers":["paper-phan-nat-rev-cancer","paper-pan-n-engl-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Risk of distant recurrence in years 5-20 for ER-positive breast cancer after 5 years of endocrine therapy, by original stage","value":"10% to 41%","source":"Pan et al., NEJM 2017","url":"https://doi.org/10.1056/NEJMoa1701830"},{"label":"Adjuvant chemotherapy use with ctDNA-guided vs standard management in stage II colon cancer (DYNAMIC), with non-inferior recurrence-free survival","value":"15% vs 28%","source":"Tie et al., NEJM 2022","url":"https://doi.org/10.1056/NEJMoa2200075"}],"causes":["Quiescent cells do not divide, so chemotherapy and many targeted agents have nothing to act on.","Dormant cells are rare and dispersed, below the resolution of any imaging modality.","The signals that hold cells dormant or awaken them (niche, immune surveillance, inflammation, surgery) are only partly characterised and have no drugs.","Adjuvant trials treat everyone because there has been no way to identify who still harbours disease.","Reimbursement and guidelines have been slow to adopt molecular residual disease assays outside haematology."],"currentEfforts":["Natera's Signatera and Foresight Diagnostics' PhasED-Seq are tumour-informed ctDNA assays used in interventional trials of MRD-guided therapy.","DYNAMIC and CIRCULATE-Japan (GALAXY/VEGA/ALTAIR) test ctDNA-guided escalation and de-escalation of adjuvant chemotherapy in colon cancer.","IMvigor011 randomises ctDNA-positive patients after cystectomy to atezolizumab or placebo, a phase 3 that acts only on a molecular signal.","The CAMBRIA trials in breast cancer test switching to an oral SERD in patients at risk of late recurrence.","clonoSEQ (Adaptive Biotechnologies) is FDA-cleared for MRD in myeloma and leukaemia, and MRD negativity has been accepted by the FDA as an endpoint for accelerated approval in myeloma.","Cancer Grand Challenges funds teams working specifically on the biology of tumour dormancy."],"successLooksLike":"A positive molecular residual disease test after curative therapy triggers a treatment that is proven to prevent relapse and improve survival, and MRD-negative patients are safely spared adjuvant therapy in the major solid tumours."},{"id":"b-negative-results","kind":"bottleneck","name":"Failures are hidden","aka":[],"tldr":"Negative trials, failed drugs and abandoned programmes are rarely published, so the same mistakes are repeated.","summary":"A large share of completed clinical trials never report results, and negative trials are published later and less often than positive ones. In the first years after the US results-reporting mandate took effect, only about one in eight applicable trials posted results on ClinicalTrials.gov within the required year, and fewer than half of NIH-funded trials were published within thirty months of completion. Commercial programmes that fail in phase 1 or 2 are usually terminated silently, and preclinical negative results are almost never shared, so companies and academic groups repeat the same experiments and the same mistakes. The costs are duplicated trials, patients exposed to hypotheses already refuted, and meta-analyses biased toward benefit. Enforcement of existing law, journals that publish nulls, and registries of terminated programmes and preclinical negatives would fix most of it.","asOf":"2026-09-08","links":[{"label":"Anderson et al., Compliance with results reporting at ClinicalTrials.gov (NEJM 2015)","url":"https://doi.org/10.1056/NEJMsa1409364"},{"label":"Ross et al., Publication of NIH funded trials registered in ClinicalTrials.gov (BMJ 2012)","url":"https://doi.org/10.1136/bmj.d7292"},{"label":"ClinicalTrials.gov results reporting requirements (FDAAA 801)","url":"https://clinicaltrials.gov/policy/reporting-requirements"},{"label":"AllTrials","url":"https://www.alltrials.net/"}],"tags":[],"related":["clinicaltrials-gov","eudract-ctis","pubmed-europepmc","idea-tr2-irb-results-gate","idea-tr2-material-failure-disclosure","idea-tr2-writeup-grants","idea-tr2-null-result-reporting","idea-tr2-negative-plenaries","idea-tr2-off-label-outcome-registry","idea-tr2-reversal-registry","idea-tr2-sponsor-transparency-score","idea-moon-negative-results-ledger"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":["iniparib","rovalpituzumab-tesirine","epacadostat","magrolimab","bempegaldesleukin","tiragolumab","eprenetapopt","melflufen","trastuzumab-duocarmazine","tusamitamab-ravtansine"],"companies":[],"institutions":["nci","asco"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-anderson-n-engl-j-med","paper-ross-bmj"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Applicable trials reporting results on ClinicalTrials.gov within 12 months of completion, 2008-2013","value":"13.4%","source":"Anderson et al., NEJM 2015","url":"https://doi.org/10.1056/NEJMsa1409364"},{"label":"NIH-funded trials published in a peer-reviewed journal within 30 months of completion","value":"Fewer than half (46%)","source":"Ross et al., BMJ 2012","url":"https://doi.org/10.1136/bmj.d7292"}],"causes":["Journals and careers reward positive, novel findings over nulls.","Sponsors have commercial reasons to keep failed programmes and their data private.","Reporting requirements are weakly enforced; the FDA has issued few penalties under FDAAA 801.","Preclinical work has no registration or reporting infrastructure at all.","Writing up a failure costs time and money that no one funds."],"currentEfforts":["The FDAAA 801 Final Rule (2017) requires results posting for applicable trials, and the EU Clinical Trials Regulation makes CTIS results public.","The AllTrials campaign and the TrialsTracker (EBM DataLab, Oxford) publicly name sponsors and institutions that fail to report.","Registered Reports (Center for Open Science) and journals such as PLOS ONE and eLife commit to publishing regardless of result.","The Reproducibility Project: Cancer Biology published every replication attempt, positive or negative.","ASCO's TAPUR and NCI-MATCH publish negative cohorts as a matter of policy."],"successLooksLike":"More than 90% of completed trials post results within a year, every terminated drug programme has a public record of why it stopped, and preclinical negative results are routinely deposited in a searchable registry."},{"id":"b-care-fragmentation","kind":"bottleneck","name":"Fragmented care and guideline gaps","aka":[],"tldr":"Patients fall between specialists, wait for referrals and often do not get the treatment guidelines say they should.","summary":"The distance between what guidelines recommend and what patients receive is one of the largest and least glamorous sources of avoidable death. Every four-week delay from diagnosis to treatment increases mortality by roughly 6-13% across common cancers, and delays accumulate at each hand-off between primary care, imaging, biopsy, pathology, molecular testing, multidisciplinary review and treatment. Genomic testing that determines eligibility for targeted therapy is incomplete in a large share of eligible lung cancer patients, and results arrive after first-line treatment has begun. Variation between hospitals in surgical quality, radiotherapy technique and systemic therapy use exceeds the effect size of most new drugs. The causes are organisational: no single owner of the patient pathway, information that does not travel with the patient, capacity constraints, and payment systems that reward activity rather than timeliness or adherence.","asOf":"2026-09-08","links":[{"label":"Hanna et al., Mortality due to cancer treatment delay: systematic review and meta-analysis (BMJ 2020)","url":"https://doi.org/10.1136/bmj.m4087"},{"label":"NHS England Cancer Waiting Times statistics","url":"https://www.england.nhs.uk/statistics/statistical-work-areas/cancer-waiting-times/"},{"label":"ASCO Quality Oncology Practice Initiative (archived copy)","url":"https://web.archive.org/web/20231029084627/https://practice.asco.org/quality-improvement/quality-programs/quality-oncology-practice-initiative"}],"tags":[],"related":["nccn","esmo-guidelines","nci-cancer-centers","idea-acc-one-stop-breast-diagnostic-clinic","idea-acc-28-day-diagnosis-standard","idea-acc-national-virtual-mdt-rare-complex","idea-acc-primary-care-shared-care-agreements","idea-acc-tumour-board-implementation-audit","idea-acc-diagnostic-interval-public-registry","idea-acc-reflex-biomarker-panels-by-tumour","idea-acc-rapid-diagnostic-centres-vague-symptoms","idea-acc-embedded-decision-aids","idea-acc-financial-toxicity-screening-at-diagnosis","idea-acc-universal-asynchronous-second-opinion","idea-acc-episode-payment-tied-to-concordance","idea-acc-transition-handoff-medication-reconciliation","idea-acc-pathway-capacity-simulation","idea-cost-subcutaneous-community","idea-cost-gold-card-oncology","idea-cost-real-time-pa-fhir","idea-cost-travel-teleoncology"],"cancers":["nsclc","colorectal","pancreatic","head-and-neck","esophageal"],"sections":[],"technologies":["cgp","liquid-biopsy","companion-diagnostic","ai-trial-matching"],"targets":[],"drugs":[],"companies":["foundation-medicine","guardant-health","tempus"],"institutions":["asco","esmo","nccn-org","the-christie","royal-marsden"],"pathways":[],"terms":["standard-of-care","first-line","neoadjuvant-adjuvant","companion-diagnostic-term"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-hanna-bmj"],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"major","metrics":[{"label":"Increase in mortality per four-week delay in cancer treatment (surgery, systemic, radiotherapy) across seven cancers","value":"6% to 13% per 4 weeks","source":"Hanna et al., BMJ 2020","url":"https://doi.org/10.1136/bmj.m4087"},{"label":"Patients with metastatic non-squamous NSCLC in a US community network tested for all five guideline-recommended biomarkers before first-line therapy (2018-2020)","value":"Under half","source":"Robert et al., Lung Cancer 2022","url":"https://doi.org/10.1016/j.lungcan.2022.01.021"},{"label":"NHS England standard for time from urgent referral to first treatment","value":"62 days (target 85% of patients)","source":"NHS England Cancer Waiting Times","url":"https://www.england.nhs.uk/statistics/statistical-work-areas/cancer-waiting-times/"}],"causes":["No single clinician or team owns the pathway from symptom to treatment.","Records, images and test results do not follow the patient between organisations.","Diagnostic capacity (imaging, endoscopy, pathology) is the rate-limiting step in most systems.","Molecular testing is ordered sequentially and late rather than reflexively at diagnosis.","Payment rewards volume of procedures rather than adherence to guidelines or speed.","Guidelines are long, frequently updated and not embedded in the tools clinicians use."],"currentEfforts":["NHS Rapid Diagnostic Centres and the Faster Diagnosis Standard target 28 days from referral to diagnosis or exclusion of cancer.","The NHS Genomic Medicine Service provides a national test directory and reflex testing for cancer, and ESMO and NCCN maintain guidelines with explicit biomarker testing recommendations.","ASCO's Quality Oncology Practice Initiative (QOPI) certifies practices against guideline-concordance measures.","Mandatory multidisciplinary tumour boards in the UK, EU cancer centres and NCI-designated centres reduce variation in treatment decisions.","Patient navigation programmes, now reimbursable under US Medicare (2024), assign a person to guide patients through the pathway.","Comprehensive genomic profiling and liquid biopsy at diagnosis (Foundation Medicine, Guardant, Tempus) shorten the time to actionable results."],"successLooksLike":"Median time from suspicion to first treatment is under six weeks in every high-income system, more than 90% of eligible patients have complete biomarker results before first-line therapy, and guideline-concordant treatment rates exceed 90% with little inter-hospital variation."},{"id":"b-funding-allocation","kind":"bottleneck","name":"Funding follows fashion, not burden","aka":[],"tldr":"Research money follows visibility, not burden: breast, prostate and leukaemia receive far more funding per death or year of life lost than lung, pancreatic, liver, oesophageal, gastric, bladder and uterine cancers, and metastasis research gets an estimated 5% of funding despite causing most deaths. Advocacy strength and peer review that rewards mechanism explain the skew.","summary":"Research funding across cancers correlates poorly with burden: breast cancer, leukaemia and prostate cancer receive far more NCI and charity funding per death or per year of life lost than lung, pancreatic, liver, oesophageal, gastric, bladder and uterine cancers, and the pattern is similar in the UK and Europe. Across research types the skew is stronger: metastasis, prevention, implementation, surgery, radiotherapy, supportive care and research in low-income settings are all under-funded relative to their share of deaths or their potential to prevent them, while cell-intrinsic biology and drug discovery are over-represented. The causes are the visibility and advocacy strength of some cancers, peer review that rewards mechanistic novelty, the absence of a burden-weighted portfolio strategy at most funders, and a commercial sector that follows patentability. Transparent portfolio mapping, burden-weighted funding targets, and mission-oriented programmes are the corrective levers.","asOf":"2026-09-08","links":[{"label":"Carter & Nguyen, A comparison of cancer burden and research spending (BMC Cancer 2012)","url":"https://doi.org/10.1186/1471-2407-12-526"},{"label":"NCI Funded Research Portfolio","url":"https://fundedresearch.cancer.gov/"},{"label":"International Cancer Research Partnership","url":"https://www.icrpartnership.org/"},{"label":"Cancer Grand Challenges","url":"https://cancergrandchallenges.org/"}],"tags":[],"related":["globocan","seer","idea-fund-survivorship-endowment-levy","idea-moon-cure-prizes"],"cancers":["pancreatic","hcc","gastric","esophageal","nsclc","urothelial","endometrial","glioblastoma"],"sections":["prevention","surgery","radiation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":["nci","cruk","iarc","aacr","asco"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-sun-bmc-cancer"],"journals":[],"dependsOn":[],"notes":[],"stage":"funding-incentives","severity":"major","metrics":[{"label":"Cancers under-funded relative to incidence, mortality and years of life lost in an analysis of NCI, charity and industry funding","value":"Lung, pancreas, liver, oesophagus, stomach, bladder and uterine cancers under-funded; breast, prostate and leukaemia over-funded","source":"Carter & Nguyen, BMC Cancer 2012","url":"https://doi.org/10.1186/1471-2407-12-526"},{"label":"Estimated share of cancer research funding directed at metastasis, which causes most deaths","value":"~5% (estimate)","source":"Sleeman & Steeg, European Journal of Cancer 2010","url":"https://doi.org/10.1016/j.ejca.2010.02.039"},{"label":"Share of cancer deaths occurring in low- and middle-income countries, where a small fraction of research is conducted","value":"~70%","source":"WHO fact sheet, Cancer","url":"https://www.who.int/news-room/fact-sheets/detail/cancer"}],"causes":["Advocacy strength and public visibility differ enormously between cancers.","Peer review rewards mechanistic novelty and feasibility over expected health gain.","Most funders have no explicit burden- or impact-weighted portfolio strategy.","Industry funding follows patentable, high-price opportunities, which excludes prevention, surgery and repurposing.","Investigator supply is self-reinforcing: fields with money train the next generation, which applies for more."],"currentEfforts":["The NCI Funded Research Portfolio and the International Cancer Research Partnership (ICRP) publish funding by cancer type and Common Scientific Outline category, making the skew measurable.","Cancer Grand Challenges (Cancer Research UK and NCI) funds large teams on defined neglected problems including cachexia, dormancy and early-onset cancers.","Cancer Research UK's research strategy prioritises prevention, early detection and cancers of unmet need including lung, pancreatic, oesophageal and brain.","The Lancet Oncology Commissions on global cancer surgery and radiotherapy quantified the neglected return on investment in those modalities.","The US Cancer Moonshot and ARPA-H fund mission-directed programmes outside conventional peer review.","Disease-specific foundations for pancreatic, lung and brain cancers have grown rapidly to correct the advocacy imbalance."],"successLooksLike":"Funders publish burden-weighted portfolio analyses annually, funding per year of life lost varies less than two-fold across the major cancers, and metastasis, prevention, surgery and radiotherapy each receive a share of research funding commensurate with their contribution to deaths and cures."},{"id":"b-incentive-misalignment","kind":"bottleneck","name":"Incentives reward me-too drugs and marginal gains","aka":[],"tldr":"The system pays the same for a drug that adds two months as for a cure, so companies race to copy rather than to cure.","summary":"The economics of oncology reward being the fifth entrant to a validated target more reliably than being the first to attempt an unsolved one. Thousands of trials have tested PD-1/PD-L1 antibodies, more than a dozen of which are approved with near-identical activity, and TROP2 and HER2 ADCs, KRAS G12C inhibitors and BCMA-directed therapies each have crowded fields, while first-in-class attempts on the hardest problems (pancreatic cancer, glioblastoma, metastasis prevention, cachexia) are few. Prices are set with no relation to benefit, so a drug adding two months of median survival can be priced like a cure, and regulatory precedent makes the follow-on path cheaper and more predictable. Patents reward molecules rather than outcomes; nothing pays for the trial that shows a drug can be stopped, given for less time, or replaced by a generic. Aligning reward with magnitude of benefit, through value-based pricing, benefit-graded exclusivity, prizes and public development of unattractive assets, is the structural fix.","asOf":"2026-09-08","links":[{"label":"Upadhaya et al., PD1/PDL1 inhibitor clinical trial landscape (Nat Rev Drug Discov 2022)","url":"https://doi.org/10.1038/d41573-022-00030-4"},{"label":"Mailankody & Prasad, Five years of cancer drug approvals: innovation, efficacy, and costs (JAMA Oncology 2015)","url":"https://doi.org/10.1001/jamaoncol.2015.0373"},{"label":"Vokinger et al., Prices and clinical benefit of cancer drugs (Lancet Oncology 2020)","url":"https://doi.org/10.1016/S1470-2045(20)30139-X"},{"label":"ARPA-H","url":"https://arpa-h.gov/"}],"tags":[],"related":["fda-approvals","esmo-guidelines","trop2-adc-roadmap","idea-fund-public-option-generic-oncology"],"cancers":["pancreatic","glioblastoma","nsclc","tnbc","multiple-myeloma"],"sections":[],"technologies":["checkpoint-inhibitor","adc","kras-inhibitors"],"targets":["pd1","pdl1","trop2","her2","bcma","kras"],"drugs":["pembrolizumab","nivolumab","tislelizumab","toripalimab","camrelizumab","serplulimab","sacituzumab-govitecan","datopotamab-deruxtecan","sacituzumab-tirumotecan","imatinib"],"companies":[],"institutions":["esmo","asco","nci"],"pathways":[],"terms":["accelerated-approval","breakthrough-designation","standard-of-care","first-line","project-frontrunner"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-upadhaya-nat-rev-drug-discov","paper-vokinger-lancet-oncol","paper-mailankody-jama-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"funding-incentives","severity":"critical","metrics":[{"label":"Clinical trials of PD-1/PD-L1 inhibitors registered by 2021","value":"5,683 trials","source":"Upadhaya et al., Nature Reviews Drug Discovery 2022","url":"https://doi.org/10.1038/d41573-022-00030-4"},{"label":"Median overall survival gain of cancer drugs approved by FDA 2002-2014","value":"2.1 months","source":"Fojo, Mailankody & Lo, JAMA Otolaryngology 2014","url":"https://doi.org/10.1001/jamaoto.2014.1570"},{"label":"Correlation between cancer drug prices and clinical benefit (ESMO-MCBS, ASCO framework) in the US and Europe","value":"No significant association","source":"Vokinger et al., Lancet Oncology 2020","url":"https://doi.org/10.1016/S1470-2045(20)30139-X"},{"label":"Relationship between price and novelty (first-in-class vs next-in-class) or benefit for cancer drugs approved by FDA 2009-2013","value":"No difference in price by novelty; no relation to benefit","source":"Mailankody & Prasad, JAMA Oncology 2015","url":"https://doi.org/10.1001/jamaoncol.2015.0373"}],"causes":["Price is unrelated to benefit, so a marginal drug earns as much per patient as a transformative one.","Validated targets carry lower scientific and regulatory risk, which capital markets prefer.","Regulatory precedent and surrogate endpoints make follow-on approvals faster and cheaper.","Patent exclusivity rewards new molecules, not new evidence about old ones or about how to use less.","The hardest problems have long, uncertain, expensive development paths with no interim commercial return."],"currentEfforts":["ESMO-MCBS and the ASCO Value Framework grade benefit so that payers and guidelines can distinguish transformative from marginal drugs.","The US Inflation Reduction Act (2022) introduces Medicare price negotiation, weakening the link between market entry and guaranteed price.","The EU pharmaceutical legislation reform proposal (2023) modulates regulatory data protection according to unmet need and comparative trials.","ARPA-H and the Cancer Moonshot fund high-risk programmes that private capital avoids.","FDA Project FrontRunner encourages first-line development of genuinely novel agents rather than late-line me-too positioning.","Cancer Grand Challenges and philanthropic funders (Stand Up To Cancer) support first-in-class attempts on neglected problems."],"successLooksLike":"Reward for a new cancer drug scales with its measured benefit, so that the expected return on a curative or first-in-class programme for an unsolved cancer exceeds that of a sixth entrant to a crowded class, and the share of new approvals that are first-in-class with substantial benefit rises measurably."},{"id":"b-hereditary-risk","kind":"bottleneck","name":"Inherited risk is mostly unidentified","aka":[],"tldr":"Most people who carry a high-risk cancer gene do not know it until they or a relative gets cancer.","summary":"Pathogenic germline variants in BRCA1/2, the Lynch syndrome mismatch-repair genes, TP53, PALB2, CDH1 and others confer lifetime cancer risks of 40-80%, and effective risk reduction exists (risk-reducing surgery, intensified surveillance, aspirin, PARP inhibitors when cancer occurs). Yet most carriers are identified only after a cancer diagnosis, and often not even then: fewer than a fifth of US women with a history of breast or ovarian cancer who met testing criteria had been tested. Cascade testing of relatives, the cheapest way to find healthy carriers, reaches a minority of eligible family members. Germline testing is also inequitable, with reference databases dominated by European ancestry, so variants of uncertain significance are more common in other populations. Polygenic risk scores could stratify screening for common cancers but are not yet implemented or validated across ancestries.","asOf":"2026-09-08","links":[{"label":"Childers et al., National estimates of genetic testing in women with breast or ovarian cancer (JCO 2017)","url":"https://doi.org/10.1200/JCO.2017.73.6314"},{"label":"Win et al., Prevalence and penetrance of major genes for colorectal cancer (CEBP 2017)","url":"https://doi.org/10.1158/1055-9965.EPI-16-0693"},{"label":"ClinVar","url":"https://www.ncbi.nlm.nih.gov/clinvar/"}],"tags":[],"related":["clinvar","idea-interception-vaccines","idea-prev-cascade-direct-contact-relatives","idea-prev-deferred-disclosure-newborn-genomes","idea-prev-prs-screening-start-age","idea-prev-prs-ancestry-portability-standard","idea-prev-reflex-germline-testing","idea-prev-traceback-deceased-probands","idea-prev-vus-saturation-editing-consortium","idea-prev-li-fraumeni-mri-plus-cfdna","idea-prev-genetic-counselling-chatbot","idea-prev-family-history-auto-match","idea-prev-genetic-non-discrimination-insurance"],"cancers":["ovarian","tnbc","breast-hr-positive","colorectal","endometrial","pancreatic","prostate","gastric"],"sections":[],"technologies":["germline-testing","risk-reducing-salpingectomy","colorectal-screening","chemoprevention"],"targets":["brca","tp53"],"drugs":["olaparib"],"companies":["natera","illumina"],"institutions":[],"pathways":[],"terms":["germline-vs-somatic","lynch-syndrome","vus","men1-hereditary-net","hrd"],"trials":["olympia"],"people":[],"bottlenecks":[],"keyPapers":["paper-childers-j-clin-oncol","paper-win-cancer-epidemiol-biomarkers-prev"],"journals":[],"dependsOn":[],"notes":[],"stage":"prevention-detection","severity":"major","metrics":[{"label":"US women with a history of breast or ovarian cancer meeting NCCN criteria who had undergone genetic testing (2005-2015 NHIS)","value":"Fewer than 20%","source":"Childers et al., JCO 2017","url":"https://doi.org/10.1200/JCO.2017.73.6314"},{"label":"Estimated population prevalence of Lynch syndrome","value":"About 1 in 279","source":"Win et al., Cancer Epidemiology, Biomarkers & Prevention 2017","url":"https://doi.org/10.1158/1055-9965.EPI-16-0693"},{"label":"Share of participants in genome-wide association studies of European ancestry (2016)","value":"81%","source":"Popejoy & Fullerton, Nature 2016","url":"https://doi.org/10.1038/538161a"}],"causes":["Testing is triggered by family history, which is often unknown, incomplete or not asked about.","Cascade testing depends on the index patient informing relatives, with no systematic follow-through.","Genetic counselling capacity is limited and testing criteria are complex.","Reference databases under-represent non-European ancestries, producing more uncertain variants.","Population screening for germline variants has been considered too costly, although sequencing costs have fallen sharply."],"currentEfforts":["The NHS Jewish BRCA Testing Programme offers population-based BRCA testing to people with Jewish ancestry in England.","Universal tumour screening for Lynch syndrome in all colorectal and endometrial cancers is guideline-recommended and being audited in the UK and US.","NCCN and ESMO guidelines have broadened germline testing criteria, including all patients with pancreatic, ovarian and metastatic prostate cancer.","The WISDOM trial tests risk-based breast screening incorporating polygenic risk scores against annual mammography.","ClinVar and ENIGMA/InSiGHT expert panels curate variant classification and reduce uncertain variants.","Mainstreaming programmes let oncologists order germline tests directly, bypassing the counselling bottleneck."],"successLooksLike":"A majority of carriers of high-penetrance variants are identified before a cancer diagnosis, cascade testing reaches most first-degree relatives, and variant classification and polygenic risk perform equally well across ancestries."},{"id":"b-knowledge-diffusion","kind":"bottleneck","name":"Knowledge reaches practice too slowly","aka":[],"tldr":"Knowledge diffusion is slow: it takes years for a proven result to change what most patients receive, and no one can keep up with the literature.","summary":"The often-quoted estimate that it takes 17 years for research evidence to reach clinical practice is an average across medicine, and oncology, with more than a million and a half new biomedical citations indexed in MEDLINE each year and guidelines that change several times a year in the commonest cancers, is at the demanding end. Oncologists in community practice cannot read the primary literature, guideline documents run to hundreds of pages, and knowledge is fragmented across specialties, so uptake of new standards (for example, biomarker testing, de-escalation, geriatric assessment) lags years behind evidence and varies widely between hospitals. Patients, meanwhile, cannot find trustworthy, current, plain-language information about their own disease and options. Living guidelines, curated knowledge bases embedded in clinical systems, decision support at the point of care, and open plain-language resources are the mechanisms for closing the gap.","asOf":"2026-09-08","links":[{"label":"Morris, Wooding & Grant, The answer is 17 years, what is the question (JRSM 2011)","url":"https://doi.org/10.1258/jrsm.2011.110180"},{"label":"NCI PDQ Cancer Information Summaries","url":"https://www.cancer.gov/publications/pdq"},{"label":"OncoKB","url":"https://www.oncokb.org/"},{"label":"MEDLINE citation counts by year","url":"https://www.nlm.nih.gov/bsd/medline_cit_counts_yr_pub.html"}],"tags":[],"related":["nccn","esmo-guidelines","oncokb","civic","cancer-gov-pdq","pubmed-europepmc","idea-data-thirty-day-practice-change-learning","idea-data-patient-explainers-per-recommendation","idea-data-evidence-to-adoption-tracker","idea-data-retraction-propagation","idea-data-open-licensed-guidelines","idea-data-unanswered-questions-registry","idea-data-living-llm-oncology-benchmark","idea-data-rapid-guideline-translation","idea-data-offline-cds-lmic-generalists","idea-data-toxicity-cds-for-nurses","idea-data-72-hour-second-opinion-network","idea-data-standard-of-care-change-alerts","idea-data-full-data-with-abstract","idea-data-implementation-trials-programme","idea-data-wikipedia-oncology-editors","idea-data-order-set-defaults-30-days","idea-data-standard-of-care-api","idea-moon-living-machine-readable-guidelines"],"cancers":["nsclc","breast-her2-positive","colorectal"],"sections":[],"technologies":["ai-trial-matching","cgp","geriatric-assessment"],"targets":[],"drugs":[],"companies":[],"institutions":["mskcc","nccn-org","esmo","asco","nci"],"pathways":[],"terms":["standard-of-care","her2-low","companion-diagnostic-term"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-morris-j-r-soc-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"major","metrics":[{"label":"Estimated average time for research evidence to reach clinical practice","value":"17 years","source":"Morris, Wooding & Grant, Journal of the Royal Society of Medicine 2011","url":"https://doi.org/10.1258/jrsm.2011.110180"},{"label":"Citations added to MEDLINE per year (recent years)","value":"More than 1.5 million","source":"NLM MEDLINE citation counts by year of publication","url":"https://www.nlm.nih.gov/bsd/medline_cit_counts_yr_pub.html"},{"label":"Patients with metastatic non-squamous NSCLC in a US community network tested for all five guideline-recommended biomarkers before first-line therapy (2018-2020)","value":"Under half","source":"Robert et al., Lung Cancer 2022","url":"https://doi.org/10.1016/j.lungcan.2022.01.021"}],"causes":["The volume of publication exceeds any clinician's capacity to read and appraise it.","Guidelines are long documents rather than decision tools embedded in the systems clinicians use.","Community oncologists treat every cancer and cannot track sub-specialty evidence in each.","Continuing education, reimbursement and institutional protocols lag guideline changes.","Patient-facing information is either too technical, out of date, or commercially or ideologically biased."],"currentEfforts":["NCCN Guidelines are updated continuously and ESMO living guidelines update on evidence rather than on a fixed cycle.","OncoKB (MSKCC) and CIViC curate the clinical actionability of genomic alterations, and OncoKB was the first somatic variant knowledge base partially recognised by the FDA (2021).","NCI PDQ summaries provide expert-maintained, plain-language and professional cancer information free of charge.","Clinical decision support embedded in electronic records (for example, guideline-driven order sets and molecular tumour board tools) brings recommendations to the point of care.","Cochrane and the ASCO and ESMO guideline programmes provide systematic, graded syntheses of evidence."],"successLooksLike":"A practice-changing trial result is reflected in guidelines within months and in the majority of eligible patients' care within a year, and any patient can find a current, plain-language, sourced explanation of their disease and options."},{"id":"b-preclinical-models","kind":"bottleneck","name":"Lab models that fail to predict what happens in patients","aka":[],"tldr":"Nine in ten cancer drugs that work in mice fail in humans. Our models are the reason.","summary":"Oncology has the lowest probability of success of any therapeutic area: only a few percent of agents entering phase 1 reach approval, and most failures are for lack of efficacy that the preclinical package did not anticipate. Immortalised cell lines have drifted for decades and lack a microenvironment; subcutaneous xenografts grow in immunodeficient mice with no human immune system, stroma or metastatic pattern; genetically engineered mouse tumours evolve with far less heterogeneity than human disease; and patient-derived organoids capture epithelial biology but not vessels, immune cells or drug pharmacokinetics. Preclinical studies are also small, unblinded and rarely replicated. The result is a pipeline that spends billions in humans to learn what the models could not tell it. Better-validated models, functional testing on fresh patient tissue, and systematic benchmarking of model predictions against clinical outcomes are the fixes.","asOf":"2026-09-08","links":[{"label":"Wong, Siah & Lo, Estimation of clinical trial success rates (Biostatistics 2019)","url":"https://doi.org/10.1093/biostatistics/kxx069"},{"label":"Begley & Ellis, Raise standards for preclinical cancer research (Nature 2012)","url":"https://doi.org/10.1038/483531a"},{"label":"Cancer Models / PDCM Finder","url":"https://www.cancermodels.org/"}],"tags":[],"related":["cancer-models","depmap","idea-organoid-guided-adc","idea-bio1-immune-matched-humanised-mice","idea-bio1-co-clinical-avatar-trials","idea-bio1-organoid-immune-coculture","idea-bio1-tumour-on-chip-penetration","idea-bio1-in-silico-trials-dose","idea-bio1-preclinical-preregistration","idea-bio1-multicentre-mouse-trials","idea-bio1-model-predictivity-benchmark","idea-bio1-negative-preclinical-repository","idea-bio1-tumour-slice-cultures","idea-bio1-metastatic-niche-models","idea-bio1-aged-comorbid-models","idea-bio1-comparative-oncology-dogs","idea-bio1-zebrafish-avatars","idea-bio1-multi-organ-chip-tox","idea-bio1-rare-cancer-organoid-bank","idea-bio1-organoid-assay-clinical-validation","idea-bio1-organoid-cell-therapy-potency","idea-bio1-somatic-crispr-gemms","idea-bio1-ctc-derived-explants","idea-bio1-model-authentication-mandate","idea-bio1-reverse-translation-resistance-models","idea-bio1-ex-vivo-perfused-tumour","idea-bio1-model-patient-matching","idea-bio1-virtual-cell-perturbation","idea-moon-self-driving-cancer-labs"],"cancers":["pancreatic","glioblastoma","colorectal"],"sections":["drug-discovery"],"technologies":["organoids","pdx-models","functional-drug-testing","pdac-organoid-pharmacotyping","crispr-screens","bh3-profiling"],"targets":[],"drugs":[],"companies":["champions-oncology","xilis","curesponse","sengine","recursion"],"institutions":["broad-institute","nci","cold-spring-harbor","francis-crick"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-wong-biostatistics","paper-begley-nature"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Probability that an oncology drug entering phase 1 is eventually approved","value":"3.4%","source":"Wong, Siah & Lo, Biostatistics 2019","url":"https://doi.org/10.1093/biostatistics/kxx069"},{"label":"Oncology likelihood of approval from phase 1 (industry data 2003-2011), lowest of all disease areas","value":"6.7%","source":"Hay et al., Nature Biotechnology 2014","url":"https://doi.org/10.1038/nbt.2786"},{"label":"Landmark preclinical cancer papers whose findings Amgen scientists could reproduce","value":"6 of 53 (11%)","source":"Begley & Ellis, Nature 2012","url":"https://doi.org/10.1038/483531a"}],"causes":["Cell lines have adapted to plastic for decades and no longer resemble the tumours they came from.","Xenografts require immunodeficient hosts, so anything involving the immune system is invisible.","Mouse tumours are clonally simpler and are treated when small, which overstates drug effect.","Organoids lack stroma, vasculature and immune cells, and organoid drug exposure is not human pharmacokinetics.","Preclinical efficacy studies are small, unblinded, unregistered and rarely include controls that match clinical practice.","There is no systematic scoring of which models predicted which clinical results, so the field cannot learn which models to trust."],"currentEfforts":["The Human Cancer Models Initiative (NCI, Cancer Research UK, Wellcome Sanger, Hubrecht) and the PDCM Finder catalogue next-generation organoid and PDX models with clinical annotation.","DepMap (Broad Institute) systematically maps genetic dependencies across more than a thousand cell lines to separate robust from model-specific findings.","Champions Oncology, Xilis and cureSponse run functional drug testing on patient-derived xenografts, micro-organospheres and ex vivo tumour explants, with prospective studies comparing model prediction to patient response.","The NCI Patient-Derived Models Repository distributes clinically annotated PDX and organoid models.","Humanised-mouse and immune-competent organoid co-culture systems are being developed to test immunotherapies preclinically.","The FDA Modernization Act 2.0 (2022) removed the statutory requirement for animal testing, opening the door for validated human-cell and computational models."],"successLooksLike":"A drug that clears a defined preclinical panel has a documented, prospectively measured probability of clinical efficacy well above today's few percent, and functional testing on a patient's own tumour cells is validated to predict that patient's response."},{"id":"b-manufacturing-cell-therapy","kind":"bottleneck","name":"Manufacturing cost and time for living and radioactive medicines","aka":[],"tldr":"Cell therapies take weeks to make for one patient and cost hundreds of thousands of dollars. Isotopes run short.","summary":"Autologous CAR-T, TIL and TCR-T are one-batch-per-patient manufacturing: leukapheresis, shipping, transduction, expansion, release testing and return take several weeks, during which some patients progress or die, a proportion of products fail specification, and list prices run to several hundred thousand dollars before hospital costs. Only a small fraction of eligible patients receive approved CAR-T because of slot availability, referral and cost. Radioligand therapy has a different supply problem: lutetium-177 depends on a handful of reactors and enrichment sources, and actinium-225 for alpha therapy has been limited to a few curies a year worldwide from thorium-229 stocks, with accelerator and thorium-based production only now scaling. ADCs and bispecifics have complex biologics supply chains with a small number of contract manufacturers. In vivo CAR generation, allogeneic products, point-of-care and automated manufacturing, and new isotope production routes are the technical answers.","asOf":"2026-09-08","links":[{"label":"Hernandez, Prasad & Gellad, Total costs of chimeric antigen receptor T-cell immunotherapy (JAMA Oncology 2018)","url":"https://doi.org/10.1001/jamaoncol.2018.0977"},{"label":"Maude et al., Tisagenlecleucel in children and young adults with B-cell lymphoblastic leukemia (ELIANA, NEJM 2018)","url":"https://doi.org/10.1056/NEJMoa1709866"},{"label":"US Department of Energy Isotope Program","url":"https://www.isotopes.gov/"}],"tags":[],"related":["cell-therapy-roadmap","radiopharma-roadmap","idea-in-vivo-car-solid","idea-bispecific-vs-car-t-second-line","idea-reg-in-vivo-cart-off-the-shelf-vial","idea-reg-two-day-cart-rapid-release","idea-reg-ipsc-master-bank-qualified-once","idea-reg-closed-manufacturing-interop-standard","idea-reg-cart-potency-reference-standards","idea-reg-vein-to-vein-public-benchmark","idea-reg-legacy-radium-recovery-ac225","idea-reg-ac225-accelerator-pharmacopoeia","idea-reg-yb176-enrichment-capacity","idea-reg-medical-isotope-reactor-reserve","idea-reg-alpha-emitter-portfolio","idea-reg-regional-radiopharmacy-hubs","idea-reg-adc-platform-designation","idea-reg-hpapi-continuous-flow","idea-reg-open-source-lentiviral-vectors","idea-reg-non-viral-cart-manufacturing","idea-reg-manufacturing-slot-exchange","idea-reg-hospital-exemption-harmonised","idea-reg-ai-process-control-cell-manufacturing","idea-reg-early-apheresis-banking","idea-reg-public-cell-therapy-foundries","idea-reg-comparability-by-design-digital-twin","idea-reg-isotope-supply-observatory","idea-moon-in-vivo-cart-generic-price"],"cancers":["dlbcl","all-leukemia","multiple-myeloma","prostate","neuroendocrine","melanoma"],"sections":[],"technologies":["car-t","in-vivo-car-t","allogeneic-cell-therapy","til-therapy","tcr-t","car-nk-macrophage","radioligand-therapy","targeted-alpha-therapy","adc"],"targets":[],"drugs":["tisagenlecleucel","ciltacabtagene-autoleucel","lifileucel","pluvicto","lutathera","ac225-psma"],"companies":["novartis","umoja-biopharma","interius","orna-therapeutics","allogene","caribou","sana-biotechnology","fate-therapeutics","terrapower-isotopes","orano-med","itm","curium","iovance"],"institutions":[],"pathways":[],"terms":["alpha-vs-beta","dosimetry","crs","icans","theranostics"],"trials":["eliana","zuma-7","cartitude-4","vision"],"people":[],"bottlenecks":[],"keyPapers":["paper-hernandez-jama-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"regulation-manufacturing","severity":"major","metrics":[{"label":"Estimated total cost of CAR-T therapy including list price and management of adverse events (2018 US analysis)","value":"Up to ~US$500,000 per patient","source":"Hernandez, Prasad & Gellad, JAMA Oncology 2018","url":"https://doi.org/10.1001/jamaoncol.2018.0977"},{"label":"Patients in ELIANA for whom tisagenlecleucel could not be manufactured","value":"7 of 92 enrolled","source":"Maude et al., NEJM 2018","url":"https://doi.org/10.1056/NEJMoa1709866"},{"label":"Historic global actinium-225 supply from thorium-229 stockpiles","value":"~63 GBq (1.7 Ci) per year","source":"Robertson et al., Current Radiopharmaceuticals 2018","url":"https://doi.org/10.2174/1874471011666180416161908"}],"causes":["Autologous products are bespoke batches with no economies of scale.","Viral vector and GMP capacity are limited and expensive.","Actinium-225 has historically come only from decay of legacy thorium-229 held by a few government laboratories.","Lutetium-177 production depends on reactor time and enriched target material from few suppliers.","Release testing, cold-chain logistics and hospital accreditation add weeks and cost that no design optimisation removes."],"currentEfforts":["Umoja, Interius and Orna are developing in vivo CAR-T that programmes T cells inside the patient, removing manufacturing entirely.","Allogene, Caribou, Sana and Fate build off-the-shelf allogeneic CAR-T and CAR-NK products from healthy donor or iPSC sources.","Novartis' T-Charge and point-of-care manufacturing programmes at academic centres cut vein-to-vein time to days.","The US Department of Energy Isotope Program is producing accelerator-based actinium-225, and TerraPower Isotopes and Orano Med are scaling thorium-derived and accelerator-based supply.","ITM and Curium have expanded non-carrier-added lutetium-177 production, and Novartis has built dedicated radioligand plants.","Automated closed-system cell manufacturing platforms (for example, Miltenyi CliniMACS Prodigy and Lonza Cocoon) reduce labour and cleanroom needs."],"successLooksLike":"A CAR-T or equivalent cell therapy is available within a week of decision at a cost comparable to a course of biologic therapy, the majority of eligible patients actually receive it, and isotope supply is no longer the limiting factor for any approved radioligand."},{"id":"b-metastasis-biology","kind":"bottleneck","name":"Metastasis is understood least and studied last","aka":[],"tldr":"Metastasis causes about nine in ten cancer deaths but gets a small fraction of research money and almost no trials of its own.","summary":"Most cancer deaths are attributable to metastatic disease, yet the biology of dissemination is studied far less than the biology of the primary tumour, and almost every drug is developed to shrink lesions rather than to stop spread. The steps that let cells leave a primary, survive shear stress and immune attack in circulation, extravasate, adapt to a foreign organ and either grow or lie dormant involve distinct programmes (epithelial-mesenchymal plasticity, pre-metastatic niche formation, organ-specific tropism) that have few validated drug targets and no approved anti-metastatic agent. Preclinical models seldom metastasise in a human-like pattern, regulatory endpoints reward tumour shrinkage rather than prevention of new lesions, and the patients who would benefit most from an anti-metastatic drug (those with resected early disease) are hard to enrol in long trials. Metastasis-directed local therapy and circulating tumour DNA-defined trials are the first designs to treat spread itself as the target.","asOf":"2026-09-08","links":[{"label":"Dillekås et al., Are 90% of deaths from cancer caused by metastases? (Cancer Medicine 2019)","url":"https://doi.org/10.1002/cam4.2474"},{"label":"Steeg, Targeting metastasis (Nature Reviews Cancer 2016)","url":"https://doi.org/10.1038/nrc.2016.25"},{"label":"Cancer Grand Challenges","url":"https://cancergrandchallenges.org/"}],"tags":[],"related":["idea-psma-pet-guided-mdt","idea-ctdna-guided-adjuvant-crc","idea-peritoneal-directed-gastric","idea-bio2-antimetastatic-endpoint","idea-bio2-ctc-clearance-go-nogo","idea-bio2-net-blockade-perioperative","idea-bio2-perioperative-betablocker-nsaid","idea-bio2-premetastatic-niche-assay","idea-bio2-inhaled-lung-niche-immunotherapy","idea-bio2-liver-niche-kupffer-reprogramming","idea-bio2-cxcr4-mobilise-then-kill","idea-bio2-rapid-autopsy-commons","idea-bio2-metastasis-screen-standard","idea-bio2-ctc-culture-functional-testing","idea-bio2-lymph-node-immune-priming","idea-bio2-organotropism-atlas","idea-bio2-metastasis-funding-floor","idea-bio2-matrix-stiffness-prevention","idea-bio2-brain-met-prevention-trials","idea-bio2-oligometastatic-signature","idea-moon-metastasis-prevention-indication"],"cancers":["breast-hr-positive","tnbc","colorectal","prostate","melanoma","nsclc"],"sections":[],"technologies":["sbrt","liquid-biopsy","mrd-testing","psma-pet","sentinel-node","whole-body-mri"],"targets":[],"drugs":[],"companies":[],"institutions":["cruk","nci","mskcc"],"pathways":["emt"],"terms":["oligometastatic","oligoprogression","peritoneal-metastasis","stage-shift","ctdna","mrd"],"trials":["mslt-ii","dynamic","circulate-japan","imvigor011"],"people":[],"bottlenecks":[],"keyPapers":["paper-dillekas-cancer-med","paper-steeg-nat-rev-cancer"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Share of cancer deaths attributable to metastatic disease (registry-based analysis)","value":"~67% directly; ~90% when metastasis contributed","source":"Dillekås, Rogers & Straume, Cancer Medicine 2019","url":"https://doi.org/10.1002/cam4.2474"},{"label":"Estimated share of cancer research funding directed at metastasis","value":"~5% (estimate)","source":"Sleeman & Steeg, European Journal of Cancer 2010","url":"https://doi.org/10.1016/j.ejca.2010.02.039"},{"label":"Adults with cancer who develop brain metastases","value":"~20%","source":"Achrol et al., Nature Reviews Disease Primers 2019","url":"https://doi.org/10.1038/s41572-018-0055-y"}],"causes":["Metastatic lesions are rarely biopsied or banked, so their biology is inferred from primaries.","Mouse models seldom metastasise spontaneously to human-relevant organs, making anti-metastatic drugs hard to test.","Regulatory endpoints (response rate, progression-free survival) reward shrinkage of existing lesions, not prevention of new ones.","Adjuvant trials to prevent metastasis require thousands of patients and years of follow-up, which industry avoids.","Funding and career incentives favour primary-tumour and cell-intrinsic biology over the multi-organ, systemic biology of spread."],"currentEfforts":["Cancer Grand Challenges (Cancer Research UK and the NCI) funds international teams working on dormancy, metastatic niche biology and the origins of spread.","Circulating tumour DNA-guided adjuvant trials (DYNAMIC, CIRCULATE-Japan, IMvigor011) treat molecular evidence of dissemination before it becomes visible metastasis.","PSMA-PET-guided metastasis-directed stereotactic radiotherapy is being tested as a curative-intent strategy in oligorecurrent prostate cancer.","The Metastasis Research Society and Grand Challenge teams are developing metastasis-competent organoid and mouse models.","MSLT-II and related sentinel-node trials define when regional metastasis needs surgery and when it does not."],"successLooksLike":"At least one approved drug is labelled for prevention of metastasis on the basis of a trial whose primary endpoint was new distant lesions, and metastasis receives a share of research funding proportionate to its share of deaths."},{"id":"b-misinformation","kind":"bottleneck","name":"Misinformation and unproven therapies","aka":[],"tldr":"Patients are sold unproven treatments and frightened away from proven ones.","summary":"Patients with cancer are a target for misinformation at the most vulnerable point in their lives. About a third of the most popular cancer articles on social media contain misinformation and most of that is potentially harmful; alternative clinics, supplement sellers and online communities promote unproven treatments, and patients who choose alternative medicine instead of conventional treatment for curable cancers die at more than twice the rate. The same ecosystem depresses uptake of proven prevention (HPV vaccination) and screening, and undermines trust in trials. Countering it requires trustworthy, current, plain-language information that is easier to find than the misinformation, clinicians who ask about and engage with what patients have read, integrative oncology that offers evidence-based complementary care within mainstream oncology rather than outside it, and regulatory action against fraudulent products.","asOf":"2026-09-08","links":[{"label":"Johnson et al., Cancer misinformation and harmful information on Facebook and other social media (JNCI 2022)","url":"https://doi.org/10.1093/jnci/djab141"},{"label":"Johnson et al., Use of alternative medicine for cancer and its impact on survival (JNCI 2018)","url":"https://doi.org/10.1093/jnci/djx145"},{"label":"NCI PDQ Integrative, Alternative, and Complementary Therapies summaries","url":"https://www.cancer.gov/about-cancer/treatment/cam"},{"label":"WHO infodemic management","url":"https://www.who.int/health-topics/infodemic"}],"tags":[],"related":["cancer-gov-pdq","nccn","pubmed-europepmc","idea-moon-verified-information-layer","idea-moon-clinician-creator-programme","idea-moon-unproven-clinic-registry","idea-moon-ban-unproven-therapy-ads","idea-moon-ai-cancer-answer-audit","idea-moon-wikipedia-oncology-fellowships","idea-moon-hype-index","idea-moon-patient-community-moderators","idea-moon-living-plain-evidence-summaries","idea-moon-treatment-refusal-pathway"],"cancers":["cervical","breast-hr-positive","colorectal","prostate","nsclc"],"sections":[],"technologies":["hpv-vaccine"],"targets":[],"drugs":[],"companies":["cancer-commons"],"institutions":["asco","nci","cruk","iarc"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-johnson-j-natl-cancer-inst"],"journals":[],"dependsOn":[],"notes":[],"stage":"people-culture","severity":"moderate","metrics":[{"label":"Most popular cancer articles on social media (2018-2019) containing misinformation, and share of those with potentially harmful content","value":"32.5% misinformation; 76.9% of it harmful","source":"Johnson et al., JNCI 2022","url":"https://doi.org/10.1093/jnci/djab141"},{"label":"Risk of death for patients with curable cancers choosing alternative medicine instead of conventional treatment (US National Cancer Database)","value":"2.5 times higher","source":"Johnson et al., JNCI 2018","url":"https://doi.org/10.1093/jnci/djx145"}],"causes":["Fear, uncertainty and the harshness of conventional treatment create demand for hopeful alternatives.","Social media algorithms amplify emotionally engaging content regardless of accuracy.","Commercial actors profit from supplements, alternative clinics and unregulated tests.","Trustworthy information is often technical, out of date or hard to find.","Clinicians have little time to discuss what patients have read, and patients fear judgement for asking."],"currentEfforts":["NCI PDQ and Cancer Research UK maintain expert-reviewed, plain-language information on treatments, including complementary and alternative therapies.","The Society for Integrative Oncology and ASCO publish joint evidence-based guidelines on integrative therapies, bringing complementary care within mainstream oncology.","The FDA and FTC issue warning letters and enforcement actions against products marketed with fraudulent cancer cure claims.","WHO's infodemic management programme develops methods for monitoring and countering health misinformation.","Platform policies at YouTube and Meta remove or label certain medical misinformation, with cancer treatment claims covered under YouTube's medical misinformation policy.","Programmes such as Gavi and WHO's HPV vaccination communication work address vaccine hesitancy in cervical cancer prevention."],"successLooksLike":"A patient searching for their diagnosis finds accurate, current, sourced information before they find misinformation, the share of harmful cancer content among the most-viewed material falls below 10%, and fewer than 1% of patients with curable cancers forgo conventional treatment."},{"id":"b-global-access","kind":"bottleneck","name":"Most of the world has almost no cancer care","aka":[],"tldr":"Seven in ten cancer deaths happen in low- and middle-income countries, where radiotherapy, pathology, surgery and drugs are scarce.","summary":"About 70% of cancer deaths occur in low- and middle-income countries, and the gap is widening as incidence shifts toward them. Dozens of countries have no radiotherapy machine at all; fewer than a quarter of patients in low-income countries who need cancer surgery can get safe, affordable, timely surgery; pathology turnaround is measured in weeks; and essential cancer medicines on the WHO list are unavailable or unaffordable in much of Africa and South Asia. The consequence is that survival for the same cancer differs more between countries than any drug has ever achieved: childhood cancer survival exceeds 80% in high-income countries and is below 30% in many low-income ones. Most of the deficit is not in frontier technology but in basic infrastructure, workforce, financing and referral systems, and in the fact that cancer control receives a small share of global health funding.","asOf":"2026-09-08","links":[{"label":"WHO fact sheet, Cancer","url":"https://www.who.int/news-room/fact-sheets/detail/cancer"},{"label":"Sullivan et al., Global cancer surgery: delivering safe, affordable, and timely cancer surgery (Lancet Oncology Commission 2015)","url":"https://doi.org/10.1016/S1470-2045(15)00223-5"},{"label":"Atun et al., Expanding global access to radiotherapy (Lancet Oncology Commission 2015)","url":"https://doi.org/10.1016/S1470-2045(15)00222-3"},{"label":"IAEA Rays of Hope","url":"https://www.iaea.org/services/rays-of-hope"}],"tags":[],"related":["globocan","idea-single-dose-hpv-self-sampling-elimination","idea-acc-open-hardware-linac","idea-acc-linac-uptime-contracts","idea-acc-hypofractionation-default-lmic","idea-acc-modern-cobalt-bridge","idea-acc-pooled-procurement-radiotherapy","idea-acc-telepathology-district-network","idea-acc-eml-listing-price-commitments","idea-acc-biosimilar-prequalification-pool","idea-acc-childhood-medicines-platform-scale","idea-acc-hub-spoke-national-networks","idea-acc-mobile-see-and-treat-units","idea-acc-gbci-country-dashboards","idea-acc-infusion-sparing-regimens","idea-acc-metronomic-lmic-platform-trial","idea-acc-global-cancer-financing-window","idea-acc-frugal-hdr-brachytherapy","idea-acc-registries-as-aid-condition","idea-acc-abandonment-stipends","idea-acc-diaspora-tele-tumour-boards","idea-acc-solar-microgrids-for-cancer-units","idea-moon-radiotherapy-for-everyone-2040","idea-moon-pooled-procurement-licensing-pool","idea-cost-pooled-procurement-essential","idea-cost-voluntary-licensing-oncology"],"cancers":["cervical","hcc","gastric","esophageal","all-leukemia","hodgkin-lymphoma","breast-her2-positive","head-and-neck"],"sections":["radiation","surgery"],"technologies":["imrt-igrt","brachytherapy","histopathology-ihc","digital-pathology-ai","hpv-vaccine","hpv-testing","precancer-ablation"],"targets":[],"drugs":["trastuzumab-biosimilars","imatinib","gardasil-9"],"companies":[],"institutions":["iarc","tata-memorial","sysucc","cams-cancer-hospital","west-china"],"pathways":[],"terms":["biosimilar","hbv-hcv"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-sullivan-lancet-oncol","paper-atun-lancet-oncol"],"journals":["journal-of-cancer-policy"],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"critical","metrics":[{"label":"Share of cancer deaths occurring in low- and middle-income countries","value":"~70%","source":"WHO fact sheet, Cancer","url":"https://www.who.int/news-room/fact-sheets/detail/cancer"},{"label":"Childhood cancer survival in high-income vs many low- and middle-income countries","value":"More than 80% vs less than 30%","source":"WHO fact sheet, Cancer in children","url":"https://www.who.int/news-room/fact-sheets/detail/cancer-in-children"},{"label":"Patients needing cancer surgery in low-income countries with access to safe, affordable, timely surgery","value":"Less than 25% (fewer than 5% in some settings)","source":"Sullivan et al., Lancet Oncology Commission on Global Cancer Surgery 2015","url":"https://doi.org/10.1016/S1470-2045(15)00223-5"},{"label":"Patients in low-income countries with access to radiotherapy relative to need (2015 estimate)","value":"Under 10%","source":"Atun et al., Lancet Oncology Commission 2015","url":"https://doi.org/10.1016/S1470-2045(15)00222-3"}],"causes":["Cancer control receives a small fraction of development assistance for health, which has focused on infectious disease.","Radiotherapy, pathology and surgery require capital, maintenance and trained staff that low-income systems cannot sustain.","Essential medicines are priced for high-income markets and supply chains for cold-chain biologics do not reach many countries.","Weak primary care and referral systems mean patients present late and travel far.","Out-of-pocket payment for treatment pushes families into poverty or abandonment of care."],"currentEfforts":["The WHO Global Initiative for Childhood Cancer aims for 60% survival for childhood cancer globally by 2030 and the CureAll framework guides national programmes.","The IAEA Rays of Hope initiative expands radiotherapy in countries with none or little capacity, building on the DIRAC registry of machines.","The UICC-led Access to Oncology Medicines (ATOM) Coalition works with industry to make essential cancer medicines available in low- and lower-middle-income countries.","City Cancer Challenge (C/Can) builds city-level cancer care systems in low- and middle-income countries.","Tata Memorial Centre's National Cancer Grid connects hundreds of Indian centres to standardise care and run low-cost trials.","Biosimilar trastuzumab, rituximab and bevacizumab and generic imatinib have brought core therapies within reach of many national formularies."],"successLooksLike":"Every country has at least one radiotherapy machine per million people, pathology turnaround under two weeks, and all WHO essential cancer medicines available at affordable prices; survival gaps between countries for curable cancers (childhood leukaemia, Hodgkin lymphoma, cervical, breast) close to within ten percentage points."},{"id":"b-generic-repurposing","kind":"bottleneck","name":"No incentive to repurpose cheap drugs","aka":[],"tldr":"Old, cheap drugs with anti-cancer signals never get the trials they need because no one profits from the result.","summary":"Hundreds of licensed non-cancer drugs, including metformin, aspirin, statins, beta-blockers, antihistamines, antifungals and antihelminthics, have preclinical or observational anti-cancer signals, and the ReDO project has catalogued more than 300 of them. Because they are off-patent, no company will fund the phase 3 trials needed to change practice or the label, and academic funders have limited budgets for large pragmatic trials. When trials are run, results can go either way: metformin failed to improve survival in breast cancer (MA.32), while aspirin halved colorectal cancer in Lynch syndrome (CAPP2) and eflornithine, an old antiparasitic, was approved in 2023 for neuroblastoma maintenance after a single-arm trial with an external control. The regulatory route to a new indication for a generic, and the reimbursement of an off-label use, remain unclear in most countries, so even positive results diffuse slowly.","asOf":"2026-09-08","links":[{"label":"Pantziarka et al., The Repurposing Drugs in Oncology (ReDO) Project (ecancer 2014)","url":"https://doi.org/10.3332/ecancer.2014.442"},{"label":"Goodwin et al., Effect of metformin vs placebo on invasive disease-free survival in breast cancer (MA.32, JAMA 2022)","url":"https://doi.org/10.1001/jama.2022.6147"},{"label":"Burn et al., Cancer prevention with aspirin in hereditary colorectal cancer (CAPP2, Lancet 2020)","url":"https://doi.org/10.1016/S0140-6736(20)30366-4"},{"label":"NHS England Medicines Repurposing Programme","url":"https://www.england.nhs.uk/medicines-2/medicines-repurposing-programme/"}],"tags":[],"related":["drugbank-chembl","open-targets","idea-reg-global-repurposing-fund","idea-reg-third-party-label-update","idea-reg-repurposing-prize","idea-reg-repurposing-social-impact-bond","idea-reg-recovery-style-platform-repurposing","idea-reg-perioperative-propranolol-etodolac","idea-reg-statin-hcc-prevention-trial","idea-reg-evidence-gate-before-repurposing-phase3","idea-reg-target-trial-emulation-pipeline","idea-reg-new-indication-exclusivity-off-patent","idea-reg-delinked-market-entry-reward","idea-reg-generic-targeted-therapy-drup","idea-reg-antihistamine-plus-io-trial","idea-reg-losartan-pancreatic-stroma","idea-reg-regulator-ready-repurposing-dossiers","idea-reg-guideline-fast-track-repurposed","idea-reg-payer-coverage-off-label-generic-commitment","idea-reg-nonprofit-marketing-authorisation-holder","idea-reg-metronomic-lmic-phase3-to-label","idea-reg-generic-chemo-strategic-reserve","idea-reg-factorial-addon-arms-cooperative-trials","idea-reg-preregistered-ai-repurposing-scoring","idea-moon-generics-for-cancer-fund"],"cancers":["colorectal","breast-hr-positive","prostate","neuroblastoma"],"sections":[],"technologies":["chemoprevention","ai-drug-design"],"targets":[],"drugs":["eflornithine"],"companies":["us-worldmeds","recursion","cctg"],"institutions":["cruk","nci"],"pathways":[],"terms":["lynch-syndrome","real-world-evidence"],"trials":["nmtrc003"],"people":[],"bottlenecks":[],"keyPapers":["paper-pantziarka-ecancermedicalscience","paper-goodwin-jama"],"journals":[],"dependsOn":[],"notes":[],"stage":"regulation-manufacturing","severity":"major","metrics":[{"label":"Licensed non-cancer drugs with published evidence of anticancer activity catalogued in ReDO_DB","value":"More than 300","source":"Pantziarka et al., ecancermedicalscience 2018","url":"https://doi.org/10.3332/ecancer.2018.886"},{"label":"Effect of metformin on invasive disease-free survival in high-risk operable breast cancer (MA.32, 3,649 patients)","value":"HR 1.01 (no benefit)","source":"Goodwin et al., JAMA 2022","url":"https://doi.org/10.1001/jama.2022.6147"},{"label":"Colorectal cancer reduction with aspirin in Lynch syndrome carriers at 10-year follow-up (CAPP2, per-protocol)","value":"HR 0.56","source":"Burn et al., Lancet 2020","url":"https://doi.org/10.1016/S0140-6736(20)30366-4"}],"causes":["Off-patent drugs cannot recoup trial costs through exclusivity, so no commercial sponsor exists.","Academic and charity funders cannot afford many multi-thousand-patient phase 3 trials.","Observational signals are confounded, and many repurposing hypotheses fail when tested properly, discouraging funders.","Regulatory and reimbursement pathways for new indications of generics are unclear and vary by country.","Clinicians are reluctant to prescribe off-label without a label change and guideline endorsement."],"currentEfforts":["The Anticancer Fund's ReDO project maintains the ReDO_DB database and co-funds repurposing trials.","Add-Aspirin (Cancer Research UK, MRC) randomised more than 10,000 patients across breast, colorectal, gastro-oesophageal and prostate cancer to adjuvant aspirin.","The NHS England Medicines Repurposing Programme identifies and supports off-patent drugs for new indications, including licensing routes.","The NCI Cancer Therapy Evaluation Program and cooperative groups (for example, CCTG MA.32) run the large pragmatic trials industry will not.","US WorldMeds obtained FDA approval of eflornithine (DFMO) for neuroblastoma in 2023 on the basis of an externally controlled trial, a precedent for repurposed agents.","The European Commission's STAMP expert group and the 2023 EU pharmaceutical legislation proposal include incentives for repurposing."],"successLooksLike":"A funded pipeline runs at least ten adequately powered phase 3 repurposing trials at any time, positive results lead to label changes and guideline inclusion within two years, and at least three repurposed generics become standard of care in a major cancer."},{"id":"b-immunotherapy-response","kind":"bottleneck","name":"No one can predict who responds to immunotherapy","aka":[],"tldr":"Checkpoint drugs cure some patients and do nothing for most. We still cannot tell the two apart before treating.","summary":"PD-L1 immunohistochemistry, tumour mutational burden and mismatch-repair deficiency are the approved selection biomarkers for checkpoint inhibitors, and all are imperfect: PD-L1-negative patients respond, PD-L1-high patients progress, TMB thresholds vary by tumour type and assay, and only dMMR/MSI-high is a strong discriminator. As a result checkpoint inhibitors are given to nearly half of US patients with cancer while an estimated one in eight benefits, at a cost of tens of billions of dollars and a burden of immune-related adverse events in non-responders. Better predictors will need to integrate the tumour (antigenicity, interferon signalling, antigen presentation), the microenvironment (T-cell infiltration, myeloid state), the host (HLA genotype, microbiome) and early on-treatment dynamics (ctDNA clearance, CD8 PET, radiomics). The regulatory and reimbursement systems also need to accept biomarkers that deselect patients, which no sponsor is motivated to develop.","asOf":"2026-09-08","links":[{"label":"Haslam & Prasad (JAMA Network Open 2019)","url":"https://doi.org/10.1001/jamanetworkopen.2019.2535"},{"label":"Cristescu et al., Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy (Science 2018)","url":"https://doi.org/10.1126/science.aar3593"},{"label":"Hirsch et al., Blueprint PD-L1 IHC assay comparison project (JTO 2017)","url":"https://doi.org/10.1016/j.jtho.2016.11.2228"}],"tags":[],"related":["idea-cd8-pet-io","idea-multimodal-foundation-model","immunotherapy-roadmap","idea-bio2-tcr-repertoire-early-readout","idea-bio2-spatial-signature-cdx","idea-bio2-ctdna-six-week-io-switch","idea-bio2-io-biomarker-data-commons","idea-bio2-til-reactivity-selection","idea-bio2-antigen-presentation-triage","idea-bio2-epigenetic-priming-cold-tumours","idea-bio2-neoadjuvant-biomarker-engine","idea-bio2-steroid-sparing-irae"],"cancers":["nsclc","melanoma","head-and-neck","urothelial","gastric","colorectal"],"sections":[],"technologies":["checkpoint-inhibitor","immuno-pet","liquid-biopsy","pathology-foundation-model","digital-pathology-ai","cgp"],"targets":["pd1","pdl1"],"drugs":["pembrolizumab","nivolumab","atezolizumab","durvalumab"],"companies":["owkin","tempus","paige","foundation-medicine"],"institutions":[],"pathways":[],"terms":["tps","cps","tmb","msi","tils","cold-vs-hot","irae","ctdna"],"trials":["keynote-024-189","keynote-048","checkmate-067","keynote-177"],"people":[],"bottlenecks":[],"keyPapers":["paper-haslam-jama-netw-open","paper-hirsch-j-thorac-oncol","paper-cristescu-science"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"major","metrics":[{"label":"US patients with cancer eligible for checkpoint inhibitors vs estimated responders (2018)","value":"43.6% eligible, 12.5% respond","source":"Haslam & Prasad, JAMA Network Open 2019","url":"https://doi.org/10.1001/jamanetworkopen.2019.2535"},{"label":"Objective response rate to pembrolizumab vs chemotherapy in untreated NSCLC selected by PD-L1 TPS 50% or more (KEYNOTE-024): even the best-selected group has a majority of non-responders","value":"44.8% vs 27.8%","source":"Reck et al., NEJM 2016","url":"https://doi.org/10.1056/NEJMoa1606774"}],"causes":["Response depends on several partly independent axes (antigenicity, T-cell infiltration, antigen presentation, host factors) that no single marker captures.","PD-L1 expression is heterogeneous, dynamic and measured with several non-interchangeable assays and cut-offs.","TMB is a proxy for neoantigen load that ignores immunogenicity and antigen presentation.","Sponsors are not rewarded for developing biomarkers that shrink the eligible population.","Early on-treatment signals (ctDNA, imaging) are rarely used to stop ineffective therapy."],"currentEfforts":["The Friends of Cancer Research TMB Harmonization Project and the Blueprint PD-L1 IHC Comparison Project have standardised assay reporting.","Immuno-PET tracers such as 89Zr-labelled CD8 minibodies are in trials as early pharmacodynamic readouts of T-cell infiltration.","ctDNA dynamics in the first weeks of checkpoint therapy are being validated as an early marker of benefit in lung cancer and melanoma.","Multimodal foundation models trained on pathology, genomics and radiology (Owkin, Tempus, Paige) aim to predict immunotherapy benefit from routine data.","FDA approval of pembrolizumab for TMB-high solid tumours (2020) and dMMR tumours (2017) established tumour-agnostic immune biomarkers."],"successLooksLike":"A validated pre-treatment or early on-treatment test identifies the large majority of non-responders before they are exposed to months of ineffective immunotherapy, and the share of treated patients who benefit at least doubles."},{"id":"rejuv-agenda-screening-without-a-trial","kind":"bottleneck","name":"No randomised trial shows that any survivorship screening programme reduces death","aka":[],"tldr":"Survivors are screened for second cancers on the strength of how large their risk is, not on the strength of a trial showing that screening them saves lives. For most organs there is no programme at all.","summary":"This is the sharpest gap the second-cancers facet of this round found, and it stated it in one sentence: no randomised trial has shown that any survivorship screening programme reduces death from a second cancer.\n\nWhere a programme exists, it rests on inference. The United Kingdom's very high risk breast screening protocol offers annual magnetic resonance imaging to women irradiated to the chest when young; the case for it is the size of the measured risk, the youth of the women, and general evidence that magnetic resonance imaging finds breast cancer earlier in high-risk women. It is not a trial in irradiated survivors, and, as the paediatric facet put it, there will not be one. That is a defensible decision rather than an oversight: the population is small, the latency is decades, and randomising a woman with a measured several-fold risk to no surveillance is not something an ethics committee would approve.\n\nWhere no programme exists, the reasoning has never been done at all. No country screens for lung cancer on the basis of having had chest radiotherapy, even though measured second-primary lung cancer rates in some survivor groups exceed the rate in the control arm of the trial that established that lung screening saves lives. No country screens survivors' skin as a programme. There is no bladder screening for people who have had cyclophosphamide anywhere. For adults treated as adults there is no organised colonoscopy programme after abdominal or pelvic radiotherapy in the United Kingdom, the United States or Europe. For thyroid cancer after neck radiotherapy the international guideline panel compared ultrasound against feeling the neck, found neither superior, and published a decision aid instead of a recommendation.\n\nThe reason this matters more than it looks is that a survivor cannot tell the difference between a gap and a decision. Being told there is no programme sounds like reassurance. On several of these organs it is not reassurance, it is an unanswered question, and the records on this front say which is which.","asOf":"2026-10-02","links":[{"label":"Oeffinger et al., Chronic health conditions in adult survivors of childhood cancer (NEJM 2006;355:1572-1582)","url":"https://doi.org/10.1056/NEJMsa060185"},{"label":"NHS England: very high risk breast screening protocol","url":"https://www.gov.uk/government/publications/breast-screening-very-high-risk-women-surveillance-protocols"}],"tags":["rejuvenation","survivorship","open-problem","screening","second-cancers"],"related":["rejuv-second-screening-after-treatment-compared","rejuv-second-uk-very-high-risk-breast-screening","second-primary-lung-after-chest-radiotherapy","second-primary-skin-cancer-after-cancer-treatment","second-primary-thyroid-after-neck-radiotherapy","second-primary-bowel-after-abdominal-radiotherapy","second-primary-bladder-after-cyclophosphamide","rejuv-paed-breast-after-chest-radiotherapy","idea-rejuv-registry-randomised-screening-in-survivors","rejuvenation-roadmap"],"cancers":["hodgkin-lymphoma","breast-cancer"],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"prevention-detection","severity":"major","metrics":[{"label":"Randomised trials showing a survivorship screening programme reduces death from a second cancer","value":"None","source":"OnCo second-cancers facet, 2 October 2026","url":"https://doi.org/10.1056/NEJMsa060185"},{"label":"Organs with an organised survivor screening programme in any country, among those with a measured excess risk on this front","value":"Breast after chest radiotherapy given young; none for lung, skin, bladder, bowel in adults","source":"NHS England very high risk breast screening protocol and the second-cancer records on this front"}],"causes":["Latency between treatment and a second cancer is often two to four decades, longer than a screening trial can run.","Each survivor group is small, so a trial powered for mortality would need international recruitment nobody is funding.","Randomising people with a known elevated risk to no surveillance is difficult to justify to an ethics committee.","Screening programmes are organised by population risk factors such as age and smoking, and treatment history is not a field those programmes can query.","Radiotherapy fields and doses are frequently not recorded anywhere a survivor or a general practitioner can reach, so risk cannot be computed even where it is known."],"currentEfforts":["NHS England runs a very high risk breast screening protocol for women irradiated to the chest when young, and names total body irradiation as a group it will review for eligibility.","The Children's Oncology Group Long-Term Follow-Up Guidelines and the International Guideline Harmonization Group define risk-based surveillance after childhood cancer by exposure rather than diagnosis.","Registry-linked cohorts, including the Childhood Cancer Survivor Study and the British Childhood Cancer Survivor Study, continue to measure the risks that any programme would be built on."],"successLooksLike":"At least one survivorship surveillance programme has been evaluated against an outcome that matters, by registry-based randomisation or by a cohort with a contemporaneous unscreened comparison, and the organs where no programme exists have an explicit, published reason rather than a silence."},{"id":"rejuv-agenda-nothing-restores-cognition","kind":"bottleneck","name":"No treatment restores the thinking that cancer treatment takes","aka":[],"tldr":"The memory and concentration problems after chemotherapy and cranial radiotherapy are real and measurable. Every drug tested against them has failed, including one tested properly in 276 people, and nothing restores lost processing speed.","summary":"Two facets of this round reached the same conclusion from different directions. The body facet's record on thinking and memory says there is no drug that speeds recovery, and records that donepezil was tested properly in 276 breast cancer survivors across 87 practices with treatment groups not differing at 24 weeks on any domain. The modafinil result often quoted for cognition came from a secondary analysis inside a fatigue trial, using an open-label run-in and randomised withdrawal of responders, which is not evidence that it treats this. The paediatric facet's record on neurocognitive late effects states the same thing for children in four words: no treatment restores the lost processing speed.\n\nThe measurement problem compounds the treatment problem. Self-reported cognitive difficulty has a consistently absent or weak association with neuropsychological test scores, which does not mean the difficulty is imagined; it means the test measures something else, usually in a quiet room with no competing demands. A field that cannot agree what to measure cannot power a trial on it.\n\nWhat exists is partial. Exercise is mixed and should be described as mixed: a meta-analysis of eleven trials in 890 breast cancer survivors found significant effects on attention, working memory and perceived cognitive ability, while the largest single trial, in 253 survivors, found no difference from a health and wellness comparison on either primary outcome. Cognitive rehabilitation, meaning strategy training rather than restoration, is the approach with the most plausible evidence and has no trial large enough to establish an effect on function. The International Guideline Harmonization Group still lists neurocognitive problems among its guidelines in development, so there is no harmonised surveillance standard after childhood cancer either.\n\nThe third component is the one nobody has separated. After CAR-T cell therapy, no study has disentangled the contribution of the cell therapy itself from the contribution of everything that came before it, which in that population is several lines of chemotherapy, sometimes a transplant, sometimes central nervous system disease and sometimes cranial radiotherapy. The best long-term study has 40 patients and uses self-report rather than formal neuropsychological testing.","asOf":"2026-10-02","links":[{"label":"Management of fatigue in adult survivors of cancer: ASCO and Society for Integrative Oncology guideline update (JCO 2024), which covers the agents also tried for cognition","url":"https://doi.org/10.1200/JCO.24.00541"}],"tags":["rejuvenation","survivorship","open-problem","cognition"],"related":["cognitive-impairment-after-cancer-treatment","rejuv-paed-neurocognitive","rejuv-tx-icans-and-neurocognition","idea-acc-cognitive-rehabilitation-after-chemotherapy","idea-rejuv-core-outcome-set-for-late-effects","rejuv-trial-ex-cipn","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"major","metrics":[{"label":"Drugs shown in an adequately sized randomised trial to improve cognition after cancer treatment","value":"None","source":"OnCo body facet, cognitive-impairment-after-cancer-treatment"},{"label":"Donepezil trial: breast cancer survivors randomised, and the result at 24 weeks","value":"276 across 87 practices; treatment groups did not differ on any domain at any timepoint","source":"OnCo body facet, cognitive-impairment-after-cancer-treatment"},{"label":"Largest single randomised trial of exercise for cancer-related cognitive impairment","value":"253 survivors; no difference from health and wellness comparison on either primary outcome","source":"OnCo body facet, cognitive-impairment-after-cancer-treatment"}],"causes":["Self-report and neuropsychological testing measure different things and correlate weakly, so there is no agreed primary endpoint.","The deficits are small on average and large for a minority, which makes an unselected trial underpowered by design.","No mechanism is established well enough to pick a drug target, so candidates have been repurposed on analogy with dementia and stimulant use.","Nobody has separated the contribution of chemotherapy, cranial radiotherapy, cell therapy, anaesthesia, sleep loss, fatigue and depression, and they usually occur together.","There is no commercial sponsor: the condition has no approvable endpoint and the candidate drugs are generic."],"currentEfforts":["Cognitive rehabilitation and strategy training are offered in some centres and are the approach ASCO-adjacent guidance describes as reasonable without an evidence base that supports a recommendation.","Proton therapy is being used in children partly to reduce the dose to healthy brain, though the comparison against photons is across studies rather than within a randomised trial.","The corpus holds a proposal for trials of structured cognitive rehabilitation after chemotherapy and one for exercise and cognitive rehabilitation in older survivors is recruiting."],"successLooksLike":"A core outcome set for cancer-related cognitive impairment that pairs a functional measure with a test, and at least one adequately powered randomised trial of an intervention against it that does not fail for want of an endpoint."},{"id":"rejuv-agenda-biological-age-as-an-untested-target","kind":"bottleneck","name":"Nobody has tested whether reversing measured biological ageing changes anything","aka":[],"tldr":"Cancer treatment measurably accelerates several markers of biological ageing. No trial has ever asked whether moving one of those markers makes any difference to a person, and an entire retail industry rests on the assumption that it would.","summary":"The frontier facet of this round documented the acceleration and then documented the hole underneath it. Epigenetic clocks run fast in survivors; p16INK4a, a marker of cellular senescence, rises after chemotherapy; physiological frailty appears decades early in childhood cancer survivors. Those are measurements, and they are reasonably well made.\n\nWhat does not exist is the next step. No clock is validated as a clinical test, none is used to make a treatment decision, and no trial has shown that moving a clock changes what happens to a person. The same applies one marker down: no trial has shown that lowering p16INK4a changes a person's health, and no intervention has been shown to lengthen telomeres in people in a way that changes health. No trial has shown that an intervention reverses established frailty in survivors.\n\nMeanwhile the market has not waited. Senolytics have no completed randomised trial in cancer survivors and the one properly randomised trial of adequate design in another disease missed its primary endpoint. The trial designed to test metformin as a geroprotector has been proposed for a decade and has not started. A 48-week randomised trial of intermittent rapamycin in healthy ageing adults left its primary endpoint unchanged. There is no randomised trial of an NAD+ precursor in cancer survivors for recovery, fatigue or biological age, and a Europe PMC search for randomised trials of intravenous NAD+ in cancer survivors returns none. Clocks sold direct to consumers are not the clocks in this literature and are not interpretable against it.\n\nThe design problem is real and is worth stating rather than treating the absence as negligence. A trial with a clock as its primary endpoint proves nothing a person would notice. A trial with a clinical endpoint in survivors needs years and hundreds of participants. The way through is a trial with a function endpoint and the clock as a prespecified secondary, so that the two can be compared in the same people; one such trial is recruiting now, and it is the reason the fisetin trial record exists on this front.","asOf":"2026-10-02","links":[{"label":"Qin et al., Epigenetic age acceleration and chronic health conditions among adult survivors of childhood cancer (JNCI 2021;113:597-605)","url":"https://doi.org/10.1093/jnci/djaa147"},{"label":"Ness et al., Physiologic frailty as a sign of accelerated aging among adult survivors of childhood cancer: a report from the St Jude Lifetime cohort study (JCO 2013;31:4496-4503)","url":"https://doi.org/10.1200/JCO.2013.52.2268"},{"label":"ClinicalTrials.gov NCT06113016","url":"https://clinicaltrials.gov/study/NCT06113016"}],"tags":["rejuvenation","survivorship","open-problem","biological-ageing"],"related":["rejuv-age-epigenetic-clocks","rejuv-age-senescent-cells-after-treatment","rejuv-age-telomere-length","rejuv-age-frailty-and-late-effects","rejuv-frontier-senolytics","rejuv-frontier-metformin-ageing","rejuv-frontier-rapamycin-ageing","rejuv-frontier-nad-precursors","rejuv-frontier-what-works","idea-rejuv-biological-age-as-a-randomised-endpoint","rejuv-trial-proffi","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Trials showing that moving a biological-age marker changes a clinical outcome in cancer survivors","value":"None","source":"OnCo frontier facet, rejuv-age-epigenetic-clocks and rejuv-age-telomere-length"},{"label":"Randomised trials of an NAD+ precursor or intravenous NAD+ in cancer survivors for any rejuvenation outcome","value":"None found in Europe PMC","source":"OnCo frontier facet, rejuv-frontier-nad-precursors and rejuv-frontier-nad-infusions"},{"label":"Status of TAME, the trial designed to test metformin as a geroprotector","value":"Proposed for a decade, not started","source":"OnCo frontier facet, rejuv-frontier-metformin-ageing"}],"causes":["A biomarker endpoint is cheap and meaningless on its own; a clinical endpoint is meaningful and expensive, and almost nobody funds the second.","The clocks disagree with each other, so a trial would have to prespecify which one it believes before anyone has shown that any of them matters.","Several candidate interventions are supplements or off-patent drugs with no sponsor who would recover the cost of a trial.","Clearing senescent cells could in principle remove a barrier to tumour regrowth, which no human trial has addressed, so a trial in survivors carries a theoretical oncological risk nobody has quantified.","The market does not need the trial: the measurement alone is enough to sell a product, which removes the commercial pressure to run one."],"currentEfforts":["PROFFi, a randomised placebo-controlled phase 2 trial, is testing fisetin with and without tailored exercise against frailty in breast cancer survivors with six-minute walk distance as the primary endpoint.","The childhood cancer cohorts continue to characterise which exposures accelerate which markers and by how much.","Several regulators have acted against the marketed end of this field: FDA notifications on exosome products and regenerative medicine, and the compounding category 2 list."],"successLooksLike":"At least one randomised trial in cancer survivors with a function or event endpoint and a prespecified biological-age secondary, so the field can say whether the marker tracks the thing that matters or merely tracks itself."},{"id":"rejuv-agenda-nobody-owns-the-follow-up","kind":"bottleneck","name":"Nobody owns the follow-up once the oncology clinic lets go","aka":[],"tldr":"There are guidelines saying what a survivor should have checked and when. There is usually no mechanism that tells an individual survivor, ten years out, which checks they are due this year, or anyone whose job it is to notice they were missed.","summary":"The transplant facet named this one precisely: there is a published list of what should be done, an inspected standard for the programmes that should do it, and registries that record outcomes, and there is no mechanism in either country that reliably tells an individual survivor, ten years out and living two hundred miles from the transplant centre, which tests they are due this year.\n\nThe same failure appears on every facet. After childhood cancer the largest survey of its kind concluded that despite a significant risk of late effects after cancer therapy, the majority of childhood cancer survivors do not receive recommended risk-based care; among physicians, half felt confident providing survivorship care and 94 per cent had unmet information needs about survivors' late-effect risks. After transplant the commonest failure in revaccination is a handover gap: the transplant centre assumes the general practice will do it, the general practice assumes the centre has. After anthracyclines, surveillance echocardiography is often not arranged once a survivor leaves paediatric care, and blood pressure, the most treatable determinant of whether a survivor develops heart failure, is routinely not measured at a cancer follow-up appointment. Survivors frequently do not have their radiotherapy fields and doses recorded anywhere they can reach, which makes a risk-based guideline unusable even where someone is willing to follow it.\n\nThe instrument designed to close this is the survivorship care plan, and the honest position on it is on the history record for the 2006 report: a 2018 systematic review of thirteen randomised and eleven non-randomised studies concluded that existing research provides little evidence that such plans improve health outcomes and health care delivery. The authors' recommendation was that future research focus on ensuring that the recommendations in a plan are subsequently acted on, which is a restatement of this bottleneck rather than a solution to it.\n\nThis is separable from the commissioning problem. A service can exist and still not be used, because nobody knows the person is due. Fixing it is an information problem first, an accountability problem second, and a money problem third.","asOf":"2026-10-02","links":[{"label":"Jacobsen et al., Systematic review of the impact of cancer survivorship care plans on health outcomes and health care delivery (JCO 2018;36:2088-2100)","url":"https://doi.org/10.1200/JCO.2018.77.7482"},{"label":"Landier et al., Development of risk-based guidelines for pediatric cancer survivors: the Children's Oncology Group Long-Term Follow-Up Guidelines (JCO 2004;22:4979-4990)","url":"https://doi.org/10.1200/JCO.2004.11.032"},{"label":"ClinicalTrials.gov NCT06281496","url":"https://clinicaltrials.gov/study/NCT06281496"}],"tags":["rejuvenation","survivorship","open-problem","follow-up","handover"],"related":["rejuv-tx-long-term-follow-up-frameworks","rejuv-paed-transition-to-adult-care","rejuv-paed-uk-long-term-follow-up","rejuv-paed-cog-ltfu-guidelines","rejuv-tx-revaccination","rejuv-paed-heart","survivorship-care-plan","rejuv-history-lost-in-transition","idea-rejuv-exposure-record-a-machine-can-read","idea-acc-risk-stratified-follow-up","idea-acc-auto-generated-survivorship-plans","rejuv-trial-allocare","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":["survivorship-care-plan"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"critical","metrics":[{"label":"Physicians confident providing survivorship care, and with unmet information needs about late-effect risks","value":"About half confident; 94% with unmet information needs","source":"OnCo paediatric facet, rejuv-paed-transition-to-adult-care"},{"label":"Evidence that survivorship care plans improve health outcomes and health care delivery","value":"\"Existing research provides little evidence that SCPs improve health outcomes and health care delivery.\"","source":"Jacobsen et al., JCO 2018","url":"https://doi.org/10.1200/JCO.2018.77.7482"}],"causes":["Responsibility is shared between an oncology service that discharges and a primary care service that was not given the exposure history, so it sits with neither.","Treatment exposures are recorded in hospital systems in a form no other system can query, and often not at all for people treated decades ago.","Transition from children's to adult services is the point at which follow-up most often stops, and it coincides with leaving home and changing doctor.","The guidelines are risk-based, so they need an exposure record to be usable, and the exposure record is the thing survivors most often do not have.","No one is measured on whether a survivor got their checks, so nobody is accountable for them not happening."],"currentEfforts":["The Children's Oncology Group Long-Term Follow-Up Guidelines and the International Guideline Harmonization Group define what should be done, by exposure.","Long-term follow-up frameworks from EBMT, CIBMTR and FACT-JACIE accredit programmes, though accreditation inspects programmes rather than whether an individual received their checks.","The corpus holds proposals for risk-stratified follow-up, nurse-led follow-up clinics, automatically generated survivorship plans and a survivorship passport for adolescents and young adults.","AlloCare, a randomised trial in Denmark, is testing a stepped-care late-effects service after allogeneic transplant against usual care."],"successLooksLike":"A survivor, or a clinician they have never met, can get an accurate list of this year's due surveillance from their own treatment record, and somebody is accountable when it does not happen."},{"id":"b-workforce","kind":"bottleneck","name":"Not enough oncologists, nurses, pathologists, physicists","aka":[],"tldr":"The number of people with cancer is rising faster than the workforce trained to treat them.","summary":"Cancer incidence is rising with ageing populations while the workforce that delivers care is growing slowly, ageing itself and burning out. ASCO projected a shortfall of more than 2,000 US oncologists by 2025; the US and Canadian pathologist workforce shrank by about a sixth between 2007 and 2017 while case complexity rose; radiotherapy physicists, radiation therapists and oncology nurses are short even in wealthy systems; and in many low- and middle-income countries a single oncologist serves thousands of new patients a year. Training pipelines take a decade, maldistribution concentrates specialists in cities, and burnout affects around half of oncologists. Every other bottleneck, from trial enrolment to genomic testing to palliative care, is rate-limited by people. Task-shifting, advanced practice roles, AI triage and reporting, tele-oncology and international training programmes are the levers.","asOf":"2026-09-08","links":[{"label":"Yang et al., Projected supply of and demand for oncologists and radiation oncologists through 2025 (JOP 2014)","url":"https://doi.org/10.1200/JOP.2013.001319"},{"label":"Metter et al., Trends in the US and Canadian pathologist workforces 2007-2017 (JAMA Netw Open 2019)","url":"https://doi.org/10.1001/jamanetworkopen.2019.4337"},{"label":"WHO Global Strategy on Human Resources for Health: Workforce 2030","url":"https://www.who.int/publications/i/item/9789241511131"}],"tags":[],"related":["nci-cancer-centers","idea-acc-clinical-officer-oncology-track","idea-acc-remote-medical-physics-qa","idea-acc-echo-tele-mentoring-oncology","idea-acc-retention-packages-oncology","idea-acc-regional-oncology-training-hubs","idea-acc-ambient-ai-documentation-oncology","idea-acc-advanced-practice-radiation-therapists","idea-acc-pathologist-assistants-and-ai-triage","idea-acc-open-oncology-workforce-model","idea-acc-fast-track-relicensing-oncologists","idea-acc-mainstream-tele-genetic-counselling","idea-acc-district-surgeon-oncology-mentorship","idea-moon-task-shifting-ai-oncology-capacity"],"cancers":[],"sections":["ai-computation"],"technologies":["radiology-ai-screening","digital-pathology-ai","pathology-foundation-model","ai-trial-matching","dermoscopy-ai"],"targets":[],"drugs":[],"companies":["lunit","paige","pathai","aidoc"],"institutions":["asco","esmo","tata-memorial","iarc"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-yang-j-oncol-pract","paper-metter-jama-netw-open"],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"major","metrics":[{"label":"Projected US shortage of oncologists by 2025 (ASCO workforce model)","value":"2,258 full-time equivalents","source":"Yang et al., Journal of Oncology Practice 2014","url":"https://doi.org/10.1200/JOP.2013.001319"},{"label":"Change in the US pathologist workforce 2007-2017","value":"-17.5%","source":"Metter et al., JAMA Network Open 2019","url":"https://doi.org/10.1001/jamanetworkopen.2019.4337"},{"label":"US oncologists reporting at least one symptom of burnout (2013 survey)","value":"44.7%","source":"Shanafelt et al., JCO 2014","url":"https://doi.org/10.1200/JCO.2013.51.8480"}],"causes":["Specialist training takes ten or more years and training places have not expanded with demand.","Ageing of the workforce and early retirement, accelerated by burnout, remove capacity faster than it is replaced.","Specialists concentrate in cities and high-income countries, leaving rural and poor regions unserved.","Administrative burden, electronic record documentation and prior authorisation consume clinical time.","Pathology and physics are under-recruited because they are less visible and, in some systems, less well paid."],"currentEfforts":["The NHS Long Term Workforce Plan (2023) commits to expanding training places, including in oncology and diagnostics.","The WHO Global Strategy on Human Resources for Health: Workforce 2030 sets targets for health workforce density.","AI tools for mammography triage (Lunit, and the MASAI trial), pathology screening (Paige, PathAI) and radiotherapy auto-contouring extend the reach of scarce specialists.","Advanced practice providers and oncology pharmacists take on follow-up, toxicity management and survivorship visits.","The ESMO/ASCO Global Curriculum and the IAEA training programmes standardise oncology and physics training internationally.","Project ECHO and tele-oncology models connect community clinicians to specialist centres for case review."],"successLooksLike":"Oncology, pathology, nursing and physics workforce growth keeps pace with incidence in every region, burnout prevalence falls below a quarter, and no country has fewer than one oncologist per 500 new cancer cases."},{"id":"rejuv-agenda-thymus-does-not-regrow","kind":"bottleneck","name":"Nothing in routine use rebuilds an adult's thymus","aka":[],"tldr":"After a transplant or intensive chemotherapy, the part of the immune system that makes new kinds of T cell recovers slowly in adults and sometimes not at all. The deficit is well described, well measured and currently not correctable.","summary":"The transplant facet of this round put the gap in a sentence: there is no randomised evidence that any intervention accelerates thymic recovery in adults, and the agents trialled for it, including thymic peptides, keratinocyte growth factor, growth hormone and sex steroid blockade, have not produced a treatment in routine use.\n\nThe consequence is specific rather than general. Cell counts come back; the repertoire does not. Normal numbers of T cells can coexist with a narrowed range of things those T cells can recognise, which is why a normal count can reassure falsely, and why repertoire measurement remains a research tool rather than a clinical test. For a person this shows up as susceptibility to infections years after counts normalised, as vaccine responses that are weaker than the schedule assumes, and as the reason revaccination exists as a programme at all.\n\nIt is a moderate rather than a critical bottleneck because the problem is largely managed around rather than solved: prophylaxis by phase, immunoglobulin replacement where antibodies are low, and revaccination schedules do most of the work, and the honest framing is that these substitute for an immune system rather than rebuild one. The commonest failure in practice is not biological at all, it is that nobody owns the revaccination schedule between the transplant centre and primary care.\n\nWhat would change it is a trial, and the reason there has not been one is partly that the readout is contested. Thymic output can be estimated from T-cell receptor excision circles, repertoire diversity can be sequenced, and neither has been accepted as an endpoint that a regulator would license a drug against. An intervention that moved a repertoire measure without changing infection rates would not be worth having, and an intervention that reduced infections would be adopted whatever it did to the thymus, which is the design problem.","asOf":"2026-10-02","links":[{"label":"Cordonnier et al., Vaccination of haemopoietic stem cell transplant recipients: ECIL 7 guidelines (Lancet Infectious Diseases 2019)","url":"https://doi.org/10.1016/S1473-3099(18)30600-5"}],"tags":["rejuvenation","survivorship","open-problem","immune"],"related":["rejuv-tx-immune-reconstitution-timeline","rejuv-tx-revaccination","rejuv-tx-b-cell-aplasia-and-immunoglobulin","rejuv-tx-infection-by-phase","rejuv-frontier-immune-reconstitution","idea-rejuv-thymic-regeneration-in-adults","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"moderate","metrics":[{"label":"Interventions in routine use that accelerate thymic recovery in adults","value":"None","source":"OnCo transplant facet, rejuv-tx-immune-reconstitution-timeline"},{"label":"Agents trialled without producing a routine treatment","value":"Thymic peptides, keratinocyte growth factor, growth hormone, sex steroid blockade","source":"OnCo transplant facet, rejuv-tx-immune-reconstitution-timeline"}],"causes":["The thymus involutes with age, so adults start from a much smaller organ than the children in whom recovery is fastest.","No agreed clinical endpoint exists: repertoire diversity and thymic output are research measures, infection rates need very large trials.","The population is small and heterogeneous in conditioning regimen, graft source and graft-versus-host disease.","Agents with a plausible mechanism are old or off-patent, so nobody is funding a licensing programme.","Chronic graft-versus-host disease and its treatment themselves suppress recovery, confounding any trial run in the population that needs it most."],"currentEfforts":["Phase-based infection prophylaxis and revaccination schedules from ECIL and the CDC manage the consequence rather than the cause.","Immunoglobulin replacement covers the antibody deficit after B-cell-directed cell therapy, on evidence the transplant facet grades moderate because the strongest study is a 37-patient retrospective comparison.","Repertoire sequencing is being used in cohorts to characterise what recovers and what does not."],"successLooksLike":"An intervention that shortens the period of susceptibility after transplant or intensive therapy on an endpoint a clinician would act on, with a measure of immune recovery that regulators accept."},{"id":"b-aging-comorbidity","kind":"bottleneck","name":"Older and multimorbid patients are excluded and undertreated","aka":[],"tldr":"Most people with cancer are over 65 but most trial patients are younger and fitter. We guess how to treat the majority.","summary":"More than half of cancers and most cancer deaths occur in people over 65, yet older adults have been under-represented in registration trials for decades, and those enrolled are fitter than the general older population. Frailty, renal and hepatic function, polypharmacy, cognitive impairment and competing causes of death all change the balance of benefit and harm, but they are rarely measured in trials or in clinics, so older patients are both over-treated (full-dose regimens they cannot tolerate) and under-treated (denied curative therapy on the basis of age alone). Geriatric assessment with management is proven in randomised trials to cut severe toxicity by a fifth or more without compromising survival, and is recommended by ASCO, yet it is done in a minority of clinics. Age-specific dosing, endpoints that capture function and independence, and trial designs that include the patients who actually get cancer are the gap.","asOf":"2026-09-08","links":[{"label":"Mohile et al., Evaluation of geriatric assessment and management on toxic effects of cancer treatment (GAP70+, Lancet 2021)","url":"https://doi.org/10.1016/S0140-6736(21)01789-X"},{"label":"Sedrak et al., Older adult participation in cancer clinical trials: systematic review (CA 2021)","url":"https://doi.org/10.3322/caac.21638"},{"label":"Mohile et al., ASCO guideline: practical assessment and management of vulnerabilities in older patients receiving chemotherapy (JCO 2018)","url":"https://doi.org/10.1200/JCO.2018.78.8687"},{"label":"Cancer and Aging Research Group","url":"https://www.mycarg.org/"}],"tags":[],"related":["idea-shortened-venetoclax","fda-approvals","idea-acc-broad-eligibility-as-regulatory-default","idea-acc-pragmatic-trials-over-75-function-endpoints","idea-acc-home-chemotherapy-older-patients","idea-acc-hospital-at-home-oncology","idea-acc-electronic-frailty-index-oncology","idea-acc-geriatric-co-management-surgery","idea-acc-post-approval-evidence-over-75","idea-acc-caregiver-training-as-covered-service","idea-acc-pharmacist-interaction-review-oral-anticancer","idea-acc-time-toxicity-reporting","idea-acc-dementia-and-cancer-pathway"],"cancers":["aml","cll","dlbcl","multiple-myeloma","prostate","nsclc","colorectal"],"sections":[],"technologies":["geriatric-assessment","exercise-oncology","cardio-oncology"],"targets":[],"drugs":["venetoclax"],"companies":["alliance-oncology","ecog-acrin","swog"],"institutions":["asco","esmo"],"pathways":[],"terms":["standard-of-care","hazard-ratio"],"trials":["viale-a","cll14"],"people":[],"bottlenecks":[],"keyPapers":["paper-mohile-j-clin-oncol","paper-sedrak-ca-cancer-j-clin"],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"major","metrics":[{"label":"Patients aged 65 or older in SWOG treatment trials 1993-1996 vs their share of US cancer incidence","value":"25% vs 63%","source":"Hutchins et al., NEJM 1999","url":"https://doi.org/10.1056/NEJM199912303412706"},{"label":"Grade 3-5 toxicity with geriatric assessment-guided management vs usual care in patients aged 70 or older starting chemotherapy (GAP70+)","value":"51% vs 71%","source":"Mohile et al., Lancet 2021","url":"https://doi.org/10.1016/S0140-6736(21)01789-X"},{"label":"Grade 3-5 chemotherapy toxicity with geriatric assessment-driven intervention vs standard care (GAIN)","value":"50.5% vs 60.6%","source":"Li et al., JAMA Oncology 2021","url":"https://doi.org/10.1001/jamaoncol.2021.4158"}],"causes":["Eligibility criteria on organ function and performance status exclude many older patients from trials.","Sponsors prefer younger, fitter populations to maximise the apparent effect and minimise toxicity signals.","Chronological age is used as a proxy for fitness in clinical decisions instead of measured frailty.","Geriatric assessment takes time and is not reimbursed in most systems.","Trial endpoints (progression-free survival, response) do not capture what matters most to older patients: function, independence and cognition."],"currentEfforts":["GAP70+ and GAIN randomised trials showed geriatric assessment-guided management reduces severe toxicity, and the ASCO geriatric oncology guideline (2018, updated 2023) recommends it for all patients aged 65 or older receiving systemic therapy.","FDA guidance on inclusion of older adults in cancer clinical trials (2022) encourages removal of age-based exclusions and enrolment of the very old.","NIH's Inclusion Across the Lifespan policy (2019) requires justification for any age-based exclusion.","The Cancer and Aging Research Group (CARG) provides validated toxicity prediction tools and runs trials designed for older adults.","The International Society of Geriatric Oncology (SIOG) publishes age-specific treatment recommendations.","Venetoclax-based and other lower-intensity regimens (VIALE-A in AML, CLL14 in CLL) were developed specifically for patients unfit for standard therapy."],"successLooksLike":"Enrolment of patients over 70 in pivotal trials matches their share of incidence, geriatric assessment is performed for most older patients before systemic therapy, and age-specific treatment recommendations exist for every major cancer."},{"id":"b-overdiagnosis","kind":"bottleneck","name":"Overdiagnosis and false alarms","aka":[],"tldr":"Finding more cancer is not the same as saving lives. Screening also finds cancers that would never have hurt anyone, and treats them.","summary":"Overdiagnosis is the detection of a cancer that would never have caused symptoms or death in the person's lifetime. It is intrinsic to screening for indolent lesions and is largest where the reservoir of subclinical disease is big: thyroid, prostate, ductal carcinoma in situ of the breast, some small lung nodules and low-grade renal masses. South Korea's ultrasound-driven fifteen-fold rise in thyroid cancer incidence with unchanged mortality is the textbook case. Every overdiagnosed cancer is treated, with surgery, radiotherapy or lifelong hormone replacement, and every false positive generates anxiety, biopsies and cost. New multi-cancer and AI-based detection tests inherit the same problem unless they are evaluated on mortality and paired with active surveillance and risk-adapted management pathways. Reversing overdiagnosis requires changing the definition of what is called cancer, and changing what happens after a positive test.","asOf":"2026-09-08","links":[{"label":"Welch & Black, Overdiagnosis in cancer (JNCI 2010)","url":"https://doi.org/10.1093/jnci/djq099"},{"label":"Ahn, Kim & Welch, Korea's thyroid-cancer epidemic (NEJM 2014)","url":"https://doi.org/10.1056/NEJMp1409841"},{"label":"Independent UK Panel on Breast Cancer Screening (Lancet 2012)","url":"https://doi.org/10.1016/S0140-6736(12)61611-0"}],"tags":[],"related":["idea-thyroid-overdiagnosis-reversal","idea-prev-thyroid-no-screening-no-small-biopsy","idea-prev-ptmc-active-surveillance-default","idea-prev-low-risk-dcis-surveillance-pathway","idea-prev-indolent-lesion-nomenclature-body","idea-prev-gleason6-terminology-rct","idea-prev-prostate-as-triggered-biopsy","idea-prev-small-renal-mass-surveillance","idea-prev-lung-nodule-ai-discharge","idea-prev-blood-precursor-watch-registry","idea-prev-annual-overdiagnosis-reporting","idea-prev-screening-stop-by-life-expectancy","idea-prev-barrett-surveillance-deescalation","idea-prev-melanoma-ai-thick-melanoma-metric","idea-prev-incidentaloma-natural-history-cohort","idea-prev-ipmn-surveillance-stop-rule","idea-prev-pathology-ai-borderline-anchor","idea-prev-frail-elderly-primary-endocrine-breast","idea-prev-omit-radiotherapy-low-risk-breast-default","idea-prev-modern-autopsy-reservoir-studies","idea-moon-indolence-classifiers-with-screening"],"cancers":["thyroid","prostate","breast-hr-positive","nsclc","rcc"],"sections":[],"technologies":["active-surveillance","active-surveillance-thyroid","mp-mri","mammography","low-dose-ct-screening","partial-nephrectomy-active-surveillance","mced","thyroid-fna-molecular"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":["low-risk-dtc","psa","gleason-grade-group","ppv","bethesda-category"],"trials":["protect","precision-mri","nlst-nelson"],"people":[],"bottlenecks":[],"keyPapers":["paper-welch-j-natl-cancer-inst","paper-independent-uk-panel-on-breast-cancer-lancet","paper-ahn-n-engl-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"prevention-detection","severity":"major","metrics":[{"label":"Increase in thyroid cancer incidence in South Korea 1993-2011 with no change in mortality","value":"15-fold","source":"Ahn, Kim & Welch, NEJM 2014","url":"https://doi.org/10.1056/NEJMp1409841"},{"label":"Estimated breast cancers overdiagnosed for each breast cancer death prevented by UK mammography screening","value":"About 3 overdiagnosed per death prevented","source":"Independent UK Panel on Breast Cancer Screening, Lancet 2012","url":"https://doi.org/10.1016/S0140-6736(12)61611-0"},{"label":"Prostate cancer-specific mortality at 15 years with active monitoring vs surgery vs radiotherapy in PSA-detected disease (ProtecT)","value":"3.1% vs 2.2% vs 2.9% (no significant difference)","source":"Hamdy et al., NEJM 2023","url":"https://doi.org/10.1056/NEJMoa2214122"}],"causes":["Many organs harbour a large reservoir of indolent lesions that meet histological criteria for cancer but never progress.","Higher-resolution imaging and lower biopsy thresholds detect ever-smaller lesions.","Pathology cannot yet reliably distinguish indolent from aggressive lesions at diagnosis.","Clinicians and patients find it hard to leave a diagnosed cancer untreated, even when surveillance is safe.","Screening programmes are judged on cancers found rather than deaths prevented."],"currentEfforts":["ProtecT and PRECISION established active monitoring and MRI-first diagnosis as safe standards for low-risk prostate cancer, and active surveillance is guideline-endorsed for papillary thyroid microcarcinoma.","The COMET trial randomised low-risk DCIS to active monitoring vs surgery and reported non-inferior two-year invasive cancer rates.","Lung-RADS and PI-RADS structured reporting reduce unnecessary biopsies after CT and MRI.","South Korea and the US Preventive Services Task Force recommend against screening asymptomatic adults for thyroid cancer.","The Preventing Overdiagnosis conference series and BMJ Too Much Medicine campaign maintain a research and policy agenda.","Proposals to rename low-risk lesions (for example, IDLE tumours) aim to remove the word cancer from indolent findings."],"successLooksLike":"Every screening programme reports overdiagnosis and false-positive rates alongside deaths prevented, active surveillance is the default for defined low-risk lesions in thyroid, prostate and DCIS, and thyroid cancer incidence falls back toward its mortality-justified baseline."},{"id":"b-palliative","kind":"bottleneck","name":"Pain relief and palliative care are unavailable to most","aka":[],"tldr":"Most people who die of cancer worldwide do so without adequate pain relief.","summary":"Palliative care improves quality of life, reduces depression and, in at least one landmark randomised trial, lengthened survival when introduced at diagnosis of metastatic lung cancer, yet WHO estimates that only about 14% of people who need palliative care worldwide receive it. Access to opioids, the essential and inexpensive treatment for cancer pain, is restricted in most low- and middle-income countries by regulation, fear of diversion, and lack of trained prescribers, so that the poorest half of the world uses a tiny fraction of the morphine-equivalent opioids consumed by the richest countries. Even in high-income systems, palliative care is introduced late, often in the last weeks of life, and is still confused with hospice. The Lancet Commission's essential package of palliative care costs a few dollars per capita; the bottleneck is policy, training and the position of palliative care within oncology, not science.","asOf":"2026-09-08","links":[{"label":"WHO fact sheet, Palliative care","url":"https://www.who.int/news-room/fact-sheets/detail/palliative-care"},{"label":"Knaul et al., Alleviating the access abyss in palliative care and pain relief (Lancet Commission 2018)","url":"https://doi.org/10.1016/S0140-6736(17)32513-8"},{"label":"Temel et al., Early palliative care for patients with metastatic NSCLC (NEJM 2010)","url":"https://doi.org/10.1056/NEJMoa1000678"},{"label":"Ferrell et al., ASCO guideline on integration of palliative care into standard oncology care (JCO 2017)","url":"https://doi.org/10.1200/JCO.2016.70.1474"}],"tags":[],"related":["globocan","idea-acc-neighbourhood-palliative-networks","idea-acc-earmarked-palliative-budget-share","idea-acc-pain-control-public-audit","idea-acc-serious-illness-conversation-prompts","idea-acc-single-fraction-palliative-radiotherapy-default","idea-acc-essential-palliative-package-in-uhc","idea-acc-paediatric-palliative-in-every-childhood-unit","idea-acc-home-based-end-of-life-kits","idea-acc-methadone-and-alternatives-where-morphine-fails","idea-cost-early-palliative-default"],"cancers":["pancreatic","hcc","glioblastoma","esophageal","gastric","nsclc","cervical"],"sections":["supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":["asco","esmo","iarc"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-knaul-lancet","paper-ferrell-j-clin-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"major","metrics":[{"label":"People who need palliative care worldwide who receive it (WHO estimate)","value":"~14%","source":"WHO fact sheet, Palliative care","url":"https://www.who.int/news-room/fact-sheets/detail/palliative-care"},{"label":"People who die each year with serious health-related suffering, most in low- and middle-income countries","value":"More than 25 million (2015)","source":"Knaul et al., Lancet Commission on Global Access to Palliative Care and Pain Relief 2018","url":"https://doi.org/10.1016/S0140-6736(17)32513-8"},{"label":"Median overall survival with early palliative care vs standard care in metastatic NSCLC","value":"11.6 vs 8.9 months","source":"Temel et al., NEJM 2010","url":"https://doi.org/10.1056/NEJMoa1000678"}],"causes":["Opioid regulations designed to prevent diversion block legitimate medical use in most low- and middle-income countries.","Palliative care is not integrated into oncology training or cancer plans in many countries.","Clinicians and patients equate palliative care with giving up, so referral happens late.","Palliative services are funded poorly relative to disease-directed treatment.","Workforce shortages leave no one to prescribe, dispense or deliver home-based care."],"currentEfforts":["The Lancet Commission on Global Access to Palliative Care and Pain Relief defined an essential package and costed it for low- and middle-income countries.","World Health Assembly resolution WHA67.19 (2014) commits member states to integrate palliative care into health systems.","Hospice Africa Uganda's affordable oral morphine model has been adopted across several African countries.","ASCO's clinical practice guideline (2017) recommends integration of palliative care with standard oncology care from diagnosis of advanced cancer.","ESMO Designated Centres of Integrated Oncology and Palliative Care accredit hospitals that embed palliative care within oncology.","The International Narcotics Control Board and WHO work with countries to balance opioid availability with control."],"successLooksLike":"Every country has oral morphine available and prescribed for cancer pain, palliative care is introduced at diagnosis of incurable disease as a routine part of oncology, and the share of people with serious health-related suffering who receive palliative care exceeds 80%."},{"id":"b-patient-voice","kind":"bottleneck","name":"Patients lack understanding, navigation and agency","aka":[],"tldr":"Most patients cannot understand their options, find trials, or push back, so decisions are made for them.","summary":"A cancer diagnosis drops people into a system of jargon, probabilities, time-pressed consultations and fragmented hand-offs at the moment they are least able to process it. More than a third of US adults have basic or below-basic health literacy, so consent forms, trial descriptions and even standard patient leaflets are beyond many patients; navigating referrals, insurance and appointments falls on them and their families; and the systematic collection of what patients themselves report about symptoms, which in a randomised trial improved survival, is still not routine. Patients rarely know that a trial exists for them, cannot evaluate conflicting advice, and often do not feel entitled to ask for a second opinion or decline a recommendation. Navigation services, plain-language guidelines, patient-reported outcome monitoring, shared decision aids and patient-controlled data are all evidence-based and all unevenly delivered.","asOf":"2026-09-08","links":[{"label":"Basch et al., Overall survival results of a trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment (JAMA 2017)","url":"https://doi.org/10.1001/jama.2017.7156"},{"label":"NCCN Guidelines for Patients","url":"https://www.nccn.org/patientresources/patient-resources/guidelines-for-patients"},{"label":"NCI PRO-CTCAE","url":"https://healthcaredelivery.cancer.gov/pro-ctcae/"},{"label":"FDA Patient-Focused Drug Development guidance series","url":"https://www.fda.gov/drugs/development-approval-process-drugs/fda-patient-focused-drug-development-guidance-series-enhancing-incorporation-patients-voice-medical"}],"tags":[],"related":["cancer-gov-pdq","nccn","clinicaltrials-gov","idea-moon-navigation-as-a-right","idea-moon-plain-language-consent","idea-moon-results-in-plain-words","idea-moon-certified-decision-aids","idea-moon-patient-designed-trials","idea-moon-financial-toxicity-screening","idea-moon-goals-conversation-trigger","idea-moon-patient-held-cancer-record","idea-moon-funded-second-opinion","idea-moon-recorded-consultations","idea-moon-question-prompt-lists","idea-moon-patient-experience-label","idea-moon-results-back-to-participants","idea-moon-peer-navigator-workforce","idea-moon-time-toxicity-reporting","idea-moon-preference-studies-before-phase3","idea-moon-caregiver-in-the-plan","idea-moon-literacy-first-design-standard","idea-moon-trial-eligibility-at-every-decision"],"cancers":[],"sections":[],"technologies":["ai-trial-matching"],"targets":[],"drugs":[],"companies":["cancer-commons","more-health","patient-data-vault","tempus"],"institutions":["asco","nccn-org","nci"],"pathways":[],"terms":["orr","pfs","os","standard-of-care"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"people-culture","severity":"major","metrics":[{"label":"US adults with basic or below-basic health literacy (National Assessment of Adult Literacy 2003)","value":"36%","source":"Kutner et al., National Center for Education Statistics 2006","url":"https://nces.ed.gov/pubs2006/2006483.pdf"},{"label":"Median overall survival with routine patient-reported symptom monitoring vs usual care during chemotherapy for metastatic cancer (randomised)","value":"31.2 vs 26.0 months; not replicated","source":"Basch et al., JAMA 2017; PRO-TECT, JAMA 2022, found no survival difference","url":"https://doi.org/10.1001/jama.2017.7156"},{"label":"Patients who agree to join a trial when one is offered to them, showing the deficit is in offering, not willingness","value":"55%","source":"Unger et al., JNCI 2021","url":"https://doi.org/10.1093/jnci/djaa155"}],"causes":["Clinical information is written for clinicians and consultations are too short to translate it.","No one in the system owns the job of guiding the patient between services.","Symptoms and preferences are not systematically collected, so care is driven by what clinicians observe.","Power asymmetry and fear discourage patients from questioning recommendations or seeking second opinions.","Trial information is scattered across registries written in regulatory language."],"currentEfforts":["PRO-CTCAE (NCI) and the PRO-TECT trial embed patient-reported symptom monitoring into routine care, after Basch's randomised trial showed a survival benefit.","US Medicare began reimbursing principal illness navigation services in 2024, funding patient navigators for cancer.","NCCN Guidelines for Patients translate every major guideline into plain language free of charge.","The FDA Patient-Focused Drug Development guidance series requires sponsors to incorporate patient experience data in development.","AI-based trial matching and patient-facing platforms (Cancer Commons, More Health, Tempus) help patients find trials and second opinions.","Patient-controlled record platforms (Patient Data Vault) give patients their own complete data to share with any clinician or researcher."],"successLooksLike":"Every patient with cancer has a named navigator, plain-language information about their disease and options at a reading level they can use, routine patient-reported outcome monitoring, and a documented shared decision at each treatment choice."},{"id":"b-reproducibility","kind":"bottleneck","name":"Preclinical results do not reproduce","aka":[],"tldr":"Fewer than half of landmark cancer biology findings reproduce when someone else tries.","summary":"Systematic attempts to replicate published cancer biology have failed more often than they have succeeded. Amgen scientists could confirm only 6 of 53 landmark findings; the Reproducibility Project: Cancer Biology found that effect sizes in replications were on average 85% smaller than in the original papers and that fewer than half of the effects replicated on strict criteria, while most original papers lacked the information needed to attempt a replication at all. Misidentified and contaminated cell lines contaminate tens of thousands of publications. The estimated cost of irreproducible preclinical research in the US is tens of billions of dollars a year, and the human cost is drugs taken into patients on foundations that do not hold. The causes are incentives for novelty over rigour, small underpowered experiments, flexible analysis, lack of blinding and randomisation in animal work, and no funding or credit for replication.","asOf":"2026-09-08","links":[{"label":"Errington et al., Investigating the replicability of preclinical cancer biology (eLife 2021)","url":"https://doi.org/10.7554/eLife.71601"},{"label":"Begley & Ellis, Raise standards for preclinical cancer research (Nature 2012)","url":"https://doi.org/10.1038/483531a"},{"label":"Freedman, Cockburn & Simcoe, The economics of reproducibility in preclinical research (PLOS Biology 2015)","url":"https://doi.org/10.1371/journal.pbio.1002165"},{"label":"NIH Rigor and Reproducibility","url":"https://grants.nih.gov/policy-and-compliance/policy-topics/reproducibility"}],"tags":[],"related":["depmap","hpa","cancer-models","pubmed-europepmc","idea-tr2-timestamped-eln","idea-tr2-tenure-replication-credit","idea-tr2-random-audit","idea-tr2-paid-statistical-review","idea-tr2-reference-model-panels","idea-tr2-ai-external-validation-registry","idea-tr2-two-lab-rule","idea-tr2-preclinical-living-reviews","idea-tr2-phd-replication-year","idea-moon-independent-replication-institute"],"cancers":[],"sections":["drug-discovery"],"technologies":["organoids","pdx-models","crispr-screens","functional-drug-testing"],"targets":[],"drugs":[],"companies":[],"institutions":["broad-institute","cold-spring-harbor","aacr","nci"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-begley-nature","paper-freedman-plos-biol"],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"major","metrics":[{"label":"Landmark preclinical cancer findings reproduced by Amgen","value":"6 of 53 (11%)","source":"Begley & Ellis, Nature 2012","url":"https://doi.org/10.1038/483531a"},{"label":"Median reduction in effect size in replications vs original experiments (Reproducibility Project: Cancer Biology)","value":"85% smaller","source":"Errington et al., eLife 2021","url":"https://doi.org/10.7554/eLife.71601"},{"label":"Estimated annual US spend on irreproducible preclinical research","value":"~US$28 billion","source":"Freedman, Cockburn & Simcoe, PLOS Biology 2015","url":"https://doi.org/10.1371/journal.pbio.1002165"},{"label":"Published articles using misidentified or contaminated cell lines","value":"More than 32,000","source":"Horbach & Halffman, PLOS ONE 2017","url":"https://doi.org/10.1371/journal.pone.0186281"}],"causes":["Careers and grants reward novel positive results, not confirmation or null results.","Experiments are underpowered and analysed flexibly until a significant result appears.","Cell lines are misidentified or contaminated and reagents such as antibodies are poorly validated.","Animal studies are rarely randomised, blinded or pre-registered.","Methods sections omit the detail needed to repeat the work, and raw data are not shared."],"currentEfforts":["The Reproducibility Project: Cancer Biology (Center for Open Science and Science Exchange) published open replication attempts of 50 high-impact papers.","NIH rigour and reproducibility policy (2016) requires authentication of key biological resources including cell lines in grant applications.","The ARRIVE 2.0 guidelines and journal checklists set reporting standards for animal experiments.","Registered Reports at more than 300 journals commit to publication before results are known.","DepMap and the Human Protein Atlas provide systematic, reproducible baseline datasets that individual findings can be checked against.","ICLAC (International Cell Line Authentication Committee) maintains the register of misidentified cell lines and STR authentication standards."],"successLooksLike":"A majority of high-impact preclinical cancer findings replicate when tested independently, replication is funded and credited as first-class science, and no drug enters human trials on the basis of a single unreplicated preclinical result."},{"id":"b-prevention-adoption","kind":"bottleneck","name":"Prevention we already have is not deployed","aka":[],"tldr":"Around four in ten cancers are preventable with tools we already own: vaccines, tobacco control, weight, alcohol, sun and infection control.","summary":"About 40% of cancer cases and nearly half of cancer deaths in high-income countries are attributable to modifiable risk factors, tobacco above all, followed by excess body weight, alcohol, infections, diet, inactivity and ultraviolet exposure. Proven interventions exist for most of them: HPV vaccination (a single dose is now WHO-endorsed) can eliminate cervical cancer as a public health problem; hepatitis B vaccination and hepatitis C treatment prevent liver cancer; Helicobacter pylori eradication reduces gastric cancer; tobacco taxation, plain packaging and smoke-free laws are the most cost-effective measures in all of medicine; aspirin halves colorectal cancer in Lynch syndrome carriers. Yet global HPV coverage remains well below target, tobacco still kills more than eight million people a year, and prevention receives a small share of cancer research and health spending. The bottleneck is political, financial and organisational rather than scientific.","asOf":"2026-09-08","links":[{"label":"Islami et al., Proportion of cancer attributable to modifiable risk factors (CA 2018)","url":"https://doi.org/10.3322/caac.21440"},{"label":"WHO Cervical Cancer Elimination Initiative","url":"https://www.who.int/initiatives/cervical-cancer-elimination-initiative"},{"label":"Bruni et al., HPV vaccination coverage estimates 2010-2019 (Lancet Global Health 2021)","url":"https://doi.org/10.1016/S2214-109X(21)00347-3"},{"label":"WHO Framework Convention on Tobacco Control","url":"https://fctc.who.int/"}],"tags":[],"related":["idea-single-dose-hpv-self-sampling-elimination","globocan","idea-prev-therapeutic-hpv-vaccine-cin","idea-prev-hbv-treat-all-hcc","idea-prev-hcv-test-treat-surveillance-linkage","idea-prev-hpylori-family-test-and-treat","idea-prev-hpylori-stool-resistance-guided","idea-prev-hpylori-childhood-vaccine","idea-prev-very-low-nicotine-mandate","idea-prev-cytisine-essential-medicine","idea-prev-minimum-unit-pricing-cancer-endpoints","idea-prev-opportunistic-salpingectomy-default","idea-prev-commercial-sunbed-ban","idea-prev-radon-testing-at-property-sale","idea-prev-pay-for-prevention-outcomes","idea-prev-precursor-endpoints-for-approval","idea-prev-ebv-vaccine","idea-moon-eliminate-infection-cancers"],"cancers":["cervical","hcc","gastric","nsclc","head-and-neck","esophageal","colorectal","melanoma"],"sections":["prevention"],"technologies":["hpv-vaccine","chemoprevention","hpv-testing","precancer-ablation"],"targets":[],"drugs":["gardasil-9"],"companies":["merck"],"institutions":["iarc","cruk","nci"],"pathways":[],"terms":["hbv-hcv","hpv-p16","lynch-syndrome","cin-hsil"],"trials":["ken-she"],"people":[],"bottlenecks":[],"keyPapers":["paper-islami-ca-cancer-j-clin"],"journals":[],"dependsOn":[],"notes":[],"stage":"prevention-detection","severity":"critical","metrics":[{"label":"Share of US cancer cases and deaths attributable to potentially modifiable risk factors (2014)","value":"42% of cases, 45% of deaths","source":"Islami et al., CA 2018","url":"https://doi.org/10.3322/caac.21440"},{"label":"Girls worldwide fully vaccinated against HPV by 2019","value":"~15%","source":"Bruni et al., Lancet Global Health 2021","url":"https://doi.org/10.1016/S2214-109X(21)00347-3"},{"label":"Deaths caused by tobacco each year worldwide","value":"More than 8 million","source":"WHO fact sheet, Tobacco","url":"https://www.who.int/news-room/fact-sheets/detail/tobacco"},{"label":"Reduction in colorectal cancer with daily aspirin in Lynch syndrome carriers at 10 years (CAPP2, per-protocol)","value":"HR 0.56","source":"Burn et al., Lancet 2020","url":"https://doi.org/10.1016/S0140-6736(20)30366-4"}],"causes":["Prevention benefits accrue decades later and to no identifiable individual, so it attracts less political and commercial investment than treatment.","Tobacco, alcohol and food industries lobby against effective regulation.","Vaccine hesitancy and misinformation depress HPV and hepatitis B coverage.","Health systems and research funders are organised around treatment of disease rather than population health.","Prevention interventions have no patent, so no sponsor pays for their trials or promotion."],"currentEfforts":["The WHO Cervical Cancer Elimination Initiative sets 90-70-90 targets (vaccination, screening, treatment) for 2030, and WHO endorsed single-dose HPV schedules in 2022.","Gavi finances HPV vaccine introduction in low-income countries, with a target of vaccinating 86 million girls by 2025.","The WHO Framework Convention on Tobacco Control and MPOWER measures (taxation, smoke-free laws, plain packaging) are implemented in most countries to varying degrees.","CAPP2 and CAPP3 define aspirin chemoprevention in Lynch syndrome, and NICE recommends aspirin for carriers.","Population H. pylori screen-and-treat programmes in Taiwan, Japan and Korea are reducing gastric cancer incidence.","IARC Monographs and the IARC World Cancer Report provide the evidence base for carcinogen regulation."],"successLooksLike":"Global HPV vaccination coverage exceeds 90% of girls, adult smoking prevalence falls below 5% in high-income countries and is declining everywhere, and age-standardised incidence of cervical, liver, gastric and tobacco-related cancers falls measurably in registry data."},{"id":"b-drug-pricing","kind":"bottleneck","name":"Prices and value","aka":[],"tldr":"New cancer drugs routinely cost over $150,000 a year, often for months of benefit. Systems cannot afford them and patients go bankrupt.","summary":"Launch prices of new cancer drugs in the US have risen far faster than inflation or benefit, and analyses across the US and Europe find no correlation between price and clinical benefit as graded by ESMO-MCBS or the ASCO framework. The median survival gain of drugs approved in the early 2010s was about two months, while annual prices exceeded US$100,000 and now often exceed US$150,000-200,000. Financial toxicity is a measurable side-effect: US patients with cancer are more than twice as likely to declare bankruptcy as matched controls, and bankruptcy itself predicts earlier death. Publicly funded systems respond by delay, restriction or refusal, so a drug's availability depends on the country and payer rather than on the evidence. Value-based pricing, reference pricing, negotiation, biosimilar competition and transparent benefit grading are the levers; none has yet changed the launch-price trajectory.","asOf":"2026-09-08","links":[{"label":"Prasad, De Jesús & Mailankody, The high price of anticancer drugs: origins, implications, barriers, solutions (Nat Rev Clin Oncol 2017)","url":"https://doi.org/10.1038/nrclinonc.2017.31"},{"label":"Vokinger et al., Prices and clinical benefit of cancer drugs in the USA and Europe (Lancet Oncology 2020)","url":"https://doi.org/10.1016/S1470-2045(20)30139-X"},{"label":"Ramsey et al., Washington State cancer patients found to be at greater risk for bankruptcy (Health Affairs 2013)","url":"https://doi.org/10.1377/hlthaff.2012.1263"},{"label":"CMS Medicare Drug Price Negotiation Program","url":"https://www.cms.gov/inflation-reduction-act-and-medicare/medicare-drug-price-negotiation"}],"tags":[],"related":["fda-approvals","esmo-guidelines","nccn","idea-reg-io-subscription-national","idea-reg-indication-specific-pricing","idea-reg-pooled-procurement-adult-essentials","idea-reg-biosimilar-no-efficacy-trial","idea-reg-pd1-biosimilar-advance-commitment","idea-reg-patent-pool-oncology-eml","idea-reg-price-anchored-to-mcbs","idea-reg-weight-based-io-dosing-label","idea-reg-public-deescalation-trials-for-cost","idea-reg-financial-toxicity-vital-sign","idea-reg-net-price-transparency-registry","idea-reg-rd-cost-disclosure-condition","idea-reg-provisional-price-until-os","idea-reg-patent-buyout-prize","idea-reg-secondary-patent-thicket-limits","idea-reg-nonprofit-generic-chemo-manufacturer","idea-reg-academic-cart-at-cost-coverage","idea-reg-combination-price-attribution","idea-reg-middle-income-negotiation-bloc","idea-reg-tiered-pricing-for-exclusivity","idea-reg-biosimilar-first-default-switching","idea-reg-annuity-payment-durable-therapies","idea-reg-social-impact-bond-switching","idea-reg-market-expansion-repricing","idea-moon-pay-for-cure-contracts","idea-moon-neoantigen-vaccines-at-scale","idea-cost-asp-flat-fee","idea-cost-site-neutral-infusion","idea-cost-340b-pass-through","idea-cost-indication-based-pricing","idea-cost-inflation-rebates-commercial","idea-cost-part-b-cap","idea-cost-biosimilar-default-substitution","idea-cost-interchangeable-oncology-biosimilars","idea-cost-nonprofit-generic-oncology","idea-cost-transparent-generic-pricing","idea-cost-vial-sharing-dose-rounding","idea-cost-federal-oral-parity","idea-cost-value-scale-in-coverage","idea-cost-managed-access-everywhere"],"cancers":[],"sections":[],"technologies":["car-t","checkpoint-inhibitor","adc"],"targets":[],"drugs":["trastuzumab-biosimilars","imatinib","ibrutinib","pembrolizumab","trastuzumab-deruxtecan","tisagenlecleucel"],"companies":[],"institutions":["esmo","asco","nccn-org"],"pathways":[],"terms":["biosimilar","accelerated-approval","real-world-evidence","standard-of-care"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-prasad-nat-rev-clin-oncol","paper-vokinger-lancet-oncol","paper-ramsey-health-aff-millwood"],"journals":[],"dependsOn":[],"notes":[],"stage":"regulation-manufacturing","severity":"critical","metrics":[{"label":"Median overall survival gain of cancer drugs approved by FDA 2002-2014","value":"2.1 months","source":"Fojo, Mailankody & Lo, JAMA Otolaryngology 2014","url":"https://doi.org/10.1001/jamaoto.2014.1570"},{"label":"Likelihood of bankruptcy among US patients with cancer vs matched controls (Washington State)","value":"2.65 times higher","source":"Ramsey et al., Health Affairs 2013","url":"https://doi.org/10.1377/hlthaff.2012.1263"},{"label":"Correlation between monthly treatment cost and ESMO-MCBS or ASCO clinical benefit grade for cancer drugs in the US and four European countries","value":"No significant association","source":"Vokinger et al., Lancet Oncology 2020","url":"https://doi.org/10.1016/S1470-2045(20)30139-X"},{"label":"Total cost of cancer in Europe in 2018, including health expenditure, informal care and productivity loss","value":"€199 billion","source":"Hofmarcher et al., European Journal of Cancer 2020","url":"https://doi.org/10.1016/j.ejca.2020.01.011"}],"causes":["Patent monopolies and inelastic demand allow prices to be set at what payers will bear rather than at value.","US Medicare was prohibited from negotiating drug prices until 2022, anchoring global reference prices.","Approval does not require demonstration of benefit relative to price, and surrogate-based approvals are priced like cures.","Fragmented payers lack bargaining power and information about comparative value.","Follow-on products are priced at or above the incumbent rather than competing on price."],"currentEfforts":["The US Inflation Reduction Act (2022) allows Medicare to negotiate prices for selected high-spend drugs, with the first negotiated prices effective in 2026 and oncology drugs (ibrutinib, others) in early rounds.","ESMO-MCBS and the ASCO Value Framework give payers and clinicians a standardised grade of clinical benefit.","NICE, Germany's AMNOG, and EU joint clinical assessments tie reimbursement to demonstrated added benefit.","Biosimilars of trastuzumab, bevacizumab and rituximab have cut prices substantially in Europe and the US.","The Experts in Chronic Myeloid Leukemia letter (Blood 2013) and Common Sense Oncology are clinician-led campaigns against price-benefit disconnect.","The WHO Essential Medicines List and Access to Oncology Medicines (ATOM) Coalition push affordable access to essential cancer medicines in low- and middle-income countries."],"successLooksLike":"Prices track measured clinical benefit across countries, no patient with cancer in a high-income country faces bankruptcy because of treatment, and every drug on the WHO Essential Medicines List is affordable in every country that needs it."},{"id":"b-rare-cancers","kind":"bottleneck","name":"Rare and paediatric cancers without markets","aka":[],"tldr":"Taken together rare cancers are a fifth of all cancers, but each one alone is too small for a company to invest in.","summary":"Rare cancers (incidence below 6 per 100,000 per year in the European definition) together make up about a fifth of all cancer diagnoses and have worse survival than common cancers, yet each individual entity is too small to support a conventional development programme. Sarcomas, paediatric solid tumours, rare haematological malignancies and rare molecular subtypes of common cancers depend on academic cooperative groups, repurposing of adult drugs, and basket trials. Children have historically received new agents a decade after adults; the RACE for Children Act (2020) now obliges sponsors to test molecularly relevant drugs in paediatric populations, and tumour-agnostic approvals (NTRK, RET, BRAF, dMMR) create a route for rare subtypes. Diagnosis is also a bottleneck: pathology expertise for rare entities is concentrated in a few reference centres.","asOf":"2026-09-08","links":[{"label":"Gatta et al., Rare cancers are not so rare: the rare cancer burden in Europe (EJC 2011)","url":"https://doi.org/10.1016/j.ejca.2011.08.008"},{"label":"FDA, Pediatric Oncology and the RACE for Children Act","url":"https://www.fda.gov/about-fda/oncology-center-excellence/pediatric-oncology"},{"label":"WHO Global Initiative for Childhood Cancer","url":"https://www.who.int/initiatives/the-global-initiative-for-childhood-cancer"}],"tags":[],"related":["idea-menin-infant-all","idea-gd2-car-t-frontline-consolidation","idea-tcr-t-beyond-hla-a2","idea-bio2-rare-cancer-umbrella-platform","idea-bio2-n-of-1-framework","idea-bio2-paediatric-first-development","idea-bio2-rare-cancer-telepathology-network","idea-bio2-functional-precision-rare","idea-bio2-rare-cancer-prize-fund","idea-bio2-registry-embedded-randomisation","idea-bio2-patient-partnered-rare-commons","idea-bio2-mechanism-defined-baskets","idea-bio2-shared-compound-access-pool","idea-bio2-bayesian-borrowing-acceptance","idea-bio2-rare-tumour-model-bank","idea-moon-rare-cancer-global-network","idea-moon-open-source-oncology-drug-discovery"],"cancers":["sarcoma","neuroblastoma","mesothelioma","cholangiocarcinoma","neuroendocrine","all-leukemia","thyroid"],"sections":[],"technologies":["mibg-theranostics","limb-salvage-surgery","tandem-transplant"],"targets":["ntrk","gd2","ret","kmt2a"],"drugs":["tovorafenib","dinutuximab","naxitamab","nirogacestat","eflornithine","selpercatinib","revumenib"],"companies":["y-mabs","day-one-biopharmaceuticals","springworks","us-worldmeds","childrens-oncology-group","curie-nki-eortc"],"institutions":["nci","cruk"],"pathways":[],"terms":["tumour-agnostic","basket-umbrella-platform","sarcoma-histotype-tailoring","inrg-staging","mycn-amplification","fnclcc-grade"],"trials":["defi","gd2-cart01","euro-ewing-2012","anbl0032","hr-nbl1","interfant-06","naxitamab-201","nmtrc003","aall1731"],"people":[],"bottlenecks":[],"keyPapers":["paper-gatta-eur-j-cancer"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"major","metrics":[{"label":"Share of all cancer diagnoses in Europe that are rare cancers (RARECARE)","value":"22%","source":"Gatta et al., European Journal of Cancer 2011","url":"https://doi.org/10.1016/j.ejca.2011.08.008"},{"label":"Five-year relative survival for rare vs common cancers in Europe (RARECARE)","value":"47% vs 65%","source":"Gatta et al., European Journal of Cancer 2011","url":"https://doi.org/10.1016/j.ejca.2011.08.008"},{"label":"Children (0-14) diagnosed with cancer worldwide each year","value":"~400,000","source":"WHO fact sheet, Cancer in children","url":"https://www.who.int/news-room/fact-sheets/detail/cancer-in-children"}],"causes":["Small patient numbers make randomised trials slow or impossible and the commercial return too small to justify development.","Paediatric cancers have distinct biology (fusion-driven, low mutational burden) so adult drugs often do not apply.","Reference pathology and molecular diagnosis for rare entities is concentrated in a few centres, delaying correct diagnosis.","Regulatory precedent requires disease-specific evidence, which rare cancers cannot easily generate.","Preclinical models for many rare and paediatric tumours do not exist."],"currentEfforts":["The RACE for Children Act (FDARA 2017, in force 2020) requires paediatric investigation plans for adult oncology drugs directed at relevant molecular targets.","Tumour-agnostic approvals (larotrectinib and entrectinib for NTRK fusions, selpercatinib for RET, dabrafenib-trametinib for BRAF V600E) and basket trials such as NCI-MATCH and Pediatric MATCH reach rare subtypes.","The Children's Oncology Group, SIOPEN, Euro Ewing and ITCC run international trials that make paediatric randomisation possible.","The International Rare Cancers Initiative (Cancer Research UK, NCI, EORTC) designs trials for individual rare cancers.","Y-mAbs (naxitamab), Day One (tovorafenib), SpringWorks (nirogacestat) and US WorldMeds (eflornithine) show that small companies can bring rare-cancer drugs to approval.","The EU Orphan Regulation and US Orphan Drug Act provide exclusivity, fee waivers and tax credits."],"successLooksLike":"Every rare cancer entity has a reference diagnostic pathway and at least one open trial, children receive new molecularly matched drugs within two years of adult approval, and survival for rare cancers converges with common cancers."},{"id":"rejuv-agenda-late-effects-are-not-counted","kind":"bottleneck","name":"Registries count diagnoses and deaths, and not what treatment left behind","aka":[],"tldr":"Cancer registries are good at incidence and mortality and record almost nothing about late effects, so the scale of the problem is estimated from a handful of cohorts rather than counted. Europe cannot say how many survivors it has.","summary":"Almost everything this front knows about what treatment leaves behind comes from a small number of cohorts, most of them in childhood cancer, most of them North American, and most of them describing treatment given decades ago. The Childhood Cancer Survivor Study, the St Jude Lifetime Cohort, the British Childhood Cancer Survivor Study and the PanCare consortium carry a disproportionate share of the evidence on this entire front. They are excellent and they are not a surveillance system.\n\nThe consequences show up as honest refusals throughout this round's records. For Europe, the survivor-cohort record quotes a 2024 scoping review finding that information on prevalence is fragmented and inconsistent, and states that nobody knows accurately how many European survivors there are to plan services for. For adults, there is no cohort of comparable depth at all, and the frailty record says plainly that the childhood cohorts are mostly childhood cancer and mostly White, so they do not transfer to adults or to other populations. For newer treatments there is nothing: mortality after immunotherapy and cell therapy in children has no long-term cohort yet. For skin cancer after treatment, any figure for how many survivors develop one is an undercount because of how such cancers are recorded.\n\nThe mechanism of the gap is simple. Registries were built to measure whether a population's cancer incidence and mortality are changing, which is the question public health asked of them, and they do it well. Morbidity after treatment was never in their remit, has no reporting pathway, and appears years later in a different service that does not know the registry exists. Late effects are therefore invisible in the official statistics of every country, which is why they are chronically under-resourced: the number that would justify the budget has never been produced.\n\nThis is the bottleneck under several of the others. A screening programme cannot be designed without risk estimates; a service cannot be commissioned for a population of unknown size; a trial cannot be powered on a prevalence nobody has measured.","asOf":"2026-10-02","links":[{"label":"Robison et al., Study design and cohort characteristics of the Childhood Cancer Survivor Study (Medical and Pediatric Oncology 2002;38:229-239)","url":"https://doi.org/10.1002/mpo.1316"},{"label":"NCI Office of Cancer Survivorship: statistics and graphs","url":"https://cancercontrol.cancer.gov/ocs/statistics"},{"label":"Hudson et al., Clinical ascertainment of health outcomes among adults treated for childhood cancer (JAMA 2013;309:2371)","url":"https://doi.org/10.1001/jama.2013.6296"}],"tags":["rejuvenation","survivorship","open-problem","registries","data"],"related":["ccss","sjlife","bccss","pancaresurfup","ighg","rejuv-paed-chronic-disease-burden","rejuv-paed-late-mortality","rejuv-age-frailty-and-late-effects","rejuv-history-counting-survivors","idea-acc-national-late-effects-registry","idea-moon-adult-late-effects-registry","idea-rejuv-exposure-record-a-machine-can-read","rejuvenation-roadmap","seer"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"critical","metrics":[{"label":"Childhood Cancer Survivor Study: eligible five-year survivors identified, and participants completing the baseline questionnaire","value":"20,276 eligible from 25 institutions; 14,054 participants","source":"Robison et al., Medical and Pediatric Oncology 2002","url":"https://doi.org/10.1002/mpo.1316"},{"label":"European prevalence of childhood cancer survivors","value":"\"Information on prevalence of CCS in Europe is fragmented and inconsistent.\"","source":"2024 scoping review, quoted on the PanCareSurFup record"},{"label":"Cancer registries that routinely record late effects of treatment","value":"None found by OnCo","source":"OnCo, 2 October 2026"}],"causes":["Registries were designed to measure incidence and mortality, and late morbidity was never in their remit or their funding.","Late effects appear years after treatment, in a different service, which has no route to report them back.","There is no standard coding for a late effect of treatment that distinguishes it from the same condition arising ordinarily.","Cohorts that do measure it are research-funded, so they follow the populations the grants cover and stop when the grants do.","Adults, who are the overwhelming majority of survivors, have no cohort with the exposure detail the paediatric cohorts have."],"currentEfforts":["The Childhood Cancer Survivor Study, St Jude Lifetime Cohort, British Childhood Cancer Survivor Study and PanCareSurFup continue to follow tens of thousands of survivors with abstracted treatment records.","The International Guideline Harmonization Group reconciles surveillance guidance across those cohorts' findings.","The NCI Office of Cancer Survivorship publishes national survivor prevalence annually; the corpus holds proposals for a national late-effects registry and an adult equivalent."],"successLooksLike":"A country can state how many of its cancer survivors have a treatment-attributable chronic condition, by exposure, from routine data rather than from a cohort study, and can say whether that number is falling."},{"id":"b-regulatory-fragmentation","kind":"bottleneck","name":"Regulatory divergence between regions","aka":[],"tldr":"Regulatory divergence means a drug approved in one country can take years to reach another, or never arrive.","summary":"Each of the large national regulators (FDA, EMA, MHRA, PMDA, NMPA, TGA, Health Canada) requires its own dossier, applies its own evidence standards, and reviews on its own timeline, and approval is followed by a separate reimbursement process in each country. The result is a staggered global launch in which the same drug can be available in the US a year before the EU and years before Japan or China, or never filed at all in small or low-income markets. Sponsors also run duplicate or bridging trials to satisfy region-specific requirements, and divergent decisions on accelerated approvals create confusion about what the evidence actually shows. Harmonisation through ICH, collaborative review (Project Orbis, the Access Consortium), reliance pathways for smaller regulators, and the EU's joint clinical assessment are narrowing the gap, but the sequence of separate national decisions remains the norm.","asOf":"2026-09-08","links":[{"label":"FDA Oncology Center of Excellence, Project Orbis","url":"https://www.fda.gov/about-fda/oncology-center-excellence/project-orbis"},{"label":"Downing et al., Regulatory review of novel therapeutics: comparison of three regulatory agencies (NEJM 2012)","url":"https://doi.org/10.1056/NEJMsa1200223"},{"label":"International Council for Harmonisation (ICH)","url":"https://www.ich.org/"},{"label":"EU Regulation on Health Technology Assessment","url":"https://health.ec.europa.eu/health-technology-assessment/regulation-health-technology-assessment_en"}],"tags":[],"related":["fda-approvals","esmo-guidelines","eudract-ctis","idea-reg-orbis-work-sharing","idea-reg-reliance-90-day-lmic","idea-reg-single-global-dossier","idea-reg-joint-scientific-advice-default","idea-reg-realtime-streaming-submission","idea-reg-conditional-approval-sunset","idea-reg-confirmatory-trial-escrow","idea-reg-public-assessment-reports-all","idea-reg-mutual-recognition-inspections-atmp","idea-reg-global-trial-application-portal","idea-reg-diagnostic-reliance-pathway","idea-reg-single-paediatric-plan","idea-reg-tumour-agnostic-reliance","idea-reg-label-divergence-index","idea-reg-approval-access-lag-tracker","idea-reg-joint-regulator-hta-plan","idea-reg-mrct-no-bridging","idea-reg-african-medicines-agency-oncology","idea-reg-open-source-dossier-tools","idea-reg-shared-ai-review-assistant","idea-reg-shared-postmarketing-registry-condition","idea-reg-manufacturing-change-passport","idea-reg-orphan-designation-reciprocity","idea-reg-publish-all-rejections","idea-moon-reciprocal-tumour-agnostic-approvals"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":["trastuzumab-biosimilars","toripalimab","tislelizumab","serplulimab","camrelizumab"],"companies":["hengrui","beone","akeso","henlius"],"institutions":["esmo","asco","ncc-japan","cams-cancer-hospital"],"pathways":[],"terms":["accelerated-approval","breakthrough-designation","biosimilar","standard-of-care"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-downing-n-engl-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"regulation-manufacturing","severity":"major","metrics":[{"label":"Median total review time for novel therapeutics at FDA vs EMA vs Health Canada, 2001-2010","value":"303 vs 366 vs 352 days","source":"Downing et al., NEJM 2012","url":"https://doi.org/10.1056/NEJMsa1200223"},{"label":"Countries participating in FDA Project Orbis concurrent review of oncology products","value":"7 partner regulators alongside FDA (Australia, Brazil, Canada, Israel, Singapore, Switzerland, UK)","source":"FDA Project Orbis","url":"https://www.fda.gov/about-fda/oncology-center-excellence/project-orbis"}],"causes":["National sovereignty over drug approval means each agency must reach its own decision on its own dossier.","Evidence standards differ, especially on surrogate endpoints, single-arm trials and accelerated approval.","Reimbursement and health technology assessment are separate from approval and differ in every country.","Small and low-income markets are filed last or not at all because expected revenue does not cover the cost of filing.","Requirements for local trial data (historically in Japan and China) forced duplicate bridging studies."],"currentEfforts":["FDA Project Orbis coordinates concurrent review of oncology applications with partner regulators, cutting the lag to approval in participating countries.","The Access Consortium (Australia, Canada, Singapore, Switzerland, UK) shares work on dossiers, and the MHRA International Recognition Procedure relies on trusted regulators' decisions.","ICH guidelines (including E17 on multi-regional trials) harmonise technical requirements, and China's NMPA accepts overseas data since 2017 reforms.","The EU Health Technology Assessment Regulation began joint clinical assessments for oncology medicines in January 2025.","The WHO Collaborative Registration Procedure and prequalification let low-income regulators rely on stringent-authority approvals.","The EMA and FDA run parallel scientific advice so sponsors can design one trial that satisfies both."],"successLooksLike":"A single pivotal package supports simultaneous approval in all major regions within a few months of each other, and the median lag between first global approval and availability in any country with a functioning health system falls below one year."},{"id":"rejuv-agenda-rehabilitation-not-commissioned","kind":"bottleneck","name":"Rehabilitation is recommended everywhere and commissioned almost nowhere","aka":[],"tldr":"The interventions with the best evidence after cancer are supervised exercise, psychological therapy and specialist rehabilitation. The commonest finding across this whole front is that the evidence exists and the service does not.","summary":"This is the gap that appears most often across the two hundred records on this front, and it is not a gap in knowledge. Structured exercise has randomised evidence of improved survival in colon cancer; cognitive behavioural therapy has randomised evidence for insomnia and for fatigue; decongestive therapy and supervised resistance training have evidence in lymphoedema; vocational rehabilitation has trial evidence for return to work. What they have in common is that each needs a salaried person to deliver it, and most services do not employ one.\n\nThe facets of this round recorded the same finding independently. On muscle: few services employ anyone to deliver supervised resistance training. On fatigue: behavioural treatments need weeks of effort and are poorly funded. On exercise: supervision helps for mood and is hard to fund. On bowel function after pelvic radiotherapy: referral pathways from oncology to gastroenterology rarely exist. On teeth after head and neck radiotherapy: lifelong specialist dental care is not reliably paid for in most health systems and the cost falls on the survivor. On psychological care: what is missing is not evidence but a commissioned route, and the mind facet calls that the single clearest finding of its facet. On fear of recurrence, the most reported unmet need in survivorship, OnCo could not find a health service anywhere that commissions a named service for it with a referral route.\n\nEngland measured the scale of this once. Its 2013 national report found that 24 per cent of people were offered a written assessment and care plan, averaged across trusts, which is the least demanding component of the package it was recommending.\n\nThe structural reason is that none of these interventions is a product. There is no marketing authorisation to obtain, no company to lobby for a payment code, and no tariff that follows a person out of the oncology clinic. A drug with CHALLENGE's hazard ratio would be in every guideline and on every formulary within a year; an exercise programme with the same hazard ratio is in the guidelines and in almost no budgets.","asOf":"2026-10-02","links":[{"label":"Courneya et al., Structured exercise after adjuvant chemotherapy for colon cancer, the CHALLENGE trial (NEJM 2025;393:13-25)","url":"https://doi.org/10.1056/NEJMoa2502760"},{"label":"Campbell et al., Exercise guidelines for cancer survivors: consensus statement from international multidisciplinary roundtable (Medicine and Science in Sports and Exercise 2019;51:2375-2390)","url":"https://doi.org/10.1249/MSS.0000000000002116"},{"label":"Mustian et al., Comparison of pharmaceutical, psychological and exercise treatments for cancer-related fatigue: a meta-analysis (JAMA Oncology 2017;3:961)","url":"https://doi.org/10.1001/jamaoncol.2016.6914"},{"label":"Department of Health, Living With and Beyond Cancer: Taking Action to Improve Outcomes (England, 29 March 2013)","url":"https://www.gov.uk/government/publications/living-with-and-beyond-cancer-taking-action-to-improve-outcomes"}],"tags":["rejuvenation","survivorship","open-problem","rehabilitation","commissioning"],"related":["exercise-prescription-after-cancer","rejuv-frontier-what-works","cancer-related-fatigue-management","muscle-recovery-after-cancer-treatment","rejuv-mind-access-to-psychological-care","rejuv-mind-fear-of-recurrence-treatment","rejuv-life-return-to-work","bowel-after-pelvic-radiotherapy","dry-mouth-teeth-after-head-neck-radiotherapy","idea-rejuv-rehabilitation-prescription-at-discharge","idea-acc-return-to-work-rehabilitation","idea-bio2-exercise-reimbursement","rejuv-trial-canwork","rejuv-history-ncsi-england","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":["exercise-oncology","structured-exercise-survivorship","prehabilitation"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"critical","metrics":[{"label":"People in England offered a written assessment and care plan, averaged across trusts","value":"24%","source":"Department of Health, Living With and Beyond Cancer (2013), England cancer survivorship in numbers","url":"https://www.gov.uk/government/publications/living-with-and-beyond-cancer-taking-action-to-improve-outcomes"},{"label":"Health services found by OnCo that commission a named service for fear of recurrence with a referral route","value":"None in any country","source":"OnCo mind facet, rejuv-mind-fear-of-recurrence-treatment"},{"label":"CHALLENGE: eight-year overall survival with and without a coached exercise programme after colon cancer chemotherapy","value":"90.3% against 83.2%","source":"Courneya et al., NEJM 2025","url":"https://doi.org/10.1056/NEJMoa2502760"}],"causes":["Rehabilitation is a staffed service rather than a product, so it has no sponsor, no payment code and no tariff that follows the patient.","The benefit accrues years later and often to a different budget than the one that would pay, which is the classic prevention financing problem.","Oncology services are commissioned and measured around active treatment, and discharge is the point at which measurement stops.","Workforce is the binding constraint for several of these: physiotherapists, clinical psychologists, lymphoedema specialists and dietitians are already short.","Trials of these interventions report function and quality of life, which are not the endpoints health systems were built to pay for."],"currentEfforts":["The American College of Sports Medicine roundtable states an exercise dose for cancer survivors specific enough to be prescribed and audited.","England's Recovery Package defined a minimum bundle at the end of treatment; the corpus holds proposals for exercise reimbursement, prehabilitation as standard and vocational rehabilitation integrated into cancer care.","CanWork, a randomised trial of an occupational-therapy return-to-work intervention, includes a cost-effectiveness analysis explicitly so that a commissioner can act on it."],"successLooksLike":"Every person finishing curative treatment has a named route to supervised exercise, psychological therapy and the specialist rehabilitation their treatment implies, and the proportion who use it is audited rather than assumed."},{"id":"b-ip-collaboration","kind":"bottleneck","name":"Secrecy and intellectual property block collaboration","aka":[],"tldr":"Companies with complementary drugs rarely test them together, and data that could answer questions stays locked up.","summary":"The combinations with the strongest biological rationale in oncology often pair agents owned by different companies, and the negotiation over supply, data rights, liability and eventual pricing is slow enough that most such combinations are never tested; where they are, it is usually because one party has a checkpoint inhibitor to defend. Individual patient data from completed trials, the raw material for meta-analysis, biomarker discovery and honest re-analysis, are shared rarely and slowly despite journal and funder policies: in one audit, data could actually be obtained for about 1% of trials with sharing statements. Academic competition adds its own secrecy, delaying publication of negative and confirmatory results. Pre-competitive consortia, trusted data-sharing platforms, standardised combination agreements and public sponsors of cross-company trials are the working models; they cover a small fraction of the field.","asOf":"2026-09-08","links":[{"label":"Danchev et al., Evaluation of data sharing after implementation of the ICMJE data sharing statement requirement (JAMA Netw Open 2021)","url":"https://doi.org/10.1001/jamanetworkopen.2020.33972"},{"label":"Project Data Sphere","url":"https://www.projectdatasphere.org/"},{"label":"Vivli","url":"https://vivli.org/"},{"label":"Structural Genomics Consortium","url":"https://www.thesgc.org/"}],"tags":[],"related":["clinicaltrials-gov","genie","open-targets","pubmed-europepmc","idea-fund-federated-learning-consortium"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":[],"companies":["curie-nki-eortc","big","gbg","ecog-acrin","cancer-commons","patient-data-vault"],"institutions":["aacr","cruk","nci"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-danchev-jama-netw-open"],"journals":[],"dependsOn":[],"notes":[],"stage":"funding-incentives","severity":"major","metrics":[{"label":"Trials with ICMJE data-sharing statements whose individual participant data could be obtained by requesters","value":"6 of 487 (about 1%)","source":"Danchev et al., JAMA Network Open 2021","url":"https://doi.org/10.1001/jamanetworkopen.2020.33972"},{"label":"Clinical trials of PD-1/PD-L1 inhibitors registered by 2021, mostly combinations and mostly within single-sponsor portfolios","value":"5,683 trials","source":"Upadhaya et al., Nature Reviews Drug Discovery 2022","url":"https://doi.org/10.1038/d41573-022-00030-4"}],"causes":["Cross-company combination trials require negotiation of supply, data ownership, liability and future pricing between competitors.","Individual patient data are treated as proprietary assets and shared only under restrictive agreements.","Journal and funder data-sharing policies are rarely enforced.","Academic credit depends on priority and exclusivity of data.","Antitrust and confidentiality concerns discourage pre-competitive discussion of pipelines."],"currentEfforts":["Project Data Sphere, Vivli and the YODA Project host de-identified trial data from many sponsors for secondary research.","Cancer Research UK's Combinations Alliance and NCI's Cancer Therapy Evaluation Program broker cross-company combination trials under standard agreements.","Cooperative groups and academic consortia (EORTC, BIG, GBG, ECOG-ACRIN) sponsor trials combining agents from different companies, as in many neoadjuvant breast cancer platform trials.","The ICMJE requirement for data-sharing statements (2018) and the EU Clinical Trials Regulation increase disclosure of trial data.","The Structural Genomics Consortium and Open Targets run open, pre-competitive target validation with industry partners.","AACR Project GENIE demonstrates institution-level sharing of clinical-grade genomic data for common use."],"successLooksLike":"Individual patient data from every completed registrational trial are available to qualified researchers within two years of publication, and any rational combination of two approved or late-stage agents can be tested in a public trial under a standard agreement within a year of proposal."},{"id":"b-surgery-radiation-innovation","kind":"bottleneck","name":"Surgery and radiotherapy cure most, get least","aka":[],"tldr":"Surgery and radiotherapy cure more people than drugs do, but attract a fraction of the research investment.","summary":"Surgery remains the principal curative treatment for most solid tumours, and radiotherapy is indicated in roughly half of all patients with cancer and contributes to a large share of cures, yet both attract a small fraction of research funding and almost no industry investment because techniques cannot be patented in the way molecules can. Trials in surgery and radiotherapy are academic, small, slow to recruit and hard to blind, so innovations spread by training lineage and habit rather than evidence, and important questions (extent of resection, de-escalation of radiotherapy dose, hypofractionation, omission of surgery after complete response) wait decades for answers. When trials are done they change practice dramatically: MSLT-II ended routine completion lymphadenectomy in melanoma, LACC reversed minimally invasive surgery for cervical cancer, and hypofractionation trials halved radiotherapy courses in breast and prostate cancer. Dedicated funding streams, trial-quality assurance networks and device-industry partnerships are the routes to more of this.","asOf":"2026-09-08","links":[{"label":"Sullivan et al., Global cancer surgery (Lancet Oncology Commission 2015)","url":"https://doi.org/10.1016/S1470-2045(15)00223-5"},{"label":"Atun et al., Expanding global access to radiotherapy (Lancet Oncology Commission 2015)","url":"https://doi.org/10.1016/S1470-2045(15)00222-3"},{"label":"Barton et al., Estimating the demand for radiotherapy from the evidence (Radiotherapy and Oncology 2014)","url":"https://doi.org/10.1016/j.radonc.2014.03.024"},{"label":"McCulloch et al., No surgical innovation without evaluation: the IDEAL recommendations (Lancet 2009)","url":"https://doi.org/10.1016/S0140-6736(09)61116-8"}],"tags":[],"related":["idea-organ-preservation-esophageal","idea-bladder-preservation-mibc","idea-fund-rt-planning-ai-capacity","idea-fund-ablation-versus-surgery-trials","idea-fund-radiotherapy-innovation-pathway","idea-fund-technique-innovation-prize","idea-moon-image-guided-surgery-everywhere","idea-moon-compact-flash-proton","idea-cost-hypofractionation-payment"],"cancers":["melanoma","cervical","mesothelioma","esophageal","prostate","rcc","urothelial"],"sections":["surgery","radiation"],"technologies":["robotic-surgery","fluorescence-guided-surgery","sentinel-node","sbrt","proton-therapy","flash-rt","mr-linac","imrt-igrt","brachytherapy","carbon-ion","pleurectomy-decortication"],"targets":[],"drugs":[],"companies":["intuitive-surgical","varian","elekta","reflexion","nrg-oncology","curie-nki-eortc","jcog"],"institutions":["cruk","royal-marsden"],"pathways":[],"terms":[],"trials":["mslt-ii","lacc","mars-2","sano","protect","cross"],"people":[],"bottlenecks":[],"keyPapers":["paper-sullivan-lancet-oncol","paper-atun-lancet-oncol","paper-barton-radiother-oncol","paper-mcculloch-lancet"],"journals":[],"dependsOn":[],"notes":[],"stage":"funding-incentives","severity":"major","metrics":[{"label":"Evidence-based optimal radiotherapy utilisation rate across all cancers (proportion of patients who should receive radiotherapy at least once)","value":"~48-52%","source":"Barton et al., Radiotherapy and Oncology 2014","url":"https://doi.org/10.1016/j.radonc.2014.03.024"},{"label":"Patients with cancer who will need surgery at some point in their disease (global estimate, 2015)","value":"More than 80%","source":"Sullivan et al., Lancet Oncology Commission on Global Cancer Surgery 2015","url":"https://doi.org/10.1016/S1470-2045(15)00223-5"},{"label":"Melanoma-specific survival benefit of completion lymphadenectomy after a positive sentinel node (MSLT-II), a practice used for decades before it was tested","value":"None (86% in both arms at 3 years)","source":"Faries et al., NEJM 2017","url":"https://doi.org/10.1056/NEJMoa1613210"}],"causes":["Surgical and radiotherapy techniques are not patentable, so no company funds their trials.","Procedures are operator-dependent and hard to standardise or blind in trials.","Surgeons and radiation oncologists have less protected research time and fewer trial units than medical oncology.","Funders and journals give more weight to molecular novelty than to technique or process innovation.","Equipment vendors fund device-specific studies, not comparative trials of strategy."],"currentEfforts":["NRG Oncology, EORTC, JCOG and the UK's National Radiotherapy Trials Quality Assurance group run and quality-assure multicentre radiotherapy trials.","The IDEAL framework provides a staged methodology for evaluating surgical innovation analogous to drug phases.","Cancer Research UK's RadNet network funds radiotherapy research centres across the UK.","The Lancet Oncology Commissions on global cancer surgery and radiotherapy quantified the return on investment in both modalities.","Trials such as MSLT-II, LACC, MARS 2, SANO and PROTECT have shown that testing established procedures changes practice.","FLASH, MR-linac, proton and biology-guided radiotherapy (RefleXion) are being evaluated in academic-industry consortia."],"successLooksLike":"Surgery and radiotherapy each have dedicated funding streams proportionate to their contribution to cure, every major procedure or fractionation schedule in common use has randomised evidence, and time from a practice-changing surgical or radiotherapy trial to guideline adoption is under two years."},{"id":"b-survivorship","kind":"bottleneck","name":"Survivorship and late effects are neglected","aka":[],"tldr":"Tens of millions of people live after cancer with heart damage, infertility, second cancers and fear, and few services.","summary":"More than 18 million people in the US alone are living after a cancer diagnosis, and the number grows every year as treatment improves. Many carry the consequences of that treatment: anthracycline and trastuzumab cardiotoxicity, radiation-induced second cancers, infertility, neuropathy, cognitive change, endocrine failure, lymphoedema, and fear of recurrence, with the burden greatest in those treated as children, of whom about two-thirds have a chronic health condition by adulthood and more than a quarter a severe one. Survivorship care is fragmented between oncology, which discharges, and primary care, which often lacks the information or guidance to follow up. Late effects are poorly tracked because registries record incidence and death but not morbidity, and because survivorship research and services have no billing code or sponsor. Structured risk-based follow-up, cardio-oncology and exercise-oncology services, and lifelong registries are the evidence-based components that exist but are unevenly delivered.","asOf":"2026-09-08","links":[{"label":"NCI Office of Cancer Survivorship statistics","url":"https://cancercontrol.cancer.gov/ocs/statistics"},{"label":"Oeffinger et al., Chronic health conditions in adult survivors of childhood cancer (NEJM 2006)","url":"https://doi.org/10.1056/NEJMsa060185"},{"label":"Children's Oncology Group Long-Term Follow-Up Guidelines","url":"http://www.survivorshipguidelines.org/"},{"label":"Childhood Cancer Survivor Study","url":"https://ccss.stjude.org/"}],"tags":[],"related":["seer","idea-exercise-as-adjuvant","idea-acc-auto-generated-survivorship-plans","idea-acc-risk-stratified-cardio-oncology-pathway","idea-acc-default-fertility-preservation-referral","idea-acc-national-late-effects-registry","idea-acc-tailored-second-cancer-screening","idea-acc-return-to-work-rehabilitation","idea-acc-cognitive-rehabilitation-after-chemotherapy","idea-acc-cardiometabolic-clinic-hormone-therapy","idea-acc-aya-survivorship-passport","idea-acc-survivor-biobank-late-effect-prediction","idea-acc-lymphoedema-prospective-surveillance","idea-moon-survivor-lifelong-care-model"],"cancers":["hodgkin-lymphoma","all-leukemia","neuroblastoma","breast-hr-positive","breast-her2-positive","thyroid","prostate"],"sections":["supportive-care"],"technologies":["cardio-oncology","exercise-oncology","scalp-cooling","fertility-sparing-endometrial"],"targets":[],"drugs":[],"companies":["childrens-oncology-group"],"institutions":["nci","asco"],"pathways":[],"terms":["trastuzumab-cardiotoxicity","late-recurrence","tsh-suppression"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-oeffinger-n-engl-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"access-delivery","severity":"moderate","metrics":[{"label":"People living in the US with a history of cancer (2022 estimate)","value":"More than 18 million","source":"NCI Office of Cancer Survivorship statistics","url":"https://cancercontrol.cancer.gov/ocs/statistics"},{"label":"Adult survivors of childhood cancer with at least one chronic health condition, and with a severe or life-threatening condition (CCSS)","value":"62.3% and 27.5%","source":"Oeffinger et al., NEJM 2006","url":"https://doi.org/10.1056/NEJMsa060185"}],"causes":["Oncology services are funded and organised around active treatment, not decades of follow-up.","Primary care receives incomplete treatment summaries and has little guidance on late-effect surveillance.","Registries capture diagnosis and death but not late morbidity, so the burden is invisible in official statistics.","Late effects emerge years after trials end, so trial data rarely capture them.","Survivorship interventions have no commercial sponsor and little dedicated research funding."],"currentEfforts":["The Childhood Cancer Survivor Study and St Jude LIFE cohorts follow tens of thousands of survivors for decades to quantify late effects.","The Children's Oncology Group Long-Term Follow-Up Guidelines define risk-based surveillance after childhood cancer.","The 2005 Institute of Medicine report From Cancer Patient to Cancer Survivor established survivorship care plans, now an ASCO and Commission on Cancer standard.","PanCareFollowUp and PanCareSurFup coordinate European survivorship guidelines and cohorts.","Cardio-oncology services and ESC cardio-oncology guidelines (2022) formalise surveillance for anthracycline and HER2-therapy cardiotoxicity.","The NCI Office of Cancer Survivorship funds survivorship research and maintains national statistics."],"successLooksLike":"Every patient completing curative treatment receives a risk-stratified follow-up plan shared with primary care, late effects are recorded in cancer registries, and the excess morbidity of survivors relative to matched peers is measured and declining."},{"id":"b-brain-delivery","kind":"bottleneck","name":"The brain: barrier and sanctuary","aka":[],"tldr":"The blood-brain barrier's tight junctions and efflux pumps keep antibodies, antibody-drug conjugates and most kinase inhibitors out of the brain, so glioblastoma treatment has barely changed since 2005 and brain metastases, which develop in about a fifth of adults with cancer, are usually left to radiotherapy alone.","summary":"The blood-brain barrier excludes most large molecules, and small ones that are actively pumped back out, by tight junctions and efflux transporters, so antibodies, ADCs and a large share of kinase inhibitors reach the brain at a fraction of plasma concentration. Glioblastoma median survival with the Stupp regimen has been about 15 months since 2005 and no drug since has meaningfully moved it; temozolomide, lomustine and tumour-treating fields remain the standard, with vorasidenib the first real advance in a decade for IDH-mutant low-grade glioma. Brain metastases develop in around a fifth of adults with cancer, are excluded or under-represented in most systemic-therapy trials, and are usually managed with radiotherapy alone. Newer agents (tucatinib, osimertinib, lorlatinib, trastuzumab deruxtecan) show that CNS activity is achievable when designed for; focused ultrasound, convection-enhanced delivery, intrathecal cell therapy and receptor-mediated transcytosis are the delivery technologies under test.","asOf":"2026-09-08","links":[{"label":"Stupp et al., Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma (NEJM 2005)","url":"https://doi.org/10.1056/NEJMoa043330"},{"label":"Achrol et al., Brain metastases (Nature Reviews Disease Primers 2019)","url":"https://doi.org/10.1038/s41572-018-0055-y"},{"label":"Arvanitis, Ferraro & Jain, The blood-brain barrier and blood-tumour barrier in brain tumours and metastases (Nature Reviews Cancer 2020)","url":"https://doi.org/10.1038/s41568-019-0205-x"}],"tags":[],"related":["idea-fus-plus-adc-glioma","idea-neoadjuvant-io-glioblastoma","idea-cns-first-adc-strategy","idea-mri-surveillance-replaces-pci","idea-bio2-sonobiopsy","idea-bio2-intraventricular-car-t","idea-bio2-robotic-convection-delivery","idea-bio2-transferrin-shuttle-adc","idea-bio2-implanted-ultrasound-repeat-dosing","idea-bio2-cns-cohort-mandate","idea-bio2-csf-ctdna-monitoring","idea-bio2-brain-exposure-disclosure","idea-bio2-intranasal-delivery","idea-bio2-blood-csf-barrier-model","idea-bio2-fus-for-cell-entry","idea-bio2-brain-penetrant-glue-degraders","idea-moon-brain-delivery-platform"],"cancers":["glioblastoma","breast-her2-positive","nsclc","melanoma","sclc"],"sections":[],"technologies":["bbb-focused-ultrasound","litt","glioma-car-t","ttfields","prophylactic-cranial-irradiation","sbrt","mri"],"targets":["idh","egfr","her2"],"drugs":["temozolomide","vorasidenib","tucatinib","optune","rindopepimut","dordaviprone","tovorafenib"],"companies":["insightec","novocure","servier"],"institutions":["city-of-hope","stanford","mgh","heidelberg-nct"],"pathways":[],"terms":["blood-brain-barrier","her2-brain-metastases","mgmt","egfrviii","h3k27m"],"trials":["eortc-26981","ef-14","indigo","her2climb","destiny-breast12","checkmate-548","act-iv"],"people":[],"bottlenecks":[],"keyPapers":["paper-arvanitis-nat-rev-cancer","paper-achrol-nat-rev-dis-primers"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"major","metrics":[{"label":"Median overall survival in newly diagnosed glioblastoma with radiotherapy plus temozolomide (Stupp regimen), still the standard of care","value":"14.6 months","source":"Stupp et al., NEJM 2005","url":"https://doi.org/10.1056/NEJMoa043330"},{"label":"Adults with cancer who develop brain metastases","value":"~20%","source":"Achrol et al., Nature Reviews Disease Primers 2019","url":"https://doi.org/10.1038/s41572-018-0055-y"},{"label":"Median overall survival with tumour-treating fields added to temozolomide (EF-14), the only phase 3 survival gain in glioblastoma since 2005","value":"20.9 vs 16.0 months","source":"Stupp et al., JAMA 2017","url":"https://doi.org/10.1001/jama.2017.18718"}],"causes":["Tight endothelial junctions and P-glycoprotein efflux exclude most antibodies, ADCs and many small molecules.","Glioma cells infiltrate diffusely along white-matter tracts, so surgery and focal radiotherapy cannot remove them all.","Patients with brain metastases are excluded from most trials, so CNS activity is rarely measured until after approval.","Glioma is immunologically cold and the brain has limited lymphatic drainage, blunting immunotherapy.","Tumour heterogeneity in glioblastoma is extreme and single-target agents (EGFRvIII vaccine) have failed."],"currentEfforts":["Insightec's MR-guided focused ultrasound transiently opens the blood-brain barrier and is in trials to deliver chemotherapy and antibodies to glioma.","HER2CLIMB and DESTINY-Breast12 show that tucatinib and trastuzumab deruxtecan have meaningful activity against HER2-positive brain metastases, changing trial eligibility norms.","INDIGO established vorasidenib for IDH-mutant grade 2 glioma, the first targeted therapy to delay progression in diffuse glioma.","Intraventricular and locoregional CAR-T (IL13Ra2, GD2 for H3K27M diffuse midline glioma) at City of Hope and Stanford bypass the barrier entirely.","Novocure's Optune tumour-treating fields device is approved for glioblastoma and is being tested in brain metastases and other sites.","The FDA and ASCO-Friends of Cancer Research recommendations now call for inclusion of patients with treated or stable brain metastases in trials."],"successLooksLike":"Median survival in glioblastoma exceeds three years in a phase 3 trial, and patients with brain metastases are routinely included in systemic-therapy trials with CNS response measured as a standard endpoint."},{"id":"rejuv-agenda-no-agreed-outcome-measures","kind":"bottleneck","name":"The field cannot pool its own studies, because it has not agreed what to measure","aka":[],"tldr":"On subject after subject here, the studies exist and cannot be combined, because each used a different definition, a different questionnaire or a different threshold. A prevalence that ranges from 0 to 84 per cent is a measurement problem, not a biological one.","summary":"This gap is methodological and it is why several records on this front print a named gap where a number should be. The clearest example is kidney function after childhood cancer: reported prevalence ranges from 0 to 84 per cent across studies with no agreed outcome measure, and the paediatric facet's conclusion is that the field has not agreed what to measure, so the studies cannot be compared and no single prevalence figure is defensible.\n\nThe same pattern recurs. On time to diagnosis in adolescents and young adults, a systematic review found that the skewed distribution of the data meant comparisons between studies based on medians were difficult and combining studies within a meta-analysis was not appropriate, and the consequence is that there is no defensible average time to diagnosis for that age group. On quality of life after a stoma or limb loss, the approaches used to measure health-related quality of life were inconsistent and outcome scores varied substantially. On fear of recurrence, a major review concluded that the field had expanded somewhat haphazardly over the preceding twenty years and that consensus definitions and well-validated measures were still needed. On body image after head and neck cancer there is no pooled prevalence figure at all. On bowel function after pelvic radiotherapy, prevalence is reported very differently across studies and no single cohort figure was defensible. On nails after chemotherapy, an incidence range of 0 to 44 per cent says more about how it is measured than how often it happens. On cognition, self-report and neuropsychological testing diverge, which is a disagreement about what the outcome even is.\n\nThe fix is known and unglamorous. Core outcome sets, which specify a minimum set of outcomes and the instruments to measure them, have been developed for other fields and change what can be pooled within about a decade. They cost little, need no new biology, and require that a critical mass of investigators agree to use them. The reason they are missing here is the same reason most things on this front are missing: no sponsor, no mandate and no funder asking for one.","asOf":"2026-10-02","links":[{"label":"Hudson et al., Clinical ascertainment of health outcomes among adults treated for childhood cancer (JAMA 2013;309:2371)","url":"https://doi.org/10.1001/jama.2013.6296"}],"tags":["rejuvenation","survivorship","open-problem","outcomes","methods"],"related":["rejuv-paed-kidneys","rejuv-ayac-diagnostic-delay","rejuv-mind-body-after-stoma-and-limb-loss","rejuv-mind-fear-of-recurrence","rejuv-mind-visible-difference-head-and-neck","nail-changes-after-chemotherapy","bowel-after-pelvic-radiotherapy","cognitive-impairment-after-cancer-treatment","idea-rejuv-core-outcome-set-for-late-effects","idea-moon-pro-ctcae-in-every-pivotal-trial","ighg","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"major","metrics":[{"label":"Reported prevalence of kidney impairment after childhood cancer across studies","value":"0 to 84 per cent, with no agreed outcome measure","source":"OnCo paediatric facet, rejuv-paed-kidneys"},{"label":"Reported incidence of nail changes after chemotherapy","value":"0 to 44 per cent","source":"OnCo body facet, nail-changes-after-chemotherapy"},{"label":"Meta-analysis of time to diagnosis in adolescents and young adults","value":"Not appropriate to combine studies, per the systematic review's own conclusion","source":"OnCo paediatric facet, rejuv-ayac-diagnostic-delay"}],"causes":["Each specialty measuring a late effect brought its own instrument from its own field, and none was designed for survivors.","No funder requires a core outcome set, and no journal requires that a study report one.","Patient-reported and clinician-measured outcomes diverge for several of these effects, and the field has not decided which is the outcome.","Studies are small and single-centre, so the cost of an idiosyncratic measure is invisible to the people who bear it.","Retrospective cohorts measure whatever was recorded at the time, which was not chosen for this purpose."],"currentEfforts":["The International Guideline Harmonization Group reconciles surveillance recommendations internationally, which is the guideline-level version of the same work.","PRO-CTCAE and similar instruments have standardised symptom reporting inside trials, and the corpus holds a proposal to require it in every pivotal trial.","Several of this round's records state the gap rather than printing an indefensible pooled figure, which is the minimum honest response until the measures agree."],"successLooksLike":"Core outcome sets exist for the main late effects, journals and funders ask for them, and the next systematic review of any of these topics can pool rather than narrate."},{"id":"b-early-detection","kind":"bottleneck","name":"The hardest cancers are found late","aka":[],"tldr":"Screening exists for only a few cancers. Pancreatic, ovarian, liver, oesophageal and most lung cancers are found when cure is unlikely.","summary":"Stage at diagnosis is the largest single determinant of survival: localised pancreatic cancer has a five-year relative survival around 44% in SEER data, distant disease about 3%. Organised population screening exists only for breast, cervical and colorectal cancer, low-dose CT for heavy smokers, and prostate in some systems, and uptake is incomplete even where it exists. The cancers that kill most people at a late stage (pancreas, ovary, liver, oesophagus, stomach, most lung cancers in never-smokers) have no screening test, and the largest randomised trial of ovarian screening (UKCTOCS) found no mortality benefit despite a stage shift. Multi-cancer early detection blood tests (Galleri, Shield, fragmentomics) can detect signals from dozens of cancers but have modest sensitivity for stage I disease and have not yet shown reduced mortality; the NHS-Galleri trial is the first randomised test of the concept. Risk-stratified surveillance (new-onset diabetes for pancreas, cirrhosis for liver, Barrett's for oesophagus) is the intermediate strategy.","asOf":"2026-09-08","links":[{"label":"Crosby et al., Early detection of cancer (Science 2022)","url":"https://doi.org/10.1126/science.aay9040"},{"label":"Menon et al., UKCTOCS ovarian screening mortality results (Lancet 2021)","url":"https://doi.org/10.1016/S0140-6736(21)00731-5"},{"label":"NLST, Reduced lung-cancer mortality with low-dose CT screening (NEJM 2011)","url":"https://doi.org/10.1056/NEJMoa1102873"},{"label":"NCI SEER Cancer Stat Facts","url":"https://seer.cancer.gov/statfacts/"}],"tags":[],"related":["early-detection-roadmap","idea-mced-new-onset-diabetes","idea-mced-plus-fapi","idea-hcc-blood-surveillance","idea-non-endoscopic-barretts-screening","seer","globocan","idea-prev-mced-change-control-plan","idea-prev-interval-cancer-audit-blood-tests","idea-prev-ldct-ai-negative-triage","idea-prev-opportunistic-ct-ai-registry","idea-prev-pancreas-ai-prediagnostic-ct","idea-prev-new-onset-diabetes-pancreas-pathway","idea-prev-breath-voc-symptomatic-ruleout","idea-prev-urine-dna-haematuria-triage","idea-prev-weight-loss-auto-alert","idea-prev-cgm-glycaemic-drift-pancreas","idea-prev-ehr-symptom-signature-prompts","idea-prev-mced-non-specific-symptom-triage","idea-prev-blood-count-trend-flags","idea-prev-blood-crc-test-for-non-responders","idea-prev-capsule-sponge-pharmacy","idea-prev-gastric-serology-migrant-screening","idea-prev-hcc-blood-surveillance-cirrhosis","idea-prev-mri-first-prostate-screening-prs","idea-prev-oral-cancer-community-smartphone-ai","idea-prev-lmic-five-cancer-methylation-test","idea-prev-fragmentomics-first-tier","idea-prev-live-stage-dashboard","idea-prev-open-stage-shift-microsimulation","idea-prev-emergency-department-cancer-test","idea-prev-ebv-dna-npc-screening-scaleup","idea-prev-dental-oral-exam-standard","idea-prev-bundled-cancer-check-at-60","idea-moon-population-interception"],"cancers":["pancreatic","ovarian","hcc","esophageal","gastric","nsclc","colorectal"],"sections":["early-detection"],"technologies":["mced","liquid-biopsy","fragmentomics","low-dose-ct-screening","colorectal-screening","hcc-surveillance","pancreatic-surveillance","methylation-profiling","radiology-ai-screening"],"targets":[],"drugs":["galleri","shield"],"companies":["grail","guardant-health","exact-sciences","freenome","delfi-diagnostics","harbinger-health"],"institutions":["nci","cruk","iarc"],"pathways":[],"terms":["stage-shift","ppv","ca19-9","afp","barretts-esophagus","fit-test"],"trials":["nhs-galleri","pathfinder-2","nlst-nelson","ukctocs"],"people":[],"bottlenecks":[],"keyPapers":["paper-crosby-science","paper-menon-lancet"],"journals":[],"dependsOn":[],"notes":[],"stage":"prevention-detection","severity":"critical","metrics":[{"label":"Five-year relative survival for pancreatic cancer diagnosed at localised vs distant stage (SEER, US)","value":"~44% vs ~3%","source":"NCI SEER Cancer Stat Facts: Pancreatic Cancer","url":"https://seer.cancer.gov/statfacts/html/pancreas.html"},{"label":"Reduction in lung cancer mortality with low-dose CT screening vs chest X-ray in heavy smokers (NLST)","value":"20%","source":"National Lung Screening Trial Research Team, NEJM 2011","url":"https://doi.org/10.1056/NEJMoa1102873"},{"label":"Ovarian cancer deaths prevented by annual multimodal screening over 16 years of follow-up (UKCTOCS)","value":"No significant reduction despite stage shift","source":"Menon et al., Lancet 2021","url":"https://doi.org/10.1016/S0140-6736(21)00731-5"},{"label":"New cancer cases and cancer deaths worldwide, 2022","value":"~20 million cases, 9.7 million deaths","source":"Bray et al., CA 2024 (GLOBOCAN 2022)","url":"https://doi.org/10.3322/caac.21834"}],"causes":["Most cancers that prove fatal arise in deep organs without symptoms until they are locally advanced or metastatic.","Screening tests for low-prevalence cancers need very high specificity to avoid harm, which few biomarkers achieve.","A stage shift does not guarantee a mortality reduction, so each test needs a long randomised trial that few sponsors fund.","Screening uptake is incomplete and unequal even for tests that work, because of access, cost and awareness.","Health systems lack the diagnostic capacity (imaging, endoscopy, pathology) to work up the positives a population test would generate."],"currentEfforts":["NHS-Galleri randomised about 140,000 people to GRAIL's multi-cancer blood test, the first trial of MCED on stage shift and mortality.","The NCI Cancer Screening Research Network's Vanguard study is piloting MCED evaluation in the US.","Guardant's Shield blood test was FDA-approved in 2024 for colorectal screening, and Exact Sciences' Cologuard has expanded stool DNA testing.","The USPSTF broadened lung screening eligibility in 2021 and the NHS is rolling out targeted lung health checks nationally.","Cirrhosis-based HCC surveillance and new-onset-diabetes pancreatic surveillance cohorts (NCI/NIDDK New-Onset Diabetes cohort) target high-risk groups.","Delfi, Freenome and Harbinger are developing fragmentomic and methylation-based tests aimed at earlier-stage sensitivity."],"successLooksLike":"A randomised trial shows that a multi-cancer or organ-specific test reduces late-stage incidence and cancer-specific mortality for at least one cancer without a current screening programme, and the share of pancreatic, ovarian, liver and oesophageal cancers diagnosed at stage I-II at least doubles."},{"id":"rejuv-agenda-latency-outruns-the-evidence","kind":"bottleneck","name":"The newest treatments have not existed long enough for their late effects to appear","aka":[],"tldr":"A second cancer after radiotherapy can take forty years to appear. Immunotherapy and antibody-drug conjugates have been in first-line use for a few. Being told a new drug has no late effects usually means nobody has been followed long enough to see one.","summary":"The second-cancers facet of this round wrote the sentence that this record exists to generalise: a reader who is told that a new drug has no second cancer risk should know that the usual reason is that nobody has been followed long enough to see one.\n\nThe arithmetic cuts both ways. Whether the smaller radiotherapy fields of the last twenty years have lowered the forty-year second cancer rate is not known, because nobody treated with them has yet been followed for forty years, and saying otherwise would be a guess. The risk for a girl irradiated to the chest today should be lower than for one irradiated in 1975, though by how much is not measurable because the latency has not elapsed. Whether the immunotherapies and antibody-drug conjugates now in first-line use carry any such risk is not yet measurable for the same reason. Follow-up of commercial CAR-T cohorts is still short relative to the latency of second cancers, and the FDA's 2024 labelling action on secondary T-cell malignancy states no incidence and does not attribute these malignancies to vector integration. Much of the long-term data after transplant comes from older conditioning regimens and may not describe current practice, and the largest survival cohort describes transplants performed before 2004.\n\nThis cuts against both optimism and alarm. It is the reason a survivor cannot be reassured that a new drug is clean, and equally the reason a scare about a new drug's late effects usually cannot be substantiated. The regulatory response in gene-modified cell therapy has been to require fifteen years of long-term follow-up, and the transplant facet notes the honest limit of that too: no randomised evidence exists, or could exist, on whether being followed up for fifteen years improves an individual's outcome, and loss to follow-up is the known failure mode.\n\nWhat would change it is not a trial but an infrastructure: cohorts assembled at the time of treatment with the exposure recorded in a form that can still be queried in thirty years, and a registry linkage that does not depend on anyone remembering to follow up.","asOf":"2026-10-02","links":[{"label":"Schaapveld et al., Second cancer risk up to 40 years after treatment for Hodgkin's lymphoma (NEJM 2015)","url":"https://doi.org/10.1056/NEJMoa1505949"}],"tags":["rejuvenation","survivorship","open-problem","second-cancers","latency"],"related":["rejuv-second-radiotherapy-dose-field-and-age","rejuv-second-platinum-and-parp-inhibitors","rejuv-tx-secondary-t-cell-malignancy","rejuv-tx-gene-modified-follow-up","rejuv-tx-second-cancers","rejuv-paed-second-cancers","idea-rejuv-second-cancer-latency-cohort-for-new-drugs","rejuv-agenda-late-effects-are-not-counted","rejuvenation-roadmap"],"cancers":[],"sections":["rejuvenation","supportive-care"],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Mandated long-term follow-up for gene-modified cell therapy products","value":"Up to 15 years, as non-binding guidance","source":"OnCo transplant facet, rejuv-tx-gene-modified-follow-up"},{"label":"What the FDA's 2024 CAR-T labelling action states about incidence of secondary T-cell malignancy","value":"No incidence is stated, and the malignancies are not attributed to vector integration","source":"OnCo transplant facet, rejuv-tx-secondary-t-cell-malignancy"},{"label":"Whether smaller modern radiotherapy fields have lowered the forty-year second cancer rate","value":"Not known: nobody treated with them has been followed for forty years","source":"OnCo second-cancers facet, rejuv-second-radiotherapy-dose-field-and-age"}],"causes":["Solid second cancers after radiotherapy have latencies of two to four decades, which exceeds the working life of most research programmes.","Treatment changes faster than the latency, so by the time a regimen's late effects are measurable it is no longer the regimen in use.","Pharmacovigilance reporting describes reporting rather than incidence, so a signal cannot be turned into a rate.","Trial follow-up ends when the registration endpoint is met, and extension studies are voluntary and lose people.","Linking a modern treatment record to a cancer registry thirty years later requires identifiers and consent arrangements most countries have not built."],"currentEfforts":["Fifteen-year long-term follow-up is recommended for gene-modified cell products, and registries including CIBMTR and EBMT collect long-term outcomes after transplant.","The paediatric cohorts have shown the method works when exposures are abstracted at entry: the falling second cancer rate after de-escalated childhood regimens is measurable precisely because the old cohorts exist.","Several records on this front name the unmeasurable period explicitly rather than implying safety."],"successLooksLike":"Every new systemic therapy enters a registry linkage at approval that will still report second cancers and cardiovascular events thirty years later without depending on the sponsor or on individual recall."},{"id":"b-undruggable-targets","kind":"bottleneck","name":"The undruggable drivers","aka":[],"tldr":"The proteins that drive most cancers, such as MYC, mutant p53 and most RAS variants, still have no good drug.","summary":"KRAS G12C showed that 'undruggable' is a technology problem, not a law: a covalent pocket that exists only in one mutant state yielded sotorasib and adagrasib within a decade of its discovery. But most of the oncogenic burden of human cancer sits on proteins with no enzymatic pocket, flat protein-protein interfaces, or intrinsically disordered regions: MYC, most RAS alleles, mutant and wild-type TP53, transcription factors such as the fusion oncoproteins of sarcomas and leukaemias, and phosphatases. The therapies we have act on downstream nodes with narrow therapeutic windows and rapid feedback reactivation. New modalities, including pan-RAS(ON) inhibitors, molecular glues and degraders, antisense, and peptide or vaccine strategies presenting mutant epitopes, are the first serious assault on this class. RAS mutations alone occur in roughly a fifth of all cancers, so each success reshapes several diseases at once.","asOf":"2026-09-08","links":[{"label":"Dang et al., Drugging the 'undruggable' cancer targets (Nature Reviews Cancer 2017)","url":"https://doi.org/10.1038/nrc.2017.36"},{"label":"NCI RAS Initiative","url":"https://www.cancer.gov/research/key-initiatives/ras"},{"label":"Prior et al., The frequency of Ras mutations in cancer (Cancer Research 2020)","url":"https://doi.org/10.1158/0008-5472.CAN-19-3682"}],"tags":[],"related":["kras-roadmap","idea-ras-inhibitor-neoadjuvant-pdac","idea-shared-kras-vaccine-adjuvant","depmap","open-targets","idea-bio1-myc-max-molecular-glue","idea-bio1-omomyc-mrna","idea-bio1-p53-mutant-reactivator-expansion","idea-bio1-mutant-p53-degrader","idea-bio1-pan-ras-covalent-g12d","idea-bio1-covalent-ligandability-atlas","idea-bio1-glue-degrader-atlas","idea-bio1-fusion-tf-degraders","idea-bio1-condensate-disruptors","idea-bio1-tumour-restricted-e3-atlas","idea-bio1-antibody-degrader-conjugates","idea-bio1-lytac-surface-degraders","idea-bio1-rna-targeting-small-molecules","idea-bio1-epigenetic-silencing-in-vivo","idea-bio1-paralog-synthetic-lethality","idea-bio1-pp2a-activators","idea-bio1-wrn-msi-programme","idea-bio1-macrocycle-ppi-campaign","idea-bio1-arv7-degrader","idea-bio1-ai-binders-disordered-regions","idea-bio1-pmhc-bispecifics-public-drivers","idea-bio1-mrna-intrabodies","idea-bio1-translation-dependency-myc","idea-bio1-undruggable-open-consortium","idea-bio1-undruggable-market-commitment","idea-moon-synthetic-lethality-map-every-driver","idea-moon-open-degrader-consortium"],"cancers":["pancreatic","colorectal","nsclc","aml","sarcoma"],"sections":[],"technologies":["kras-inhibitors","protac-degrader","antisense-sirna","synthetic-lethality-approaches","ai-drug-design","shared-antigen-vaccine"],"targets":["kras","tp53","menin","braf","kmt2a","npm1"],"drugs":["sotorasib","adagrasib","daraxonrasib","mrtx1133","zoldonrasib","elironrasib","revumenib","ziftomenib","eli-002-7p","eprenetapopt"],"companies":["revolution-medicines","arvinas","kymera","nurix","c4-therapeutics","monte-rosa","quanta-therapeutics","frontier-medicines","elicio-therapeutics"],"institutions":["nci","broad-institute"],"pathways":["ras-mapk","p53-cell-cycle"],"terms":[],"trials":["rasolute-302","codebreak-200","krystal-12","amplify-7p"],"people":[],"bottlenecks":[],"keyPapers":["paper-dang-nat-rev-cancer","paper-prior-cancer-res"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Cancers carrying a RAS mutation (US incidence-weighted estimate)","value":"~19% of all cancers","source":"Prior, Hood & Hartley, Cancer Research 2020","url":"https://doi.org/10.1158/0008-5472.CAN-19-3682"},{"label":"TP53 mutation frequency across 12 major tumour types (TCGA pan-cancer)","value":"42% of tumours","source":"Kandoth et al., Nature 2013","url":"https://doi.org/10.1038/nature12634"}],"causes":["Transcription factors and RAS proteins lack deep hydrophobic pockets, so conventional small molecules cannot bind them tightly.","Picomolar affinity of RAS for GTP defeats competitive nucleotide-site inhibitors.","Downstream inhibition (MEK, ERK, PI3K) triggers feedback reactivation and toxicity in normal tissue before it suppresses the tumour.","Loss-of-function tumour suppressors such as TP53 cannot be inhibited; they must be restored, degraded in their mutant form, or exploited through synthetic lethality.","Fusion oncoproteins and MYC are intrinsically disordered, which frustrates structure-based design."],"currentEfforts":["The NCI RAS Initiative at the Frederick National Laboratory works as an open consortium on RAS structure, biochemistry and screening.","Revolution Medicines' daraxonrasib (RMC-6236), a pan-RAS(ON) tri-complex inhibitor, is in phase 3 in pancreatic cancer (RASolute 302) after showing activity across KRAS alleles.","MRTX1133 and other non-covalent KRAS G12D inhibitors extend the covalent-pocket lesson to the most common pancreatic allele.","Arvinas, Kymera, Nurix, C4 Therapeutics and Monte Rosa are building PROTAC degraders and molecular glues that remove proteins without needing a functional pocket.","Menin inhibitors (revumenib, ziftomenib) show that a transcriptional complex can be drugged at a protein-protein interface in KMT2A-rearranged and NPM1-mutant leukaemia.","Off-the-shelf mutant-KRAS vaccines (ELI-002) and TCR therapies present undruggable proteins to the immune system instead of blocking them."],"successLooksLike":"Direct inhibitors or degraders with a usable therapeutic window exist for every major RAS allele, for MYC and for mutant TP53, and at least one of them extends overall survival in pancreatic or colorectal cancer."},{"id":"b-translational-valley","kind":"bottleneck","name":"The valley of death between lab and product","aka":[],"tldr":"Most academic discoveries die before anyone tests them in people because nobody funds the middle step.","summary":"Between a validated target or lead compound in an academic laboratory and a first-in-human trial lie medicinal chemistry optimisation, GMP manufacturing, toxicology, formulation and an IND or CTA dossier, work that costs several million dollars, is not publishable, and is funded by neither research grants nor, until the asset is de-risked, by companies or venture capital. Of highly promising basic science claims published in leading journals, a small minority reach licensed use decades later. The consequence is that promising ideas stall for years or vanish, and that the ideas that do cross are selected for commercial attractiveness rather than for medical need. Public translational programmes (NCI Experimental Therapeutics, NCATS), charity drug-development units, academic-industry alliances and non-dilutive philanthropic capital exist but are small relative to the flow of discoveries.","asOf":"2026-09-08","links":[{"label":"Butler, Translational research: crossing the valley of death (Nature 2008)","url":"https://doi.org/10.1038/453840a"},{"label":"Contopoulos-Ioannidis et al., Translation of highly promising basic research into clinical applications (Am J Med 2003)","url":"https://doi.org/10.1016/S0002-9343(03)00013-5"},{"label":"NCI Experimental Therapeutics (NExT) Program","url":"https://next.cancer.gov/"},{"label":"NCATS","url":"https://ncats.nih.gov/"}],"tags":[],"related":["open-targets","depmap","idea-fund-national-drug-development-office","idea-fund-academic-sponsor-indemnity-pool","idea-moon-public-phase1-factory"],"cancers":[],"sections":["drug-discovery"],"technologies":["ai-drug-design","protac-degrader","crispr-screens"],"targets":[],"drugs":[],"companies":["recursion","insilico-medicine","isomorphic-labs","generate-biomedicines"],"institutions":["nci","cruk","broad-institute","francis-crick","dkfz","icr-london","nki"],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-butler-nature","paper-contopoulos-ioannidis-am-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"funding-incentives","severity":"major","metrics":[{"label":"Highly promising basic science findings (101 claims in top journals, 1979-1983) that led to a licensed clinical use within about 20 years","value":"5 of 101; 1 in extensive use","source":"Contopoulos-Ioannidis, Ntzani & Ioannidis, American Journal of Medicine 2003","url":"https://doi.org/10.1016/S0002-9343(03)00013-5"},{"label":"Probability that an oncology drug entering phase 1 is eventually approved","value":"3.4%","source":"Wong, Siah & Lo, Biostatistics 2019","url":"https://doi.org/10.1093/biostatistics/kxx069"}],"causes":["Research grants fund discovery, not the GMP manufacturing, toxicology and regulatory work needed to reach humans.","Companies and investors want assets de-risked to a stage academic labs cannot reach alone.","Academic incentives reward publications, not INDs, and translational work is not publishable.","Universities lack the specialised staff (CMC, regulatory, clinical operations) and technology transfer is slow.","Ideas for rare, paediatric or low-income indications have no commercial pull at any stage."],"currentEfforts":["The NCI Experimental Therapeutics (NExT) programme provides free drug development services from lead optimisation through IND-enabling studies and early trials.","NCATS and the Clinical and Translational Science Award hubs fund translational infrastructure at US academic centres.","Cancer Research UK's Centre for Drug Development and Commercial Partnerships take academic assets into phase 1 and license them onward.","Stand Up To Cancer Dream Teams and the Cancer Grand Challenges fund multi-institution translational teams with milestone-based awards.","ARPA-H funds high-risk translational programmes with active programme management.","Academic-industry alliances (for example, AstraZeneca's and Bayer's open innovation programmes and the Structural Genomics Consortium) share tools and compounds pre-competitively."],"successLooksLike":"Any target or lead with strong validation can access funded, professional IND-enabling development regardless of commercial attractiveness, and the median time from a validated academic discovery to first-in-human dosing falls below five years."},{"id":"b-combination-space","kind":"bottleneck","name":"Too many combinations to test","aka":[],"tldr":"There are thousands of possible drug pairs and sequences. Trials can test a few dozen a year.","summary":"With well over a hundred approved oncology agents and hundreds in development, the space of pairwise combinations runs to tens of thousands, and sequences and schedules multiply it further, while the field can run only a few dozen adequately powered combination trials a year. The combinations that are tested are chosen by commercial ownership and precedent rather than biology: thousands of PD-1/PD-L1 combination trials have been launched, most adding an agent to a checkpoint inhibitor without a predictive biomarker. Analyses of historical combination trials suggest that many 'successes' reflect independent action in different patients rather than synergy, which means better patient selection would achieve the same benefit with fewer drugs. Platform trials, factorial and adaptive designs, ex vivo functional testing, and computational prioritisation from dependency maps and combination screens are the only ways to explore the space at a useful rate.","asOf":"2026-09-08","links":[{"label":"Palmer & Sorger, Combination cancer therapy can confer benefit via patient-to-patient variability without drug additivity or synergy (Cell 2017)","url":"https://doi.org/10.1016/j.cell.2017.11.009"},{"label":"Upadhaya et al., Challenges and opportunities in the PD1/PDL1 inhibitor clinical trial landscape (Nat Rev Drug Discov 2022)","url":"https://doi.org/10.1038/d41573-022-00030-4"},{"label":"Menden et al., Community assessment to advance computational prediction of cancer drug combinations (Nature Communications 2019)","url":"https://doi.org/10.1038/s41467-019-09799-2"}],"tags":[],"related":["depmap","drugbank-chembl","idea-payload-switching","idea-dual-payload-first","idea-organoid-guided-adc","idea-tr2-shared-control-network","idea-tr2-organoid-coclinical-arms","idea-tr2-alternating-vs-concurrent","idea-tr2-rwe-combination-emulation","idea-tr2-factorial-adjuvant-generics","idea-tr2-platform-single-ethics","idea-tr2-resistance-mechanism-baskets","idea-tr2-combination-utility","idea-tr2-combination-patent-pool","idea-tr2-sequence-registry","idea-tr2-bandit-allocation","idea-tr2-qsp-combo-dosing","idea-tr2-combo-readiness-dossier","idea-tr2-rt-drug-platform","idea-tr2-paediatric-combo-prea","idea-tr2-combination-forecast-tournament","idea-tr2-antagonism-surveillance","idea-tr2-contribution-of-components-mandate","idea-tr2-window-of-opportunity-triplets","idea-moon-automated-combination-discovery"],"cancers":["prostate","tnbc","nsclc","melanoma"],"sections":[],"technologies":["ai-drug-design","functional-drug-testing","organoids","bh3-profiling","dual-payload-adc","crispr-screens"],"targets":[],"drugs":[],"companies":[],"institutions":["nci","broad-institute"],"pathways":[],"terms":["basket-umbrella-platform","adc-sequencing","first-line"],"trials":["stampede","myelomatch"],"people":[],"bottlenecks":[],"keyPapers":["paper-palmer-cell","paper-upadhaya-nat-rev-drug-discov","paper-menden-nat-commun"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Clinical trials of PD-1/PD-L1 inhibitors registered by 2021, the large majority as combinations","value":"5,683 trials","source":"Upadhaya et al., Nature Reviews Drug Discovery 2022","url":"https://doi.org/10.1038/d41573-022-00030-4"},{"label":"Approved combination therapies whose clinical benefit is explained by independent drug action (patient-to-patient variability) without synergy","value":"Most of 15 combinations analysed","source":"Palmer & Sorger, Cell 2017","url":"https://doi.org/10.1016/j.cell.2017.11.009"}],"causes":["Combinatorial growth: n drugs give n(n-1)/2 pairs before doses, schedules and sequences are considered.","Companies preferentially combine their own assets and rarely cross-license for early testing.","Conventional two-arm trials test one combination each and take years.","Preclinical synergy poorly predicts clinical benefit, so prioritisation is weak.","Regulatory paths for combinations of two unapproved agents are complex."],"currentEfforts":["I-SPY 2 (Quantum Leap Healthcare Collaborative) and STAMPEDE run adaptive platform trials that add and graduate combination arms continuously.","NCI ComboMATCH tests biomarker-directed combinations across cooperative groups.","DREAM Challenges and the NCI-DREAM drug combination prediction challenge benchmark computational prioritisation of pairs.","The AstraZeneca-Sanger drug combination screen and DepMap provide public combination and dependency data for hypothesis generation.","Ex vivo functional testing (organoids, explants, BH3 profiling) is used to select combinations for individual patients.","Payload-class switching and dual-payload ADCs address the sequencing problem structurally instead of trial by trial."],"successLooksLike":"Platform trials in each major cancer test dozens of combination arms a year against shared controls, and computational or ex vivo prioritisation has a demonstrated positive predictive value for which combinations succeed in the clinic."},{"id":"b-toxicity-qol","kind":"bottleneck","name":"Toxicity and quality of life are undervalued","aka":[],"tldr":"Trials measure how long people live, not how they live. Side-effects are under-reported and under-treated.","summary":"Oncology trials are designed around survival and progression, and the harms and lived experience of treatment are secondary at best. Clinicians systematically under-report symptomatic toxicities compared with patients, quality-of-life data are collected in a minority of trials, analysed inconsistently, published late or not at all, and rarely influence approval, guidelines or price: more than half of EMA approvals in 2009-13 had no evidence of quality-of-life benefit at approval and most still did not years later. Chronic low-grade toxicity, which drives discontinuation of oral therapies, is invisible in the grade 3-4 tables that trials report. Immune-related adverse events, ADC-associated interstitial lung disease, and CAR-T neurotoxicity have created new categories of harm that patients must weigh against uncertain benefit. Patient-reported outcome instruments, standardised analysis, mandatory reporting and benefit frameworks that grade quality of life are the corrective tools.","asOf":"2026-09-08","links":[{"label":"Di Maio et al., Symptomatic toxicities experienced during anticancer treatment: agreement between patient and physician reporting (JCO 2015)","url":"https://doi.org/10.1200/JCO.2014.57.9334"},{"label":"Davis et al., Availability of evidence of benefits on overall survival and quality of life of cancer drugs approved by EMA 2009-13 (BMJ 2017)","url":"https://doi.org/10.1136/bmj.j4530"},{"label":"EORTC Quality of Life Group","url":"https://qol.eortc.org/"}],"tags":[],"related":["fda-approvals","esmo-guidelines","idea-moon-toxicity-first-endpoints","idea-moon-supportive-care-arpa","idea-moon-neuropathy-prevention-programme","idea-moon-cardioprotection-by-default","idea-moon-olanzapine-antiemetic-everywhere","idea-moon-quality-adjusted-pricing","idea-moon-adult-late-effects-registry","idea-moon-cognitive-toxicity-programme","idea-moon-irae-prediction-and-prevention","idea-moon-fertility-preservation-default","idea-moon-hearing-protection-cisplatin","idea-moon-oncology-hospital-at-home","idea-moon-supportive-care-platform-trial","idea-moon-open-pro-data-commons","idea-moon-mucositis-taste-programme","idea-moon-sexual-health-as-toxicity-domain","idea-moon-quality-adjusted-survival-standard","idea-cost-financial-toxicity-screening"],"cancers":[],"sections":["supportive-care"],"technologies":["scalp-cooling","cardio-oncology","adc","checkpoint-inhibitor","car-t"],"targets":[],"drugs":[],"companies":["curie-nki-eortc"],"institutions":["esmo","asco","nci"],"pathways":[],"terms":["irae","crs","icans","ild","orr","pfs","os","hazard-ratio"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-davis-bmj","paper-di-maio-j-clin-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"people-culture","severity":"major","metrics":[{"label":"Cancer drug indications approved by the EMA 2009-2013 with no evidence of quality-of-life or survival benefit at approval","value":"57%","source":"Davis et al., BMJ 2017","url":"https://doi.org/10.1136/bmj.j4530"},{"label":"Agreement between physician and patient reporting of symptomatic toxicities (anorexia, nausea, vomiting, constipation, diarrhoea, hair loss) in three randomised trials","value":"Physicians under-reported all six symptoms","source":"Di Maio et al., JCO 2015","url":"https://doi.org/10.1200/JCO.2014.57.9334"},{"label":"Median overall survival with routine patient-reported symptom monitoring vs usual care during chemotherapy (randomised)","value":"31.2 vs 26.0 months; not replicated","source":"Basch et al., JAMA 2017; PRO-TECT, JAMA 2022, found no survival difference","url":"https://doi.org/10.1001/jama.2017.7156"}],"causes":["Regulatory approval and pricing are based on efficacy endpoints, so sponsors invest little in quality-of-life measurement.","Clinician-graded adverse event scales miss what patients feel and ignore chronic low-grade toxicity.","Quality-of-life analyses are heterogeneous, unpublished or reported long after the primary paper.","Trial designs treat quality of life as exploratory rather than as a co-primary endpoint.","Toxicity management is under-resourced and supportive-care research is under-funded."],"currentEfforts":["PRO-CTCAE (NCI) provides a validated patient-reported version of the toxicity scale now used in many trials.","The FDA guidance on core patient-reported outcomes in cancer clinical trials (2021 draft) specifies which PRO domains should be measured.","SISAQOL-IMI is developing consensus standards for analysing and reporting quality-of-life data in cancer trials.","ESMO-MCBS awards credit for demonstrated quality-of-life improvement, and EORTC QLQ-C30 and its modules are the standard instruments.","Project Optimus and dose-optimisation trials aim to reduce chronic toxicity by testing lower doses and shorter durations.","Common Sense Oncology campaigns for endpoints that matter to patients, including quality of life and overall survival."],"successLooksLike":"Every registrational oncology trial collects and publishes patient-reported outcomes with the primary results, quality of life is graded in every value framework and reimbursement decision, and discontinuation for toxicity on chronic oral therapy falls below 10%."},{"id":"b-trial-design","kind":"bottleneck","name":"Trial design, endpoints and cost","aka":[],"tldr":"A phase 3 trial takes years and hundreds of millions of dollars, and often answers a question that has already moved on.","summary":"The randomised phase 3 trial remains the evidentiary standard, but the way oncology runs it is slow, expensive and often uninformative. Surrogate endpoints such as progression-free survival and response rate correlate weakly with overall survival for most indications, yet they underpin most approvals; more than half of cancer drugs approved by the EMA in 2009-13 had no evidence of survival or quality-of-life benefit at approval, and most still did not years later. Control arms are frequently obsolete by readout, crossover and post-progression therapy blur survival differences, and duration and dose of therapy are almost never tested. Median clinical development costs run to hundreds of millions of dollars per approved drug, which prices out academic questions. Platform and adaptive designs, ctDNA and MRD endpoints, pragmatic registry-based randomisation and clear value frameworks are the tools for faster, cheaper and more honest trials.","asOf":"2026-09-08","links":[{"label":"Davis et al., Availability of evidence of benefits on survival and quality of life of cancer drugs approved by EMA 2009-13 (BMJ 2017)","url":"https://doi.org/10.1136/bmj.j4530"},{"label":"Prasad et al., Strength of association between surrogate endpoints and survival in oncology (JAMA Intern Med 2015)","url":"https://doi.org/10.1001/jamainternmed.2015.2829"},{"label":"FDA Oncology Center of Excellence, Project FrontRunner","url":"https://www.fda.gov/about-fda/oncology-center-excellence/project-frontrunner"},{"label":"Wouters et al., Estimated research and development investment needed to bring a new medicine to market (JAMA 2020)","url":"https://doi.org/10.1001/jama.2020.1166"}],"tags":[],"related":["fda-approvals","esmo-guidelines","clinicaltrials-gov","idea-mrd-guided-stop-myeloma","idea-tr1-external-control-rulebook","idea-tr1-bayesian-borrowing-smaller-controls","idea-tr1-seamless-2-3-with-prespecified-go","idea-tr1-standing-platform-per-cancer","idea-tr1-crossover-adjusted-survival-standard","idea-tr1-ttf-and-qol-coprimary","idea-tr1-public-de-escalation-trial-fund","idea-tr1-immunotherapy-stop-trials","idea-tr1-ctdna-guided-stop-in-metastatic-disease","idea-tr1-mams-for-sequencing-questions","idea-tr1-post-approval-pragmatic-trial-in-excluded","idea-tr1-shared-control-arms-across-sponsors","idea-tr1-tolerability-estimands","idea-tr1-registry-linked-os-followup","idea-tr1-shrinkage-subgroup-analysis","idea-tr1-ai-central-imaging-reads","idea-tr1-public-trial-cost-benchmarks","idea-tr1-esource-ehr-to-edc","idea-tr1-aggregated-n-of-1-supportive-care","idea-tr1-aggressive-futility-boundaries","idea-tr1-window-of-opportunity-default","idea-tr1-randomised-phase-2-before-phase-3","idea-tr1-smart-designs-for-adaptive-strategies","idea-tr1-target-trial-emulation-to-prioritise-rcts","idea-tr1-tumour-agnostic-approval-standard","idea-tr1-open-protocol-and-sap-library","idea-tr1-validate-real-world-progression-endpoints"],"cancers":["prostate","multiple-myeloma","breast-hr-positive"],"sections":[],"technologies":["mrd-testing","ai-trial-matching"],"targets":[],"drugs":[],"companies":[],"institutions":["esmo","asco","nci"],"pathways":[],"terms":["pfs","os","orr","hazard-ratio","accelerated-approval","basket-umbrella-platform","real-world-evidence","mrd-negativity-myeloma","pcr","efs","project-frontrunner"],"trials":["stampede","myelomatch"],"people":[],"bottlenecks":[],"keyPapers":["paper-davis-bmj","paper-wouters-jama"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"critical","metrics":[{"label":"Median estimated R&D investment to bring a new drug to market (all areas, 2009-2018), capitalised","value":"Median US$985 million, mean US$1.3 billion; oncology among the highest","source":"Wouters, McKee & Luyten, JAMA 2020","url":"https://doi.org/10.1001/jama.2020.1166"},{"label":"Cancer drug indications approved by the EMA 2009-2013 with no evidence of survival or quality-of-life benefit at approval","value":"57%","source":"Davis et al., BMJ 2017","url":"https://doi.org/10.1136/bmj.j4530"},{"label":"Trial-level correlations between surrogate endpoints and overall survival in oncology rated weak vs strong (systematic review)","value":"52% weak, 23% strong","source":"Prasad et al., JAMA Internal Medicine 2015","url":"https://doi.org/10.1001/jamainternmed.2015.2829"},{"label":"Median overall survival gain of cancer drugs approved by FDA 2002-2014","value":"2.1 months","source":"Fojo, Mailankody & Lo, JAMA Otolaryngology 2014","url":"https://doi.org/10.1001/jamaoto.2014.1570"}],"causes":["Regulators accept surrogate endpoints that were never validated for the specific indication.","Sponsors choose control arms, endpoints and populations to maximise the chance of a positive trial rather than to answer the clinical question.","Trials take so long that the standard of care changes before readout.","Duration, dose and sequencing questions have no commercial sponsor.","Per-patient costs, site set-up, monitoring and data management have inflated far faster than the value of the information produced."],"currentEfforts":["The FDA Oncology Center of Excellence's Project FrontRunner and Project Confirm push earlier randomised development and enforce confirmatory trials after accelerated approval.","ESMO-MCBS and the ASCO Value Framework grade trials by magnitude of clinical benefit, exposing marginal results.","STAMPEDE, I-SPY 2 and myeloMATCH demonstrate multi-arm multi-stage platform trials that add and drop arms without restarting.","FDA accepted MRD negativity as an endpoint for accelerated approval in multiple myeloma (2024) and is evaluating ctDNA as an early endpoint in solid tumours.","Registry-based and pragmatic randomised trials (for example, in Scandinavian cardiology and now oncology) cut per-patient cost by an order of magnitude.","The Common Sense Oncology movement advocates overall survival and quality of life as the default endpoints."],"successLooksLike":"New approvals are based on validated endpoints (survival, quality of life or a surrogate proven for that indication), a typical phase 3 answers its question in under three years at a fraction of today's cost, and platform trials become the default in the major cancers."},{"id":"b-trial-diversity","kind":"bottleneck","name":"Trials do not represent the people who get cancer","aka":[],"tldr":"Older, Black, Hispanic, Asian, rural, poor and multimorbid patients are under-represented, so results may not apply to them.","summary":"Trials that lead to cancer drug approvals enrol predominantly younger, white, fitter patients treated at academic centres in a handful of high-income countries. In pivotal trials supporting US approvals from 2008 to 2018, Black patients made up about 3% and Hispanic patients about 6% of participants, and race was not even reported in a large share of trials; patients over 65 have been under-represented for decades relative to their share of incidence. Pharmacogenomic differences (for example, in DPYD, UGT1A1 and EGFR mutation prevalence), differences in comorbidity, body composition and social context all mean that efficacy and toxicity can differ, and the evidence to know does not exist. The geographic concentration also means most of the world's patients are treated on the basis of trials in populations unlike them. Regulators now require diversity action plans, but enforcement, infrastructure and trust are the real constraints.","asOf":"2026-09-08","links":[{"label":"Loree et al., Disparity of race reporting and representation in trials leading to cancer drug approvals (JAMA Oncology 2019)","url":"https://doi.org/10.1001/jamaoncol.2019.1870"},{"label":"FDA Oncology Center of Excellence, Project Equity (archived copy)","url":"https://web.archive.org/web/20240415200043/https://www.fda.gov/about-fda/oncology-center-excellence/project-equity"},{"label":"Hutchins et al., Underrepresentation of patients 65 years of age or older in cancer-treatment trials (NEJM 1999)","url":"https://doi.org/10.1056/NEJM199912303412706"}],"tags":[],"related":["fda-approvals","clinicaltrials-gov","idea-tr1-diversity-plans-with-consequences","idea-tr1-ancestry-aware-pharmacology-programme","idea-tr1-duffy-null-neutrophil-threshold","idea-tr1-mandatory-older-adult-cohort","idea-tr1-parallel-comorbidity-cohorts","idea-tr1-lmic-sites-in-pivotal-trials","idea-tr1-community-health-worker-recruitment","idea-tr1-language-access-in-trials","idea-tr1-site-equity-index","idea-tr1-include-pregnancy-capable-people-sensibly","idea-tr1-age-floor-twelve-for-adult-trials","idea-tr1-sex-stratified-pk-and-dosing","idea-tr1-disability-and-mental-illness-inclusion","idea-tr1-representativeness-in-the-label","idea-tr1-diverse-investigator-pipeline","idea-tr1-paid-community-advisory-boards","idea-tr1-post-marketing-safety-by-ancestry-and-sex","idea-tr1-ecog-2-dedicated-cohorts","idea-tr1-trial-desert-map"],"cancers":["prostate","multiple-myeloma","tnbc","nsclc","hcc","gastric"],"sections":[],"technologies":["geriatric-assessment","ai-trial-matching"],"targets":[],"drugs":[],"companies":["swog","ecog-acrin","hengrui","akeso","beone"],"institutions":["nci","asco","sysucc","tata-memorial","cams-cancer-hospital"],"pathways":[],"terms":["egfr-exon19-l858r","hazard-ratio"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-loree-jama-oncol","paper-hutchins-n-engl-j-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Black and Hispanic participants in trials supporting FDA cancer drug approvals 2008-2018","value":"3.1% and 6.1%","source":"Loree et al., JAMA Oncology 2019","url":"https://doi.org/10.1001/jamaoncol.2019.1870"},{"label":"Patients aged 65 or older in SWOG treatment trials 1993-1996 vs their share of US cancer incidence","value":"25% vs 63%","source":"Hutchins et al., NEJM 1999","url":"https://doi.org/10.1056/NEJM199912303412706"},{"label":"Share of genome-wide association study participants of European ancestry (2016)","value":"81%","source":"Popejoy & Fullerton, Nature 2016","url":"https://doi.org/10.1038/538161a"}],"causes":["Trial sites are concentrated at academic centres in wealthy areas and countries.","Eligibility criteria on organ function, comorbidity and performance status disproportionately exclude older and minority patients.","Historical abuses and present-day experiences of discrimination reduce trust in research.","Costs of participation (travel, time off work, childcare) are highest for poorer patients.","Sponsors face no penalty for unrepresentative enrolment and prefer fast accrual at established sites."],"currentEfforts":["The Food and Drug Omnibus Reform Act (2022) requires Diversity Action Plans for pivotal trials, with FDA draft guidance issued in 2024.","FDA Project Equity in the Oncology Center of Excellence coordinates representation in oncology trials.","The ASCO-ACCC initiative provides site self-assessment and Just ASK implicit-bias training to increase enrolment of under-represented groups.","NIH policy on inclusion across the lifespan (2019) removed age-based exclusions from NIH-funded trials.","NCORP places NCI trials in minority and underserved community sites.","Asian and Chinese cooperative groups and companies (Hengrui, Akeso, BeOne) now run registrational trials in Asian populations, changing the geographic mix of evidence."],"successLooksLike":"Enrolment in pivotal trials matches the age, ancestry and geographic distribution of the disease within a few percentage points, and subgroup data are sufficient to detect clinically meaningful differences in efficacy or toxicity."},{"id":"b-trial-enrolment","kind":"bottleneck","name":"Trials enrol too few, too slowly","aka":[],"tldr":"Fewer than one in ten adults with cancer joins a trial. Trials close for lack of patients, not lack of ideas.","summary":"Around 8% of adults with cancer participate in a treatment trial, and a fifth of trials fail to complete, most often for poor accrual. The barriers are structural before they are personal: most patients are treated at community practices with no open trial, eligibility criteria exclude patients with brain metastases, organ dysfunction, prior cancers or HIV, travel and time costs fall on patients, and physicians have no time or incentive to screen and refer. When a trial is actually offered, more than half of patients say yes. Slow accrual lengthens trials, raises costs, delays answers and kills questions that industry will not fund. Broadened eligibility, decentralised and community-based trial infrastructure, automated matching from the electronic record, and paying patients' costs are the fixes with evidence.","asOf":"2026-09-08","links":[{"label":"Unger et al., Systematic review of barriers to cancer trial participation (JNCI 2019)","url":"https://doi.org/10.1093/jnci/djy221"},{"label":"Kim et al., Broadening eligibility criteria to make clinical trials more representative (JCO 2017)","url":"https://doi.org/10.1200/JCO.2017.73.7916"},{"label":"NCI Community Oncology Research Program","url":"https://ncorp.cancer.gov/"}],"tags":[],"related":["clinicaltrials-gov","eudract-ctis","idea-tr1-justify-every-exclusion","idea-tr1-brain-mets-default-included","idea-tr1-pk-informed-organ-thresholds","idea-tr1-prior-cancer-hiv-hepatitis-inclusion","idea-tr1-ehr-point-of-care-trial-alert","idea-tr1-tumour-board-trial-line-item","idea-tr1-ngs-report-live-trial-match","idea-tr1-just-in-time-site-network","idea-tr1-national-master-trial-agreement","idea-tr1-mutual-recognition-ethics-review","idea-tr1-home-infusion-trial-drugs","idea-tr1-community-site-quota","idea-tr1-travel-lodging-in-every-budget","idea-tr1-pay-participants-for-time","idea-tr1-lay-trial-navigators","idea-tr1-opt-out-research-contact-register","idea-tr1-registry-embedded-randomisation","idea-tr1-rare-cancer-trial-in-a-box","idea-tr1-pre-consented-cohort-randomisation","idea-tr1-public-screen-fail-reasons","idea-tr1-eligibility-impact-statement","idea-tr1-live-trial-slot-api","idea-tr1-pathology-triggered-referral","idea-tr1-evening-weekend-trial-clinics","idea-tr1-universal-routine-cost-coverage","idea-tr1-open-trials-inside-guidelines","idea-tr1-mobile-research-units","idea-tr1-drop-non-essential-biopsies","idea-moon-global-open-trials-os","idea-cost-trial-enrolment-cost-lever"],"cancers":[],"sections":[],"technologies":["ai-trial-matching"],"targets":[],"drugs":[],"companies":["swog","ecog-acrin","alliance-oncology","nrg-oncology","cctg","jcog","unicancer","tempus"],"institutions":["nci","asco"],"pathways":[],"terms":["basket-umbrella-platform","standard-of-care"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-kim-j-clin-oncol"],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"critical","metrics":[{"label":"Adults with cancer participating in treatment trials (US, meta-analysis)","value":"~8%","source":"Unger et al., JNCI 2019","url":"https://doi.org/10.1093/jnci/djy221"},{"label":"Patients who agree to participate when a trial is offered to them","value":"55%","source":"Unger et al., JNCI 2021","url":"https://doi.org/10.1093/jnci/djaa155"},{"label":"Adult cancer trials in ClinicalTrials.gov that failed to complete, most often for poor accrual","value":"~20%","source":"Stensland et al., JNCI 2014","url":"https://doi.org/10.1093/jnci/dju229"}],"causes":["Most patients are treated in community settings where no relevant trial is open.","Restrictive eligibility criteria exclude a large share of real patients for reasons unrelated to safety.","Travel, lost income and out-of-pocket costs fall on patients and are rarely reimbursed.","Physicians lack time, incentives and tools to identify eligible patients.","Trial activation takes months per site because of contracting and ethics review, so windows of opportunity close."],"currentEfforts":["ASCO and Friends of Cancer Research broadened eligibility recommendations have been adopted in FDA guidance on brain metastases, organ dysfunction, prior malignancy and HIV.","The NCI Community Oncology Research Program (NCORP) brings NCI trials to community practices across the US.","AI-based trial matching from the electronic record (Tempus, TrialJectory and others) screens patients continuously.","FDA guidance on decentralised clinical trials (2023) allows remote visits, local labs and telehealth consent.","Just-in-time site activation models and central IRBs cut the months of set-up before a site can enrol.","Programmes that reimburse patient travel and lodging (for example, Lazarex Cancer Foundation) directly remove the cost barrier."],"successLooksLike":"At least a quarter of adults with cancer are enrolled in a trial, eligibility criteria are justified by safety alone, and median phase 3 accrual time halves so that fewer than 5% of trials close for poor accrual."},{"id":"b-tumor-heterogeneity","kind":"bottleneck","name":"Tumour heterogeneity and clonal evolution","aka":[],"tldr":"Every tumour is a population of genetically distinct clones: in multi-region sequencing of kidney tumours, roughly two thirds of mutations were missing from at least one region. Treatment kills the dominant clones and leaves resistant minor clones to grow back, yet a single diagnostic biopsy is still treated as the whole disease.","summary":"Every cancer is a population of clones that differ by lesion, by region within a lesion, and over time. Multi-region sequencing of renal and lung tumours shows that a majority of somatic mutations are not shared by every region, so a single biopsy at diagnosis systematically under-samples the disease it is used to treat. Therapy then acts as a selection pressure: pre-existing minor subclones carrying resistance alleles expand, and new lesions can be driven by different clones from the primary. Clinical practice still assumes one genotype per patient, re-biopsies at progression are uncommon, and few trials adapt treatment to the clone that is actually growing. Longitudinal circulating tumour DNA, single-cell and spatial profiling, and evolutionary trial designs are the tools that could turn heterogeneity from an excuse into a target.","asOf":"2026-09-08","links":[{"label":"Gerlinger et al., Intratumor heterogeneity and branched evolution (NEJM 2012)","url":"https://doi.org/10.1056/NEJMoa1113205"},{"label":"TRACERx lung cancer evolution (Nature 2023)","url":"https://doi.org/10.1038/s41586-023-05783-5"},{"label":"McGranahan & Swanton, Clonal heterogeneity and tumor evolution (Cell 2017)","url":"https://doi.org/10.1016/j.cell.2017.01.018"}],"tags":[],"related":["genie","cbioportal","idea-ctdna-switch-generalised","idea-bio1-multiregion-blocks-default","idea-bio1-rapid-autopsy-network","idea-bio1-truncal-branch-labelling","idea-bio1-outlier-lesion-biopsy","idea-bio1-barcoded-avatars-clonal-fitness","idea-bio1-evolvability-index","idea-bio1-adaptive-therapy-platform","idea-bio1-ctdna-adaptive-tki","idea-bio1-first-strike-second-strike","idea-bio1-collateral-sensitivity-atlas","idea-bio1-evolutionary-double-bind","idea-bio1-spatial-clone-immune-map","idea-bio1-methylation-clone-tracking","idea-bio1-evolution-forecasting","idea-bio1-baseline-ultradeep-resistant-clones","idea-bio1-cfrna-plasticity-tracking","idea-bio1-evolutionary-tumour-boards","idea-bio1-apobec-inhibitor-adjunct","idea-bio1-cin-vulnerability-kif18a","idea-bio1-federated-evolution-atlas","idea-bio1-clone-to-lesion-mapping","idea-bio1-rebiopsy-before-switch","idea-moon-open-cancer-cell-state-model"],"cancers":["rcc","nsclc","glioblastoma","pancreatic","breast-hr-positive"],"sections":[],"technologies":["single-cell-spatial","liquid-biopsy","cgp","wes-wgs","mrd-testing"],"targets":[],"drugs":[],"companies":["10x-genomics","guardant-health","natera"],"institutions":["francis-crick","cruk","moffitt"],"pathways":[],"terms":["resistance","ctdna","vaf","mutational-signature","oligoprogression","esr1-mutation"],"trials":["serena-6","circulate-japan"],"people":[],"bottlenecks":[],"keyPapers":["paper-mcgranahan-cell"],"journals":[],"dependsOn":[],"notes":[],"stage":"biology","severity":"critical","metrics":[{"label":"Somatic mutations not detectable across every region of a renal tumour (multi-region sequencing)","value":"63-69%","source":"Gerlinger et al., NEJM 2012","url":"https://doi.org/10.1056/NEJMoa1113205"},{"label":"Association of elevated copy-number intratumour heterogeneity with recurrence or death in resected NSCLC (TRACERx)","value":"Hazard ratio 4.9","source":"Jamal-Hanjani et al., NEJM 2017","url":"https://doi.org/10.1056/NEJMoa1616288"},{"label":"Patients and tumour regions in the TRACERx evolutionary cohort","value":"421 patients, 1,644 regions","source":"Frankell et al., Nature 2023","url":"https://doi.org/10.1038/s41586-023-05783-5"}],"causes":["Genomic instability generates diversity continuously, so any large tumour contains rare clones that are already resistant to the next drug.","A single diagnostic biopsy is treated as representative of all lesions for the whole course of disease.","Metastases are rarely biopsied, so the clones that kill patients are the least characterised.","Trials are designed around a fixed genotype and rarely re-sample or adapt treatment as clones shift.","Sequencing costs, tissue access and reimbursement rules discourage repeat and multi-region profiling."],"currentEfforts":["TRACERx (Cancer Research UK) follows hundreds of lung cancer patients with multi-region and longitudinal sequencing to map how clones evolve under treatment.","SERENA-6 showed that switching endocrine therapy on detection of an emerging ESR1 clone in ctDNA, before radiological progression, extends progression-free survival.","Single-cell and spatial transcriptomics platforms (10x Genomics and others) resolve clonal architecture within a lesion.","AACR Project GENIE and cBioPortal aggregate sequential sequencing from thousands of patients to expose recurrent evolutionary routes.","Adaptive therapy trials at Moffitt Cancer Center test dosing designed to keep sensitive clones alive so they suppress resistant ones."],"successLooksLike":"Every patient with advanced disease has serial molecular profiling (tissue or blood) that changes management, and trials routinely randomise on emerging clones rather than on the diagnostic biopsy. Resistance is anticipated and pre-empted rather than discovered on a scan."},{"id":"b-real-world-evidence","kind":"bottleneck","name":"Weak real-world evidence and registries","aka":[],"tldr":"We do not reliably know what happens to patients after approval, so we cannot tell which drugs deliver in practice.","summary":"Regulatory approval, increasingly on surrogate endpoints and single-arm trials, is meant to be followed by confirmation in practice, but the systems to do that are weak. Of cancer drugs given accelerated approval in the US over a quarter-century, only a minority later showed an overall survival benefit in a confirmatory trial, and many confirmatory studies used the same surrogate as the original approval. Post-marketing commitments are often delayed or unmet. Population cancer registries record incidence, stage and death but rarely treatment or recurrence, and treatment datasets such as the NHS SACT are the exception rather than the rule. Real-world effectiveness in older, sicker, more diverse patients is frequently lower than trial efficacy, and without linked outcome data neither payers nor clinicians can tell which drugs deliver. Regulatory frameworks for real-world evidence, national treatment registries and clinico-genomic databases are the building blocks; completeness and outcome capture remain the gap.","asOf":"2026-09-08","links":[{"label":"Gyawali, Hey & Kesselheim, Assessment of the clinical benefit of cancer drugs receiving accelerated approval (JAMA Intern Med 2019)","url":"https://doi.org/10.1001/jamainternmed.2019.0462"},{"label":"FDA Real-World Evidence","url":"https://www.fda.gov/science-research/science-and-research-special-topics/real-world-evidence"},{"label":"FDA Oncology Center of Excellence, Project Confirm","url":"https://www.fda.gov/about-fda/oncology-center-excellence/project-confirm"},{"label":"NHS Systemic Anti-Cancer Therapy dataset","url":"https://digital.nhs.uk/ndrs/data/data-sets/sact"}],"tags":[],"related":["fda-approvals","seer","globocan","esmo-guidelines","clinicaltrials-gov","idea-data-rwe-rct-calibration-library","idea-data-accelerated-approval-registry-condition","idea-data-rwpfs-validation-programme","idea-data-registry-based-rcts","idea-data-off-label-outcomes-registry","idea-data-real-world-dose-intensity","idea-data-excluded-populations-rwe-mandate","idea-data-linked-pharmacovigilance-interactions","idea-data-access-programme-outcomes","idea-data-registry-genomics-linkage-programme","idea-data-observational-study-registration","idea-data-gcp-for-real-world-data","idea-data-registry-follow-up-of-trial-participants","idea-data-wearable-endpoints-validation","idea-data-external-control-arm-standard","idea-data-paediatric-rwe-consortium","idea-data-sequencing-analysis-per-approval","idea-data-outcome-based-price-reassessment","idea-moon-federated-rwe-network"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":[],"companies":["foundation-medicine","tempus","guardant-health"],"institutions":["asco","esmo","nci"],"pathways":[],"terms":["real-world-evidence","accelerated-approval","os","pfs"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":["paper-gyawali-jama-intern-med"],"journals":[],"dependsOn":[],"notes":[],"stage":"data-knowledge","severity":"major","metrics":[{"label":"Cancer drug indications granted FDA accelerated approval 1992-2017 that later demonstrated an overall survival benefit in a confirmatory trial","value":"19 of 93 (20%)","source":"Gyawali, Hey & Kesselheim, JAMA Internal Medicine 2019","url":"https://doi.org/10.1001/jamainternmed.2019.0462"},{"label":"Cancer drug indications approved by the EMA 2009-2013 still lacking evidence of survival or quality-of-life benefit after a median 5.4 years","value":"About half","source":"Davis et al., BMJ 2017","url":"https://doi.org/10.1136/bmj.j4530"}],"causes":["Population registries were designed for incidence and mortality surveillance, not treatment and outcomes.","Sponsors have little incentive to complete confirmatory trials quickly once a drug is on the market.","Real-world data lack structured capture of progression, response and toxicity.","Confounding by indication makes non-randomised comparisons unreliable without careful design.","Regulators have historically been reluctant to withdraw approvals on the basis of missing or negative confirmatory data."],"currentEfforts":["The FDA Real-World Evidence Framework (2018, under the 21st Century Cures Act) and Oncology Center of Excellence Project Confirm track and publish the status of accelerated approvals and confirmatory trials.","The Flatiron Health-Foundation Medicine clinico-genomic database links sequencing to longitudinal outcomes for tens of thousands of patients and has supported FDA decisions.","The EMA's DARWIN EU network runs real-world studies across European data sources for regulatory questions.","The NHS Systemic Anti-Cancer Therapy (SACT) dataset records every systemic therapy episode in England with outcomes, and the Cancer Drugs Fund uses it for managed-access re-evaluation.","SEER-Medicare linkage and the NCI SEER programme provide population-level treatment and survival for US patients over 65.","ASCO's TAPUR and registry trials generate prospective evidence for off-label targeted therapy."],"successLooksLike":"Every approved cancer drug has post-approval effectiveness and safety data published within three years from national treatment registries, accelerated approvals are confirmed or withdrawn on schedule, and real-world survival is tracked and reported alongside trial results in guidelines."},{"id":"b-dose-optimisation","kind":"bottleneck","name":"Wrong doses","aka":[],"tldr":"Most drug doses were chosen as the highest a person can tolerate, which is often more than they need.","summary":"Oncology inherited from cytotoxic chemotherapy the assumption that more drug is better, so phase 1 trials escalate to the maximum tolerated dose over a few weeks in a few dozen patients and that dose becomes the label. Targeted agents, antibodies, ADCs and immunotherapies saturate their targets far below toxicity, and chronic dosing exposes patients to months of grade 2 toxicity that the 28-day dose-limiting-toxicity window never measured. The consequences are dose reductions and discontinuations in a third or more of patients on many oral targeted agents, avoidable cost, and occasionally drugs that fail because nobody can stay on them. The FDA's Project Optimus (2021) and its 2024 guidance now require randomised comparison of at least two doses before registration, and trials such as PERSEPHONE show that duration can be halved without loss of efficacy. Dosimetry-based individualisation is the parallel challenge in radioligand therapy.","asOf":"2026-09-08","links":[{"label":"FDA Oncology Center of Excellence, Project Optimus","url":"https://www.fda.gov/about-fda/oncology-center-excellence/project-optimus"},{"label":"Shah, Rahman, Theoret & Pazdur, The drug-dosing conundrum in oncology: when less is more (NEJM 2021)","url":"https://doi.org/10.1056/NEJMp2109826"},{"label":"Earl et al., 6 versus 12 months of adjuvant trastuzumab (PERSEPHONE, Lancet 2019)","url":"https://doi.org/10.1016/S0140-6736(19)30650-6"}],"tags":[],"related":["fda-approvals","idea-shortened-venetoclax","idea-desmoid-intermittent-dosing","idea-net-dosimetry-prrt","idea-total-body-pet-dosimetry","idea-tr1-public-dose-reduction-trials-of-approved-drugs","idea-tr1-weight-based-vs-flat-dosing-trials","idea-tr1-extended-interval-checkpoint-dosing","idea-tr1-therapeutic-drug-monitoring-for-oral-tkis","idea-tr1-intermittent-dosing-to-delay-resistance","idea-tr1-adaptive-therapy-randomised-phase-2","idea-tr1-start-low-titrate-up-oral-agents","idea-tr1-pre-emptive-pharmacogenomic-testing","idea-tr1-model-based-dose-finding-with-late-toxicity","idea-tr1-adc-dose-and-schedule-optimisation","idea-tr1-dose-evidence-statement-in-label","idea-tr1-patient-reported-toxicity-dose-algorithms","idea-tr1-food-effect-dose-reduction","idea-tr1-factorial-dose-finding-for-combinations","idea-tr1-organ-impairment-pk-before-approval","idea-tr1-low-dose-immunotherapy-for-lmic","idea-tr1-response-adapted-dose-reduction","idea-moon-individualised-dosing-all-oral-drugs","idea-cost-low-dose-abiraterone-food","idea-cost-low-dose-immunotherapy-trials","idea-cost-optimus-legacy-drugs","idea-cost-extended-interval-default","idea-cost-shorter-course-trials"],"cancers":["breast-her2-positive","breast-hr-positive","nsclc","aml","neuroendocrine","sarcoma"],"sections":[],"technologies":["radioligand-therapy","prrt","adc","kinase-inhibitors","cdk46-inhibitor"],"targets":[],"drugs":["sotorasib","abemaciclib","nirogacestat","venetoclax","trastuzumab"],"companies":["finaldose","amgen"],"institutions":[],"pathways":[],"terms":["dosimetry","ild","alpha-vs-beta"],"trials":["monarch-3","persephone","apt-trial"],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"stage":"trials","severity":"major","metrics":[{"label":"Patients requiring dose reduction of abemaciclib for adverse events in first-line MONARCH 3","value":"~43%","source":"Goetz et al., JCO 2017","url":"https://doi.org/10.1200/JCO.2017.75.6155"},{"label":"Six vs twelve months of adjuvant trastuzumab in HER2-positive breast cancer (PERSEPHONE): 4-year disease-free survival","value":"89.4% vs 89.8% (non-inferior)","source":"Earl et al., Lancet 2019","url":"https://doi.org/10.1016/S0140-6736(19)30650-6"}],"causes":["Phase 1 designs optimise for speed and identify the maximum tolerated dose rather than the optimal biological dose.","Dose-limiting toxicity is assessed in one cycle, missing cumulative and chronic toxicities.","Sponsors have no commercial incentive to study lower doses or shorter durations after approval.","Regulators historically accepted a single dose in registration trials.","Pharmacokinetic variability between patients is large but dosing is rarely individualised."],"currentEfforts":["FDA Project Optimus and the 2024 final guidance on dose optimisation require sponsors to compare multiple doses before registration.","PERSEPHONE, SOLD and Short-HER established that shorter adjuvant trastuzumab is non-inferior for many patients.","The CodeBreaK 100 randomised 960 mg vs 240 mg sotorasib comparison, required by the FDA, is the model for post-approval dose re-evaluation.","Dosimetry-personalised PRRT and radioligand therapy trials replace fixed activity with absorbed-dose-based dosing.","Intermittent and stop-and-restart dosing is being tested for nirogacestat in desmoid tumours and for venetoclax in AML.","Finaldose and academic pharmacometrics groups apply model-informed precision dosing to oncology drugs."],"successLooksLike":"Every new oncology drug is registered with a dose chosen by randomised comparison, the rate of dose reductions and discontinuations for toxicity on oral targeted agents falls below 20%, and duration of therapy is tested rather than assumed."}]