Loading
Triple-negative breast cancer was named for what it lacks, the three receptors other breast cancers are treated through. This roadmap follows it from the receptor discoveries and the basal-like signature of 2000, through chemotherapy, platinum, PARP inhibitors, immunotherapy and antibody-drug conjugates, to the trials asking who can have less and who needs more, with registry dates to 2030.
For most of its history triple-negative breast cancer was a remainder. Oestrogen receptor testing, then HER2 (Slamon 1987), sorted breast cancers into those with a target; the tumours negative for all three were left with chemotherapy alone. Gene expression profiling gave the remainder a biology (Perou 2000; Sørlie 2001, 2003), linked it to germline BRCA1 (Sørlie 2003; Foulkes 2003; Atchley 2008) and to the founder mutations that concentrate hereditary disease in particular populations (Struewing 1997; Górski 2000), and registry studies of 2006 to 2007 gave it a name, a natural history (relapse peaking at three years, then subsiding) and a demography: younger women, Black and Hispanic women, poorer women, worse survival at every stage. In the United Kingdom the POSH cohort found the same excess in young Black women within a health service with equal access.
Chemotherapy was the whole of treatment until 2017, and it worked: anthracycline and taxane regimens cut breast cancer deaths by a third (EBCTCG 2012), tumours disappeared before surgery twice as often as in other subtypes (Liedtke 2008), and those that disappeared were largely cured (Cortazar 2014). Residual disease was the problem, graded from 2007 by the residual cancer burden score. Platinum raised the response rate (GeparSixto and CALGB 40603, 2014 to 2015) and later relapse-free survival (BrighTNess, GeparSixto follow-up), and capecitabine after residual disease became the first post-neoadjuvant treatment (CREATE-X 2017). Molecular subtyping (Lehmann 2011, 2016; Burstein 2015) showed the disease was several diseases, but pathway-targeted small molecules failed, most recently the AKT inhibitor capivasertib (CAPItello-290, 2026).
Three classes then arrived in six years. PARP inhibitors for germline BRCA carriers (OlympiAD, EMBRACA 2017 to 2018; OlympiA 2021, with a survival benefit sustained at six years). Immunotherapy: atezolizumab first (IMpassion130, 2018, later withdrawn after IMpassion131 and an assay dispute), then pembrolizumab first line for PD-L1 combined positive score of 10 or more (KEYNOTE-355) and before and after surgery for stage II to III disease (KEYNOTE-522, with a 4.9-point overall survival gain at five years, 86.6 against 81.7 percent, reported in 2024). Antibody-drug conjugates: sacituzumab govitecan after two lines (ASCENT 2021), trastuzumab deruxtecan for the third of triple-negative tumours that are HER2-low (DESTINY-Breast04, 2022), then first line for all comers (ASCENT-03, ASCENT-04, TROPION-Breast02, 2025 to 2026), with median survival approaching two years against 13 months in 2008.
The open questions are now about selection and quantity. Who can have less: pembrolizumab omission after pathological complete response (OptimICE-pCR), anthracycline omission (SCARLET), chemotherapy omission in lymphocyte-rich stage I tumours. Who needs more: residual disease trials with antibody-drug conjugates (ASCENT-05, TROPION-Breast03) and the ctDNA-guided designs that c-TRAK TN showed must test earlier and more sensitively. And what the trials have not settled: PD-L1 assay concordance, the order of two topoisomerase-payload antibody-drug conjugates, brain metastases in half of metastatic patients, HER2-low scoring reproducibility, and a disparity in incidence and outcome that trial enrolment has not reflected. UK and NHS specifics are on the UK and NHS page for triple-negative breast cancer.
Oestrogen receptor assays sorted breast cancers into those that would respond to endocrine therapy and those that would not; in 1987 Slamon and colleagues found HER-2/neu amplified in 30 percent of 189 tumours and predictive of early relapse, adding a third test. The tumours negative for all three were a remainder with chemotherapy as their only treatment. The thresholds that define the remainder are conventions: the 2010 ASCO and CAP guideline fixed oestrogen and progesterone receptor positivity at 1 percent of nuclei after finding up to 20 percent of tests worldwide might be wrong, and the 1 to 10 percent low-positive band it created still behaves like triple-negative disease in many series.
Every HER2 test, every trastuzumab prescription and the whole HER2-positive breast cancer category trace back to this observation. It is the model for how a genomic marker of bad prognosis became a drug target and then the basis for one of the largest survival gains in solid tumour oncology.
Triple-negative is a laboratory definition; this guideline wrote the oestrogen and progesterone half of it, and the 1 to 10 percent low-positive band it created is still argued over.
Perou and colleagues read 8,102 genes in 65 tumours in 2000 and found breast cancer fell into groups, one of them basal epithelial-like; Sørlie showed in 2001 that the basal-like group had the worst outcome and in 2003 that the subtypes reproduced in other laboratories' data and that tumours from BRCA1 carriers fell into the basal group. Foulkes confirmed the BRCA1 link the same year with a cytokeratin 5/6 stain (odds ratio 9.0). The remainder now had a biology and a hereditary cause; basal-like and triple-negative overlap but are not the same set.
Struewing's 1997 study of 5,318 Ashkenazi Jewish volunteers put the breast cancer risk of the three founder mutations carried by over 2 percent of that population at 56 percent by age 70; Górski found in 2000 that three BRCA1 changes accounted for 82 percent of the mutations in 66 Polish families, one of them (5382insC) shared with the Ashkenazi set. Atchley's 2008 MD Anderson series showed 57 percent of BRCA1 carriers' tumours were triple-negative against 14 percent in non-carriers, which made a triple-negative diagnosis itself a reason to test. Founder panels make population testing cheap where they exist; where they do not, full sequencing is needed and uptake lags.
Dent's Toronto cohort (2007) found triple-negative disease in 11.2 percent of 1,601 patients with a distant relapse hazard ratio of 2.6 that peaked at three years and faded after five; Bauer's California registry study (2007) of 6,370 cases fixed its demography as younger, Black, Hispanic and poorer women with worse survival at every stage. Carey's Carolina Breast Cancer Study (2006) had found the basal-like subtype in 39 percent of premenopausal African American women against 16 percent of others, and Lin (2008) showed 46 percent of metastatic patients developed brain metastases with a median survival of 13.3 months. National counts followed: 12.2 percent of US cases in 2010 (Howlader 2014), odds ratio 2.27 for Black women in 1.15 million cases (Scott 2019). In the UK the POSH cohort of women under 41 found 26.1 percent triple-negative disease in Black women against 18.6 percent in White women and five-year survival of 71.1 versus 82.4 percent despite equal chemotherapy use.
The Oxford overview of 123 trials (EBCTCG 2012) showed taxane-plus-anthracycline regimens cut breast cancer deaths by about a third regardless of receptor status, so chemotherapy's absolute benefit was largest in the high-risk remainder. Liedtke's MD Anderson series (2008) found triple-negative tumours disappeared completely before surgery twice as often as others (22 versus 11 percent) yet survival was worse, because women with residual disease relapsed early. Symmans graded residual disease with the residual cancer burden score in 2007; the CTNeoBC pooled analysis (2014) found pathological complete response predicted survival most strongly in triple-negative disease (event-free survival hazard ratio 0.24) but could not validate it as a trial-level surrogate. The US regulator nonetheless built an accelerated approval pathway on it.
Lehmann's 2011 analysis of 587 tumours defined six subtypes (basal-like 1 and 2, immunomodulatory, mesenchymal, mesenchymal stem-like, luminal androgen receptor), each with candidate drugs from cell line work; the 2016 refinement showed two came from immune and stromal cells, cut the scheme to four, and found pathological complete response ranged from 41 percent in basal-like 1 to 18 percent in basal-like 2. Burstein's Baylor study (2015) reached a similar four-way split and showed immune activation within basal-like tumours separated longer from shorter survival. The subtypes shaped trial design (androgen receptor antagonists for the luminal androgen receptor group, platinum for basal-like 1) but no subtype-directed small molecule has succeeded in phase 3.
GeparSixto (2014) and CALGB 40603 (2015) showed carboplatin raised pathological complete response, in CALGB 40603 from 41 to 54 percent in breast and axilla. Survival followed: GeparSixto's final analysis (2018) gave a disease-free survival hazard ratio of 0.56 in triple-negative disease and found 70 percent of tumours homologous recombination deficient whether or not BRCA was mutated, with the deficiency predicting response but not carboplatin benefit; BrighTNess at 4.5 years (2022) gave an event-free survival hazard ratio of 0.57 for carboplatin over paclitaxel alone and nothing for added veliparib. The TNT trial showed carboplatin's advantage in metastatic disease was confined to germline BRCA carriers. St Gallen 2025 made platinum a consensus recommendation for early triple-negative disease.
CREATE-X randomised 910 Japanese and Korean women with HER2-negative residual disease after neoadjuvant chemotherapy to six months of capecitabine or nothing: five-year overall survival 89.2 versus 83.6 percent (hazard ratio 0.59), and in the triple-negative group 78.8 versus 70.3 percent (0.52), with hand-foot syndrome in 73 percent. It was the first proof that the post-neoadjuvant window could be used, the design OlympiA, ASCENT-05 and TROPION-Breast03 inherited, and capecitabine remains the option for residual disease without a BRCA variant in ESMO, NCCN and UK practice. Whether it adds to adjuvant pembrolizumab has never been tested.
OlympiAD (2017) and EMBRACA (2018) showed olaparib and talazoparib beat chemotherapy for progression-free survival in metastatic germline BRCA-mutated breast cancer, without a clear survival gain. OlympiA (2021) gave a year of adjuvant olaparib to 1,836 high-risk carriers, 82 percent with triple-negative disease, and improved invasive disease-free survival; the 2022 analysis showed an overall survival hazard ratio of 0.68 and the 2026 six-year update 0.72 (six-year survival 87.5 versus 83.2 percent) with no excess of leukaemia and fewer new BRCA-related cancers. Neoadjuvant PARP inhibition in unselected disease failed (BrighTNess veliparib arm). The 2020 ASCO, ASTRO and SSO hereditary guideline sets the surgical and systemic rules for carriers; whether olaparib adds to pembrolizumab or capecitabine in the same patient is untested.
IMpassion130 (2018) showed atezolizumab with nab-paclitaxel lengthened progression-free survival in PD-L1-positive metastatic disease by the SP142 assay; IMpassion131 with paclitaxel was negative, the US indication was withdrawn in 2021, and Rugo's assay comparison showed SP142, SP263 and 22C3 called 46, 75 and 73 percent of the same tumours positive with 69 percent concordance. KEYNOTE-355 (2020, survival 2022) established pembrolizumab with chemotherapy for 22C3 combined positive score of 10 or more. KEYNOTE-522 (2020) added pembrolizumab before and after surgery for stage II to III disease: pathological complete response 64.8 versus 51.2 percent, event-free survival gain in 2022, and in 2024 a 4.9-point five-year overall survival gain (86.6 against 81.7 percent), the first survival benefit of immunotherapy in early breast cancer. ASCO reversed its 2021 guideline within 15 months; ESMO, NCCN and St Gallen followed. Leon-Ferre (2024) showed lymphocyte-rich stage I tumours do well without any chemotherapy, and OptimICE-pCR now asks whether adjuvant pembrolizumab can be dropped after complete response.
ASCENT (2021) doubled survival with sacituzumab govitecan after two or more lines (12.1 versus 6.7 months); DESTINY-Breast04 (2022) showed trastuzumab deruxtecan worked in HER2-low tumours, which are 36.6 percent of triple-negative disease though biologically indistinguishable from HER2-zero (Schettini 2021). ASCENT-03 and ASCENT-04 (2025 to 2026) and TROPION-Breast02 (2026: progression-free survival 10.8 versus 5.6 months, overall survival 23.7 versus 18.7) moved TROP2 antibody-drug conjugates to first line, alone or with pembrolizumab, and the bispecific EGFR and HER3 conjugate izalontamab brengitecan posted a positive phase 3 in pretreated disease. The same year the AKT inhibitor capivasertib failed to improve survival first line (CAPItello-290), the latest pathway-targeted small molecule to do so. Median survival in first-line trials now approaches two years against 13.3 months in the 2008 Dana-Farber series.
Radovich (2020) showed ctDNA after neoadjuvant chemotherapy in 142 residual-disease patients carried a distant relapse hazard ratio of 2.99 and a death hazard ratio of 4.16. The UK c-TRAK TN trial (2023) then tested acting on it: 27 percent of 161 women turned ctDNA-positive within a year, but 72 percent of those already had metastases on staging and none of five given pembrolizumab cleared their DNA, so later designs test earlier, with tumour-informed assays, and use drugs with more single-agent activity. Meanwhile the residual cancer burden score, validated across 5,161 patients (Yau 2022), became the entry criterion for antibody-drug conjugate trials after neoadjuvant therapy: ASCENT-05 (sacituzumab govitecan with pembrolizumab, 1,514 patients) and TROPION-Breast03 (datopotamab deruxtecan with or without durvalumab, 1,174 patients).
The de-escalation trials are large and slow: OptimICE-pCR (pembrolizumab versus observation after pathological complete response, 1,295 estimated, primary completion May 2033) and SCARLET (anthracycline-free chemo-immunotherapy, 2,400 estimated, March 2033). The escalation trials report sooner: ASCENT-05 (June 2027) and TROPION-Breast03 (September 2027) for residual disease. First-line combinations follow: TROPION-Breast05 (datopotamab deruxtecan with durvalumab against chemotherapy with pembrolizumab in PD-L1-positive disease, 625 estimated, July 2027), IZABRIGHT-Breast01 (izalontamab brengitecan first line in immunotherapy-ineligible disease, 600, March 2028) and the PD-L1 and VEGF bispecific PM8002 with nab-paclitaxel (392, July 2027). KEYNOTE-522 completed on the registry in October 2025; OlympiA's study completion is listed for May 2029.
Four things the trials have not settled. PD-L1 assays disagree on a quarter of patients and only one, 22C3 combined positive score of 10, has an approved drug attached. Two TROP2 antibody-drug conjugates and trastuzumab deruxtecan share a topoisomerase I payload and no randomised trial has asked which to give first or whether the second works after the first. Brain metastases occur in about half of metastatic patients (Lin 2008) and most trials exclude active brain disease. And the disease is twice as common in Black women in the United States, with worse survival in the UK POSH cohort despite equal access, yet trial enrolment does not reflect it. Each has an idea on this page; the UK-specific gaps are set out on the UK and NHS page for triple-negative breast cancer.
The clearest demonstration that PD-L1 positive in triple-negative breast cancer means different things depending on the kit; with atezolizumab withdrawn, pembrolizumab's 22C3 combined positive score of 10 is the surviving standard, and roughly a quarter of patients get a different answer depending on which assay their laboratory runs.
The 13.3-month median is the pre-immunotherapy, pre-antibody-drug conjugate benchmark against which first-line trials now reporting medians near two years are measured; the brain metastasis rate is why trials that exclude active brain disease leave the question unanswered.
The most recent national figure behind the statement that Black women in the United States have about twice the incidence of triple-negative breast cancer; trial enrolment has not matched it.
The UK's own evidence that the disparity is not only American and not only about access: within the NHS, with equal chemotherapy use, young Black women had more triple-negative disease and worse survival. It anchors the disparities idea on the triple-negative page and the UK and NHS page.
Every era's records, trial outcomes and papers, and every watch item, as JSON.
Readouts, decisions and registry completion dates ahead. Each date is quoted from its source, not inferred; a missing date means no source states one.
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Loading this step…
Answers the two questions that hung over adjuvant olaparib, durability and late leukaemia, in its favour; the remaining question is whether carriers who also received pembrolizumab or capecitabine, whom the trial did not study, get the same benefit.
The first-line overall survival result that made an antibody-drug conjugate the standard for PD-L1-negative or immunotherapy-ineligible metastatic triple-negative disease, and the reason the sequencing question (which TROP2 drug first, what after it) is now urgent.
The latest in a line of negative targeted-therapy trials in triple-negative disease (EGFR, VEGF, iniparib, now AKT): a modest delay in progression that did not translate into survival, in the same year that an antibody-drug conjugate did. It is why the roadmap treats pathway-targeted small molecules as the road not taken.
NCCN is the source of the category grades on the triple-negative breast cancer page and the first major guideline to place an antibody-drug conjugate first line for the disease.
Platinum and immunotherapy are now consensus for early triple-negative disease; the open votes have moved to who can safely receive less.
A second publication from the KEYNOTE-522 trial, later than the first and described in its title as an update or longer-term analysis. Read it with the primary publication linked from the trial page; the record was linked automatically and its figures have not been checked by hand.
The evidence behind the stage I de-escalation statement on the triple-negative page: an ordinary haematoxylin and eosin slide identifies a fifth of patients whose outcome without chemotherapy matches treated cohorts. A prospective trial of chemotherapy omission is the missing step.
This is the European standard the UK page for triple-negative breast cancer is compared against; NICE guidance covers the same ground with a narrower set of funded drugs.
Shares ASCENT, DESTINY-Breast04, HER2-low and HER2-ultralow metastatic breast cancer, HER2-low and HER2-ultralow and the tags tnbc, breast.
Shares IMpassion130, KEYNOTE-355, Combined positive score (CPS), Tumour-infiltrating lymphocytes (TILs) and the tags tnbc, breast.
Shares GeparSixto, Residual cancer burden (RCB), Homologous recombination deficiency (HRD), Pathologic complete response (pCR) and the tags tnbc, breast.
Shares IMpassion130, KEYNOTE-355, PD-L1 Immunohistochemistry Assay Comparison in Atezolizumab Plus nab-Paclitaxel-Treated Advanced Triple-Negative Breast Cancer, Combined positive score (CPS) and the tags tnbc, breast.
Shares OlympiA, Residual cancer burden (RCB), KEYNOTE-522, Pathologic complete response (pCR) and the tags tnbc, breast.
Shares Residual cancer burden (RCB), Pathologic complete response (pCR), HER2-positive breast cancer, Triple-negative breast cancer (TNBC) and the tags tnbc, breast.
Shares HER2-low and HER2-ultralow, Pathologic complete response (pCR), Triple-negative breast cancer (TNBC), HR-positive / HER2-negative breast cancer and the tags tnbc, breast.
Shares Pathologic complete response (pCR), Androgen receptor, Triple-negative breast cancer (TNBC) and the tags tnbc, breast.