Triple-negative breast cancer (TNBC)

What remains unsolved: residual disease after KEYNOTE-522 (RCB II-III) still carries ~40-50% relapse risk; ADC sequencing (TOP1-payload cross-resistance) is unstudied; PD-L1-negative early disease has immunotherapy proven only in KEYNOTE-522's unselected population; brain metastases; and the biology of the mesenchymal/claudin-low subtype resists every class. The most promising directions are ADC + IO first line, de-escalation guided by TILs and pCR, ctDNA-guided escalation, TROP2 PET selection, next-generation ADCs with non-TOP1 or dual payloads, and personalised vaccines in the adjuvant setting.

How common it is, world and United States. No registry publishes a world count of triple-negative disease; the share has to be applied to breast cancer as a whole. The IARC breast fact sheet, served today with the GLOBOCAN 2024 estimates, gives 2,434,087 new breast cancers a year (the second commonest cancer, age-standardised rate 47.8 per 100,000) and 693,660 deaths (fourth, 12.9 per 100,000), with 45.1 percent of cases in Asia, 22.3 percent in Europe, 13.2 percent in Northern America, 9.3 percent in Africa and 9.0 percent in Latin America and the Caribbean. The triple-negative share in registries with receptor data runs from 8.4 percent of 1,151,724 US cases in 2010 to 2014 (Scott 2019) through 10.8 percent of US cases with a known subtype in SEER 21, 2019 to 2023 (the SEER subtype fact sheet), 12.2 percent of cases with known status in 2010 (Howlader 2014) and 12 percent in 2012 to 2016 (Howard and Olopade 2021), to around 15 percent on the Cancer Research UK patient page and 17 percent of 15,204 women treated at NCCN centres in 2000 to 2006 (Lin 2012). Applied to 2.4 million, a share of 10 to 15 percent would be about 240,000 to 370,000 people a year; that is arithmetic on the figures above, not a published count. In the United States the age-adjusted rate of hormone receptor-negative, HER2-negative breast cancer is 14.3 per 100,000 women a year against 132.5 for all breast cancer (SEER 21, 2019 to 2023), and 321,910 new female breast cancers of all subtypes are projected for 2026.

Who gets it. Age: triple-negative tumours are diagnosed younger than hormone receptor-positive tumours. Women under 40 had 1.53 times the odds of the triple-negative phenotype in the California registry (Bauer 2007) and 1.95 times the odds relative to women aged 50 to 64 in US Cancer Statistics (Scott 2019); Howlader found triple-negative patients 10 to 30 percent less likely to be diagnosed at older ages and 6.4 to 20 times more likely to present with high-grade disease; Cancer Research UK says the type develops more often in women under 40. Ethnicity: in SEER 21 (2019 to 2023) the rate of hormone receptor-negative, HER2-negative disease is 25.7 per 100,000 in non-Hispanic Black women against 13.0 in non-Hispanic White, 12.5 in Hispanic, 11.4 in American Indian and Alaska Native and 10.5 in Asian and Pacific Islander women, roughly a two-fold difference; non-Hispanic Black women had 2.27 times the odds of a triple-negative diagnosis in 2010 to 2014 (Scott 2019); the basal-like subtype was found in 39 percent of premenopausal African American women against 16 percent of non-African American women in the Carolina Breast Cancer Study (Carey 2006); and in 507 women in Nigeria and Senegal, triple-negative disease (basal-like 27 percent plus unclassified 28 percent) predominated, with a mean age of 44.8 (Huo 2009). In the UK, the POSH cohort of women diagnosed at 40 or under found triple-negative tumours in 26.1 percent of Black against 18.6 percent of White patients (Copson 2014), and an analysis of 116,500 women in England (2013 to 2018) found greater adjusted odds of ER-negative, high-grade and high-stage tumours in ethnic minority women, most marked in Black African and Black Caribbean women (Gathani 2021). Even after adjustment for stage, treatment delay and socioeconomic factors, African American women with triple-negative disease remain nearly twice as likely to die of it (Howard and Olopade 2021). Inherited risk: 57.1 percent of breast cancers in BRCA1 carriers were triple-negative against 23.3 percent in BRCA2 carriers and 13.8 percent in non-carriers (Atchley 2008); in 6,893 carriers the triple-negative share fell with age at diagnosis in BRCA1 and rose in BRCA2 carriers, and medullary tumours were BRCA1-related (Mavaddat 2012). Looked at from the other side, 14.6 percent of 1,824 triple-negative patients unselected for family history carried a deleterious mutation in one of 17 genes, 8.5 percent BRCA1, 2.7 percent BRCA2 and 1.2 percent PALB2 (Couch 2015); a prospective US registry found 15.4 percent with BRCA1 or BRCA2 mutations, 27.6 percent among those diagnosed at 50 or under, 11.4 percent at 51 to 60 and 4.9 percent at 61 or over (Sharma 2014); and 12 percent of 2,733 UK women with breast cancer at 40 or under carried a BRCA mutation (Copson 2018). Cancer Research UK puts the lifetime breast cancer risk of a BRCA1 or BRCA2 carrier at 45 to 65 percent by age 70, with about 1 in 450 women carrying a mutation; PALB2 carries a relative risk of 7.18 and a 53 percent risk to age 80 (Yang 2020).

United Kingdom (Cancer Research UK, drawing on the four national registries). Breast cancer is the commonest cancer, with 59,413 new cases a year (2019 and 2021 to 2022), about 59,000 in females and 420 in males, 15 percent of all cancers; a quarter are diagnosed at 75 or over; incidence rates have risen 20 percent since the early 1990s and are projected to fall 1 percent by 2038 to 2040 while the count rises to about 70,300 a year; incidence rates are lower in the Asian and Black ethnic groups than the White group in England (2013 to 2017), which sits beside the higher triple-negative share in Black women noted above. There are about 11,200 deaths a year (2022 to 2024), the fourth commonest cause of cancer death and the second in women; mortality rates have fallen 46 percent since the early 1970s. In England in 2022, 85 percent of cases with a known stage were stage I or II; in 2019, 53 percent were diagnosed through an urgent suspected cancer referral, 32 percent through screening and 4 percent after an emergency presentation. About 23 percent of UK breast cancers are judged preventable, 8 percent attributed to alcohol and 1 percent to ionising radiation. Cancer Research UK does not publish a separate UK incidence count for triple-negative disease; the around 15 percent share on its patient page, and the 20 percent share in the young-onset POSH cohort (558 of 2,733), are the UK figures found for this record.

Risk factors that separate triple-negative from other breast cancers. Reproduction runs the other way from hormone receptor-positive disease: in 23,353 cases and 71,072 controls pooled across 31 studies, parous women had a lower risk of every luminal and HER2-enriched-like subtype but a higher risk of triple-negative disease immediately after a birth (odds ratio 3.12 for multiparous women right after their last birth), attenuating over decades (Jung 2022); in African American women parity raised triple-negative risk (odds ratio 1.37) while breastfeeding lowered ER-negative risk (0.81), and ER-negative risk rose with each birth in women who had not breastfed (1.68 for four or more births) (Palmer 2014). Cancer Research UK's summary that triple-negative disease is not associated with parity, from an older meta-analysis, is recorded beside these newer pooled analyses. Obesity: triple-negative cases were more often obese than hormone receptor-positive cases (odds ratio 1.20 case-case, 1.24 case-control), a signal confined to premenopausal women (1.43) in an 11-study meta-analysis (Pierobon 2013), and 1.89 in an Atlanta case series (Trivers 2009); Cancer Research UK notes the postmenopausal body-mass effect is largely limited to ER- and PR-positive tumours (adult weight gain raised ER- and PR-negative risk by 34 percent against 133 percent for positive tumours). Age at first birth and at menarche, and oral contraceptive or HRT use, are associated mainly with hormone receptor-positive tumours (CRUK risk page). A first-degree relative with breast cancer doubles breast cancer risk of any type, and the germline factors above (BRCA1 above all, then PALB2 and BRCA2) are the strongest known triple-negative risk factors. What is shared with all breast cancer: alcohol (8 percent of UK cases), ionising radiation (1 percent) and breast density (3 to 5 times the risk in the most dense breasts).

How it presents. Symptoms are those of any breast cancer: a lump or thickening in the breast or armpit, a change in size, shape or feel, skin puckering, dimpling, a rash or redness, fluid from the nipple, or a nipple that turns in (Cancer Research UK triple-negative and symptoms pages). What differs is the route and the tempo. Triple-negative tumours are less often found by a screening mammogram (29 percent against 48 percent of hormone receptor-positive, HER2-negative tumours in NCCN centres) and present with a higher tumour category but are less likely to involve the lymph nodes (Lin 2012); they are over-represented among interval cancers, which appear between screens and carry worse prognostic features than screen-detected cancers (Houssami 2017); and they are more often high grade (Howlader 2014). Relapse is early: in 1,601 Toronto patients the triple-negative group had 2.6 times the hazard of distant recurrence and 3.2 times the hazard of death within five years but not afterwards, with the risk of distant recurrence peaking at about three years and falling rapidly (Dent 2007); recurrence and death rates were higher only in the first three years in the MD Anderson neoadjuvant series, which also found more visceral and fewer bone relapses (Liedtke 2008); and in 116 patients with metastatic disease treated in 2000 to 2006, 46 percent developed brain metastases before death and median survival from the metastatic diagnosis was 13.3 months (Lin 2008), a figure that the later trials in the treatment rows have moved. NICE NG12 (1.4) asks GPs to refer on the suspected cancer pathway anyone aged 30 or over with an unexplained breast lump, anyone 50 or over with discharge, retraction or other change in one nipple, and to consider it for skin changes or an unexplained axillary lump at 30 or over.

How it is diagnosed. In the UK the GP refers to a one-stop breast clinic for triple assessment: examination, mammogram, ultrasound and a needle core (or vacuum-assisted) biopsy, with axillary ultrasound and needle biopsy of any abnormal node (Cancer Research UK tests page). NICE NG101 (1.3.1 to 1.3.4) requires ER, PR and HER2 status to be assessed simultaneously on every invasive breast cancer at the initial histopathological diagnosis, by standardised, quality-assured immunohistochemistry reported quantitatively, and recorded at the preoperative and postoperative multidisciplinary meetings. Triple-negative is then a laboratory definition: ER negative when fewer than 1 percent of tumour nuclei stain, PR likewise, with 1 to 10 percent reported as a new ER Low Positive category on which endocrine benefit is uncertain (ASCO/CAP 2020, Allison); HER2 negative when immunohistochemistry is 0 or 1+, or 2+ with no amplification on in situ hybridisation (ASCO/CAP 2018, Wolff). The 2023 ASCO-CAP update kept those categories but added a reporting comment because the difference between HER2 0 and 1+ now decides eligibility for trastuzumab deruxtecan (Wolff 2023). NG101 defines triple-negative simply as ER, PR and HER2 each scored negative under local multidisciplinary team rules. Where the receptors are low rather than absent the behaviour is the same: ER 1 to 10 percent tumours matched ER-negative tumours for germline BRCA prevalence, response and survival in a US registry (Yoder 2022) and in an Italian series (Dieci 2021), and Sweden treats ER 1 to 9 percent HER2-negative tumours as triple-negative, with 9.9 percent of 5,655 such cancers in 2008 to 2020 falling in that band (Acs 2024). Breast MRI is not routine before surgery but is offered when the extent of disease is unclear, when breast density defeats mammography, or to size an invasive lobular cancer (NG101 1.2). Staging tests beyond the breast and axilla, where the stage or symptoms warrant, are CT, bone scan, liver ultrasound or PET-CT, and blood tests including CA 15-3 which cannot diagnose or direct treatment on its own (Cancer Research UK tests page). NG101 (1.3.6) adds one test that applies to this subtype alone: offer germline BRCA1 and BRCA2 testing to women under 50 with triple-negative breast cancer, including those with no family history.

What the pathologist sees. Most triple-negative cancers are invasive carcinomas of no special type (the term the 2019 WHO classification, 5th edition, uses for what was called ductal carcinoma) of grade 3, with a high proliferation index: the Swedish cohort's median Ki-67 was 63 percent in ER-zero and 60 percent in ER-low tumours, and 80.8 percent of ER-zero tumours were grade III (Acs 2024). By gene expression, 78.6 percent of triple-negative tumours are basal-like and 21.4 percent belong to other intrinsic subtypes, while 31.5 percent of basal-like tumours are not triple-negative (Prat 2013); basal-like tumours share many molecular features with high-grade serous ovarian cancer, and TP53 is the mutation that recurs across them (TCGA 2012). Within the triple-negative group, Lehmann's 2011 analysis of 587 cases found six expression subtypes (basal-like 1 and 2, immunomodulatory, mesenchymal, mesenchymal stem-like and luminal androgen receptor), refined in 2016 to four tumour-intrinsic subtypes once the immunomodulatory and stem-like signals were traced to infiltrating lymphocytes and stroma; Burstein's four (luminal androgen receptor, mesenchymal, basal-like immunosuppressed and basal-like immune-activated) overlap them. Each has a page linked from this record. The WHO classification also names special histological types that are usually triple-negative: metaplastic carcinoma (70.3 percent triple-negative in 2,451 US cases, Ong 2018), carcinoma with medullary pattern (no longer a separate type in the 5th edition, Thomas 2023; Cserni 2020), adenoid cystic carcinoma (95 percent localised at diagnosis, 15-year relative survival 91.4 percent, Ghabach 2010), apocrine carcinoma (ER and PR negative, androgen receptor positive, Vranic 2013) and secretory carcinoma (ETV6-NTRK3 fusion in 92 percent, Tognon 2002). Rare subtypes together are 5 to 10 percent of triple-negative cancers (Thomas 2023), and the European Working Group for Breast Screening Pathology concludes that patients with the indolent pure forms (classical adenoid cystic, low-grade adenosquamous, fibromatosis-like metaplastic, secretory, acinic cell, tall cell with reversed polarity) are unlikely to benefit from chemotherapy (Cserni 2021). Stromal tumour-infiltrating lymphocytes are scored on the biopsy, and PD-L1 combined positive score is tested when metastatic treatment is planned (the glossary terms carry the thresholds).

How it is staged. The TNM 8th edition (2017) is the usual system in the UK, and Cancer Research UK sets it out in plain words. T1 is 2 cm or less (T1mi 0.1 cm or less, T1a to 0.5 cm, T1b to 1 cm, T1c to 2 cm), T2 over 2 to 5 cm, T3 over 5 cm, T4 into the chest wall (a), skin (b), both (c) or inflammatory carcinoma (d). Nodes are staged clinically (cN) before any treatment and pathologically (pN) from the sentinel node or clearance: pN1 one to three axillary nodes (pN1mi micrometastases 0.2 to 2 mm), pN2 four to nine, pN3 ten or more or infraclavicular, or axillary plus internal mammary, or supraclavicular nodes. Stage 1A is T1 N0; 1B adds micrometastases; 2A is T0 to T1 N1 or T2 N0; 2B is T2 N1 or T3 N0; 3A is N2 with T0 to T3 or T3 N1; 3B is T4 with N0 to N2; 3C is any N3; stage 4 is any M1. The American Joint Committee on Cancer's 8th edition kept this anatomic stage but added a prognostic stage that folds grade, ER, PR and HER2 status (and, for some hormone receptor-positive tumours, a low Oncotype DX score) into the stage group (Giuliano 2017); in a 3,327-patient MD Anderson cohort the prognostic stage moved 29.5 percent of patients up and 28.1 percent down, and predicted disease-specific survival better than the anatomic stage (C index 0.84 against 0.74), with the same result in 54,727 Californian patients (Weiss 2018). Because grade 3 and receptor negativity count against a tumour, a triple-negative cancer generally carries a higher prognostic than anatomic stage; the UK pathway pages quote the anatomic stage. When chemotherapy is given before surgery, as it is for most stage II and III triple-negative disease, the response at surgery (pathological complete response or residual cancer burden) becomes the stronger prognostic marker (Cortazar 2014; Symmans 2017; the glossary terms).

Screening and prevention. There is no screening test for triple-negative disease as such. The NHS Breast Screening Programme invites everyone registered as female from age 50 (first invitation between 50 and 53) every three years until their 71st birthday, and women 71 or over can ask for screening (NHS). Because triple-negative tumours arrive younger and grow between rounds, a smaller share is screen-detected (Lin 2012; Houssami 2017), which is one reason the family history pathway matters here. NICE CG164 sets the UK rules: GPs refer to a specialist breast or genetics clinic on defined family patterns, and clinicians should seek genetics advice for families containing triple-negative breast cancer under 40, Jewish ancestry, sarcoma under 45 or glioma; genetic testing is offered when the combined BRCA1 and BRCA2 carrier probability is 10 percent or more (calculated with CanRisk, formerly BOADICEA, or the Manchester score); known BRCA1 or BRCA2 carriers are offered annual MRI from 30 to 49 and annual mammography from 40 to 69, TP53 carriers MRI from 20, and MRI is not offered to carriers aged 50 to 69 unless the breasts are dense. In the UK MARIBS study of 649 women at high familial risk, MRI sensitivity was 77 percent against 40 percent for mammography, 92 against 23 percent in BRCA1 carriers (Leach 2005). Risk-reducing surgery: bilateral mastectomy cut breast cancer risk by about 90 percent in BRCA1 and BRCA2 carriers in PROSE (1.9 percent against 48.7 percent over a mean 6.4 years; Rebbeck 2004) and pooled relative risk 0.114 in a 15-study meta-analysis, with contralateral mastectomy 0.072 and salpingo-oophorectomy 0.552 for breast cancer overall (Li 2016), though a prospective study of 3,722 carriers found oophorectomy did not lower breast cancer risk in BRCA1 carriers (hazard ratio 0.96) (Kotsopoulos 2017). CG164 restricts risk-reducing mastectomy to a small proportion of women from high-risk families, managed by a multidisciplinary team with genetic counselling, verified family history and reconstruction options discussed. Chemoprevention has a gap that matters for this subtype: NICE offers tamoxifen (premenopausal) or anastrozole (postmenopausal) for five years to women at high risk, and considers them at moderate risk, but tamoxifen had no effect on invasive ER-negative breast cancer in IBIS-I (hazard ratio 1.05 over 16 years; Cuzick 2015) or NSABP P-1 (Fisher 1998), and anastrozole's 49 percent reduction in IBIS-II was in ER-positive disease (Cuzick 2020); Cancer Research UK says the trial evidence is unclear for women with gene changes. One observational signal runs the other way: tamoxifen after a first breast cancer was associated with a lower contralateral breast cancer risk in BRCA1 (hazard ratio 0.38) and BRCA2 (0.33) carriers regardless of the first tumour's ER status (Phillips 2013). Breastfeeding is the modifiable factor with a specific ER-negative signal (Palmer 2014), and keeping a healthy weight and being physically active are the general measures Cancer Research UK names.