Skip to content
Bending the Curve Trial by Trial
?

Breast

Breast cancer

How breast cancer detection, surgery and drug treatment have changed, from early descriptions to biomarker-guided care.

5-year relative survival · United States, SEER 8 registries

76.16%→93.86%

1975 (observed value) to 2018 (observed value; last observed year in the table). Read these with the caveats under “Then and now”.Source

35 milestones, by decade

  • In US registry data on the SEER Cancer Stat Facts page, five-year relative survival for female breast cancer was 76.16% in 1975 and 93.86% in 2018 (SEER 8 series; the page does not say whether the year is the year of diagnosis), while the age-adjusted US death rate was highest at 33.25 per 100,000 women in 1989 and was 18.65 per 100,000 in 2024. The page does not say how much of the change is due to detection, screening or treatment.
  • A randomised trial of high-dose chemotherapy for metastatic breast cancer, published in the Journal of Clinical Oncology in 1995, was audited on site: records could be found for only 61 of the 90 patients reported as randomised, and the journal retracted the paper in 2001. None of its outcome data are quoted on this site.
  • Bevacizumab with paclitaxel received accelerated FDA approval for previously untreated HER2-negative metastatic breast cancer on 22 February 2008. FDA's withdrawal of the indication took effect on 18 November 2011, and a labelling supplement approved on 20 December 2011 removed it from the label. An FDA briefing document from November 2023 states that withdrawal of an accelerated-approval indication has generally been initiated voluntarily by the sponsor, and that FDA removing an indication itself has occurred only once, for bevacizumab in metastatic breast cancer.

Evidence current to October 2026 · Checked against primary sources to the extent possible

A guided tour of 11 milestones. You can pause or step through it yourself.

Filter milestonesshowing all 35
Treatment type
Setting

Antiquity & classical

  1. c. 1600 BCE · Origin · UnderstandingEdwin Smith Papyrus: Case 45, bulging breast tumoursAn Egyptian surgical text, copied around 1600 BCE, describes a man with cool, bulging swellings on the breast. Its 1930 translator reads the treatment line as 'there is no treatment', with the word supplied in brackets. The text does not name the condition, so calling it cancer is an interpretation.Case 45 of Breasted's 1930 translation of the Edwin Smith Surgical Papyrus describes a man with bulging tumours spreading over the breast: very cool, no fever, no granulation, no fluid or secretions. The verdict is 'an ailment with which I will contend' (doubtful). The treatment line reads 'There is no [treatment]', the bracket being Breasted's. Breasted dates the copy to the seventeenth century BCE and does not state that the disease is cancer.

Anatomists & surgeons

  1. 1894 · Treatment · SurgeryHalsted's radical mastectomy and its published resultsIn 1894 the surgeon William Halsted published results of breast cancer operations carried out at the Johns Hopkins Hospital. The radical mastectomy he pioneered removes the breast, the chest muscles and the lymph nodes in the armpit. It remained the standard operation until the late twentieth century, when randomised trials tested less extensive surgery.Case-series report (Ann Surg 1894) of operations for breast cancer at Johns Hopkins Hospital from June 1889 to January 1894, per its title. Its opening sentence states that in fifty cases operated on by what Halsted calls the complete method, only three local recurrences could be traced. Historians describe the operation as removal of breast, pectoral muscles and axillary nodes. Other counts and survival figures were not extracted.

The age of rays

  1. 1896 · Treatment · EndocrineBeatson reports ovarian removal for inoperable breast cancerIn 1896 the Glasgow surgeon George Beatson described removing the ovaries of two younger women whose breast cancer could not be operated on, after learning that removing a cow's ovaries keeps up its milk supply. Both improved, and one had no remaining sign of cancer eight months later. He offered it as an experiment, not a new standard.Case series of three patients read to the Medico-Chirurgical Society of Edinburgh on 20 May 1896. Bilateral oophorectomy in two younger women (aged 33 and 40) with recurrent or inoperable breast carcinoma, with thyroid extract: one had all visible disease gone at eight months; the other improved, then worsened when she did not keep taking treatment. The third, postmenopausal, received thyroid extract alone without marked effect. No controls or survival data; the author called the method experimental.

Chemotherapy

  1. 1971 · Understanding · Understanding⚑ needs reviewThe estrogen receptor and receptor-guided endocrine therapyElwood Jensen's laboratory work from the late 1950s led to the idea of an oestrogen receptor in target tissues. In 1971 he and colleagues reported on oestrogen receptors and breast cancer response to adrenal removal. Measuring the receptor in tumour tissue later became a way to predict who would benefit from hormone treatment.Qualitative history milestone; no trial data. Jensen et al. (Natl Cancer Inst Monogr 1971) addressed estrogen receptors in relation to breast cancer response to adrenalectomy (title only; no abstract available). Jensen's receptor work dates from the late 1950s (O'Malley and Khan 2013). Reviews describe receptor measurement as a test predicting outcome of antihormonal therapy and as the target for developing tamoxifen.
  2. 1971 · Detection · Detection & prevention⚑ needs reviewFirst screening RCTs: HIP (1971) and Swedish counties (1985)Two early randomised trials invited women to mammography and compared deaths from breast cancer with women not invited. The New York HIP trial and a Swedish trial in two counties each reported fewer breast cancer deaths in the invited group. Both reports had limits, and later trials and analyses have debated how large the benefit is.HIP (women aged 40-64; mammography plus clinical examination, four rounds): about 25% lower breast cancer mortality at 18 years in women aged 40-49 and 50-59 at entry, with the author questioning utility in the forties. Swedish county trial (162,981 enrolled; 134,867 aged 40-74 analysed): interim results to end of 1984 showed 31% lower breast cancer mortality and 25% fewer stage II or more advanced cancers in the invited group. The 1971 HIP paper has no abstract.
  3. 1971 · Treatment · EndocrineTamoxifen: from ICI 46474 to adjuvant standardIn 1971 a small Manchester study gave an anti-oestrogen tablet, later known as tamoxifen, to forty-six women with advanced breast cancer. About one in five clearly responded and side effects were mild. Later trials and pooled analyses showed that about five years of tamoxifen after surgery lowers recurrence and death in oestrogen-receptor-positive disease.Single-arm study (Cole 1971): ICI46474 in 46 postmenopausal women with advanced disease; clear response in 10 (22%). NSABP B-14 (N=2644, node-negative, ER-positive): disease-free survival 83% vs 77% with tamoxifen vs placebo (P<0.00001). EBCTCG 2011 (ER-positive disease, n=10,645): recurrence RR 0.53 in years 0-4 and 0.68 in years 5-9; breast cancer mortality RR 0.71, 0.66 and 0.68 in successive 5-year periods.

Molecular origins

  1. 1976 · Treatment · ChemotherapyAdjuvant CMF chemotherapy after mastectomy (Milan, 1976)In Milan, women with cancer in the armpit lymph nodes after mastectomy were randomly given either no further treatment or a year of combination chemotherapy. The first report, after twenty-seven months, described an advantage for chemotherapy in treatment failure. A twenty-year follow-up found better relapse-free and overall survival.Milan trial of 386 women with axillary node-positive breast cancer after radical mastectomy: 12 monthly cycles of CMF vs no further treatment. 1976 report at 27 months: treatment failure in 5.3% with CMF vs 24% of controls (term as worded in the journal abstract). 20-year follow-up (median 19.4 years): relative risk of relapse 0.71 (95% CI 0.56-0.90; P=0.004) and of death 0.78 (95% CI 0.62-0.99; P=0.04); benefit in all subgroups except postmenopausal women.
  2. 1981 · De-escalation · SurgeryBreast conservation vs mastectomy: Milan I, NSABP B-04 and B-06For most of the twentieth century, removing the whole breast was the standard operation for early breast cancer. Randomised trials from Milan and from the NSABP group found that removing only the tumour area, with radiotherapy, gave survival that did not differ significantly from mastectomy. Local recurrence was more common after the breast-conserving operation in the Milan trial.Milan I (701 women, tumours up to 2 cm, no palpable axillary nodes): at median 20 years, breast cancer death 26.1% with quadrantectomy plus radiotherapy vs 24.3% with Halsted mastectomy (P=0.8); local recurrence 8.8% vs 2.3% (P<0.001). NSABP B-06 at 20 years: overall survival HR 1.05 (95% CI 0.90-1.23) for lumpectomy and 0.97 (0.83-1.14) for lumpectomy plus irradiation vs total mastectomy; ipsilateral recurrence 14.3% with vs 39.2% without irradiation.
  3. 1987 · Understanding · UnderstandingHER-2/neu amplification predicts relapse and survivalIn 1987 a laboratory study of 189 breast cancers found extra copies of the HER-2/neu gene in about three in ten tumours. Having the extra copies predicted both time to relapse and overall survival, and this held after accounting for other known prognostic factors. The study did not test any treatment.Study of 189 primary breast cancers: HER-2/neu was amplified 2- to greater than 20-fold in 30% of tumours. Amplification was a significant predictor of overall survival and time to relapse, retained after adjustment for other prognostic factors, and had greater prognostic value than most established factors, including hormone receptor status, in node-positive disease.
  4. 1994 · Understanding · UnderstandingCloning of BRCA1 (1994) and BRCA2 (1995)In 1994 a research team reported a strong candidate for BRCA1, a gene on chromosome 17 in which inherited changes raise the risk of breast and ovarian cancer. In 1995 another team identified BRCA2, a second such gene on chromosome 13. Later large studies estimated how high the risk is for carriers.Miki et al. (Science, 1994): positional cloning of a strong candidate for BRCA1 on chromosome 17q; probable predisposing mutations in five of eight kindreds; predicted protein of 1863 amino acids. Wooster et al. (Nature, 1995): six different germline mutations in breast cancer families likely due to BRCA2 (13q12-q13). Later prospective cohort (Kuchenbaecker, JAMA 2017): cumulative breast cancer risk to age 80 years 72% (BRCA1) and 69% (BRCA2).

Targeted therapy

  1. 1998 · Understanding · ChemotherapyEBCTCG overviews: what adjuvant therapy achievesThe Early Breast Cancer Trialists' Collaborative Group pools patient-level data from many randomised trials. Its overviews of 1998, 2005 and 2011 reported that years of tamoxifen for oestrogen-receptor-positive disease and months of chemotherapy lowered recurrence and long-term breast cancer death, and that radiotherapy after breast-conserving surgery lowered recurrence.Patient-level meta-analyses, each with its own trial set. 1998 tamoxifen (about 5 years vs none; ER-positive or untested): 10-year survival 61.4% vs 50.5% (node-positive), 78.9% vs 73.3% (node-negative). 1998 polychemotherapy: mortality reduced 27% (age under 50), 11% (50-69). 2005: anthracycline-based chemotherapy cut annual breast cancer death rate about 38% and 20% in those groups; tamoxifen 31% (ER-positive). 2011: radiotherapy after breast-conserving surgery cut 10-year recurrence from 35.0% to 19.3%.
  2. 1998 · Treatment · TargetedTrastuzumab approved for HER2-overexpressing metastatic breast cancerIn 1998 the US regulator approved trastuzumab, an antibody aimed at a protein called HER2 that some breast cancers make in excess. A randomised trial published in 2001 found that adding it to chemotherapy delayed progression of metastatic disease and lengthened survival, with a higher risk of heart problems.Randomised trial (469 patients) of chemotherapy with or without trastuzumab in HER2-overexpressing metastatic breast cancer. Median time to progression 7.4 vs 4.6 months (P<0.001); response rate 50% vs 32%; median survival 25.1 vs 20.3 months (P=0.01). NYHA class III or IV cardiac dysfunction: 27% with anthracycline, cyclophosphamide and trastuzumab vs 8% without; 13% with paclitaxel and trastuzumab vs 1% without.
  3. 2000 · Reversal · ChemotherapyHigh-dose chemotherapy with stem-cell rescue: a reversalVery high doses of chemotherapy with stem-cell rescue were tested in randomised trials in breast cancer. A randomised trial in metastatic disease found no survival benefit, and a review of 14 trials in earlier disease found no survival gain but more treatment-related deaths. An on-site review of one further study found its records did not match the data presented.Metastatic disease, Philadelphia BMT Group (N=199 randomised after response to induction): 3-year survival 32% vs 38% with high-dose chemotherapy plus transplantation vs conventional chemotherapy, not significant; median TTP 9.6 vs 9.0 months. High-risk primary disease, Netherlands (N=885): 5-year RFS 65% vs 59% (HR 0.83; 95% CI 0.66-1.03; P = 0.09). Cochrane, 14 trials, 5600 women: treatment-related deaths RR 7.97 (95% CI 3.99-15.92). Bezwoda records are not evidence.
  4. 2002 · Treatment · EndocrineAromatase inhibitors versus tamoxifen (ATAC, BIG 1-98)In 2002 the ATAC trial showed that, for postmenopausal women with hormone-sensitive early breast cancer, five years of the aromatase inhibitor anastrozole gave slightly better disease-free survival than tamoxifen, with a different side-effect profile. Letrozole showed a similar edge in 2005, and a pooled analysis later found fewer breast cancer deaths with aromatase inhibitors.ATAC (N=9366), anastrozole vs tamoxifen: 3-year DFS 89.4% vs 87.4%, HR 0.83 (0.71-0.96); the combination was no better than tamoxifen. At 10 years, DFS HR 0.91 (0.83-0.99) in all randomised patients. BIG 1-98 (N=8010), letrozole vs tamoxifen: DFS HR 0.81 (0.70-0.93). EBCTCG 2015 (5 years of AI vs tamoxifen): 10-year breast cancer mortality 12.1% vs 14.2%, RR 0.85 (0.75-0.96).
  5. 2003 · De-escalation · SurgerySentinel-node biopsy and less axillary surgery (B-32, Z0011)Sentinel-node biopsy removes only the first lymph nodes that drain a breast tumour instead of clearing the whole armpit. In randomised trials, leaving the armpit alone did not significantly worsen survival when the sentinel node was clear, or when one or two nodes held cancer in women having breast-conserving surgery with radiotherapy.Veronesi 2003 (516 patients, tumours up to 2 cm): sentinel-node status accuracy 96.9%; no overt axillary metastasis in 167 without dissection. NSABP B-32 (5,611 randomised; 3,986 SLN-negative analysed): 8-year OS 90.3% SLN alone vs 91.8% with ALND (HR 1.20, 95% CI 0.96-1.50). ACOSOG Z0011 (891 women, 1-2 positive SLNs, lumpectomy plus whole-breast radiotherapy): 10-year OS 86.3% SLND alone vs 83.6% ALND (HR 0.85; noninferiority P=.02).
  6. 2005 · Treatment · TargetedAdding trastuzumab to adjuvant chemotherapy (HERA, B-31/N9831)In 2005 two reports showed that adding trastuzumab to treatment after surgery for HER2-positive early breast cancer led to fewer recurrences, second cancers or deaths in early follow-up. One trial gave trastuzumab after chemotherapy; the other report, a joint analysis of two trials, gave it alongside part of the chemotherapy. Heart problems were a recognised risk.HERA (1 year of trastuzumab vs observation after chemotherapy; first interim analysis, median follow-up 1 year): disease-free survival HR 0.54 (95% CI 0.43 to 0.67; P<0.0001). Joint NSABP B-31/NCCTG N9831 analysis (chemotherapy with or without concurrent trastuzumab; first interim analysis): HR 0.48 (P<0.0001), 133 vs 261 events. Designs differ; the results are reported separately and are not comparable across trials.

Immunotherapy

  1. 2011 · Reversal · TargetedFDA withdraws bevacizumab for metastatic breast cancer (2011)In 2008 the US regulator approved bevacizumab with paclitaxel as first treatment for metastatic breast cancer because it delayed cancer growth, on condition that further trials confirm the benefit. Those trials showed smaller gains and no survival advantage. The US indication was withdrawn in November 2011. The European listing for this use remains.E2100 (N=722; paclitaxel with or without bevacizumab): median PFS 11.8 vs 5.9 months (HR 0.60; P<0.001); median OS 26.7 vs 25.2 months (HR 0.88; P = 0.16). AVADO (docetaxel; bevacizumab 15 mg/kg): median PFS 10.1 vs 8.2 months (HR 0.77; P = 0.006, unstratified). RIBBON-1: PFS HR 0.69 (capecitabine) and 0.64 (taxane or anthracycline cohort). Accelerated approval 22 February 2008; withdrawal decision 18 November 2011.
  2. 2011 · Treatment · TargetedCLEOPATRA: pertuzumab added to trastuzumab and docetaxelA large trial in people starting treatment for HER2-positive metastatic breast cancer compared adding a second antibody, pertuzumab, to trastuzumab and docetaxel against adding a placebo. Progression took longer with pertuzumab and, in later reports, survival was longer. The US regulator approved the combination in June 2012.Double-blind phase 3, 808 patients, first-line HER2-positive metastatic disease: pertuzumab vs placebo plus trastuzumab and docetaxel. Independently assessed median PFS 18.5 vs 12.4 months (HR 0.62; 95% CI 0.51 to 0.75; P<0.001). Final overall survival (median follow-up 50 months): 56.5 vs 40.8 months (HR 0.68; 95% CI 0.56 to 0.84). End-of-study descriptive analysis: median OS 57.1 vs 40.8 months; 8-year OS 37% vs 23%.
  3. 2012 · Treatment · Antibody–drug conjugateEMILIA: trastuzumab emtansine versus lapatinib plus capecitabineIn people with HER2-positive advanced breast cancer previously treated with trastuzumab and a taxane, trastuzumab emtansine, an antibody linked to a chemotherapy drug, delayed progression and lengthened survival compared with lapatinib plus capecitabine. Grade 3 or 4 side effects were less frequent overall.Randomised trial (991 patients previously treated with trastuzumab and a taxane): trastuzumab emtansine vs lapatinib plus capecitabine. Median PFS (independent review) 9.6 vs 6.4 months (HR 0.65; 95% CI 0.55 to 0.77; P<0.001). Median OS at the second interim analysis 30.9 vs 25.1 months (HR 0.68; 95% CI 0.55 to 0.85; P<0.001). Response rate 43.6% vs 30.8%. Grade 3 or 4 adverse events 41% vs 57%.
  4. 2012 · Detection · Detection & preventionScreening, overdiagnosis and the benefit-harm debateEstimates of how much mammography screening lowers breast cancer deaths, and of how many cancers it finds that would never have caused harm (overdiagnosis), differ between reviews and trials. The differences reflect the trials chosen, the methods and the assumptions made, and the balance remains debated.Estimates differ by method. UK Independent Panel 2012, meta-analysis of 11 randomised trials: breast cancer mortality RR 0.80 (95% CI 0.73-0.89); excess incidence 11% of cancers in the invited group long term. Cochrane 2013: RR 0.90 (0.79-1.02) in 3 adequately randomised trials, 0.81 across 7. CNBSS 25 years: HR 1.05 (0.85-1.30) for death from cancers diagnosed in the screening period; 22% of screen-detected invasive cancers overdiagnosed. Welch 2016 (SEER, age 40+): 132 overdiagnosed cases per 100,000 women.
  5. 2015 · Treatment · Targeted⚑ needs reviewCDK4/6 inhibitors added to first-line endocrine therapyIn 2015 the US regulator gave palbociclib accelerated approval with letrozole as initial therapy for advanced hormone-receptor-positive breast cancer. Phase 3 trials in 2016 and 2017 then tested three CDK4/6 inhibitors with an aromatase inhibitor, each delaying progression; with palbociclib, median progression-free survival was 24.8 versus 14.5 months. Longer follow-up showed a survival gain with ribociclib.Phase 3, first-line, postmenopausal HR-positive, HER2-negative advanced disease. PALOMA-2 (N=666): median PFS 24.8 vs 14.5 months, HR 0.58 (0.46-0.72). MONALEESA-2 (N=668): PFS HR 0.56 (0.43-0.72); final OS 63.9 vs 51.4 months, HR 0.76 (0.63-0.93). MONARCH 3 (N=493): PFS HR 0.54 (0.41-0.72); final OS HR 0.804 (0.637-1.015), not significant.
  6. 2015 · De-escalation · RadiationPRIME II: omitting radiotherapy in low-risk older womenPRIME II asked whether older women with small, low-risk, hormone-sensitive breast cancers who had breast-conserving surgery and endocrine treatment could skip radiotherapy. Without radiotherapy, local recurrence was higher. At ten years, survival was almost identical and distant spread was not higher.PRIME II (1,326 women aged 65 or older; hormone receptor-positive, node-negative, T1-T2 up to 3 cm, clear margins; breast-conserving surgery plus endocrine therapy): whole-breast radiotherapy vs none. 5-year IBTR 1.3% vs 4.1% (HR 5.19, 95% CI 1.99-13.52). 10-year local recurrence 0.9% vs 9.5% (HR 10.4, 95% CI 4.1-26.1; P<0.001); distant recurrence as first event 3.0% vs 1.6%; overall survival 80.7% vs 80.8%.
  7. 2016 · De-escalation · UnderstandingGenomic assays guide who can skip chemotherapyGene-expression tests on tumour tissue helped sort women with hormone-receptor-positive breast cancer by chemotherapy benefit. Women with no nodal disease and a mid-range score, and postmenopausal women with one to three positive nodes, did similarly well on hormone therapy alone. A 70-gene trial found a small gap in its low-risk group. Premenopausal women with positive nodes gained from chemotherapy.MINDACT (N=6693): high clinical risk, low genomic risk without chemotherapy had 5-year distant metastasis-free survival 94.7% (95% CI 92.5-96.2), lower limit above the 92% boundary. TAILORx (recurrence score 11-25, node-negative): iDFS HR 1.08 (0.94-1.24), endocrine vs chemoendocrine; 9-year iDFS 83.3% vs 84.3%. RxPONDER (1-3 nodes, score 25 or lower): iDFS HR (chemoendocrine vs endocrine only) 1.02 in postmenopausal and 0.60 in premenopausal women.
  8. 2018 · Treatment · Antibody–drug conjugateKATHERINE: T-DM1 for residual disease after neoadjuvant therapyFor people with HER2-positive early breast cancer who still had invasive cancer at surgery after pre-operative treatment, switching to trastuzumab emtansine instead of continuing trastuzumab lowered the risk of the cancer coming back. A later report also reported a lower risk of death.Phase 3 (1,486 patients) with residual invasive disease after neoadjuvant taxane plus trastuzumab: adjuvant T-DM1 vs trastuzumab for 14 cycles. Interim: 3-year iDFS 88.3% vs 77.0% (HR 0.50; 95% CI 0.39 to 0.64; P<0.001). Final iDFS (median follow-up 8.4 years): HR 0.54 (95% CI 0.44 to 0.66); 7-year 80.8% vs 67.1%. Second interim OS: HR 0.66 (95% CI 0.51 to 0.87; P=0.003); 7-year 89.1% vs 84.4%.
  9. 2019 · Treatment · Targeted⚑ needs reviewPI3K/AKT pathway inhibitors after endocrine therapyAfter hormone therapy stops working, adding a drug that blocks the PI3K/AKT pathway to fulvestrant can delay progression. Alpelisib helped women whose tumours carry a PIK3CA mutation; capivasertib helped overall, most clearly when the pathway was altered; inavolisib with palbociclib and fulvestrant lengthened survival in PIK3CA-mutated disease. Gedatolisib helped in tumours without the mutation.After endocrine therapy: SOLAR-1 (PIK3CA-mutated, n=341): PFS 11.0 vs 5.7 months, HR 0.65 (0.50-0.85); final OS HR 0.86 (0.64-1.15), not significant. CAPItello-291 (N=708): PFS HR 0.60; altered pathway HR 0.50. INAVO120 (N=325, endocrine-resistant, PIK3CA-mutated): PFS 15.0 vs 7.3 months, OS HR 0.67 (0.48-0.94). VIKTORIA-1 (PIK3CA wild-type, N=392): PFS HR 0.24 and 0.33 vs fulvestrant.

Precision & engineered

  1. 2020 · Treatment · TargetedAdjuvant CDK4/6 inhibition: monarchE and NATALEETwo trials showed that adding a CDK4/6 inhibitor to hormone therapy after surgery lowers the chance of recurrence in higher-risk hormone-receptor-positive breast cancer. Abemaciclib was given for two years in node-positive disease, and ribociclib for three years in a wider stage two and three group. After seven years abemaciclib has also shown a small survival gain.monarchE (N=5637, node-positive, high risk, open-label): iDFS HR 0.75 (0.60-0.93) at the first analysis; at 7 years iDFS 77.4% vs 70.9% (HR 0.734) and OS HR 0.842 (0.722-0.981; P=0.027), 7-year OS 86.8% vs 85.0%. NATALEE (N=5101, stage II-III): 3-year iDFS 90.7% vs 87.6%, HR 0.749 (0.628-0.892); overall survival immature.
  2. 2020 · Treatment · ImmunotherapyKEYNOTE-522: pembrolizumab added to chemotherapy in early TNBCAdding the immunotherapy drug pembrolizumab to chemotherapy before surgery, and continuing it afterwards, left more people with early triple-negative breast cancer with no invasive cancer at surgery. Later reports showed fewer relapses and longer survival. The immune drug also adds its own side effects.Phase 3, randomised 2:1, placebo-controlled; stage II or III TNBC (N=1174; 784 vs 390): neoadjuvant pembrolizumab with paclitaxel-carboplatin then anthracycline-cyclophosphamide, adjuvant pembrolizumab up to nine cycles. pCR 64.8% vs 51.2% (first 602 patients; P<0.001). EFS at 36 months 84.5% vs 76.8% (HR 0.63; 95% CI 0.48-0.82). OS at 60 months 86.6% vs 81.7% (P = 0.002; no HR in the abstract).
  3. 2020 · Treatment · Antibody–drug conjugateSacituzumab govitecan (ASCENT): first TROP-2 ADC in metastatic TNBCIn people whose triple-negative breast cancer had spread and come back after earlier treatment, a drug that carries chemotherapy directly to cancer cells (sacituzumab govitecan) was compared with standard single-agent chemotherapy. It roughly tripled the time before the cancer grew and extended life. Low blood counts and diarrhea were more common.Phase 3, relapsed or refractory metastatic TNBC; sacituzumab govitecan vs single-agent chemotherapy of physician's choice, 1:1. Primary analysis (235 vs 233 without brain metastases): median PFS 5.6 vs 1.7 months (HR 0.41; 95% CI 0.32-0.52; P<0.001); median OS 12.1 vs 6.7 months (HR 0.48; 95% CI 0.38-0.59); ORR 35% vs 5%. Final analysis (ITT, 267 vs 262): PFS 4.8 vs 1.7 months (HR 0.41); OS 11.8 vs 6.9 months (HR 0.51; 95% CI 0.42-0.63).
  4. 2021 · Treatment · TargetedOlympiA: one year of adjuvant olaparib for germline BRCA carriersFor people with an inherited BRCA1 or BRCA2 variant and early HER2-negative breast cancer at high risk of coming back, one year of the PARP inhibitor olaparib after standard treatment left more people alive and free of invasive cancer at three years than placebo. Longer follow-up also showed a survival advantage.Double-blind phase 3 (N=1836): HER2-negative high-risk early breast cancer with germline BRCA1/2 pathogenic or likely pathogenic variants, after local treatment and (neo)adjuvant chemotherapy; 1 year olaparib vs placebo. First interim analysis (median follow-up 2.5 years): 3-year iDFS 85.9% vs 77.1% (HR 0.58; P<0.001). Second interim OS analysis (3.5 years): HR 0.68 (98.5% CI 0.47-0.97; P = 0.009); 4-year OS 89.8% vs 86.4%.
  5. 2021 · Reversal · Immunotherapy⚑ needs reviewAtezolizumab in PD-L1+ metastatic TNBC: approval, then withdrawalIn 2019 the US regulator gave early approval to atezolizumab with nab-paclitaxel for advanced triple-negative breast cancer with PD-L1 on immune cells, based on delayed progression. A second trial, using paclitaxel, did not show a benefit, and the indication was removed in October 2021. A different immunotherapy combination stayed approved for a differently defined group.IMpassion130 (N=902; atezolizumab plus nab-paclitaxel): PD-L1-positive subgroup median PFS 7.5 vs 5.0 months (HR 0.62; 95% CI 0.49-0.78); ITT PFS 7.2 vs 5.5 months (HR 0.80); final ITT OS HR 0.87 (95% CI 0.75-1.02; P = 0.077). IMpassion131 (N=651; paclitaxel; PD-L1-positive): PFS 6.0 vs 5.7 months (HR 0.82; 95% CI 0.60-1.12; P = 0.20); OS HR 1.11. An FDA letter of 6 October 2021 approved removal of the indication.
  6. 2022 · Treatment · Antibody–drug conjugateDESTINY-Breast03: T-DXd versus T-DM1 in HER2-positive diseaseIn people with HER2-positive metastatic breast cancer previously treated with trastuzumab and a taxane, trastuzumab deruxtecan was compared directly with trastuzumab emtansine. The newer drug delayed progression for longer. Lung inflammation was a recognised risk.Phase 3 (524 patients previously treated with trastuzumab and a taxane): trastuzumab deruxtecan vs trastuzumab emtansine. Primary analysis: 12-month PFS (blinded central review) 75.8% vs 34.1% (HR 0.28; 95% CI 0.22 to 0.37; P<0.001); adjudicated drug-related ILD 10.5% vs 1.9%. Updated analysis: median PFS 28.8 vs 6.8 months (HR 0.33); OS HR 0.64 (95% CI 0.47 to 0.87), median not reached.
  7. 2022 · Treatment · Antibody–drug conjugateT-DXd in HER2-low breast cancer (DESTINY-Breast04, -06)In 2022 a trial of trastuzumab deruxtecan, an antibody linked to a chemotherapy drug, in metastatic breast cancers with low HER2 levels reported delayed progression and longer survival than the doctor's choice of chemotherapy. The report states earlier HER2-directed drugs had been ineffective in this group. Lung inflammation was a recognised risk.DESTINY-Breast04: phase 3, 2:1 randomisation of 557 patients with HER2-low (IHC 1+ or 2+/ISH-) metastatic disease after 1-2 chemotherapy lines, trastuzumab deruxtecan vs physician's choice. Hormone receptor-positive cohort (494): median PFS 10.1 vs 5.4 months (HR 0.51); OS 23.9 vs 17.5 months (HR 0.64; P=0.003). Long-term update: overall-cohort OS 22.9 vs 16.8 months (HR 0.69; 95% CI 0.55 to 0.86). Adjudicated drug-related ILD: 12.1%.
  8. 2022 · Treatment · Endocrine⚑ needs reviewElacestrant: the first oral SERD (EMERALD)Elacestrant is an oral drug that breaks down the oestrogen receptor. In the EMERALD trial it delayed progression compared with standard hormone therapy in women whose cancer had progressed after earlier treatment that included a CDK4/6 inhibitor, and the benefit was clearest when the tumour carried an ESR1 mutation. Regulators approved it for that group.EMERALD (N=477; ER-positive, HER2-negative, pretreated including a CDK4/6 inhibitor): PFS by blinded central review HR 0.70 (0.55-0.88) overall and 0.55 (0.39-0.77) in ESR1-mutated disease; ESR1-mutated median PFS 3.8 vs 1.9 months (FDA page and paper Discussion). 12-month PFS in ESR1-mutated disease 26.8% vs 8.2%. Interim overall survival HR 0.75 (0.54-1.04) overall, not significant.
  9. 2025 · Treatment · Antibody–drug conjugateDESTINY-Breast09: T-DXd plus pertuzumab in first-line therapyIn a first-line trial for HER2-positive metastatic breast cancer, trastuzumab deruxtecan plus pertuzumab was compared with a taxane, trastuzumab and pertuzumab. Progression took longer with the new combination, and the US regulator approved it in December 2025. Lung inflammation, including two deaths, was reported.Phase 3 (1157 enrolled), 1:1:1, first-line HER2-positive metastatic disease. Interim analysis (cutoff 26 February 2025): trastuzumab deruxtecan plus pertuzumab (383 patients) vs THP (387): median PFS (blinded central review) 40.7 vs 26.9 months (HR 0.56; 95% CI 0.44 to 0.71; P<0.00001). Confirmed response 85.1% vs 78.6%. Adjudicated drug-related ILD 12.1% vs 1.0%. The trastuzumab deruxtecan plus placebo arm remained blinded.
  10. 2025 · Treatment · Antibody–drug conjugateADCs reach first-line metastatic TNBC (ASCENT-03/-04, TROPION-02)Three recent trials tested antibody-drug conjugates against chemotherapy as the first treatment for advanced triple-negative breast cancer. Two drugs delayed progression for people who could not receive PD-1/PD-L1 immunotherapy, and sacituzumab govitecan with pembrolizumab did so in PD-L1-positive disease. The groups and comparators differ, so the results should not be set side by side.Three phase 3 trials, first-line advanced TNBC. ASCENT-03 (N=558; not candidates for PD-(L)1 inhibitors): PFS 9.7 vs 6.9 months, sacituzumab govitecan vs chemotherapy (HR 0.62; 95% CI 0.50-0.77). ASCENT-04 (N=443; PD-L1-positive): with pembrolizumab, PFS 11.2 vs 7.8 months (HR 0.65; 95% CI 0.51-0.84). TROPION-Breast02 (N=644; immunotherapy not an option): PFS 10.8 vs 5.6 months (HR 0.57); OS 23.7 vs 18.7 months (HR 0.79; 95% CI 0.64-0.98).

Then and now

  • 5-year relative survival · United States, SEER 8 registries

    • 76.16%

      1975 (observed value)

    Females, all races; SEER 8 five-year relative survival series, 1975-2018, on the SEER Cancer Stat Facts female breast page. Age range and stage mix are not stated for this series on the page.

    Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the modelled trend value in the adjacent column. Survival measured from diagnosis can rise through earlier detection (lead-time bias) and through detection of cancers that would not have caused harm (overdiagnosis) even when length of life is unchanged (NCI PDQ breast screening summary); the SEER page does not say how much of the change in this series is due to these effects, to treatment or to other factors, so read it beside the mortality values. SEER 8 covers a limited set of registries. The page does not label the 'Year' column as year of diagnosis (the graph note says only 'from 1975-2018'), and does not state an age range or stage mix for this series. Observed values are shown because the 2019-2023 entries in the table are modelled trend values only.

    Source

  • 5-year relative survival · United States, SEER 8 registries

    • 93.86%

      2018 (observed value; last observed year in the table)

    Females, all races; SEER 8 five-year relative survival series, 1975-2018, on the SEER Cancer Stat Facts female breast page. Age range and stage mix are not stated for this series on the page.

    Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the modelled trend value in the adjacent column. Survival measured from diagnosis can rise through earlier detection (lead-time bias) and through detection of cancers that would not have caused harm (overdiagnosis) even when length of life is unchanged (NCI PDQ breast screening summary); the SEER page does not say how much of the change in this series is due to these effects, to treatment or to other factors, so read it beside the mortality values. SEER 8 covers a limited set of registries. The page does not label the 'Year' column as year of diagnosis (the graph note says only 'from 1975-2018'), and does not state an age range or stage mix for this series. Observed values are shown because the 2019-2023 entries in the table are modelled trend values only. Compare only with the 1975 value from the same SEER 8 series.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 91.9%

      2016-2022

    Females with female breast cancer; the headline figure on the SEER Cancer Stat Facts page (SEER 21, excluding IL). Race and age range are not stated beside the figure.

    A different registry set from SEER 8 and a different time basis, so it is not a continuation of the 1975-2018 series and must not be subtracted from it. The page labels the period '2016-2022' and does not say whether this is year of diagnosis. Relative survival; the SEER page does not discuss lead-time bias or overdiagnosis, which can raise survival measured from diagnosis (NCI PDQ breast screening summary).

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 100%

      2016-2022; localized stage at diagnosis (64% of cases)

    Females, all races, localized (confined to primary site) by SEER Combined Summary Stage.

    Stage-specific value from the same SEER 21 table as the 91.9% figure. The page states that stage at diagnosis has a strong influence on the length of survival and gives the stage mix as 64% localized, 27% regional, 6% distant and 2% unstaged. It is a five-year figure and says nothing about survival beyond five years. Not comparable with SEER 8 values.

    Source

  • 5-year relative survival · United States, SEER 21 (excluding IL)

    • 33.8%

      2016-2022; distant stage at diagnosis (6% of cases)

    Females, all races, distant (metastasized) by SEER Combined Summary Stage.

    Same SEER 21 table. Stage at diagnosis, not disease subtype, is the only split shown; the page gives no breakdown by hormone receptor or HER2 status here. Not comparable with SEER 8 values.

    Source

  • Age-standardised mortality · United States

    • 31.45per 100,000

      1975 (observed)

    Females, all races; age-adjusted death rate from US mortality data (not SEER registries), 'Death Rate - U.S.' column; the page states that deaths come from U.S. Mortality.

    Death rates describe the whole US female population and are not measured from the date of diagnosis; the SEER page does not attribute changes to any cause. The NCI PDQ breast screening summary names cause-specific mortality as the end point needed to assess screening benefit apart from lead-time and overdiagnosis biases.

    Source

  • Age-standardised mortality · United States

    • 33.25per 100,000

      1989 (observed; highest value in the 1975-2024 table)

    Females, all races; age-adjusted death rate from US mortality data.

    Highest observed value in the 'Death Rate - U.S.' column of the 1975-2024 table. Same caveats as the 1975 value.

    Source

  • Age-standardised mortality · United States

    • 18.65per 100,000

      2024 (observed)

    Females, all races; age-adjusted death rate from US mortality data.

    The page headline gives 18.9 per 100,000 for 2020-2024 deaths; use one or the other, not both. Same caveats as the 1975 value.

    Source

Survival figures are affected by earlier detection (lead time) and overdiagnosis, and by changes in who is diagnosed. For screened cancers, read them alongside mortality.

Evidence chains

A lane follows one defined population and setting, where each new trial tests against what was then standard. Compare only within a lane, and read every comparison as illustrative.

HER2-positive metastatic, first line

HER2-positive metastatic breast cancer with no earlier treatment for advanced disease · Metastatic · endpoint: PFS

Results are not comparable across trials.

Eras, chemotherapy partners and prior-treatment rules differ. The 2001 trial reported time to progression (drawn as a diamond), not progression-free survival, and its control was chemotherapy alone. CLEOPATRA's control (docetaxel and trastuzumab) built on that result, and DESTINY-Breast09 compares against a taxane with trastuzumab and pertuzumab. Review methods and follow-up differ, so compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

01020304050PFS, median (months)199820112025slamon-2001-trastuzumab-mbcCLEOPATRADESTINY-Breast09
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
PFS, median, across 3 trials in this lane. slamon-2001-trastuzumab-mbc (1998): experimental 7.4 months vs control 4.6 months; CLEOPATRA (2011): experimental 18.5 months vs control 12.4 months; DESTINY-Breast09 (2025): experimental 40.7 months vs control 26.9 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
slamon-2001-trastuzumab-mbc1998TTP (differs from lane)4.6 months7.4 months
CLEOPATRA2011PFS12.4 months18.5 months
DESTINY-Breast092025PFS26.9 months40.7 months

HER2-positive metastatic, after trastuzumab and a taxane

HER2-positive advanced or metastatic breast cancer previously treated with trastuzumab and a taxane · Metastatic · endpoint: PFS

Results are not comparable across trials.

EMILIA and DESTINY-Breast03 differ in era, earlier exposure to pertuzumab, handling of brain metastases and PFS review methods. DESTINY-Breast03's primary analysis reported a 12-month landmark, so the median shown is from its labelled long-term update. T-DM1 was the experimental arm in EMILIA and the control arm in DESTINY-Breast03.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520253035PFS, median (months)20122022EMILIADESTINY-Breast03
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
PFS, median, across 2 trials in this lane. EMILIA (2012): experimental 9.6 months vs control 6.4 months; DESTINY-Breast03 (2022): experimental 28.8 months vs control 6.8 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
EMILIA2012PFS6.4 months9.6 months
DESTINY-Breast032022PFS6.8 months28.8 months

PD-L1-positive (CPS 10 or more) triple-negative, first line

Previously untreated triple-negative breast cancer with PD-L1 CPS 10 or more · Metastatic · endpoint: PFS

Results are not comparable across trials.

The KEYNOTE-355 figure is for its CPS 10 or more subgroup, while ASCENT-04 enrolled only PD-L1-positive patients (CPS 10 or more by the 22C3 assay, per the FDA notice). Chemotherapy partners, pembrolizumab in ASCENT-04's control arm, and follow-up differ. ASCENT-03 and TROPION-Breast02, for patients not eligible for PD-(L)1 inhibitors, are a different population and are not part of this lane.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

02468101214PFS, median (months)20212025KEYNOTE-355ASCENT-04
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
PFS, median, across 2 trials in this lane. KEYNOTE-355 (2021): experimental 9.7 months vs control 5.6 months; ASCENT-04 (2025): experimental 11.2 months vs control 7.8 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
KEYNOTE-3552021PFS5.6 months9.7 months
ASCENT-042025PFS7.8 months11.2 months

Standard of care, then and now

How the usual sequence of treatment changed for one setting. Choose a point in time; each step links to the milestone behind it.

HER2-positive metastatic breast cancer: lines of treatment

HER2-positive advanced or metastatic breast cancer · Metastatic

  1. Antibody–drug conjugate11st line: T-DXd + pertuzumabDESTINY-Breast09: T-DXd plus pertuzumab in first-line therapy (2025)
  2. Antibody–drug conjugate2After trastuzumab and a taxane: T-DXdDESTINY-Breast03: T-DXd versus T-DM1 in HER2-positive disease (2022)
Today. Based only on the trials cited in each step. Trastuzumab deruxtecan has been tested in both settings; the FDA actions are listed in the milestones.
Text version
  • 2000: 1st line: chemotherapy + trastuzumab
  • 2013: 1st line: taxane + trastuzumab + pertuzumab → After trastuzumab and a taxane: T-DM1
  • Today: 1st line: T-DXd + pertuzumab → After trastuzumab and a taxane: T-DXd

Reversals and dead ends

Approaches that were abandoned or withdrawn, and trials whose results were later questioned, are part of the story.

  • 2000

    High-dose chemotherapy with stem-cell rescue: a reversal

    Very high doses of chemotherapy with stem-cell rescue were tested in randomised trials in breast cancer. A randomised trial in metastatic disease found no survival benefit, and a review of 14 trials in earlier disease found no survival gain but more treatment-related deaths. An on-site review of one further study found its records did not match the data presented.

    Metastatic disease, Philadelphia BMT Group (N=199 randomised after response to induction): 3-year survival 32% vs 38% with high-dose chemotherapy plus transplantation vs conventional chemotherapy, not significant; median TTP 9.6 vs 9.0 months. High-risk primary disease, Netherlands (N=885): 5-year RFS 65% vs 59% (HR 0.83; 95% CI 0.66-1.03; P = 0.09). Cochrane, 14 trials, 5600 women: treatment-related deaths RR 7.97 (95% CI 3.99-15.92). Bezwoda records are not evidence.

  • 2011

    FDA withdraws bevacizumab for metastatic breast cancer (2011)

    In 2008 the US regulator approved bevacizumab with paclitaxel as first treatment for metastatic breast cancer because it delayed cancer growth, on condition that further trials confirm the benefit. Those trials showed smaller gains and no survival advantage. The US indication was withdrawn in November 2011. The European listing for this use remains.

    E2100 (N=722; paclitaxel with or without bevacizumab): median PFS 11.8 vs 5.9 months (HR 0.60; P<0.001); median OS 26.7 vs 25.2 months (HR 0.88; P = 0.16). AVADO (docetaxel; bevacizumab 15 mg/kg): median PFS 10.1 vs 8.2 months (HR 0.77; P = 0.006, unstratified). RIBBON-1: PFS HR 0.69 (capecitabine) and 0.64 (taxane or anthracycline cohort). Accelerated approval 22 February 2008; withdrawal decision 18 November 2011.

  • 2021

    Atezolizumab in PD-L1+ metastatic TNBC: approval, then withdrawal

    In 2019 the US regulator gave early approval to atezolizumab with nab-paclitaxel for advanced triple-negative breast cancer with PD-L1 on immune cells, based on delayed progression. A second trial, using paclitaxel, did not show a benefit, and the indication was removed in October 2021. A different immunotherapy combination stayed approved for a differently defined group.

    IMpassion130 (N=902; atezolizumab plus nab-paclitaxel): PD-L1-positive subgroup median PFS 7.5 vs 5.0 months (HR 0.62; 95% CI 0.49-0.78); ITT PFS 7.2 vs 5.5 months (HR 0.80); final ITT OS HR 0.87 (95% CI 0.75-1.02; P = 0.077). IMpassion131 (N=651; paclitaxel; PD-L1-positive): PFS 6.0 vs 5.7 months (HR 0.82; 95% CI 0.60-1.12; P = 0.20); OS HR 1.11. An FDA letter of 6 October 2021 approved removal of the indication.

Today and tomorrow

Standard of care today

Screening
The USPSTF 2024 recommendation statement (JAMA) recommends biennial screening mammography for women aged 40 to 74 years (grade B) and states that current evidence is insufficient to assess the balance of benefits and harms of screening mammography at age 75 or older, and of supplemental breast ultrasonography or MRI in women identified to have dense breasts on an otherwise negative screening mammogram (I statements). The recommendation applies to cisgender women and other people assigned female at birth aged 40 or older at average risk of breast cancer. PubMed lists two errata for the statement (JAMA, 30 September 2024, titled 'Incorrect Study N Reported for Supporting Study'; JAMA, 12 May 2026, titled 'Incorrect Data and Reference'); their text could not be opened, so what they correct is not known. Programmes in other countries are not summarised here.
US Preventive Services Task Force 2024
Adjuvant

Hormone receptor-positive, HER2-negative early breast cancer, FDA regulatory actions:

  • olaparib for germline BRCA-mutated, HER2-negative high-risk early breast cancer previously treated with neoadjuvant or adjuvant chemotherapy (11 March 2022
  • the indication is not restricted by hormone receptor status)
  • abemaciclib with endocrine therapy for node-positive early breast cancer at high risk of recurrence, with the earlier Ki-67 requirement removed (3 March 2023)
  • ribociclib with an aromatase inhibitor for stage II or III early breast cancer at high risk of recurrence (17 September 2024). Guideline documents on this topic: ESMO early breast cancer (Loibl et al., 2024) and NCCN (JNCCN Insights, version 5.2025)
  • cited by reference only, with no recommendation summarised here
FDA 2022US Food and Drug Administration 2023US Food and Drug Administration 2024Guideline: Loibl S, André F, Bachelot T, Barrios CH, Bergh J, Burstein HJ, et al 2024Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2025
Adjuvant

HER2-positive early breast cancer, FDA regulatory actions:

  • trastuzumab within a regimen of doxorubicin, cyclophosphamide and paclitaxel for node-positive disease (16 November 2006)
  • ado-trastuzumab emtansine for residual invasive disease after neoadjuvant taxane and trastuzumab-based treatment (3 May 2019)
  • on 15 May 2026, trastuzumab deruxtecan as adjuvant treatment for residual invasive disease after neoadjuvant trastuzumab (with or without pertuzumab) and taxane-based treatment (supported by DESTINY-Breast05) and, as neoadjuvant treatment of stage II or III disease, trastuzumab deruxtecan followed by a taxane, trastuzumab and pertuzumab (supported by DESTINY-Breast11). Guideline documents on this topic: ESMO early breast cancer (Loibl et al., 2024) and NCCN (JNCCN Insights, version 5.2025)
  • cited by reference only, with no recommendation summarised here
US Food and Drug Administration 2006US Food and Drug Administration 2019FDA 2026Guideline: Loibl S, André F, Bachelot T, Barrios CH, Bergh J, Burstein HJ, et al 2024Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2025
Neoadjuvant

Early triple-negative breast cancer, FDA regulatory actions:

  • pembrolizumab with chemotherapy as neoadjuvant treatment, then as a single agent after surgery, for high-risk early-stage disease (26 July 2021)
  • olaparib as adjuvant treatment for germline BRCA-mutated, HER2-negative high-risk early breast cancer previously treated with neoadjuvant or adjuvant chemotherapy (11 March 2022
  • not restricted to triple-negative disease). Guideline documents on this topic: ESMO early breast cancer (Loibl et al., 2024) and NCCN (JNCCN Insights, version 5.2025)
  • cited by reference only, with no recommendation summarised here
FDA 2021FDA 2022Guideline: Loibl S, André F, Bachelot T, Barrios CH, Bergh J, Burstein HJ, et al 2024Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2025
Metastatic

Hormone receptor-positive, HER2-negative advanced or metastatic disease, FDA actions with endocrine therapy include:

  • palbociclib with letrozole as initial endocrine-based therapy (3 February 2015, accelerated approval)
  • ribociclib with an aromatase inhibitor as initial endocrine-based therapy (13 March 2017)
  • alpelisib with fulvestrant for PIK3CA-mutated disease after an endocrine-based regimen (24 May 2019)
  • capivasertib with fulvestrant for PIK3CA/AKT1/PTEN-altered disease after endocrine-based therapy (16 November 2023)
  • inavolisib with palbociclib and fulvestrant for endocrine-resistant, PIK3CA-mutated disease after recurrence on or after adjuvant endocrine therapy (10 October 2024)
  • and gedatolisib with fulvestrant, with or without palbociclib, for disease without a detected PIK3CA mutation after progression on at least one line of endocrine therapy (14 July 2026). EMA marketing authorisations were issued for capivasertib (17 June 2024) and inavolisib (18 July 2025)
US Food and Drug Administration 2015US Food and Drug Administration 2017US Food and Drug Administration 2019US Food and Drug Administration 2023US Food and Drug Administration 2024US Food and Drug Administration 2026European Medicines Agency 2024European Medicines Agency 2025
Metastatic

Hormone receptor-positive, HER2-negative advanced or metastatic disease, later lines and ESR1-directed therapy. For ESR1-mutated disease with progression after at least one line of endocrine therapy, FDA actions include elacestrant (27 January 2023), imlunestrant (25 September 2025), imlunestrant with abemaciclib (18 September 2026) and vepdegestrant (1 May 2026). Separately, camizestrant with a CDK4/6 inhibitor received FDA accelerated approval when an ESR1 mutation is detected during aromatase inhibitor and CDK4/6 inhibitor therapy (4 September 2026). Antibody-drug conjugates:

  • sacituzumab govitecan after endocrine-based therapy and at least two additional systemic therapies in the metastatic setting (3 February 2023)
  • datopotamab deruxtecan after prior endocrine-based therapy and chemotherapy (17 January 2025)
  • and trastuzumab deruxtecan for HER2-low or HER2-ultralow disease that progressed on endocrine therapy in the metastatic setting (27 January 2025). Guideline documents on this topic: ESMO metastatic breast cancer (de Azambuja et al., 2026), NCCN (JNCCN, version 4.2026) and ABC8 (Cardoso et al., 2026)
  • cited by reference only, with no recommendation summarised here
US Food and Drug Administration 2023FDA 2025FDA 2026FDA 2026FDA 2026FDA 2023FDA 2025US Food and Drug Administration 2025Guideline: de Azambuja E, Barrios CH, Bartsch R, Curigliano G, Dent R, Gennari A, et al 2026Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2026Guideline: Cardoso F, Ribeiro JM, Schumacher-Wulf E, Swain SM, Paluch-Shimon S, Penault-Llorca F, et al 2026
Metastatic

HER2-positive advanced or metastatic disease, FDA actions include:

  • trastuzumab deruxtecan with pertuzumab as first-line treatment (15 December 2025)
  • tucatinib with trastuzumab and pertuzumab as maintenance treatment of unresectable locally advanced or metastatic disease following induction treatment (7 October 2026)
  • for hormone receptor-positive disease, palbociclib with trastuzumab, with or without pertuzumab, and endocrine therapy as maintenance treatment following induction treatment (24 June 2026)
  • and trastuzumab deruxtecan after a prior anti-HER2 regimen (regular approval 4 May 2022). Guideline documents on this topic: ESMO metastatic breast cancer (de Azambuja et al., 2026), NCCN (JNCCN, version 4.2026) and ABC8 (Cardoso et al., 2026)
  • cited by reference only, with no recommendation summarised here
US Food and Drug Administration 2025FDA 2026FDA 2026US Food and Drug Administration 2022Guideline: de Azambuja E, Barrios CH, Bartsch R, Curigliano G, Dent R, Gennari A, et al 2026Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2026Guideline: Cardoso F, Ribeiro JM, Schumacher-Wulf E, Swain SM, Paluch-Shimon S, Penault-Llorca F, et al 2026
Metastatic

Triple-negative advanced or metastatic disease, FDA actions include:

  • pembrolizumab with chemotherapy for PD-L1 CPS 10 or more (accelerated approval 13 November 2020, converted 26 July 2021)
  • sacituzumab govitecan after two or more prior systemic therapies, at least one for metastatic disease (converted 7 April 2021)
  • datopotamab deruxtecan for people not candidates for PD-1 or PD-L1 inhibitor therapy (22 May 2026)
  • and first-line sacituzumab govitecan alone (not candidates for PD-1/PD-L1 inhibitor-based therapy) or with pembrolizumab (PD-L1 CPS 10 or more) (24 June 2026). Guideline documents on this topic: ESMO metastatic breast cancer (de Azambuja et al., 2026), NCCN (JNCCN, version 4.2026) and ABC8 (Cardoso et al., 2026)
  • cited by reference only, with no recommendation summarised here
FDA 2020FDA 2021FDA 2021FDA 2026FDA 2026Guideline: de Azambuja E, Barrios CH, Bartsch R, Curigliano G, Dent R, Gennari A, et al 2026Guideline: Gradishar WJ, Moran MS, Abraham J, Abramson V, Aft R, Agnese D, et al 2026Guideline: Cardoso F, Ribeiro JM, Schumacher-Wulf E, Swain SM, Paluch-Shimon S, Penault-Llorca F, et al 2026

Trials to watch

  • CAMBRIA-2 NCT05952557

    Does adjuvant camizestrant, with or without abemaciclib, improve invasive breast cancer-free survival compared with standard endocrine therapy (aromatase inhibitor or tamoxifen, with or without abemaciclib) in ER-positive, HER2-negative early breast cancer at intermediate-high or high risk of recurrence? Registry status: active, not recruiting.

  • EMBER-4 NCT05514054

    Does adjuvant imlunestrant improve invasive disease-free survival compared with standard adjuvant endocrine therapy in people with ER-positive, HER2-negative early breast cancer at increased risk of recurrence who have already taken endocrine therapy for 2 to 5 years? Registry status: active, not recruiting.

  • TROPION-Breast03 NCT05629585

    Does datopotamab deruxtecan, with or without durvalumab, improve invasive disease-free survival compared with investigator's choice therapy (capecitabine or pembrolizumab) in stage I to III triple-negative breast cancer with residual invasive disease after neoadjuvant therapy? Registry status: active, not recruiting.

  • ASCENT-05 NCT05633654

    Does adjuvant sacituzumab govitecan plus pembrolizumab improve invasive disease-free survival compared with treatment of physician's choice (pembrolizumab, or pembrolizumab plus capecitabine) in triple-negative breast cancer with residual invasive disease after neoadjuvant therapy and surgery? Registry status: recruiting.

  • OptimICE-pCR NCT05812807

    In early-stage triple-negative breast cancer with a pathologic complete response after neoadjuvant chemotherapy plus pembrolizumab, does continuing pembrolizumab differ from observation in recurrence-free survival? Registry status: recruiting.

  • VIKTORIA-2 NCT06757634

    As first-line treatment of HR-positive, HER2-negative advanced breast cancer, does gedatolisib with palbociclib and endocrine therapy (fulvestrant or letrozole) improve progression-free survival compared with ribociclib and endocrine therapy? Registry status: recruiting.

Pioneers

References

Show all 183 references

Loading references… Open the list as a page

Every figure was checked against its primary source to the extent possible. Items that could not be fully confirmed are marked “needs review”. Educational history, not medical advice.

Report an error

Links open the primary source. A flag marked ⚑ needs review means a value or claim has not yet been independently confirmed.

2021ReversalImmunotherapy

majorMetastaticNeeds review

Atezolizumab in PD-L1+ metastatic TNBC: approval, then withdrawal

In 2019 the US regulator gave early approval to atezolizumab with nab-paclitaxel for advanced triple-negative breast cancer with PD-L1 on immune cells, based on delayed progression. A second trial, using paclitaxel, did not show a benefit, and the indication was removed in October 2021. A different immunotherapy combination stayed approved for a differently defined group.

IMpassion130 (N=902; atezolizumab plus nab-paclitaxel): PD-L1-positive subgroup median PFS 7.5 vs 5.0 months (HR 0.62; 95% CI 0.49-0.78); ITT PFS 7.2 vs 5.5 months (HR 0.80); final ITT OS HR 0.87 (95% CI 0.75-1.02; P = 0.077). IMpassion131 (N=651; paclitaxel; PD-L1-positive): PFS 6.0 vs 5.7 months (HR 0.82; 95% CI 0.60-1.12; P = 0.20); OS HR 1.11. An FDA letter of 6 October 2021 approved removal of the indication.

Standard before

Chemotherapy was the control: nab-paclitaxel plus placebo in IMpassion130 and paclitaxel plus placebo in IMpassion131. FDA granted accelerated approval on 8 March 2019 on the basis of progression-free survival, with overall survival data immature.

What changed

On 6 October 2021 FDA approved a supplement removing the atezolizumab triple-negative breast cancer indication. Pembrolizumab with chemotherapy for PD-L1 CPS 10 or more (KEYNOTE-355) had been approved under accelerated approval on 13 November 2020 and was converted on 26 July 2021.

Why it mattered

It shows an accelerated-approval indication granted on a progression-free survival effect being removed within three years, in a supplement that also records fulfilment of a postmarketing requirement naming a second trial of the same drug (IMpassion131), which did not improve progression-free or overall survival. It also shows how PD-L1 assay and cut-off define different populations.

Caveats

  • IMpassion130 and IMpassion131 used different chemotherapy backbones (nab-paclitaxel versus paclitaxel) and different randomisation ratios (1:1 versus 2:1); the difference in outcome cannot be attributed to one cause.
  • The FDA atezolizumab indication used PD-L1 on tumor-infiltrating immune cells in at least 1% of the tumor area (SP142); the pembrolizumab indication used a combined positive score of 10 or more. The populations and assays differ, so KEYNOTE-355 results (PFS 9.7 vs 5.6 months, CPS 10 or more) are context and not a head-to-head comparison.
  • In IMpassion130 the PD-L1-positive OS results (25.0 vs 15.5 months at the interim analysis; 25.4 vs 17.9 months at final analysis) were not formally tested because the ITT overall-survival difference was not significant (final HR 0.87; P = 0.077).
  • The FDA letter records that the sponsor's supplement (dated 2 September 2021) provides for removal of the indication and that the final report for the postmarketing requirement naming IMpassion131 was received; it does not discuss the trial results or state a reason. Unlike the bevacizumab removal, the letter shows a sponsor-submitted supplement, not an FDA-initiated withdrawal.