Synthesis: KEYNOTE-522 and the neoadjuvant chemo+IO standard

Evidence grades (GRADE-adapted): A high — multiple well-conducted RCTs or systematic reviews converge. B moderate — single pivotal RCT or consistent observational evidence. C limited — single observational study, mechanistic, or expert consensus. D preclinical / hypothesis-generating.

The pre-KEYNOTE-522 landscape

Through the mid-2010s, neoadjuvant therapy for stage II–III triple-negative breast cancer (TNBC) centered on anthracycline + taxane regimens with optional platinum addition. Two questions dominated trial design:

• Does adding carboplatin to standard chemotherapy improve outcomes in TNBC? GeparSixto reported that adding carboplatin to non-pegylated liposomal doxorubicin + paclitaxel + bevacizumab raised the pCR rate from 36.9% to 53.2% in TNBC (n=315 in the TNBC subset)^[1]A. CALGB 40603 in parallel reported a pCR boost from 41% to 54% with carboplatin addition to weekly paclitaxel → dose-dense AC, though the magnitude of long-term-outcome benefit was less clear^[2]B. These results made carboplatin a common but not universal addition to TNBC neoadjuvant regimens by 2018.
• What about PARP inhibitor or immunotherapy addition? BrighTNess tested adding veliparib + carboplatin (or carboplatin alone) to standard paclitaxel → AC and confirmed the carboplatin pCR benefit while showing no additional benefit from veliparib^[3]A. GeparNuevo tested adding durvalumab and reported a positive pCR signal in a planned subgroup (BMI < 25 or window-of-opportunity arm) but a non-significant primary endpoint^[4]B.

Two important framing observations from this era. First, pCR after neoadjuvant chemotherapy in TNBC was already known to strongly correlate with long-term outcomes via the 2014 CTNeoBC pooled analysis (event-free survival HR 0.24 for pCR vs no-pCR in TNBC across 12 trials)^[5]A. Second, residual cancer burden (RCB) classification after surgery was shown to be an even finer-grained prognostic stratifier than pCR alone (Symmans 2017)^[6]A. The trial-design community therefore had high confidence that a regimen that improved pCR would also improve long-term outcomes — but the trial-by-trial pCR boosts from platinum addition were not always followed by survival benefits, raising the question of whether chemotherapy intensification alone was hitting a ceiling.

KEYNOTE-522 — design

Schmid and colleagues randomized 1,174 patients with previously untreated stage II or III TNBC 2:1 to pembrolizumab + chemotherapy vs placebo + chemotherapy^[7]A. The chemotherapy backbone was a sequential two-phase regimen:

• Phase 1 (12 weeks): paclitaxel (80 mg/m² weekly) + carboplatin (AUC 5 every 3 weeks, or AUC 1.5 weekly).
• Phase 2 (12 weeks): doxorubicin or epirubicin + cyclophosphamide (AC or EC) every 3 weeks.
• Pembrolizumab or placebo (200 mg every 3 weeks) overlaid across both phases for a total of 8 doses neoadjuvantly.
• Surgery after completion of neoadjuvant therapy.
• Adjuvant pembrolizumab or placebo continued for an additional 9 doses (27 weeks) post-surgery for a total of approximately 1 year of pembrolizumab exposure.

Co-primary endpoints: pathologic complete response (pCR, defined as ypT0/Tis ypN0) and event-free survival (EFS). pCR was assessed at the time of definitive surgery; EFS over time. Stratification factors: nodal status, tumor size, choice of carboplatin schedule. Importantly, PD-L1 status was not used for enrollment selection — the trial was all-comer within stage II–III TNBC.

pCR results — the headline 2020 NEJM publication

The first interim analysis at a median follow-up of 15.5 months reported the pCR primary endpoint result^[7]:

• pCR in the pembrolizumab arm: 64.8% (260 of 401 patients).
• pCR in the placebo arm: 51.2% (103 of 201 patients).
• Absolute difference: 13.6 percentage points (95% CI 5.4 to 21.8; p=0.00055).

The pCR benefit was observed across all pre-specified subgroups including nodal status, tumor stage, and carboplatin schedule. The PD-L1 subgroup analysis showed numerical benefit in both PD-L1-positive and PD-L1-negative tumors, with no clear interaction — suggesting that PD-L1 status, at least at the assay-level used, is not a useful biomarker for stratifying the early-stage IO benefit (in contrast to its predictive value in metastatic disease).

EFS update — the 2022 NEJM publication

The second pre-specified interim analysis at median follow-up 39.1 months reported the EFS co-primary endpoint^[8]A:

• 36-month EFS in the pembrolizumab arm: 84.5%
• 36-month EFS in the placebo arm: 76.8%
• EFS hazard ratio: 0.63 (95% CI 0.48 to 0.82; p<0.001)

The EFS benefit was concentrated in the patients with residual disease at surgery, an unsurprising but important observation: patients who achieved pCR had excellent EFS in both arms (94.4% vs 92.5% at 36 months — numerically marginal), while patients with residual disease had a much larger absolute benefit from pembrolizumab (67.4% vs 56.8%). This pattern motivated several follow-on questions about whether the adjuvant pembrolizumab continuation provides meaningful benefit in pCR achievers (where the recurrence risk is already low) versus residual-disease patients (where it is much higher).

Adverse events and toxicity profile

Grade 3 or higher treatment-related adverse events occurred in 76.8% of the pembrolizumab arm vs 72.2% of the placebo arm, predominantly chemotherapy-attributable (neutropenia, anemia, febrile neutropenia)^[7]. Immune-mediated adverse events specifically were more frequent with pembrolizumab (43.6% vs 21.9%) but mostly grade 1–2; the most common were skin reactions (rash, pruritus), thyroid dysfunction, and infusion reactions.

Serious immune-mediated events — pneumonitis, colitis, hepatitis, adrenal insufficiency — occurred in low single-digit percentages but require active monitoring and prompt management; the ASCO guideline for immune-related adverse event (irAE) management is the operational standard^[9]A. Treatment discontinuation due to adverse events was higher with pembrolizumab (27.7% vs 14.1%), and the practical implication for community oncology practices is that running KEYNOTE-522 regimens at scale requires endocrine and rheumatology support that wasn't always co-located before. This is one of the implementation challenges that has slowed uniform adoption.

Contemporaneous trials — why pembrolizumab and not atezolizumab or durvalumab

Several other immune checkpoint inhibitors were tested in early-stage TNBC during the same window with discordant results:

• IMpassion031 (atezolizumab + chemo, n=333) — reported a positive pCR result (58% vs 41%, absolute 17 points; HR for the binary endpoint not directly comparable)^[10]A. Long-term EFS data were less favorable than KEYNOTE-522, and atezolizumab's withdrawal from the metastatic TNBC indication in the US (following IMpassion131's failure) curtailed development in the early-stage setting.
• NeoTRIPaPDL1 (atezolizumab + chemo, n=280) — did not meet its primary EFS endpoint at 5-year follow-up, despite a numerically higher pCR (43.5% vs 40.8%)^[11]B. The chemotherapy backbone (carboplatin + nab-paclitaxel without anthracycline) differed from KEYNOTE-522's, and this is hypothesized as one reason for the discordance — the anthracycline component may be necessary for the IO benefit to materialize.
• GeparNuevo (durvalumab + chemo, n=174) — numerically positive pCR with non-significant primary endpoint at initial analysis^[4], but a long-term follow-up reported substantial iDFS benefit with durvalumab (HR 0.48)^[12]B. Notably, durvalumab in GeparNuevo was given only in the neoadjuvant setting, not as adjuvant continuation; this divergence has reopened the question of whether adjuvant IO continuation is necessary or whether a shorter, neoadjuvant-only IO course would suffice.

The aggregate read: the IO + chemo benefit in early-stage TNBC appears real but is sensitive to the specific IO agent, chemotherapy backbone, and timing of IO exposure. KEYNOTE-522 became the standard because it had the largest patient population, the most pre-specified-endpoint-positive results across both pCR and EFS, and an early FDA approval (July 2021).

Evidence table — pivotal early-stage TNBC neoadjuvant trials

Standard-of-care integration

KEYNOTE-522 was incorporated into the major guidelines on a similar timeline:

• FDA approval for high-risk early-stage TNBC: July 2021 (pCR primary-endpoint based).
• NCCN: incorporated as preferred regimen in version 4.2021 (TNBC stage II–III).
• ESMO Clinical Practice Guidelines (Cardoso 2019 with subsequent updates): incorporated the KEYNOTE-522 regimen as standard of care for stage II–III TNBC^[13]A.
• St. Gallen consensus 2023: endorsed the regimen with notes on de-escalation research priorities.

For patients who do not achieve pCR, the addition of adjuvant capecitabine (per the pre-IO-era CREATE-X trial in residual TNBC after standard neoadjuvant chemo)^[14]A is widely offered in addition to the protocol-specified adjuvant pembrolizumab continuation; the evidence for capecitabine specifically after the KEYNOTE-522 regimen is still maturing but the precedent is strong enough that most centers offer it. For germline-BRCA-mutated patients with high-risk residual disease, adjuvant olaparib for 1 year per OlympiA is also offered.

Open questions and active investigation

• Is carboplatin necessary if pembrolizumab is added? KEYNOTE-522 included carboplatin in the chemo backbone, so the trial design cannot isolate the carboplatin contribution from the IO contribution. NeoTRIPaPDL1's anthracycline-free, IO-positive regimen suggests carboplatin may not be sufficient on its own to support IO benefit; whether anthracycline + IO without carboplatin (or carboplatin + IO without anthracycline) is comparable to the full KEYNOTE-522 regimen is being tested in several smaller trials. The clinical-utility question is real because carboplatin contributes meaningfully to neuropathy, marrow toxicity, and discontinuation rates.
• Is the anthracycline component necessary? The 2010s saw substantial interest in anthracycline-sparing regimens to reduce long-term cardiotoxicity. KEYNOTE-522's regimen retains AC/EC, which complicates de-escalation discussions. Trials testing taxane + carboplatin + pembrolizumab without anthracycline are ongoing; results in the next 2–3 years.
• Is adjuvant pembrolizumab continuation necessary for pCR achievers? The 2022 EFS analysis suggested most of the IO benefit accrues in residual-disease patients; pCR achievers had near-equivalent EFS regardless of arm. Several de-escalation trials (including OptimICE-pCR and DESCARTES) are testing whether the 9-dose adjuvant continuation can be safely omitted in pCR achievers. This is the highest-impact de-escalation question because the adjuvant phase is the source of substantial late immune-related adverse events.
• How to manage residual disease in the IO era. Adjuvant capecitabine (CREATE-X precedent), adjuvant olaparib (OlympiA, in BRCA-mutated), or adjuvant ADC (sacituzumab govitecan in residual TNBC, tested in ASCENT-04 and others) are all candidates; sequencing and combination are not yet established. The TRADE/SASCIA trials are testing adjuvant sacituzumab govitecan in residual TNBC.
• Predictive biomarkers beyond stage. PD-L1 did not predict the KEYNOTE-522 benefit. TIL density, gene-expression signatures, and tumor mutational burden have been tested retrospectively with mixed signals; none is currently used to gate enrollment or stratify treatment selection. The IM (Lehmann) / BLIA (Burstein) subtypes were retrospectively associated with higher pCR in some analyses but have not been prospectively validated for treatment-selection use.
• Treatment of patients who progress on neoadjuvant therapy. The trial protocol did not pre-specify management for patients who progressed during neoadjuvant treatment (a small but high-risk subgroup); current practice varies widely. A dedicated registry or trial for this subgroup would be valuable.
• Implementation in resource-limited settings. Pembrolizumab's per-cycle cost (~$10,000 in the US; lower in some international markets via tendering) is a substantial barrier to global rollout. Biosimilar pembrolizumab is in development; first approvals expected in the late 2020s. WHO Essential Medicines List inclusion would substantially accelerate global access.

For the patient-layer companion to this trial discussion, see Treatment options and the immunotherapy section of What is TNBC?. For the metastatic-setting pembrolizumab evidence base (KEYNOTE-355), the metastatic-TNBC first-line synthesis is in domain E. For the PD-L1 assay heterogeneity that complicates cross-trial interpretation of IO-related results, see PD-L1 testing and assay heterogeneity.

References

Each citation links to the original publication via DOI. The same records are searchable in the evidence library by title or DOI.

1. von Minckwitz G, Schneeweiss A, Loibl S, et al. Neoadjuvant carboplatin in patients with triple-negative and HER2-positive early breast cancer (GeparSixto; GBG 66): a randomised phase 2 trial. Lancet Oncol. 2014;15(7):747–756. doi:10.1016/S1470-2045(14)70160-3. ↩
2. Sikov WM, Berry DA, Perou CM, et al. Impact of the addition of carboplatin and/or bevacizumab to neoadjuvant once-per-week paclitaxel followed by dose-dense doxorubicin and cyclophosphamide on pathologic complete response rates in stage II to III triple-negative breast cancer: CALGB 40603 (Alliance). J Clin Oncol. 2015;33(1):13–21. doi:10.1200/JCO.2014.57.0572. ↩
3. Loibl S, O'Shaughnessy J, Untch M, et al. Addition of the PARP inhibitor veliparib plus carboplatin or carboplatin alone to standard neoadjuvant chemotherapy in triple-negative breast cancer (BrighTNess): a randomised, phase 3 trial. Lancet Oncol. 2018;19(4):497–509. doi:10.1016/S1470-2045(18)30111-6. ↩
4. Loibl S, Untch M, Burchardi N, et al. A randomised phase II study investigating durvalumab in addition to an anthracycline taxane-based neoadjuvant therapy in early triple-negative breast cancer: clinical results and biomarker analysis of GeparNuevo study. Ann Oncol. 2019;30(8):1279–1288. doi:10.1093/annonc/mdz158. ↩
5. Cortazar P, Zhang L, Untch M, et al. Pathological complete response and long-term clinical benefit in breast cancer: the CTNeoBC pooled analysis. Lancet. 2014;384(9938):164–172. doi:10.1016/S0140-6736(13)62422-8. ↩
6. Symmans WF, Wei C, Gould R, et al. Long-Term Prognostic Risk After Neoadjuvant Chemotherapy Associated With Residual Cancer Burden Classification. J Clin Oncol. 2017;35(10):1049–1060. doi:10.1200/JCO.2015.63.1010. ↩
7. Schmid P, Cortes J, Pusztai L, et al. Pembrolizumab for Early Triple-Negative Breast Cancer (KEYNOTE-522). N Engl J Med. 2020;382(9):810–821. doi:10.1056/NEJMoa1910549. ↩
8. Schmid P, Cortes J, Dent R, et al. Event-free Survival with Pembrolizumab in Early Triple-Negative Breast Cancer (KEYNOTE-522). N Engl J Med. 2022;386(6):556–567. doi:10.1056/NEJMoa2112651. ↩
9. Brahmer JR, Lacchetti C, Schneider BJ, et al. Management of Immune-Related Adverse Events in Patients Treated With Immune Checkpoint Inhibitor Therapy: ASCO Clinical Practice Guideline. J Clin Oncol. 2018;36(17):1714–1768. doi:10.1200/JCO.2017.77.6385. ↩
10. Mittendorf EA, Zhang H, Barrios CH, et al. Neoadjuvant atezolizumab in combination with sequential nab-paclitaxel and anthracycline-based chemotherapy versus placebo and chemotherapy in patients with early-stage triple-negative breast cancer (IMpassion031): a randomised, double-blind, phase 3 trial. Lancet. 2020;396(10257):1090–1100. doi:10.1016/S0140-6736(20)31953-X. ↩
11. Gianni L, Huang CS, Egle D, et al. Pathologic complete response and outcomes with neoadjuvant atezolizumab + carboplatin + nab-paclitaxel in triple-negative breast cancer: 5-year analysis of the NeoTRIPaPDL1 trial. Ann Oncol. 2022;33(5):534–543. doi:10.1016/j.annonc.2022.02.004. ↩
12. Loibl S, Schneeweiss A, Huober J, et al. Neoadjuvant durvalumab improves survival in early triple-negative breast cancer independent of pathological complete response. Ann Oncol. 2022;33(11):1149–1158. doi:10.1016/j.annonc.2022.07.1940. ↩
13. Cardoso F, Kyriakides S, Ohno S, et al. Early breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2019;30(8):1194–1220. doi:10.1093/annonc/mdz173. ↩
14. Masuda N, Lee SJ, Ohtani S, et al. Adjuvant Capecitabine for Breast Cancer after Preoperative Chemotherapy (CREATE-X). N Engl J Med. 2017;376(22):2147–2159. doi:10.1056/NEJMoa1612645. ↩

Last reviewed: 2026-05-31. Researcher-layer synthesis page. Evidence grades follow the GRADE-adapted rubric defined at the top of this page. Citations are anchored to the full bibliographic entries above; click the ↩ arrow next to any reference to return to its first citation in the prose.