Reviews/Perioperative Immunotherapy Solid Tumours/Immune checkpoint inhibitor neoadjuvant benefit inconsistent in triple-negative breast cancer
Tnbc · PERIO-TNBC-03 · v2
The claim argues that the benefit of adding immunotherapy to chemotherapy before surgery for triple-negative breast cancer depends on the specific drug, and that not every such combination improves long-term outcomes even when it shrinks tumours more completely. The corpus shows clearly that adding the immunotherapy drug atezolizumab raises the rate of complete tumour clearance ([2], [6]), and it describes pembrolizumab as the established standard of care ([8]). But the only long-term outcome figure that compares an immunotherapy arm against a control, from the IMpassion031 trial [6], actually leans in favour of the drug rather than showing it fails, and the large dedicated trial designed to answer the long-term question [16] reports no result. Across many different immunotherapy agents the tumour-clearance benefit looks broadly similar ([2], [3], [9], [11], [23]), which reads more like a shared class effect than a drug-specific one. Because the corpus has no case where one agent clearly succeeds on long-term outcome while another clearly fails, it cannot demonstrate the drug-specific-versus-class contrast the claim asserts. The panel therefore leaned against the claim.
Randomised data show atezolizumab added to chemotherapy increases pathologic complete response, from 18.8% to 55.6% in NCI 10013 [2] and as the met primary endpoint of phase 3 IMpassion031 [6]. The only comparative long-term figure in the corpus is IMpassion031's descriptive secondary event-free survival hazard ratio of 0.76 (95% CI 0.47-1.21), which favours atezolizumab and crosses 1 [6]. Pathologic complete response gains recur across pembrolizumab [3], tislelizumab [9], sintilimab-based regimens [11] and the bispecific cadonilimab [23], a broadly consistent pattern. Pembrolizumab's event-free survival benefit is named only as standard-of-care context from KEYNOTE-522 [8], not as corpus data, and the pivotal atezolizumab EFS trial NSABP B-59/GeparDouze [16] reports no efficacy result. The contested point: The division is whether the corpus is merely insufficient to establish agent-specificity or actively points the other way. The single comparative event-free survival datum [6] favours atezolizumab rather than showing failure, and no corpus trial pairs a succeeding agent against a failing one.
All 27 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Perioperative Immunotherapy Solid Tumours, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The corpus contains no positive primary EFS result for any ICI agent. Pembrolizumab's EFS benefit (KEYNOTE-522) appears only as 'standard of care' context ([8]), never as corpus data. Without a positive agent's EFS result, agent-specificity is unestablishable.
The single comparative EFS datum, [6] IMpassion031, FAVOURED atezolizumab (HR 0.76) — an underpowered descriptive secondary endpoint. It does not show atezolizumab fails to improve EFS, so it cannot support 'not all ICI+chemo improve EFS'; if anything it is consistent with benefit.
Where it remains uncertain. The one pivotal atezolizumab event-free survival trial [16] reports no result and pembrolizumab's EFS success appears only as standard-of-care context [8], so no positive comparator EFS exists to anchor agent-specificity.
Aleatory
IMpassion031's event-free survival estimate rests on few events, giving a hazard ratio of 0.76 with a 95% CI of 0.47-1.21 that crosses 1 [6]; single-arm trials enrol tens of patients ([3], [9], [11], [23]).
Epistemic
NSABP B-59/GeparDouze is the phase 3 trial with event-free survival as primary endpoint, but the corpus abstract reports no efficacy result [16].
Inferential
The corpus demonstrates pCR gains but the claim concerns event-free survival; inferring the EFS pattern from pCR is not supported for most agents here.
The load-bearing papers from a corpus of 27, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Randomized evidence that atezolizumab added to chemotherapy improves pCR
“the pCR rate was 18.8% (95% CI 4.0-45.6%) in Arm A, and 55.6% (95% CI 40.0-70.4%) in Arm B (estimated treatment difference: 36.8%, 95% CI 8.5-56.6%; p = 0.018).”
Panel auditInternal validity: medium · scope of inference: pCR and TIL effect of adding atezolizumab to carboplatin/paclitaxel in stage II-III TNBC.Key limitations: no EFS · short follow-up · small sample · no anthracycline in control
Only comparative EFS datum in corpus: non-significant (CI crosses 1) but favouring atezolizumab despite pCR improvement
“Long-term outcomes favored the atezolizumab group (event-free survival hazard ratio (HR), 0.76; 95% confidence interval (CI), 0.47-1.21; disease-free survival HR, 0.76; 95% CI, 0.44-1.30; overall survival HR, 0.56; 95% CI, 0.30-1.04).”
Panel auditInternal validity: high · scope of inference: Long-term event-free, disease-free and overall survival of adding atezolizumab to chemotherapy in stage II/III TNBC.Key limitations: EFS descriptive/underpowered secondary endpoint · wide CI · moderate sample
Pembrolizumab established as standard of care — but as context only; no primary EFS result for the positive agent is in the corpus
“The addition of pembrolizumab (pembro) to neoadjuvant chemotherapy (NAC) is standard of care for the treatment of early triple-negative breast cancer (TNBC) after KEYNOTE-522 trial demonstrated improved pathologic complete response (pCR) rates with the combination.”
Panel auditInternal validity: low · scope of inference: Clinical predictors of pCR with neoadjuvant chemotherapy plus pembrolizumab in real-world TNBC.Key limitations: retrospective · single agent only · no EFS result · no comparator
Single-arm design gives a pCR rate but no randomized comparison and no event-free survival, so it cannot speak to EFS benefit or agent-specificity.
Panel auditInternal validity: low · scope of inference: pCR achievable with pembrolizumab plus nab-paclitaxel/anthracycline in early TNBC.Key limitations: single-arm · no EFS · no comparator · protocol amended mid-trial
4 of 27 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The corpus carries broad pCR benefit and one equivocal EFS HR but no positive comparator EFS and no reported null EFS, so it leans against affirming agent-specificity and is more consistent with an unconfirmed class effect.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Reported head-to-head EFS by agent. Agent-specificity cannot be judged without comparative event-free survival outcomes for more than one agent; the corpus has only one such figure, which favours atezolizumab [6]. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Report the primary event-free survival analyses of the dedicated phase 3 trials (e.g. the awaited NSABP B-59/GeparDouze result [16]) and compare across agents with matched endpoints.
Open question: Does any ICI plus chemotherapy regimen improve pCR without improving event-free survival? Why it is still open: The only comparative EFS figure in the corpus [6] favours atezolizumab and the pivotal EFS trial [16] reports no result, so no discordant case is present.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19