Reviews/Targeted Therapy Prostate Cancer/Pembrolizumab activity in microsatellite instability-high or high tumour mutational burden prostate cancer
Resistance · PC-RES-03 · v2
Borderline: within 4 points of the Likely/Uncertain boundary, so a re-grade could land either side.
The claim bundles two patient groups together: those whose prostate tumours have a DNA-repair defect (MSI-H) and those with a high number of mutations (TMB-H). For the first group, three separate prostate datasets show that a meaningful share of patients respond to this class of immunotherapy [2][3][10], and broad trials across many cancer types confirm that MSI-H tumours tend to respond [1][4]. For the second group, the evidence points the other way: the one prostate study that separated high-mutation tumours without the repair defect found no measurable tumour shrinkage [3], and a large pooled analysis found high-mutation prostate tumours rarely respond [7]. Much of the positive signal is based on a blood-marker (PSA) response rather than imaging or survival, and the prostate studies are small and look back at past records. The panel treated the two halves differently rather than blending them, which is why the overall position sits in a middle band.
In MSI-H/dMMR metastatic castration-resistant prostate cancer, Graham [2] reported PSA50 responses in 8 of 15 evaluable (53%), Lenis [3] reported about 45% RECIST and 65% PSA50 responses, and Abida [10] reported that 6 of 11 ICB-treated patients (54.5%) had a greater than 50% PSA decline with 4 radiographic responses. Pan-tumour MSI-H/dMMR trials support general sensitivity, with an objective response rate of 34.3% in KEYNOTE-158 [1] and 30.8% in its cohort K update [4]. Against this, the TMB-H half is not supported: no patient with TMB-H/MSS prostate cancer had a RECIST response in [3], and McGrail [7] found high-TMB prostate tumours reached an ORR of only 15.3% (95% CI 9.2-23.4). Unselected mCRPC responds poorly, at 5% and 3% in KEYNOTE-199 cohorts 1 and 2 [9], underscoring why biomarker selection matters. The contested point: The split runs along the biomarker conjunction: the MSI-H/dMMR arm is independently replicated [2][3][10] while the TMB-H arm is contradicted by the one paper that isolates it [3] and by the pooled prostate data in [7]. A second fault line is whether PSA50/PSA decline, a biochemical surrogate, suffices to call the activity 'clinically meaningful' [2][10].
All 15 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 Targeted Therapy Prostate Cancer, 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 claim's 'or TMB-H' is contradicted by the corpus: TMB-H/MSS prostate tumours had 0% RECIST response [3] and prostate TMB-H failed to reach 20% ORR (15.3%) [7]. The conjunction overreaches on the TMB-H arm.
Much of the claimed 'antitumour activity' rests on PSA50/PSA decline (a biochemical surrogate) rather than RECIST or overall survival [2][10], weakening the 'clinically meaningful' assertion.
Where it remains uncertain. No TMB-H/MSS prostate patient had a RECIST response [3] and high-TMB prostate tumours reached only 15.3% ORR [7], so the 'or TMB-H' clause overreaches.
Aleatory
Response estimates rest on single-digit events (8 of 15 [2]; 6 of 11 [10]), so chance alone can move the observed rate substantially.
Epistemic
Only one paper isolates TMB-H/MSS prostate (8 ICB-treated [3]) plus a pooled signal [7]; the arm is contradicted but on very limited direct data.
Inferential
Much 'activity' is PSA50/PSA decline [2][10], a biochemical surrogate, not RECIST or overall survival, so inferring clinical meaningfulness is a step beyond the data.
The load-bearing papers from a corpus of 15, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Pan-tumor phase II evidence that MSI-H/dMMR tumours are pembrolizumab-sensitive, supporting biological plausibility for the MSI-H arm.
“Objective response rate was 34.3% (95% CI, 28.3% to 40.8%).”
Panel auditInternal validity: medium · scope of inference: Previously treated advanced noncolorectal MSI-H/dMMR solid tumours (endometrial, gastric, cholangiocarcinoma, pancreatic predominating); not prostate.Key limitations: no prostate cohort · single-arm · pan-tumour pooled ORR
Direct evidence of meaningful pembrolizumab activity in MSI-H/MMRd mCRPC.
“PSA50 responses to pembrolizumab occurred in 8 (53%) men.”
Panel auditInternal validity: low · scope of inference: MMRd/MSI-high metastatic prostate cancer patients treated with pembrolizumab at two academic centres.Key limitations: small n · retrospective · no comparator · PSA50 endpoint not RECIST
Confirms activity in the MSI-H arm by RECIST while isolating the TMB-H/MSS arm, which showed no RECIST response.
“About 45% of patients with MSI-H/dMMR had a RECIST response, and 65% had a PSA50 response. No patient with TMB-H/MSS had a RECIST response, and 50% had a PSA50 response.”
Panel auditInternal validity: medium · scope of inference: Prostate cancer patients with MSI-H/dMMR or TMB-H/MSS receiving immune checkpoint blockade at one centre.Key limitations: ICB not pembrolizumab-specific · small treated subgroups · retrospective · TMB-H/MSS no RECIST response
In prostate and similar cancers TMB-H tumours failed to reach a 20% ORR (15.3%) and did not outperform TMB-L, arguing against meaningful TMB-H-driven activity.
Panel auditInternal validity: medium · scope of inference: ICB response in TMB-H tumours across cancer types where CD8 and neoantigen load are uncorrelated, including prostate cancer.Key limitations: prostate pooled with breast/glioma · ICB not pembrolizumab-specific · retrospective · not restricted to castration-resistant disease
4 of 15 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 robust, independently replicated MSI-H signal carries the claim into the Likely band, but the contradicted TMB-H conjunction, surrogate endpoints and small retrospective samples cap confidence at 0.70.
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.
TMB-H mCRPC direct evidence. The claim's TMB-H clause rests on contradiction from one small arm [3] and a pooled analysis [7]; whether any TMB-H/MSS prostate subset benefits is unresolved. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: Prospective pembrolizumab trial enrolling TMB-H/MSS mCRPC with RECIST and OS endpoints, powered to detect or exclude a 20% ORR.
Open question: Does any TMB-H/MSS mCRPC subset derive RECIST-confirmed benefit from pembrolizumab? Why it is still open: The only isolating paper found no RECIST response in TMB-H/MSS prostate [3] and pooled prostate TMB-H reached just 15.3% ORR [7], but both are small or indirect.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 26 Oct 2026 · protocol v0.19