FDA grants accelerated approval to dostarlimab in dMMR/MSI-H locally advanced rectal cancer: can cCR12-based organ preservation move to the bladder?

Data as of 2026-10-21 (Taipei time). [Fact] statements are sourced; [Inference] statements are this article's judgment.

1. One-line conclusion

[Fact] On the basis of single-arm "12-month sustained clinical complete response (cCR12)" data, FDA granted accelerated approval to a regimen that may spare patients chemotherapy, radiotherapy and surgery (FDA). [Inference] This sets a regulatory template for an "immunotherapy + watch-and-wait" organ-preservation path. For GU, the template can inform bladder-preservation trials, but dMMR is rare in bladder cancer (in one study, MMR IHC loss was 1% in bladder urothelial carcinoma and 9% in UTUC, PubMed), so the population for a direct copy is small.

2. What happened

[Fact] On 2026-10-19 US Eastern, FDA granted accelerated approval to dostarlimab (Jemperli, GSK) for adults with MSI-H or dMMR stage II/III locally advanced rectal cancer, as determined by an FDA-approved solid tumor test (FDA).

• Trial: AZUR-1 (NCT05723562), multicenter, single-arm, open-label, 154 untreated patients; dostarlimab 500 mg every 3 weeks for 9 cycles (FDA).
• Assessment: at week 6, week 12 and about 3 weeks after end of treatment, by endoscopy, rectal MRI and chest/abdomen/pelvis CT; patients with cCR entered nonoperative management, followed every 4 months for the first 2 years and every 6 months thereafter (FDA).
• Primary endpoint: cCR12 by independent central review (ICR) (FDA).
• Results: among 116 ICR-evaluable patients, cCR12 was 47% (95% CI 27%–56%, as printed by FDA); another 38 (33%) had reached cCR but had less than 12 months of follow-up, and 44 (38%) had near-complete response; at a median follow-up of 26.3 months, no patient with cCR12, cCR or nCR had disease progression. ICR ORR was 79% (95% CI 71%–85%). Among 144 investigator-assessed patients, cCR12 was 68% (95% CI 60%–76%) and ORR 82% (FDA).
• Treatment course: of 154 patients, 110 (71%) received only dostarlimab, with no other treatment and no progression; 26 (17%) underwent surgery (FDA).
• Supportive trial: investigator-initiated multicenter single-arm trial NCT04165772, 50 patients, cCR12 93% (95% CI 82%–99%), ORR 100% (FDA).
• Conditions of approval: AZUR-1 must be completed postmarketing, describing cCR12/24/36, 3-year organ preservation, 3-year recurrence or metastasis rate and 5-year survival, with expected completion 2030-07-31 (FDA accelerated approvals list).
• Review: Project Orbis, RTOR, Priority Review and Breakthrough Therapy; approved 4 months ahead of the goal date (FDA). The application was also part of FDA's Commissioner's National Priority Review Voucher (CNPV) pilot program (FDA). [Fact] FDA estimates about 3% of locally advanced rectal cancers are dMMR/MSI-H (FDA press release).
• Single-arm trial; no HR or p value.

3. Why it matters

• [Inference] Endpoint innovation: cCR12 is a surrogate endpoint for organ preservation, and FDA accepted it from a single arm, on the condition of a very large effect (dMMR tumors respond very strongly to PD-1) and postmarketing requirements for long-term follow-up.
• [Inference] Large gap between ICR and investigator assessment: 47% versus 68% (FDA). cCR assessment (endoscopy + MRI) is highly subjective, a problem every organ-preservation trial must face.
• [Inference] Biomarker-driven: efficacy rests on dMMR, a highly predictive marker; in populations without such a marker, single-arm cCR is unlikely to be accepted.
• Safety: FDA lists warnings including immune-mediated adverse reactions and infusion reactions (FDA); AZUR-1 adverse event rates by grade: not found.

4. Regulatory and commercial implications

• FDA: [Inference] This is a precedent for an organ-preservation indication; neoadjuvant immunotherapy trials in other dMMR solid tumors (e.g., gastroesophageal, gynecologic) can cite cCR-type endpoints.
• NMPA: China review status not found; Orbis partners were Canada, Israel, Switzerland and the UK (FDA).
• SoC: [Inference] The standard for dMMR rectal cancer may shift from total neoadjuvant therapy plus surgery to immunotherapy plus watch-and-wait; universal MMR testing becomes a necessary step.
• For the next trial: [Inference] single-arm cCR endpoints are acceptable only with a very-high-effect biomarker population and committed long-term follow-up; organ-preservation trials lacking either should assume a randomized design.

5. GU implications and OncoLattice recommendations

All recommendations below are [Inference]; cited facts carry their own sources.

1. Upper tract urothelial carcinoma (UTUC) and Lynch: [Fact] UTUC is the third most common malignancy in Lynch syndrome; in one study MMR IHC loss was 9% in UTUC (n=117) versus 1% in bladder urothelial carcinoma (n=160) (PubMed). In a germline-testing registry, 32% of tested UTUC patients carried a pathogenic MMR variant versus 6.1% of bladder-only patients, with MSH2 variants in 27% of UTUC patients (full text, Results section 3.1, Eur Urol Open Sci full text; a referred-for-testing population, so selection bias applies). [Inference] Nephroureterectomy has a large impact on renal function, so "immunotherapy + observation" in dMMR UTUC is an organ-preservation trial worth designing.
2. Bladder-preservation trial design:
   • The primary endpoint can follow the cCR12 structure, with a multimodal definition using cystoscopy, biopsy, urine cytology and MRI (e.g., VI-RADS), read centrally.
   • Predefine follow-up frequency during observation and the triggers for salvage cystectomy.
   • Postmarketing requirements can be expected to include 24/36-month durable response and organ-preservation rates; plan 5-year follow-up from the outset.
3. ICR capability: sites must be able to upload standardized cystoscopy images and MRI, which was key to AZUR-1's ICR-based primary endpoint.
4. Screening funnel: dMMR is rare, so enrollment depends on universal testing or referral from Lynch clinics; select sites with genetic counseling and pathology departments that run MMR IHC routinely.
5. Patient acceptance: watch-and-wait requires patients to accept intensive follow-up; sites need dedicated coordinators.

6. Unknowns and what to watch

• AZUR-1 cCR24, cCR36 and 3-year organ preservation rate.
• The correct figures for the CI around ICR 47%.
• Timing and venue of the full data publication.
• Whether any sponsor launches a similar design in dMMR UTUC or bladder cancer.

Evidence limitations

Both trials are single-arm; the ICR–investigator gap is large; cCR12 is a surrogate endpoint, and long-term recurrence risk is unknown (as the label also states). The GU section is entirely inference; there are no organ-preservation data for dMMR urologic tumors.

Endnote: differences from the original forwarded account

1. The forwarded account said "the full Phase 2 will be presented at ESMO"; FDA's documents do not mention this, and no source was found for an ESMO presentation, so it is labeled unverified.
2. The link in the forwarded account is FDA's general approval-notification index (FDA); the specific data are on the individual approval page (FDA).
3. ICR cCR12 is 47%, but its 95% CI of 27%–56% is not centered on the point estimate; this may be a typo in the original or reflect an unusual method. Check label section 14.2 before citing.