Assessment: Is This a Good Meta-Analysis Topic?

Short Answer: Conditionally - with significant caveats and a recommended pivot

1. Impact Score: HIGH

• The Cercek et al. NEJM 2025 trial (PMID 40293177) showed 100% clinical complete response (cCR) and 100% organ preservation in 49 dMMR locally advanced rectal cancer patients treated with dostarlimab monotherapy, with 92% 2-year recurrence-free survival. This is paradigm-shifting data.
• dMMR/MSI-H accounts for roughly 5-10% of all rectal cancers - a biologically distinct and important subgroup.
• The key unresolved clinical question - does this population even need chemoradiation? - is exactly what your proposed comparison addresses. This question is being debated at ESMO, ASCO, and NCCN levels right now.
• Organ preservation and avoidance of pelvic radiation toxicity are major quality-of-life drivers.
• A well-done meta-analysis framing this comparison would attract significant citations and clinical attention.

2. Feasibility: MODERATE-LOW (Critical Issues)

Problem 1: No head-to-head RCTs exist yet

• Single-arm phase 2 trials (immunotherapy arm only, no CRT comparator within the same study)
• Retrospective cohort studies
• Historical comparisons against older CRT datasets

Problem 2: Existing systematic reviews already cover much of this ground

Problem 3: Data heterogeneity is extreme

• Different PD-1 agents used: dostarlimab, pembrolizumab, sintilimab, camrelizumab, tislelizumab, nivolumab
• Different cycle numbers (4 to 9 cycles)
• Variable primary endpoints (cCR, pCR, organ preservation, DFS)
• Different staging workups and response assessment tools
• Most studies are Chinese single-center trials; Cercek et al. is a US multicenter trial - these populations may not be comparable

3. Methodological Suitability

4. What Would Strengthen This (Recommended Pivots)

"PD-1 monotherapy vs. chemoradiation in dMMR/MSI-H LARC: a systematic review with indirect comparison and gap analysis informing future trial design"

"Surgical and pathological outcomes after neoadjuvant PD-1 monotherapy vs. chemoradiation in dMMR/MSI-H rectal cancer: a systematic review"

"Organ preservation rates with neoadjuvant immunotherapy in dMMR/MSI-H rectal cancer: a systematic review and meta-analysis"

Summary Table

Organ Preservation Rates with Neoadjuvant Immunotherapy in dMMR/MSI-H Rectal Cancer: A Systematic Review and Meta-Analysis

What "Organ Preservation" Actually Means Here

Why This Is a Distinct and Underexplored Outcome

• Pool organ preservation rate as the primary meta-analytic endpoint
• Analyze local regrowth rates after watch-and-wait in this specific dMMR + immunotherapy cohort
• Compare organ preservation specifically in rectal (not colorectal) subgroups
• Analyze what predicts successful organ preservation (number of cycles, agent type, tumor stage, ctDNA clearance)
• Report stoma-free survival or sphincter preservation as discrete outcomes
• Analyze duration of follow-up as a moderator of organ preservation durability

The Clinical Stakes: Why It Matters

What the Current Evidence Shows (The Raw Data Your Meta-Analysis Would Pool)

How the Meta-Analysis Would Be Structured

PICO Framework

Statistical Method

• Random-effects model (DerSimonian-Laird or REML) for pooling proportions
• Freeman-Tukey double arcsine transformation to stabilize proportions near 0 or 1 (critical here since rates are very high)
• I² statistic for heterogeneity
• Subgroup analyses by: ICI agent, number of cycles, tumor stage (cT3 vs cT4), tumor height (low vs mid rectal), follow-up duration
• Meta-regression for predictors of organ preservation
• Funnel plot + Egger's test for publication bias

Secondary Outcomes You Could Include

1. Local regrowth rate after watch-and-wait - what proportion of patients who achieved cCR had regrowth? And was salvage surgery still curative when performed?
2. Stoma-free survival at 1 and 2 years
3. Sphincter preservation rate among patients who did go to surgery
4. Distant metastasis-free survival in organ-preserved patients
5. Grade 3-4 immune-related adverse events - safety profile of the strategy
6. Pseudoprogression / pseudoresidue rates - a clinically important phenomenon (PMID 36791752) where imaging overestimates residual tumor, causing unnecessary surgery

Evidence Gap That Makes This Novel

• Most studies have short follow-up (8-26 months)
• The Cercek 2025 NEJM paper provides the best data (median 20 months, 92% 2-year recurrence-free survival in cohort 1, n=49)
• But we don't know: at 3, 4, 5 years, what proportion maintain organ preservation without regrowth?
• A meta-analysis that pools follow-up-stratified data would directly answer this and would be the most policy-relevant paper in this space right now

Feasibility Assessment for This Specific Design

Practical Steps to Execute This

1. Register on PROSPERO now - it confirms novelty and protects your idea
2. Search databases: PubMed, Embase, CENTRAL, Web of Science, ClinicalTrials.gov (for unpublished trial arms)
3. Inclusion criteria: prospective or retrospective studies, dMMR/MSI-H confirmed, rectal cancer only (not colorectal mixed), PD-1 monotherapy arm extractable, organ preservation outcome reported
4. Exclusion: pMMR/MSS patients, metastatic disease, combination CRT+ICI (keep it monotherapy)
5. Contact corresponding authors for studies that only report pCR without stating whether W&W was pursued
6. Subgroup the data by Asian vs Western centers (a potentially large confounder - most high-rate results come from Chinese centers with strict dMMR selection)

One Important Caveat to Acknowledge Upfront

Summary

PROSPERO Conflict Assessment - Verdict: CLEAR TO REGISTER

Document Delivered

What's Inside the Document

Protocol Sections (17 total)

1. Background and Rationale - Clinical context, why dMMR is different, the organ preservation opportunity
2. Objectives - Primary + 7 secondary objectives
3. PICO Framework - Formal population/intervention/comparator/outcome table
4. Eligibility Criteria - Detailed inclusion/exclusion with rationale
5. Search Strategy - 10 databases listed + full PubMed search string ready to copy
6. Study Selection Process - Dual-reviewer screening, kappa reporting, Covidence/Rayyan use
7. Data Extraction - Full variable list across 7 categories in a formatted table
8. Risk of Bias - RoB 2.0 for RCTs, ROBINS-I for observational studies
9. Statistical Analysis Plan - Freeman-Tukey transformation, random-effects model, I², prediction intervals
10. Subgroup Analyses - 7 pre-specified subgroups (agent, cycles, T-stage, tumor height, geography, study design, follow-up)
11. Meta-Regression - Continuous moderators
12. Sensitivity Analyses - 6 pre-specified scenarios
13. Publication Bias - Funnel plot, Egger's test, trim-and-fill
14. GRADE Certainty Assessment - SoF table planned
15. PROSPERO Conflict Table - All 5 identified registrations assessed
16. PRISMA 2020 Checklist - All 26 items, color-coded by section, with location and reported columns to fill in
17. Key References - 9 landmark citations

Key Methodological Points Built Into the Protocol

• Organ preservation as the primary pooled proportion (first protocol to do this)
• Freeman-Tukey transformation explicitly specified (handles near-100% proportions correctly)
• Rectal cancer only - colon mixed studies excluded unless rectum is separable
• PD-1 monotherapy only - combination CRT studies go to sensitivity analysis
• Author contact procedure built in for incomplete data
• Geography subgroup (East Asian vs. Western) explicitly pre-specified to address the major confounder

PROSPERO Registration — Complete Draft (All 22+ Fields)

ELIGIBILITY SCREENER (Answer these when starting a new registration)

SECTION A — ADMINISTRATIVE DETAILS

Organ Preservation Rates with Neoadjuvant PD-1 Inhibitor Monotherapy in dMMR/MSI-H Locally Advanced Rectal Cancer: A Systematic Review and Meta-Analysis of Pooled Proportions

Not applicable — review conducted in English

01 July 2026

28 February 2027

Preliminary searches: Started | Screening: Not started | Data extraction: Not started | Risk of bias: Not started | Synthesis: Not started

[REPLACE with PI name, institutional email, ORCID, department, institution, country, address]

Reviewer A (Independent Screener): [REPLACE — Name, Institution, ORCID] Reviewer B (Independent Screener): [REPLACE — Name, Institution, ORCID] Arbiter (3rd Reviewer): [REPLACE] Statistician: [REPLACE]

This systematic review receives no external funding. No pharmaceutical company is involved in its design or conduct.

All team members declare no conflicts of interest. No author has received support from manufacturers of dostarlimab, pembrolizumab, nivolumab, sintilimab, camrelizumab, or other PD-1 inhibitors.

SECTION B — REVIEW DETAILS

Among adults with dMMR/MSI-H locally advanced rectal cancer (stage II–III, non-metastatic), what proportion achieve organ preservation (avoidance of radical surgery) following neoadjuvant PD-1 inhibitor monotherapy, and what patient and treatment factors predict successful organ preservation?

Databases: PubMed/MEDLINE, Embase (Ovid), Cochrane CENTRAL, Web of Science, CNKI, Wanfang, ClinicalTrials.gov, WHO ICTRP, ASCO and ESMO congress abstracts 2020–2026. Inception to July 2026. No date restriction. English and Chinese language.

PubMed search string: ("rectal cancer" OR "rectal neoplasm" OR "rectal carcinoma") AND ("dMMR" OR "deficient mismatch repair" OR "MSI-H" OR "microsatellite instability-high") AND ("PD-1" OR "pembrolizumab" OR "dostarlimab" OR "nivolumab" OR "sintilimab" OR "camrelizumab" OR "immune checkpoint inhibitor") AND ("neoadjuvant" OR "preoperative") AND ("organ preservation" OR "watch and wait" OR "nonoperative" OR "complete response")

Locally advanced rectal cancer (stage II–III, non-metastatic) with confirmed dMMR (IHC) or MSI-H (PCR/NGS). MeSH: Rectal Neoplasms; Microsatellite Instability; DNA Mismatch Repair; Neoadjuvant Therapy; Organ Preservation

Adults ≥18 years, histologically confirmed rectal adenocarcinoma, cT3-T4 and/or cN+, M0, with dMMR confirmed by IHC (loss of MLH1/MSH2/MSH6/PMS2) OR MSI-H by PCR or NGS. Exclusions: metastatic disease; pMMR/MSS status; colon cancer only (rectal subgroup must be separately extractable); age <18 years.

Neoadjuvant PD-1 inhibitor monotherapy (any agent: dostarlimab, pembrolizumab, nivolumab, sintilimab, camrelizumab, tislelizumab, toripalimab, or other investigational anti-PD-1 antibodies) administered without concurrent chemotherapy, radiotherapy, or anti-CTLA-4 agents.

No comparator required. This is a single-arm pooled proportion meta-analysis. Where a study includes multiple arms, only the PD-1 monotherapy arm will be extracted for the primary analysis.

Phase 1b/2/3 clinical trials; prospective cohort studies; retrospective cohort studies with ≥5 dMMR/MSI-H rectal cancer patients. Minimum follow-up 3 months. Exclusions: <5 patients; case reports; narrative reviews; abstracts without extractable patient-level data.

SECTION C — OUTCOMES

1. Clinical complete response (cCR) rate
2. Pathological complete response (pCR) rate (among operated patients)
3. Local regrowth rate (among watch-and-wait patients)
4. 2-year disease-free survival (DFS)
5. 2-year overall survival (OS)
6. Sphincter preservation rate (among operated patients)
7. Grade ≥3 immune-related adverse events (irAEs)
8. Stoma-free survival at 2 years

SECTION D — DATA AND ANALYSIS

Two independent reviewers using a pre-piloted standardized form. Variables: study ID, country, design, trial registration, sample size, demographics, tumor characteristics (stage, T-category, height from anal verge), dMMR confirmation method, ICI agent/dose/cycles, response assessment method/timing, organ preservation outcome (n/total), cCR, pCR, regrowth data, survival, and adverse events. Disagreements resolved by discussion; third reviewer arbitrates. Authors contacted for missing data (minimum 2 attempts). Cohen's kappa reported for inter-rater agreement.

Randomized trials: Cochrane RoB 2.0 (5 domains). Non-randomized prospective and retrospective cohort studies: ROBINS-I (7 domains). Two independent reviewers; disagreements by consensus. High/critical RoB studies excluded in sensitivity analysis.

Primary analysis: Pooled proportion meta-analysis using random-effects model (DerSimonian-Laird; REML as sensitivity). Freeman-Tukey double arcsine transformation applied to stabilize near-unity proportions before pooling; back-transformed for presentation. Pooled proportion reported with 95% CI and 95% prediction interval. Software: R (meta, metafor packages).

Heterogeneity: I² statistic (thresholds: <25% low, 25-50% moderate, 50-75% substantial, >75% considerable), Cochran Q test (p<0.10), tau².

Pre-specified subgroup analyses (7): (1) ICI agent; (2) treatment cycles (<4 vs. 4-6 vs. >6); (3) tumor T-stage (cT2-T3 vs. cT4); (4) tumor height (<5 cm vs. 5-10 cm vs. >10 cm from anal verge); (5) geographic region (East Asian vs. Western); (6) study design (prospective/RCT vs. retrospective); (7) follow-up duration (<12 vs. 12-24 vs. >24 months).

Sensitivity analyses (6): Excluding high/critical RoB studies; prospective only; ≥12 months follow-up only; including PD-1 + anti-angiogenic combinations; Bayesian random-effects; leave-one-out.

Publication bias: Funnel plot + Egger's test (if ≥10 studies); trim-and-fill.

Certainty: GRADE framework; Summary of Findings table.

SECTION E — ADDITIONAL DETAILS

English and Chinese included. Other languages via translation.

[REPLACE: your country]

Submission to peer-reviewed oncology or colorectal surgery journal (target: Annals of Surgery, Lancet Gastroenterology & Hepatology, British Journal of Surgery, or JNCCN). Results presented at ASCO, ESMO, or ASCRS.

dMMR; MSI-H; mismatch repair deficiency; microsatellite instability; rectal cancer; locally advanced rectal cancer; PD-1 inhibitor; immune checkpoint inhibitor; neoadjuvant immunotherapy; organ preservation; watch and wait; nonoperative management; dostarlimab; pembrolizumab; meta-analysis; pooled proportion

Five related registrations/reviews identified — none duplicates this study's primary endpoint:

• CRD42024594173 (Cui et al. 2025): Mixed colorectal; primary outcome pCR/MPR, not organ preservation
• CRD420251174716 (2025): DFS comparison across dMMR CRC treatment paradigms, no organ preservation focus
• CRD42024619949 (BMC Cancer 2025): CRT + ICI combinations, mixed MMR, not monotherapy
• PMID 40889731 (Petrelli 2025): Narrative synthesis only; no pooled proportion meta-analysis
• PMID 41343101 (Mirza 2025): Prospective trials only; colon+rectum mixed; pCR primary endpoint Conclusion: This registration is novel and not duplicative.

Ongoing — protocol stage, data collection not begun

This review focuses exclusively on PD-1 monotherapy in rectal cancer. The Freeman-Tukey double arcsine transformation is pre-specified to address the methodological challenge of pooling near-unity proportions. Full protocol version 1.0 (June 2026) available on request from the corresponding author.

This Week's Submission Plan

Key rule: You can cite the CRD number in your manuscript immediately after submitting the draft — you do not need to wait for final approval.