POD1UM-303
Retifanlimab with carboplatin and paclitaxel for locally recurrent or metastatic squamous cell carcinoma of the anal canal (POD1UM-303/InterAACT-2): a global, phase 3 randomised controlled trial
Phase 3N = 308NCT04472429
- Population
- Adults aged 18 or older with inoperable locally recurrent or metastatic squamous cell carcinoma of the anal canal, ECOG performance status 0 or 1, no previous systemic therapy; HIV allowed if well controlled (CD4 count above 200 per microlitre and undetectable viral load). 308 randomised.
- Design
- Global, multicentre, double-blind, randomised phase 3 trial at 70 centres in 12 countries (EU, Australia, Japan, UK, USA). Participants were randomised 1:1 to retifanlimab 500 mg or placebo every 4 weeks with standard carboplatin-paclitaxel for up to 1 year; placebo-arm participants could cross over to retifanlimab monotherapy on confirmed progression. Primary endpoint: independently assessed progression-free survival per RECIST 1.1, analysed by intention to treat. Overall survival was the key secondary endpoint (final analysis reported separately).
- Primary endpoint
- PFS
- Arms
- Retifanlimab plus carboplatin-paclitaxel (experimental, n = 154): Retifanlimab 500 mg intravenously every 4 weeks, with carboplatin (AUC 5, day 1) and paclitaxel (80 mg/m2, days 1, 8 and 15) for 6 cycles, then retifanlimab alone; study drug for up to 12 months
- Placebo plus carboplatin-paclitaxel (control, n = 154): Placebo every 4 weeks, with carboplatin (AUC 5, day 1) and paclitaxel (80 mg/m2, days 1, 8 and 15) for 6 cycles, then placebo alone for up to 12 months; crossover to retifanlimab monotherapy allowed on confirmed progression
- People
- Rao S, POD1UM-303/InterAACT-2 study investigators
PFSHR 0.63 (95% CI 0.47–0.84)Δ median PFS+1.9 monthsOSHR 0.70 (95% CI 0.49–1.01)Δ median OS+6.2 monthsOSHR 0.75 (95% CI 0.55–1.01)Δ median OS+10.6 months
Show the numbers
| Endpoint | Analysis | Retifanlimab plus carboplatin-paclitaxel | Placebo plus carboplatin-paclitaxel | Effect |
|---|---|---|---|---|
| PFSTime from randomisation to first documented progressive disease or death from any cause, assessed by independent review per RECIST 1.1 | primaryRao 2025 | median 9.3 months (7.5–11.3) | median 7.4 months (7.1–7.7) | HR 0.63 (95% CI 0.47–0.84)P=0.0006One-sided p-value as published. The data cutoff of 15 April 2024 is from the EMA assessment report (primary PFS analysis and interim overall survival analysis); the FDA label calls this the final analysis of PFS. |
| OS | interimGandhy 2026 | median 29.2 months | median 23 months (15.1–27.9) | HR 0.70 (95% CI 0.49–1.01)Prespecified interim overall survival analysis at the time of the PFS analysis. Retifanlimab-arm 95% CI: 24.2 months to not evaluable. P=0.0273 is reported against a significance threshold of 0.0103 (FDA approval summary, full text; its text says the p-value is 'below' the threshold, but 0.0273 is numerically above 0.0103, and the summary, the FDA label and the EMA assessment report all state that the result was not statistically significant). At this analysis 69 placebo-arm patients (45%) had crossed over to retifanlimab. |
| OSFull analysis set (308 randomised) | finalRao 2026 | median 32.8 months | median 22.2 months | HR 0.75 (95% CI 0.55–1.01)P=0.0305Published as P = 0.0305; the abstract calls the result a consistent and clinically meaningful improvement. The current FDA label (05/2026) and the EMA assessment report state that overall survival did not meet the prespecified criterion for statistical significance. 95% CIs of the medians from the label: 25.7-44.5 months (retifanlimab) and 15.7-27.2 months (placebo). Data cutoff 1 August 2025 and 172 deaths (EMA assessment report); 77 patients in the placebo arm crossed over to retifanlimab monotherapy. |
| PFSRandomised patients (n = 308); analysis population not stated in the abstract | long term-updateRao 2026 | — | — | HR 0.62 (95% CI 0.47–0.81)P=0.0002Updated efficacy analysis reported with the final overall survival data; the P value is labelled nominal in the abstract. The abstract gives no medians for this update. Cutoff date from the EMA assessment report (final OS analysis). |
| ORRFull analysis set (308 randomised)Best overall confirmed response (complete or partial) by blinded independent central review per RECIST 1.1 | primarySource 2026 | 55.8% (47.6–63.8%) | 44.2% (36.2–52.4%) | Complete response in 34 (22.1%) vs 21 (13.6%) and partial response in 52 (33.8%) vs 47 (30.5%). The FDA label (S-004) gives 56% (48-64) and 44% (36-52); median duration of response 14.0 vs 7.2 months in the same label. |
| ORRRandomised patients (n = 308); analysis population not stated in the abstract | finalRao 2026 | 56.5% | 44.8% | Disease control rates in the same sentence of the abstract: 87.7% vs 80.5%. |
| OSFull analysis set (308 randomised) | exploratorySource 2026 | — | — | HR 0.63 (95% CI 0.47–0.86)Final overall survival analysis corrected for crossover to retifanlimab with the rank-preserving structural failure time (RPSFT) model. The EMA assessment report (benefit-risk section, p. 98 and p. 101) calls such analyses exploratory, subject to methodological assumptions and supportive only. |
| OSFull analysis set (308 randomised) | exploratorySource 2026 | — | — | HR 0.53 (95% CI 0.36–0.79)Final overall survival analysis corrected for crossover with inverse probability of censoring weighting (IPCW). The EMA assessment report (benefit-risk section, p. 98 and p. 101) calls such analyses exploratory, subject to methodological assumptions and supportive only. |
| Grade 3 or worse adverse eventsPatients with adverse event data in the primary analysis (safety population of 154 and 152 patients per the FDA approval summary) | primaryRao 2025 | 83.1% | 75% | Serious adverse events 47.4% vs 38.8%. Grade 3 or worse neutropenia 35.1% vs 29.6% and anaemia 19.5% vs 20.4%. Four fatal adverse events with retifanlimab (one treatment-related, pancytopenia) and one with placebo (not treatment-related). |
| Endpoint | Analysis | Retifanlimab plus carboplatin-paclitaxel | Placebo plus carboplatin-paclitaxel | Effect |
|---|---|---|---|---|
| PFSTime from randomisation to first documented progressive disease or death from any cause, assessed by independent review per RECIST 1.1 | primaryRao 2025 | median 9.3 months (7.5–11.3) | median 7.4 months (7.1–7.7) | HR 0.63 (95% CI 0.47–0.84)P=0.0006One-sided p-value as published. The data cutoff of 15 April 2024 is from the EMA assessment report (primary PFS analysis and interim overall survival analysis); the FDA label calls this the final analysis of PFS. |
| OS | interimGandhy 2026 | median 29.2 months | median 23 months (15.1–27.9) | HR 0.70 (95% CI 0.49–1.01)Prespecified interim overall survival analysis at the time of the PFS analysis. Retifanlimab-arm 95% CI: 24.2 months to not evaluable. P=0.0273 is reported against a significance threshold of 0.0103 (FDA approval summary, full text; its text says the p-value is 'below' the threshold, but 0.0273 is numerically above 0.0103, and the summary, the FDA label and the EMA assessment report all state that the result was not statistically significant). At this analysis 69 placebo-arm patients (45%) had crossed over to retifanlimab. |
| OSFull analysis set (308 randomised) | finalRao 2026 | median 32.8 months | median 22.2 months | HR 0.75 (95% CI 0.55–1.01)P=0.0305Published as P = 0.0305; the abstract calls the result a consistent and clinically meaningful improvement. The current FDA label (05/2026) and the EMA assessment report state that overall survival did not meet the prespecified criterion for statistical significance. 95% CIs of the medians from the label: 25.7-44.5 months (retifanlimab) and 15.7-27.2 months (placebo). Data cutoff 1 August 2025 and 172 deaths (EMA assessment report); 77 patients in the placebo arm crossed over to retifanlimab monotherapy. |
| PFSRandomised patients (n = 308); analysis population not stated in the abstract | long term-updateRao 2026 | — | — | HR 0.62 (95% CI 0.47–0.81)P=0.0002Updated efficacy analysis reported with the final overall survival data; the P value is labelled nominal in the abstract. The abstract gives no medians for this update. Cutoff date from the EMA assessment report (final OS analysis). |
| ORRFull analysis set (308 randomised)Best overall confirmed response (complete or partial) by blinded independent central review per RECIST 1.1 | primarySource 2026 | 55.8% (47.6–63.8%) | 44.2% (36.2–52.4%) | Complete response in 34 (22.1%) vs 21 (13.6%) and partial response in 52 (33.8%) vs 47 (30.5%). The FDA label (S-004) gives 56% (48-64) and 44% (36-52); median duration of response 14.0 vs 7.2 months in the same label. |
| ORRRandomised patients (n = 308); analysis population not stated in the abstract | finalRao 2026 | 56.5% | 44.8% | Disease control rates in the same sentence of the abstract: 87.7% vs 80.5%. |
| OSFull analysis set (308 randomised) | exploratorySource 2026 | — | — | HR 0.63 (95% CI 0.47–0.86)Final overall survival analysis corrected for crossover to retifanlimab with the rank-preserving structural failure time (RPSFT) model. The EMA assessment report (benefit-risk section, p. 98 and p. 101) calls such analyses exploratory, subject to methodological assumptions and supportive only. |
| OSFull analysis set (308 randomised) | exploratorySource 2026 | — | — | HR 0.53 (95% CI 0.36–0.79)Final overall survival analysis corrected for crossover with inverse probability of censoring weighting (IPCW). The EMA assessment report (benefit-risk section, p. 98 and p. 101) calls such analyses exploratory, subject to methodological assumptions and supportive only. |
| Grade 3 or worse adverse eventsPatients with adverse event data in the primary analysis (safety population of 154 and 152 patients per the FDA approval summary) | primaryRao 2025 | 83.1% | 75% | Serious adverse events 47.4% vs 38.8%. Grade 3 or worse neutropenia 35.1% vs 29.6% and anaemia 19.5% vs 20.4%. Four fatal adverse events with retifanlimab (one treatment-related, pancytopenia) and one with placebo (not treatment-related). |
Data table
| Endpoint | HR | 95% CI |
|---|---|---|
| PFS | 0.63 | 0.47–0.84 |
| OS | 0.7 | 0.49–1.01 |
| OS | 0.75 | 0.55–1.01 |
| PFS | 0.62 | 0.47–0.81 |
| OS | 0.63 | 0.47–0.86 |
| OS | 0.53 | 0.36–0.79 |
- Retifanlimab plus carboplatin-paclitaxel (experimental)
- Placebo plus carboplatin-paclitaxel (control)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Retifanlimab plus carboplatin-paclitaxel | Placebo plus carboplatin-paclitaxel |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 6 | 63.9% | 57.0% |
| 12 | 40.9% | 32.5% |
| 17 | 28.2% | 20.3% |
| 23 | 18.0% | 11.6% |
- Retifanlimab plus carboplatin-paclitaxel (experimental)
- Placebo plus carboplatin-paclitaxel (control)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Retifanlimab plus carboplatin-paclitaxel | Placebo plus carboplatin-paclitaxel |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 18 | 65.2% | 58.1% |
| 37 | 41.5% | 32.8% |
| 55 | 27.1% | 19.1% |
| 73 | 17.7% | 11.1% |
Appears in
Sources
Rao S, Samalin-Scalzi E, Evesque L, Ben Abdelghani M, Morano F, Roy A, et al. Retifanlimab with carboplatin and paclitaxel for locally recurrent or metastatic squamous cell carcinoma of the anal canal (POD1UM-303/InterAACT-2): a global, phase 3 randomised controlled trial. Lancet. 2025;405(10495):2144-2152.
Gandhy SU, Casak SJ, Saung MT, Mishra-Kalyani PS, Tang S, Auth D, et al. FDA Approval Summary: Retifanlimab for the Treatment of Adult Patients with Metastatic or Inoperable Locally Recurrent Squamous Cell Carcinoma of the Anal Canal. Clin Cancer Res. 2026;32(20):4514-4518.
Rao S, Samalin-Scalzi E, Evesque L, Ben Abdelghani M, Morano F, Roy A, et al. Survival outcomes in POD1UM-303/InterAACT-2: a phase III study of retifanlimab plus carboplatin-paclitaxel in first-line advanced squamous anal cancer. Ann Oncol. 2026;37(8):1157-1166.
European Medicines Agency, CHMP. Zynyz-VR-0000247788: EPAR assessment report for the extension of indication to squamous cell carcinoma of the anal canal, EMADOC-1700519818-2817991, first published 18/03/2026 (102 pages; accessed 22 October 2026).