KEYNOTE-826
Pembrolizumab or placebo plus platinum-based chemotherapy, with or without bevacizumab, as first-line treatment of persistent, recurrent or metastatic cervical cancer
Phase 3N = 617NCT03635567
- Population
- Persistent, recurrent or metastatic squamous cell, adenosquamous or adenocarcinoma of the cervix, not previously treated with chemotherapy and not amenable to curative treatment; measurable disease per RECIST v1.1; ECOG performance status 1 or less
- Design
- Double-blind, placebo-controlled phase 3 trial; 1:1 randomisation to pembrolizumab 200 mg or placebo every 3 weeks for up to 35 cycles plus platinum-based chemotherapy and, at investigator discretion, bevacizumab 15 mg/kg. Dual primary endpoints, PFS and OS, tested sequentially in the PD-L1 CPS 1 or more population, the intention-to-treat population and the CPS 10 or more population. The first report is the protocol-specified first interim analysis.
- Primary endpoint
- PFS, OS
- Arms
- Pembrolizumab plus chemotherapy (experimental, n = 308): Pembrolizumab 200 mg every 3 weeks for up to 35 cycles plus platinum-based chemotherapy, with or without bevacizumab
- Placebo plus chemotherapy (control, n = 309): Placebo every 3 weeks for up to 35 cycles plus platinum-based chemotherapy, with or without bevacizumab
- People
- Colombo N, senior author KEYNOTE-826 Investigators, KEYNOTE-826 Investigators
PFSHR 0.62 (95% CI 0.50–0.77)Δ median PFS+2.2 monthsPFSHR 0.65 (95% CI 0.53–0.79)Δ median PFS+2.2 monthsPFSHR 0.58 (95% CI 0.44–0.77)Δ median PFS+2.3 months
Show the numbers
| Endpoint | Analysis | Pembrolizumab plus chemotherapy | Placebo plus chemotherapy | Effect |
|---|---|---|---|---|
| PFSPD-L1 CPS 1 or more (n = 548)Progression-free survival, investigator-assessed per RECIST v1.1 (Lorusso 2025 abstract) | interimColombo 2021 | median 10.4 months | median 8.2 months | HR 0.62 (95% CI 0.50–0.77)P<0.001Protocol-specified first interim analysis; the data cutoff date (3 May 2021) is given in the abstract of the 2023 final-OS paper as the primary data cutoff. |
| PFS | interimColombo 2021 | median 10.4 months | median 8.2 months | HR 0.65 (95% CI 0.53–0.79)P<0.001Protocol-specified first interim analysis. |
| PFSPD-L1 CPS 10 or more (n = 317) | interimColombo 2021 | median 10.4 months | median 8.1 months | HR 0.58 (95% CI 0.44–0.77)P<0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 1 or more (n = 548) | interimColombo 2021 | 53% at 2 y | 41.7% at 2 y | HR 0.64 (95% CI 0.50–0.81)P<0.001Protocol-specified first interim analysis; the abstract gives the 24-month survival percentages without confidence intervals. |
| OS | interimColombo 2021 | 50.4% at 2 y | 40.4% at 2 y | HR 0.67 (95% CI 0.54–0.84)P<0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 10 or more (n = 317) | interimColombo 2021 | 54.4% at 2 y | 44.6% at 2 y | HR 0.61 (95% CI 0.44–0.84)P=0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 1 or more (n = 548) | finalmedian follow-up 39.1 monthsMonk 2023 | median 28.6 months | median 16.5 months | HR 0.60 (95% CI 0.49–0.74)Protocol-specified final overall-survival analysis. The 39.1 months is the median study follow-up (range 32.1-46.5 months) at the final data cutoff, as stated in the abstract. |
| OSAll-comers (N = 617) | finalmedian follow-up 39.1 monthsMonk 2023 | median 26.4 months | median 16.8 months | HR 0.63 (95% CI 0.52–0.77)The final-OS abstract calls this population 'all-comer'; the first report calls the same 617 patients the intention-to-treat population. |
| OSPD-L1 CPS 10 or more (n = 317) | finalmedian follow-up 39.1 monthsMonk 2023 | median 29.6 months | median 17.4 months | HR 0.58 (95% CI 0.44–0.78)Protocol-specified final overall-survival analysis. |
| OSPD-L1 CPS 1 or more, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.60 (95% CI 0.45–0.79)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSAll-comers, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.47–0.80)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSPD-L1 CPS 1 or more, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.56 (95% CI 0.43–0.73)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSAll-comers, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.57 (95% CI 0.45–0.73)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSPD-L1 CPS 1 or more, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.44–0.85)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSAll-comers, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.67 (95% CI 0.49–0.91)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSPD-L1 CPS 1 or more, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.44–0.85)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSAll-comers, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.69 (95% CI 0.50–0.94)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| Endpoint | Analysis | Pembrolizumab plus chemotherapy | Placebo plus chemotherapy | Effect |
|---|---|---|---|---|
| PFSPD-L1 CPS 1 or more (n = 548)Progression-free survival, investigator-assessed per RECIST v1.1 (Lorusso 2025 abstract) | interimColombo 2021 | median 10.4 months | median 8.2 months | HR 0.62 (95% CI 0.50–0.77)P<0.001Protocol-specified first interim analysis; the data cutoff date (3 May 2021) is given in the abstract of the 2023 final-OS paper as the primary data cutoff. |
| PFS | interimColombo 2021 | median 10.4 months | median 8.2 months | HR 0.65 (95% CI 0.53–0.79)P<0.001Protocol-specified first interim analysis. |
| PFSPD-L1 CPS 10 or more (n = 317) | interimColombo 2021 | median 10.4 months | median 8.1 months | HR 0.58 (95% CI 0.44–0.77)P<0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 1 or more (n = 548) | interimColombo 2021 | 53% at 2 y | 41.7% at 2 y | HR 0.64 (95% CI 0.50–0.81)P<0.001Protocol-specified first interim analysis; the abstract gives the 24-month survival percentages without confidence intervals. |
| OS | interimColombo 2021 | 50.4% at 2 y | 40.4% at 2 y | HR 0.67 (95% CI 0.54–0.84)P<0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 10 or more (n = 317) | interimColombo 2021 | 54.4% at 2 y | 44.6% at 2 y | HR 0.61 (95% CI 0.44–0.84)P=0.001Protocol-specified first interim analysis. |
| OSPD-L1 CPS 1 or more (n = 548) | finalmedian follow-up 39.1 monthsMonk 2023 | median 28.6 months | median 16.5 months | HR 0.60 (95% CI 0.49–0.74)Protocol-specified final overall-survival analysis. The 39.1 months is the median study follow-up (range 32.1-46.5 months) at the final data cutoff, as stated in the abstract. |
| OSAll-comers (N = 617) | finalmedian follow-up 39.1 monthsMonk 2023 | median 26.4 months | median 16.8 months | HR 0.63 (95% CI 0.52–0.77)The final-OS abstract calls this population 'all-comer'; the first report calls the same 617 patients the intention-to-treat population. |
| OSPD-L1 CPS 10 or more (n = 317) | finalmedian follow-up 39.1 monthsMonk 2023 | median 29.6 months | median 17.4 months | HR 0.58 (95% CI 0.44–0.78)Protocol-specified final overall-survival analysis. |
| OSPD-L1 CPS 1 or more, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.60 (95% CI 0.45–0.79)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSAll-comers, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.47–0.80)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSPD-L1 CPS 1 or more, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.56 (95% CI 0.43–0.73)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSAll-comers, received bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.57 (95% CI 0.45–0.73)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSPD-L1 CPS 1 or more, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.44–0.85)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| OSAll-comers, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.67 (95% CI 0.49–0.91)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSPD-L1 CPS 1 or more, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.61 (95% CI 0.44–0.85)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
| PFSAll-comers, did not receive bevacizumab | exploratoryLorusso 2025 | — | — | HR 0.69 (95% CI 0.50–0.94)Exploratory subgroup analysis defined by bevacizumab use (not randomised on bevacizumab); final analysis. The abstract gives no medians for these subgroups. |
Data table
| Endpoint | HR | 95% CI |
|---|---|---|
| PFS | 0.62 | 0.5–0.77 |
| PFS | 0.65 | 0.53–0.79 |
| PFS | 0.58 | 0.44–0.77 |
| OS | 0.64 | 0.5–0.81 |
| OS | 0.67 | 0.54–0.84 |
| OS | 0.61 | 0.44–0.84 |
| OS | 0.6 | 0.49–0.74 |
| OS | 0.63 | 0.52–0.77 |
| OS | 0.58 | 0.44–0.78 |
| OS | 0.6 | 0.45–0.79 |
| OS | 0.61 | 0.47–0.8 |
| PFS | 0.56 | 0.43–0.73 |
| PFS | 0.57 | 0.45–0.73 |
| OS | 0.61 | 0.44–0.85 |
| OS | 0.67 | 0.49–0.91 |
| PFS | 0.61 | 0.44–0.85 |
| PFS | 0.69 | 0.5–0.94 |
- Pembrolizumab plus chemotherapy (experimental)
- Placebo plus chemotherapy (control)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Pembrolizumab plus chemotherapy | Placebo plus chemotherapy |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 7 | 62.7% | 55.3% |
| 13 | 42.0% | 33.3% |
| 20 | 26.4% | 18.4% |
| 26 | 17.7% | 11.1% |
- Pembrolizumab plus chemotherapy (experimental)
- Placebo plus chemotherapy (control)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Pembrolizumab plus chemotherapy | Placebo plus chemotherapy |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 7 | 62.7% | 55.3% |
| 13 | 42.0% | 33.3% |
| 20 | 26.4% | 18.4% |
| 26 | 17.7% | 11.1% |
Appears in
Sources
Colombo N, Dubot C, Lorusso D, Caceres MV, Hasegawa K, Shapira-Frommer R, et al. Pembrolizumab for Persistent, Recurrent, or Metastatic Cervical Cancer. N Engl J Med. 2021;385(20):1856-1867.
Monk BJ, Colombo N, Tewari KS, Dubot C, Caceres MV, Hasegawa K, et al. First-Line Pembrolizumab + Chemotherapy Versus Placebo + Chemotherapy for Persistent, Recurrent, or Metastatic Cervical Cancer: Final Overall Survival Results of KEYNOTE-826. J Clin Oncol. 2023;41(36):5505-5511.
Lorusso D, Colombo N, Dubot C, Cáceres MV, Hasegawa K, Shapira-Frommer R, et al. Pembrolizumab plus chemotherapy for advanced and recurrent cervical cancer: final analysis according to bevacizumab use in the randomized KEYNOTE-826 study. Ann Oncol. 2025;36(1):65-75.