Pembrolizumab/Chemo With or Without Bevacizumab May Represent New Frontline SOC in Cervical Cancer

The addition of pembrolizumab (Keytruda) to chemotherapy with or without bevacizumab (Avastin) significantly improved survival and response rates in patients with persistent, recurrent, or metastatic cervical cancer, according to data from the first interim analysis of the phase 3 KEYNOTE-826 trial (NCT03635567) presented as part of the Presidential Symposium during the 2021 ESMO Congress.

In the subset of patients with a PD-L1 combined positive score (CPS) of 1 or higher, the median progression-free survival (PFS) in the investigative and control arms was 10.4 months (95% CI, 9.7-12.3) vs 8.2 months (95% CI, 6.3-8.5), respectively (HR, 0.62; 95% CI, 0.50-0.77; P < .001). In this group, the median overall survival (OS) was not reached (NR) at the time of the data cutoff in the investigative arm (95% CI, 19.8–NR) vs 16.3 months (95% CI, 14.5-19.4) in the control arm (HR, 0.64; 95% CI, 0.50-0.81; P < .001).

In the all-comer population, the median PFS was 10.4 months (95% CI, 9.1-12.1) in the investigative arm vs 8.2 months (95% CI, 6.4-8.4) in the control arm (HR, 0.65; 95% CI, 0.53-0.79; P < .001). The median OS in this population was 24.4 months (19.2–NR) with pembrolizumab vs 16.5 months (95% CI, 14.5-19.4) with placebo (HR, 0.67; 95% CI, 0.54-0.84; P < .001). 

In the group of patients who had a PD-L1 CPS of 10 or higher, the median PFS with pembrolizumab plus chemotherapy with or without bevacizumab was 10.4 months (95% CI, 8.9-15.1) vs 8.1 months (95% CI, 6.2-8.8) with chemotherapy with or without bevacizumab (HR, 0.58; 95% CI, 0.44-0.77; P < .001). The median OS was also NR (95% CI, 19.1-NR) in the investigative arm vs 16.4 months (95% CI, 14.0-25.0) in the control arm (HR, 0.61; 95% CI, 0.44-0.84; P = .001).

“Adding pembrolizumab to chemotherapy with or without bevacizumab provides statistically significant, clinically meaningful improvements in PFS and OS in patients with persistent, recurrent, or metastatic cervical cancer. The significant benefit was observed in all primary analysis populations and was generally consistent across protocol-specified subgroups,” lead study author Nicoletta Colombo, MD, PhD, director of the Gynecologic Oncology Division at the European Institute of Oncology and associate professor of Obstetrics and Gynecology at the University of Milan-Bicocca in Milan, Italy, said in a presentation of the findings. “Overall, data from KEYNOTE-826 suggest that pembrolizumab plus platinum-based chemotherapy with or without bevacizumab may be a new first-line standard of care for the treatment [of these patients].”

Platinum-based chemotherapy has served as standard treatment for patients with persistent, recurrent, or metastatic cervical cancer, with the preferred regimen being platinum, paclitaxel, and bevacizumab. Although PD-1 inhibitors, such as pembrolizumab and cemiplimab (Libtayo), have previously demonstrated efficacy in patients with previously treated cervical cancer, no data are available to indicate whether the addition of a PD-L1 inhibitor to chemotherapy with or without bevacizumab would improve outcomes.

To this end, investigators launched the double-blind KEYNOTE-826 trial, which enrolled patients with persistent, recurrent, or metastatic cervical cancer that was not amenable to curative treatment.

Eligible patients needed to have an ECOG performance status of 0 or 1 and could not have previously received systemic chemotherapy. However, prior radiotherapy and chemoradiotherapy was allowed.

A total of 617 participants were randomized 1:1 to receive intravenous (IV) pembrolizumab at 200 mg every 3 weeks for up to 35 cycles plus paclitaxel and cisplatin or carboplatin every 3 weeks for up to 6 cycles with or without IV bevacizumab at 15 mg/kg every 3 weeks (n = 308) or placebo plus chemotherapy with or without bevacizumab at the same dose and schedule (n = 309).

Stratification factors included metastatic disease at diagnosis (yes vs no), PD-L1 CPS (<1 vs 1 to <10 vs ≥10), and planned bevacizumab use (yes vs no).

The dual primary end points of the study were investigator-assessed OS and PFS per RECIST v1.1 criteria; key secondary end points comprised overall response rate (ORR), duration of response (DOR), 12-month PFS, and safety. Patient-reported outcomes (PROs) were evaluated using the EuroQoL EQ-5D-5L Visual Analog Scale (VAS) score and served as an exploratory end point.

The PFS and OS benefit of adding pembrolizumab were evaluated sequentially in 3 patient populations: those with PD-L1 CPS of 1 or higher, all comers, and those with a PD-L1 CPS of 10 or higher. The overall alpha was controlled at one-sided 2.5% across the 6 primary hypotheses and all planned analysis. The statistical analysis plan permits 2 interim analyses prior to a final analysis. 

The first interim analysis was planned to occur when approximately 370 events of disease progression or death were reported in the population of patients with a PD-L1 CPS of 1 or higher. At that time, all primary hypotheses were to be tested, noted Colombo.

As of a data cutoff of May 3, 2021, 307 patients on the investigative arm and 309 patients on the control arm received treatment. 

“The median follow-up duration was almost 3 years, and approximately twice as many patients in the pembrolizumab arm remain on, or have completed, study treatment,” Colombo said.

Baseline characteristics in the all-comer and PD-L1 CPS subsets were well balanced between the 2 treatment arms, Colombo added.

Among the all-comer population, the median age between the arms was 50.5 years (range, 22-82). Moreover, 43.3% of patients had an ECOG performance status of 1, 72.3% had squamous cell disease, and 48.3% had previously received chemoradiation or radiation. 

At initial diagnosis, 30.8% of patients had stage IVB disease, and 19.8% had metastatic disease at the time of study entry. Regarding PD-L1 status, 51.4% of patients had a CPS of 10 or higher, 37.4% had a CPS between 1 and 10, and 11.2% had a CPS of less than 1. Bevacizumab was used in 63.1% of patients during the study.

Additional data reported during the meeting showed that the 12-month PFS rates in the investigative and control arms within the subset of patients with a PD-L1 CPS of 1 or higher were 45.5% (95% CI, 39.2%-51.5%) and 34.1% (95% CI, 28.3%-40.0%), respectively. In the all-comer population, these rates were 44.7% (95% CI, 38.8%-50.4%) and 33.5% (95% CI, 28.0%-39.1%), respectively. Lastly, in the subset of patients with a PD-L1 CPS of 10 or higher, these rates were 44.6% (95% CI, 36.3%-52.5%) and 33.5% (95% CI, 25.9%-41.2%), respectively.

“The benefit of adding pembrolizumab was generally consistent across the protocol-specified subgroups, with all HRs favoring the pembrolizumab arm and all 95% CIs overlapping,” Colombo said.

The OS benefit of adding pembrolizumab was generally consistent across all protocol-specified subgroups. Notably, the HR favored the addition of pembrolizumab in both bevacizumab subgroups, suggesting that the addition of the immunotherapy to chemotherapy provides benefit irrespective of whether patients can receive bevacizumab, Colombo said.

“The relative OS benefit observed in the pembrolizumab arm appeared to increase with increasing PD-L1 expression and the HR for the PD-L1–negative population was 1.00,” Colombo added. “But, given the relatively small sample size of this subgroup and the overall study design, robust conclusions regarding the efficacy of pembrolizumab plus chemotherapy with or without bevacizumab in patients with PD-L1–negative tumors cannot be drawn.”

The addition of pembrolizumab to chemotherapy with or without bevacizumab also improved ORRs vs placebo plus chemotherapy with or without bevacizumab across the 3 primary analysis populations.

In the subset of patients with a PD-L1 CPS of 1 or higher, the ORR in the investigative arm was 68.1% (95% CI, 62.2%-73.6%) vs 50.2% (95% CI, 44.1%-56.2%) in the control arm, with a median duration of response (DOR) of 18.0 months (range, 1.3+ to 24.2+) vs 10.4 months (range, 1.5+ to 22.0+), respectively. 

In the all-comer population, the ORRs were 65.9% (95% CI, 60.3%-71.2%) vs 50.8% (95% CI, 45.1%-56.5%), respectively, with the same median DORs as were reported in the PD-L1 CPS of 1 or higher subset. Lastly, in the group of patients with a PD-L1 CPS of 10 or higher, the ORRs were 69.6% (95% CI, 61.8%-76.7%) vs 49.1% (95% CI, 41.1%-57.1%), respectively, with median DORs of 21.1 months (range, 1.3+ to 24.2+) vs 9.4 months (range, 2.1+ to 21.5+), respectively.

Investigators collected information on PROs throughout treatment on the trial. Time to deterioration in the EQ-5D-5L VAS score was improved on the pembrolizumab arm, with a HR of 0.75 (95% CI, 0.58-0.97). According to Colombo, the median time to deterioration was not reached in the investigative arm vs 7.7 months in the control arm.

The safety profile of the pembrolizumab regimen was manageable. The incidence of all-cause and treatment-related grade 3 or higher or serious adverse effects (AEs) were all numerically greater in the immunotherapy-containing arm, but patients receiving pembrolizumab were generally on treatment longer vs those receiving placebo, according to Colombo. 

“Approximately one-third of patients on the pembrolizumab arm discontinued any treatment component due to an AE,” Colombo said. “The incidence of AEs leading to death were similar in the treatment arms. As expected, the incidence of immune-related AEs was higher in the pembrolizumab arm.”

The most frequent all-cause AEs reported on the investigative and control arms included anemia (61.2% vs 53.4%, respectively), alopecia (56.4% vs 57.9%), nausea (39.7% vs 43.7%), diarrhea (35.5% vs 29.8%) and fatigue (28.7% vs 27.2%), among others. The most common immune-mediated AEs were hypothyroidism (18.2% vs 9.1%) and hyperthyroidism (7.5% vs 2.9%).

Reference

1. Colombo N, Dubot C, Lorusso D, et al. Pembrolizumab plus chemotherapy versus placebo plus chemotherapy for persistent, recurrent, or metastatic cervical cancer: randomized, double-blind, phase 3 KEYNOTE-826 study. Presented at: 2021 ESMO Congress; September 16-21, 2021; virtual. Abstract LBA2_PR.