Pembrolizumab/Chemoradiotherapy Shows Clinically Meaningful PFS Improvement in Cervical Cancer


Data from the KEYNOTE-A18 trial support the addition of pembrolizumab to chemoradiotherapy, which has been in place as standard of care for patients with newly diagnosed, previously untreated, high-risk locally advanced cervical breast cancer since 1999.

Bradley J. Monk, MD, FACS, FACOG

Bradley J. Monk, MD, FACS, FACOG

In the phase 3 KEYNOTE-A18 study (NCT04221945), adding pembrolizumab (Keytruda) to external beam radiotherapy (EBRT) and concurrent chemotherapy, followed by brachytherapy, led to statistically significant and clinically meaningful improvements in progression-free survival (PFS) vs placebo plus EBRT, chemoradiotherapy, and brachytherapy in patients with newly diagnosed, previously untreated, high-risk locally advanced cervical breast cancer.1

Findings from the first interim analysis of the trial were presented during the 2023 ESMO Congress. These data support the addition of pembrolizumab to chemoradiotherapy, which has been in place as standard of care (SOC) for this patient population since 1999, according to study principal investigator Domenica Lorusso, MD, PhD.

"This is a celebration for patients because we’re challenging a treatment paradigm that has stood for more than 2 decades,” Bradley J. Monk, MD, FACS, FACOG, a professor in the Division of Gynecologic Oncology at the University of Arizona College of Medicine in Tucson and the medical director of the US Oncology Research Network–Gynecologic Program, noted in a discussion of Lorusso’s presentation at ESMO.

At a median follow-up 17.9 months (range, 0.9-31.0), the median PFS was not yet reached (NR) for patients in the pembrolizumab plus chemoradiotherapy (n = 528) arm or patients in the placebo plus chemoradiotherapy (n = 530) arm (HR, 0.70; 95% CI, 0.55-0.89; P = .0020). The 24-month PFS rates were 67.8% (95% CI, 61.8%-73.0%) vs 57.3% (95% CI, 51.2%-62.9%), respectively.

Data presented from protocol-specified subgroups suggested consistent benefit was observed with the addition of pembrolizumab to treatment; those with an ECOG performance status of 1 (HR, 0.53; 95% CI, 0.33-0.85), aged 65 years or older (HR, 0.57; 95% CI, 0.27-1.17), and with stage III to IVA disease at screening (HR, 0.58; 95% CI, 0.42-0.80) experienced the most pronounced benefit.

An increase in overall survival (OS) was reported in the pembrolizumab arm, however, median OS was NR in either arm (HR, 0.73; 95% CI, 0.49-1.07). The 24-month OS rates was 87.2% (95% CI, 82.4%-90.8%) in the pembrolizumab arm vs 80.8% (95% CI, 74.8%-85.5%) in the placebo arm.

“OS data are not yet mature referring to less than 43% of the information fraction,” Lorusso said. “But the CI crossed the unit, and the data are not significant according to the prespecified threshold of alpha significance we defined in the protocol.”

Improvements were seen regarding overall response rate (ORR) in the pembrolizumab arm vs placebo arm as the ORR was 79.3% (75.5%-82.7%) in the pembrolizumab arm, including a complete response (CR) rate of 50.7% and a partial response (PR) rate of 28.6% vs 75.9% (72.0%-79.5%) in the placebo arm, with a CR rate of 48.7% and a PR rate of 27.2%. The 12-month OS rates were 81.4% vs 77.3%, respectively.

Enrollment Criteria and Dosing

KEYNOTE-A18 enrolled patients with treatment-naïve, stage IB2 to IIB node-positive disease or stage III to IVA node-positive or node-negative disease according to International Federation of Gynecology and Obstetrics 2014 criteria; disease needed to be measurable per RECIST 1.1. Eligible patients were randomly assigned 1:1 to the investigative and control arms and stratified based on planned EBRT type, stage at screening, and planned total radiotherapy dose.

Cisplatin was given at 40 mg/m2 once weekly for 5 cycles with ERBT followed by brachytherapy in both arms; for 5 weeks those in the investigative arm received concurrent pembrolizumab and those in the control arm received concurrent placebo. Pembrolizumab or placebo was then administered every 6 weeks for 15 cycles.

Primary end points were PFS as well as OS and key secondary end points included 24-month PFS, ORR, patient-reported outcomes, and safety. Lorusso, who is an associate professor of Obstetrics and Gynecology at the Catholic University of Rome in Italy, noted that 2 key protocol amendments occurred as well.

“In January 2021 we decided to change our primary end point from blinded independent central reviewed assessed PFS to investigator assessed PFS because cervical cancer is an asymptomatic tumor, and the evaluation of the clinician is always more informative than simply reviewing a CT scan,” Lorusso noted. Additionally, in November 2022 investigators changed the multiplicity strategy in the SAP.

At the January 9, 2023, data cutoff, patients in the investigative vs control arm were continuing treatment (57.8% vs 54.9%), completed treatment (11.0% vs 10.6%), and had discontinued the trial (31.3% vs 34.5%), respectively.

In both arms, patients received a median of 5 cycles (range, 1-7) of cisplatin and 11 cycles (range, 1-20) of pembrolizumab or placebo. The median overall radiation therapy treatment time was 52 days in the pembrolizumab (range, 12-139) and placebo (range, 2-166) arms; the median total physical cervix dose was 76 Gy (range, 14-94) and the total cervix EQD2 dose was 87 Gy (range, 14-118) in the pembrolizumab arm compared with 76 Gy (range, 3-125) and 87 Gy (range, 3-207) in the placebo arm, respectively.

Baseline Characteristics in the Global Trial

Baseline characteristics were well balanced between the arms and most patients had a PD-L1 combined positive score of greater than 1 (94.9% vs 93.8%), squamous cell carcinoma (81.9% vs 84.9%), and planned to receive EBRT in the form of IMRT or VMAT (88.7% vs 88.5%) in the pembrolizumab vs placebo arms, respectively. Additionally, patients had an ECOG performance score of 1 (28.2% vs 25.2%), stage IB2-IIB (44.4% vs 42.7%) or stage III-IVA (55.6% vs 57.3%) disease at screening and planned to receive a total radiotherapy dose of 70 Gy or greater (91.1% vs 91.3%) or less than 70 Gy (8.9% vs 8.7%).

Lymph node involvement was positive pelvic only (61.6% vs 61.0%), positive para-aortic only (2.6% vs 1.9%), positive pelvic and para-aortic (19.8% vs 19.6%), or no positive pelvic/para-aortic (15.9% vs 17.5%) in the pembrolizumab vs placebo arms, respectively.

In a discussion of the trial, Monk noted that a diverse population was enrolled. “Fifty-one percent of the patients in this study were non-White, [which is] perhaps one of the highest non-White enrollments ever,” he said.

In the pembrolizumab vs placebo arm patients were White (48.0% vs 49.7%), Asian (29.3% vs 27.9%), Multiple Races (14.7% vs 16.2%), American Indian or Alaska Native (4.5% vs 4.1%), Black (2.6% vs 1.5%), and Native Hawaiian or Other Pacific Islander (0.4% vs 0.2%), respectively.

Manageable and Expected Safety Profile

Regarding safety, grade 3 or higher treatment-related adverse effects (TRAEs) occurred in 67.0% of patients in the investigative arm vs 60.6% of those in the control arm and serious TRAEs were reported in 17.2% vs 12.3% of patients, respectively. Two patients in each arm died from TRAEs and 5 patients in the pembrolizumab arm compared with 6 patients in the placebo arm died from all-cause AEs. Any grade immune-mediated AEs occurred in 32.6% vs 11.7% of patients and at grade 3 or higher in 4.2% vs 1.1% of patients, respectively. No immune-mediated AEs led to death in either arm.

“We paid particular attention to diarrhea because there is a possible overlapping toxicity between immune colitis by pembrolizumab and the post radiotherapy enteritis, but no difference [was observed] with respect to diarrhea,” Lorusso said.

Any grade TRAEs in the pembrolizumab vs placebo arm included anemia (59.3% vs 55.1%), nausea (57.2% vs 59.4%), diarrhea (50.4% vs 51.1%), decreased white blood cell count (32.6% vs 34.2%), decreased neutrophil count (29.0% vs 27.9%), vomiting (25.0% vs 28.3%), leukopenia (23.7% vs 17.4%), decreased platelet count (22.0% vs 20.4%), and neutropenia (21.4% vs 17.4%), respectively. The most common immune-mediated AEs that occurred were hypothyroidism (19.3% vs 4.5%), hyperthyroidism (11.4% vs 2.1%), and colitis (2.7% vs 1.7%).

The EORTC Quality-of-Life Core 30 analysis revealed that there were no clinically meaningful between-group differences in changes in score from baseline to week 36 for global health status/quality of life or physical function scores.

PDUFA Date Upcoming for KEYNOTE-A18

On September 20, 2023, Merck announced that the FDA accepted a supplemental Biologics License Application for priority review seeking approval of pembrolizumab in combination with EBRT plus concurrent chemotherapy, followed by brachytherapy for the treatment of patients with newly diagnosed high-risk locally advanced cervical cancer.2

The Prescription Drug User Fee Act (PDUFA) date has been set for January 20, 2024, and the regulatory decision was supported by data from KEYNOTE-A18.

Editor’s Note: Dr Lorusso has received honoraria from AstraZeneca, Clovis, Genmab, Immunogen, Merck & Co. Inc, Roche, and Tesaro; has a consulting/advisory role with PharmaMar; has received research funding from Clovis, Merck, PharmaMar, and Tesaro; and has a role with the Speaker’s Bureau at AstraZeneca, Clovis, PharmaMar, and Tesaro.


Related Videos
Related Content