2 Clarke Drive
Cranbury, NJ 08512
© 2022 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Pembrolizumab alone and in combination with chemotherapy was associated with shorter overall survival compared with the data demonstrated in the registrational clinical trials in older Medicare patients with advanced non–small cell lung cancer, providing real-world insight into the prognosis of older patients with NSCLC who are treated with immunotherapy.
Pembrolizumab (Keytruda) alone and in combination with chemotherapy was associated with shorter overall survival (OS) compared with the data demonstrated in the registrational clinical trials in older Medicare patients with advanced non–small cell lung cancer (NSCLC), providing real-world insight into the prognosis of older patients with NSCLC who are treated with immunotherapy, according to findings from a retrospective cohort study that were published in JAMA Network Open.1
The unadjusted median OS was 11.4 months (95% CI, 10.5-12.3) with single-agent pembrolizumab, which was about 15 months shorter than the survival findings from the phase 3 KEYNOTE-024 trial (NCT02142738), a study that evaluated pembrolizumab-treated patients.2
The unadjusted median OS was 12.9 months (95% CI, 11.8-14.0) with the combination of platinum, pemetrexed, and pembrolizumab, which was about 10 months shorter than the survival findings with the same combination from the KEYNOTE-189 trial (NCT02578680).3
“These comparative effectiveness results are interesting and have to be interpreted cautiously. I don’t think this shows that immunotherapy doesn’t work for older patients at all; rather, it is showing that we are seeing shifts in treatment patterns and the populations of patients potentially getting any treatment, which is going to be important to consider as we try to evaluate the real-world effectiveness of some of these treatments,” said lead study author, Kenneth L. Kehl, MD, MPH, a physician at Dana-Farber Cancer Institute and an assistant professor of medicine at Harvard Medical School, in an interview with OncLive®.
Pembrolizumab monotherapy is approved by the FDA for the first-line treatment of patients with metastatic NSCLC who have at least 50% PD-L1 expression based on findings from the KEYNOTE-024 trial. The study demonstrated a median OS of 30.0 months with pembrolizumab vs 14.2 months with platinum-based chemotherapy (HR, 0.63; 95% CI, 0.47-0.86). In 2019, the FDA expanded the approval for pembrolizumab monotherapy for the frontline treatment of patients with stage III NSCLC, who are ineligible for surgery or definitive chemoradiation, or metastatic NSCLC, with a PD-L1 expression of at least 1%.4
Pembrolizumab in combination with platinum-based chemotherapy and pemetrexed is approved for use in patients with metastatic nonsquamous NSCLC regardless of PD-L1 expression based on findings from the KEYNOTE-189 trial.3 The study demonstrated a median OS of 22.0 months with the combination vs 10.7 months with chemotherapy plus placebo (HR, 0.56; 95% CI, 0.45-0.70).
As such, the frontline use of pembrolizumab is standard for patients with NSCLC, but the clinical utility of checkpoint inhibition in older patients who did not meet the inclusion criteria for the pivotal clinical trials was not well understood.
“The motivation for doing this study when we started it a couple of years ago was to better understand the treatment patterns and outcomes for patients with advanced NSCLC during the period when immunotherapy became part of the standard of care for first-line treatment of those patients,” Kehl said.
Overall, 19,529 United States Medicare patients with advanced lung cancer between the ages of 66 and 89 years at the time of the index claim who initiated first palliative-intent systemic therapy for lung cancer between 2016 and 2018 were included in the retrospective cohort.
First-line treatments included platinum/pemetrexed (n = 5159), platinum/taxane (n = 9866), pembrolizumab (n = 3079), and platinum/pemetrexed/pembrolizumab (n = 1425).
The majority of patients (n = 11,012; 56%) were between 70 and 79 years of age at the index lung cancer event (66-69 years, n = 4963, 25%; 80-89 years, n = 3,554, 18%). Most patients were male (n = 10,509; 54%), White/non-Hispanic (n = 17,305; 89%), and not enrolled on Medicaid (n = 14,993; 77%).
The uptake of pembrolizumab-containing regimens was rapid in the second quarter of 2016; these therapies accounted for 0.7% of first-line treatments, but increased to 42.4% by the third quarter of 2018.
Notably, pembrolizumab monotherapy was more likely to be used as first-line systemic treatment in females and older patients who had higher Risk Stratification Index (RSI) scores at the time of treatment compared with chemotherapy.
Additional results showed that patients who received single-agent pembrolizumab had inferior OS early after treatment initiation compared with patients who received chemotherapy; however, the Kaplan-Meier curves for this difference narrowed over time.
RSI scores were strongly prognostic; the unadjusted restricted mean survival time (RMST) in the highest RSI quintile was 8.1 months vs 14.2 months in the lowest RSI quintile (P < .001).
Similar adjusted RMST was observed between patients who received pembrolizumab (adjusted RMST, 11.0 months; 95% CI, 10.6-11.4) and patients who received platinum/pemetrexed (adjusted RMST, 11.1 months; 95% CI, 10.9-11.3; RMST difference, -0.2 months; 95% CI, -0.5 to 0.2; P = .30).
Patients who received pembrolizumab had statistically worse survival compared with patients who received platinum/taxane, but the difference was small (adjusted RMST difference, -0.7 months; 95% CI, -1.0 to -0.4; P < .001). Patients who received platinum/pemetrexed/pembrolizumab had statistically better survival compared with patients who received platinum/pemetrexed, but again, the difference was small (adjusted RMST difference, 0.5 months; 95% CI, 0.1-0.9; P = .02).
Findings from exploratory analyses demonstrated similar OS in men who received pembrolizumab vs men who received platinum/pemetrexed (adjusted RMST difference, 0.7 months; 95% CI, 0.2-1.2) or platinum/taxane (adjusted RMST difference, -0.4 months; 95% CI, -0.8 to -0.01). Women who received pembrolizumab had slightly worse survival compared with women who received platinum/pemetrexed (adjusted RMST difference -1.1 months; 95% CI, -1.5 to -0.6) or platinum taxane (adjusted RMST difference, -1.1; 95% CI, -1.5 to -0.7).
By age, pembrolizumab treatment yielded similar survival compared with platinum/pemetrexed across categories, slightly worse survival compared with platinum/taxane in older patients, and slightly improved survival compared with platinum/pemetrexed in the youngest age category.
Sensitivity analyses of patients with Medicare part D prescription drug coverage who did not receive oral targeted therapy with gefitinib (Iressa), erlotinib (Tarceva), afatinib (Gilotrif), dacomitinib (Vizimpro), osimertinib (Tagrisso), crizotinib (Xalkori), ceritinib (Zykadia), brigatinib (Alunbrig), alectinib (Alecensa), lorlatinib (Lorbrena), dabrafenib (Tafinlar), or trametinib (Mekinist) within 3 months of the index claim (n = 12,549) reported similar findings as the findings from the exploratory analyses. However, the difference in survival outcomes between platinum/pemetrexed/pembrolizumab and platinum/pemetrexed groups was no longer statistically significant (adjusted RMST difference, 0.4 months; 95% CI, -0.1 to 1.0; P .15).
“There is an increasing interest in using observational or real-world evidence to understand the effectiveness of some of these treatments. Each source of real-world data has pros and cons. We used Medicare data here, which gives one a 10,000-foot view of what is going on and allows one to look at a very large population of patients. The data that are available are not always as granular as one may want for making decisions for individual patients,” Kehl said.
“In addition to looking at a dataset like the Medicare population, another potentially useful way to look at these questions is to look at datasets that might be smaller but could contain more granularity on some of those clinical variables and, critically, to conduct randomized controlled trials in such a way that the results are as generalizable as possible,” concluded Kehl.