乌鸦传媒

Proton pump inhibitor (PPI) use was independently associated with worse treatment outcomes in some patients with non-small cell lung cancer (NSCLC) on immune checkpoint inhibitor (ICI) therapy with pembrolizumab (Keytruda), according to a Japanese multicenter cohort study.

Among patients with a PD-L1 tumor proportion score (TPS) of 50% or more, a history of PPI use was tied to shorter progression-free survival (PFS) in the pembrolizumab monotherapy group (HR 1.38, 95% CI 1.00-1.91), but not in the ICI plus chemotherapy group, reported Tadaaki Yamada, MD, PhD, of the Kyoto Prefectural University of Medicine, Japan, and colleagues, in .

Thus, "PPI use may thus be an important clinical factor that should be considered in choosing treatment with ICI with or without chemotherapy in this population," they suggested. "Furthermore, when considering treatment strategies for patients in whom ICI plus chemotherapy is challenging, the appropriateness and necessity of PPI prescriptions should be assessed before initiating first-line treatment."

Their retrospective cohort study of 425 patients with NSCLC showed that PD-L1 TPS 50%-89% (HR 1.65 95% CI 1.18-2.31, P=.003) was also associated with poorer PFS in the ICI monotherapy group alone.

In contrast, only smoking history (HR 0.39 95% CI 0.19-0.80) and Eastern Cooperative Oncology Group (ECOG) performance status (HR 0.26, 95% CI 0.13-0.50) were independently associated with PFS in the ICI plus chemotherapy group.

"The new hypothesis ... is that when ICIs are combined with chemotherapy, the possible detrimental outcomes of PPI appear attenuated, hinting at a futuristic approach of treatment selection based not only on clinical pathological factors, but also on concomitant medications," wrote Alessio Stefani, MD, and Emilio Bria, MD, both of the Comprehensive Cancer Center, Fondazione Policlinico Universitario A. Gemelli, IRCCS, in Rome, Italy, in an .

However, they suggested the role of PPI (as well as that of other concomitant medications) needs to be confirmed in prospective trials, taking into account certain confounding factors such as advanced age, poorer ECOG performance status, the use of steroids, and additional comorbidities.

"While waiting for conclusive results, we should consider the real necessity of PPI prescription in every patient on a daily basis when starting anticancer therapies with potential interference (such as ICI) and evaluate their interruption when feasible," they wrote.

This retrospective cohort study was conducted at 13 institutions in Japan and included patients with advanced NSCLC (stage IV, including postoperative recurrence) who had received pembrolizumab monotherapy or ICI plus chemotherapy as the initial treatment between March 2017 and December 2020. Of the 425 patients in the study, 271 were treated with pembrolizumab therapy as first-line treatment (median age 72 years, 79% men) and 154 were treated with ICI plus chemotherapy (median age 69 years, 79% men).

Compared with the ICI plus chemotherapy group, the pembrolizumab monotherapy group had significantly higher proportions of older patients, patients aged 75 years or older, patients with poor ECOG performance status, patients administered PPIs, and patients administered antibiotics. PD-L1 TPS and cancer stage also differed between the two groups.

After propensity score match weighting, there were 131 matched peers in each treatment group.

Among the matched cohort, PPI users tended to have a median PFS significantly that was longer if they were in the ICI plus chemotherapy group compared with the monotherapy group (19.3 months vs 5.7 months, HR 0.38, 95% CI 0.20-0.72). These patients also had better median overall survival (OS; not reached vs 18.4 months, HR 0.43, 95% CI 0.20-0.92)

Patients without PPI exposure had similar PFS (median 18.8 months vs 10.6 months) and OS (not reached vs 29.9 months) between the ICI plus chemotherapy and ICI monotherapy groups.

The authors acknowledged the study had several limitations. For example, its multicenter retrospective design meant the possibility of selection bias couldn't be eliminated.

In addition, since the study included Japanese patients only, and previous studies have indicated differences in the gut microbiome structure between healthy individuals of different races and ethnicities, "this diversity may preclude the generalizability of our findings to patients from other countries," Yamada's group cautioned.

Comment