TOPLINE:

A cohort study found that commonly prescribed medications did not seem to affect treatment outcomes in veterans with stage IV non-small cell lung cancer (NSCLC) who received immune checkpoint inhibitors (ICI); 14 of 20 medication classes showed no association with overall survival or time to next treatment. A negative control group of chemotherapy-treated patients helped confirm that any observed associations between medication use and survival likely reflected unmeasured prognostic differences rather than true drug-ICI interactions.

METHODOLOGY:

• ICI have transformed cancer therapy, but preclinical data suggest that common medications such as antibiotics, proton pump inhibitors (PPI), opioids, and statins might alter their efficacy. Prior observational findings were inconsistent and vulnerable to confounding. This study used a chemotherapy negative control design in a large cohort to distinguish true drug-ICI interactions from confounded associations.
• Using the Veterans Health Administration national database, researchers analyzed data of 3739 patients with newly diagnosed stage IV NSCLC who received first- or second-line ICI therapy from 2015 to 2024 and 6585 patients who received chemotherapy from 2005 to 2015.
• The ICI cohort included patients who received a limited set of ICI monotherapies (pembrolizumab or nivolumab) or combination chemoimmunotherapy. The negative control cohort included patients who received platinum‐based combination chemotherapy regimens or a limited set of monotherapy regimens (pemetrexed, docetaxel, gemcitabine, paclitaxel, or vinorelbine).
• Baseline use of 20 medication classes within 3 months before therapy was assessed along with an immunomodulatory drug score. The score assigned 1 point for recent use of antibiotics, 1 for PPI, and 2 for corticosteroids, with use within 1 month counted; the score was categorized as good (0), intermediate (1-2), or poor (3-4). The primary outcomes were overall survival and time to next treatment.
• The most common medications in the ICI group were opioid analgesics (53%), statins (43%), PPI (31%), and beta‐blockers (30%). The control cohort had similar use rates for most medications but had notably higher use rates for opioids and histamine type 2 receptor antagonists than the ICI cohort (72% and 43%, respectively).

TAKEAWAY:

• After propensity weighting, the use of 15 of 20 medication classes showed no association with overall survival, and the use of 14 of 20 classes showed no association with time to next treatment in the ICI cohort.
• The use of loop diuretics (adjusted hazard ratio [aHR], 1.25; P < .001), anticoagulants (aHR, 1.14; P = .0083), opioid analgesics (aHR, 1.28; P < .001), penicillin antibiotics (aHR, 1.17; P = .0042), and fluoroquinolone antibiotics (aHR, 1.13; P = .05) was associated with worse overall survival in the ICI cohort; similar effects were observed in the chemotherapy control cohort.
• The use of loop diuretics, anticoagulants, opioid analgesics, penicillin antibiotics, and fluoroquinolone antibiotics was associated with inferior time to next treatment outcomes in the ICI cohort. Although the same was observed in the control group, a small protective effect of statins was seen in the ICI group but not in the control group.
• In a multivariable Cox regression analysis, a poor vs good immunomodulatory drug score was associated with worse outcomes in both the ICI and chemotherapy control cohorts: aHRs for overall survival were 1.29 and 1.35, respectively, and aHRs for time to next treatment were 1.26 and 1.32, respectively (P < .001 for all).
• The “findings suggest that commonly used medications do not meaningfully influence ICI treatment outcomes in stage IV NSCLC, and prior reports of such associations likely reflect type I error and/or residual confounding,” the authors of the study wrote.

IN PRACTICE:

“Brinzevich et al. deliver one of the most convincing real‐world evaluations to date of how common medications influence ICI outcomes, demonstrating that when rigorous confounding control is paired with an explicit negative control design, many previously reported drug-ICI associations resolve into patterns consistent with prognosis rather than pharmacologic antagonism or potentiation,” said Parham Habibzadeh and Diwakar Davar of the University of Pittsburgh, Pittsburgh, in an accompanying editorial.

“Until prospective or mechanism‐anchored studies demonstrate otherwise, clinicians and patients can be reassured that appropriate concomitant medication use should proceed alongside ICIs, guided by clinical need and best practices in supportive care rather than fear of undermining treatment efficacy.”

SOURCE:

The study, led by Daria Brinzevich, MD, University of Michigan, Ann Arbor, Michigan, was published online in Cancer.

LIMITATIONS:

Limitations included the study’s observational design, inclusion of predominantly older male veterans, and possible misclassification bias. Additionally, patients selected for ICIs vs chemotherapy might differ in ways not fully corrected by propensity weighting.

DISCLOSURES:

The authors did not declare any funding information. Some authors reported receiving travel funding or consulting fees or having other ties with various sources. Detailed disclosures are noted in the original article.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.