Rapid Fire Oral Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2026

Proton pump inhibitor use and atezolizumab efficacy and safety: an individual participant data meta-analysis of 12 randomised trials (145979)

Lee Li 1 , Adel Shahnam 2 , Richard Woodman 1 , Yuan Gao 1 , Natansh Modi 3 , Ganessan Kichenadasse 1 , Lewis Murray 4 , Andrew Rowland 1 , Ashley Hopkins 1 , Michael Sorich 1
  1. College of Medicine and Public Health, Flinders University , Bedford Park, SA, Australia
  2. Memorial Sloan Kettering Cancer Center, New York, NY, United States
  3. School of Pharmacy & Biomedical Science, Adelaide University, Adelaide, SA, Australia
  4. Royal Adelaide Hospital, Adelaide, SA, Australia

Background: Proton pump inhibitors (PPIs), commonly used in cancer care, may alter drug absorption, the gut microbiome and immune function. Our earlier analyses of atezolizumab trials in urothelial carcinoma (UC) and non-small cell lung cancer (NSCLC) suggested that PPI use attenuated survival benefit. We extended this work across six cancers, concurrently assessing efficacy, toxicity, alternative acid suppression and confounding.

Methods: We conducted a two-stage individual participant data meta-analysis of 12 randomised atezolizumab trials in NSCLC, small cell lung cancer, melanoma, hepatocellular carcinoma, UC and renal cell carcinoma. Sustained baseline PPI use was evaluated. Overall survival (OS), progression-free survival (PFS) and grade ≥3 adverse events (AEs) were analysed using trial-specific Cox models containing treatment, PPI use and treatment–PPI interaction terms, followed by random-effects meta-analysis. Immune-mediated adverse events (IMAEs) were evaluated among atezolizumab-treated patients. Sensitivity analyses used covariate adjustment and inverse-probability-of-treatment weighting. Histamine-2 receptor antagonists (H2As) were explored as an alternative exposure.

Results: PPI use was common (25%); 61% of users had no documented indication. PPI use was associated with reduced relative OS benefit from atezolizumab-based therapy (HRinteraction = 1.28, 95% CI 1.13–1.46) in the intention-to-treat population (N=8526), and increased relative hazard of grade ≥3 AEs (HRinteraction =1.17, 95% CI 1.03–1.33) in the safety population (N=8336), with minimal between-trial heterogeneity. No significant PFS effect modification was observed (HRinteraction = 1.11, 95% CI 0.94–1.33), with moderate heterogeneity. PPI use was not significantly associated with IMAEs (HR 1.19, 95% CI 0.99–1.43), although power was limited. H2A use (3%) showed no evidence of effect modification (OS: HRinteraction=0.83, 95% CI 0.60-1.16).

Conclusion: PPI use was associated with reduced OS benefit and greater severe toxicity from atezolizumab-based therapy. Although causality cannot be established, these findings support careful review of PPI indications in patients commencing immune checkpoint inhibitors.