Oncology · Medical Literature
[Comment] Co-medications and gut microbiome in NSCLC immunotherapy
Abstract
In a post-hoc analysis of the PACIFIC trial published in The Lancet Oncology, Leonardo Brunetti and colleagues1 found that baseline proton pump inhibitor and antibiotic use was associated with worse progression-free survival in patients with stage III non-small-cell lung cancer (NSCLC) treated with durvalumab after chemoradiotherapy, but not in those treated with placebo. This treatment-specific effect suggests an interaction with immunotherapy that is potentially mediated by microbiome disruption.
Clinical relevance
Does baseline exposure to proton pump inhibitors (PPIs) or antibiotics within 30 days before treatment initiation adversely affect progression-free survival (PFS) and overall survival (OS) in patients with unresectable stage III NSCLC receiving consolidation durvalumab after concurrent chemoradiotherapy, and does this effect differ from that observed in placebo-treated patients? __SNAPSHOT__ {"studyDesign":"Post-hoc analysis of a randomized, double-blind, placebo-controlled phase 3 trial (PACIFIC; NCT02125461), using the final 5-year data cutoff. This is a commentary by Sridhar and Minna summarizing the Brunetti et al. analysis published in the same journal issue.","setting":"Multi-national, multicenter; the PACIFIC trial was conducted at academic and community oncology centers internationally.","sampleSize":"n=660 of 713 originally randomized patients who consented to exploratory analyses (durvalumab: n=449; placebo: n=211)","population":"Adults ≥18 years with histologically confirmed unresectable stage III squamous or non-squamous NSCLC, WHO performance status 0–1, with no disease progression after ≥2 cycles of concurrent platinum-based chemoradiotherapy. The cohort was predominantly male (68.6%), White (64.2%), with an Asian minority (23.1%).","intervention":"Consolidation durvalumab 10 mg/kg IV every 2 weeks for up to 12 months, initiated 1–42 days after completion of chemoradiotherapy. Exposure of interest: baseline PPI use (any oral or IV administration within 30 days before randomization) and baseline antibiotic use (oral or IV administration within 30 days before randomization).","comparator":"Matched placebo administered on the same schedule. The same co-medication exposures (PPI and antibiotic use) were evaluated in the placebo arm to assess treatment-specificity of any observed associations.","primaryOutcome":"Progression-free survival (PFS) and overall survival (OS) stratified by baseline PPI and antibiotic exposure, analyzed separately within the durvalumab and placebo arms. An exploratory ancillary cohort used Toposcore (a metagenomic-based microbiome dysbiosis metric) to investigate potential biological mechanisms.","followUp":"Median follow-up: 62.4 months (IQR 61.9–63.2 months)","fundingSource":"No direct funding reported. Multiple co-authors disclosed financial relationships with AstraZeneca (manufacturer of durvalumab) and other pharmaceutical companies."}
Key findings
PPI exposure was present in 40% of the 660-patient cohort (n=263) and antibiotic exposure in 10% (n=69). In the durvalumab arm, baseline PPI exposure was associated with significantly shorter PFS (9.4 months [95% CI 7.6–13.7] vs. 17.2 months [95% CI 15.4–23.2]; HR 1.57 [95% CI 1.28–1.93]; p<0.0001) and shorter OS (33.0 months [95% CI 21.9–46.7] vs. 57.9 months [95% CI 48.7–not computable]; HR 1.66 [95% CI 1.30–2.13]; p<0.0001) compared with no PPI exposure. Baseline antibiotic exposure in the durvalumab arm was similarly associated with shorter PFS (9.2 months [95% CI 4.9–18.1] vs. 15.6 months [95% CI 13.6–17.6]; HR 1.50 [95% CI 1.08–2.10]; p=0.016); the OS association for antibiotics was not statistically significant, possibly reflecting smaller group size, greater heterogeneity of antibiotic exposures, and long post-progression survival. In contrast, no significant differences in PFS or OS according to PPI or antibiotic exposure were observed in the placebo arm. Interaction analyses supported a treatment-dependent (rather than a generic prognostic) pattern. Multivariable analyses including age, sex, smoking, PD-L1 expression, histology, EGFR status, ECOG-PS, and TNM stage confirmed PPI and antibiotic exposure as independent predictors of shorter PFS in durvalumab-treated patients. NNH is not calculable from available summary data. In an exploratory ancillary cohort, baseline dysbiosis measured by Toposcore (a metagenomic-based standardized microbiome metric) was associated with inferior PFS and was directionally enriched in patients with antibiotic exposure; PPI exposure alone did not show a clear microbiome shift, though combined PPI plus antibiotic exposure suggested a numerical additive dysbiotic effect. Overall median PFS was 14.4 months (durvalumab) vs. 8.2 months (placebo), and overall median OS was 47.7 months (durvalumab) vs. 27.8 months (placebo) across the full cohort.
Practice implications
This post-hoc analysis provides the first signal, from a randomized-trial dataset, that baseline PPI and antibiotic exposure are associated with meaningfully worse PFS and OS specifically in patients receiving durvalumab in the curative-intent stage III NSCLC setting—a finding not observed in the placebo arm, implicating an immunotherapy-specific interaction rather than nonspecific prognostic confounding. The treatment-by-exposure interaction analysis strengthens this interpretation, though the post-hoc, non-prespecified nature of all analyses means causality cannot be established. The findings extend signals previously reported only in the metastatic setting to earlier-stage disease. The absolute differences in PFS (approximately 7–8 months) and OS (approximately 25 months) are clinically substantial if confirmed prospectively; however, confounding by indication is a critical unresolved concern, as indications for PPI and antibiotic use, their timing, duration, and adherence were not available. Patients exposed to PPIs were more likely to be older, have higher ECOG-PS, and have PD-L1 expression >1%, all potential confounders not fully accounted for. The Toposcore microbiome findings are exploratory, derived from a small ancillary subset, and are hypothesis-generating only. Microbiome-based mitigation strategies (dietary modulation, probiotics, fecal microbiota transplantation) remain investigational. Per the authors themselves, the results should not be interpreted as a mandate to universally withhold PPIs or antibiotics from patients receiving durvalumab; rather, thoughtful prescribing—evaluating genuine clinical necessity of each medication—is the appropriate takeaway pending prospective validation. This summary is based on the published abstract and secondary reporting; full results tables and methods are not yet reviewed in detail. NOTE: The article under review by Sridhar and Minna is a Comment piece in The Lancet Oncology, not the primary research report; the clinical data summarized here derive from the underlying Brunetti et al. analysis. __LIMITATIONS__ Post-hoc, non-prespecified analysis: all findings are hypothesis-generating and subject to type I error inflation from multiple comparisons.,Confounding by indication: indications, timing, duration, and adherence for PPIs and antibiotics were not available, making confounding by indication impossible to exclude—patients on PPIs or antibiotics may be systemically sicker or have a higher comorbidity burden driving worse outcomes.,Residual confounding: despite multivariable adjustment for standard clinical covariates, unmeasured confounders (e.g., comorbidities, performance status trajectory, polypharmacy) cannot be excluded.,PPI exposure was associated with older age and higher ECOG-PS, which are known independent prognostic factors.,Antibiotic OS association was not statistically significant, likely due to small exposed subgroup (n≈69, 10% of cohort) and heterogeneity in antibiotic class, spectrum, and duration.,Toposcore microbiome data were available in only a small ancillary subcohort with 11 patients having longitudinal samples, limiting mechanistic conclusions.,Exposure window restricted to 30 days before randomization may misclassify patients with episodic or longer-term use outside this window.,Generalizability: the cohort was predominantly White and male; representation of Black/African American patients was very low (0.7%), limiting generalizability across racial and ethnic groups.,The review is of a Comment article, not the primary research report; data are as reported in published abstracts and secondary sources.,Multiple co-authors with financial ties to AstraZeneca, the durvalumab manufacturer, represent a potential source of reporting or interpretation bias. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ Baseline PPI and antibiotic use within 30 days of treatment initiation was associated with substantially worse PFS and OS with durvalumab—but not placebo—in patients with unresectable stage III NSCLC, supporting a treatment-specific interaction potentially mediated by gut microbiome disruption. These findings are hypothesis-generating from a post-hoc analysis and must be prospectively validated before informing prescribing guidelines. Clinicians should critically evaluate the medical necessity of PPIs and antibiotics in patients receiving consolidation immunotherapy, without categorically withholding agents that are genuinely indicated.