PD1 inhibitors as well as PD1/CTLA4 combinations have shown remarkable efficacy in the first-line treatment of metastatic melanoma. The impact of many concomitant medications on the clinical outcomes from PD1 based therapies remains elusive.
We retrospectively analyzed 140 patients included in the Checkmate 069 phase II clinical trial as a discovery cohort, comparing ipilimumab monotherapy with ipilimumab combined with nivolumab. We compared response rates, progression-free survival and overall survival of patients treated or not with 11 different classes of co-medications at immune therapy initiation. Disease stage, LDH levels, BRAF status, sex, age, and body mass index were also compared. Furthermore, a protein array was performed for 440 analytes in a subset of 135 patients for whom pretreatment serum was available. We validated the impact of proton pump inhibitors in an independent cohort of 68 PD1 monotherapy (pembrolizumab or nivolumab) treated patients.
In univariate analysis, baseline proton pump inhibitor treatment almost halved the objective response rates, reduced progression-free and overall survival of patients treated with ipilimumab and nivolumab but not with ipilimumab alone. This effect was maintained when accounted for multiple comparisons and in a multivariate analysis. Pretreatment serum protein analysis showed increased NCAM1 and CSF3R levels in PPI users. We found increased baseline blood leukocyte and neutrophil levels in correlation with PPI use. The results were confirmed in an independent cohort of 68, first-line melanoma patients.
Our analysis shows that proton pump inhibitors could negatively impact on the benefit from PD1 based therapies both for monotherapy and also for ipilimumab and nivolumab combination therapy. PPIs might establish a unique inflammatory immune status, prior to immune therapy initiation that interferes with treatment efficacy. These results suggest that if possible PPIs should be avoided in patients who are destined for PD1-based immunetherapies. Also, the results will have important implication for design of future clinical trials.
Department of Oncology, CHUV, Lausanne, Switzerland.
Has not received any funding.
M. Wind-Rotolo: Employee of BMS. All other authors have declared no conflicts of interest.