[Elsevier] Association of proton pump inhibitor use with outcome of patients with recurrent glioblastoma Copy

Background: The use of proton pump inhibitors with strong aldehyde dehydrogenase 1A1-activating properties (PA-PPI) has recently been associated with inferior outcomes in patients with newly diagnosed glioblastoma. While the mechanisms mediating such outcome associations remain incompletely understood, no such data have been generated for patients with recurrent glioblastoma.

Materials and methods: We conducted a pooled analysis of four clinical trials in the setting of first recurrence of glioblastoma: DIRECTOR (NCT00941460), EORTC 1410 (NCT02343406), EORTC 1608 (arm C) (NCT03224104), and EORTC 26034 (NCT00086879). We assessed PA-PPI use at baseline and at two defined landmarks, week 8 and week 16. Cox models were used to identify associations of PA-PPI use with outcome.

Results: On univariate analysis, PA-PPI use was associated with inferior overall survival, but not progression-free survival, at baseline and at both landmarks. However, PA-PPI use was no longer prognostic in multivariable analyses controlling for major prognostic factors. Omission of distinct prognostic factors from the multivariate Cox models revealed that steroid use made PA-PPI use insignificant. Indeed, PA-PPI use was associated with steroid use.

Discussion: The present analysis of prognostic outcome associations of PA-PPI use in patients with recurrent glioblastoma provides no evidence that PA-PPI use may compromise outcome in this setting. The disconnect between clear outcome associations in the first-line setting, without a link to MGMT promoter methylation, versus no associations in the recurrent setting may be linked to changing prescriptions patterns, but requires further studies. Since alternative medications are available, the indications for PA-PPI prescription should probably be narrowed in patients with glioblastoma.