Age-stratified survival analyses of pharmacological treatments in patients with glioblastoma

Abstract Glioblastoma is the most common primary malignant brain tumor in adults. Age is a prognostic factor for grade IV glioma (GIV) survival, where older adults 65 and older, tend to have quicker mortality compared to adults 18-64. This may be due to intrinsic aspects of the aged host, including age-dependent changes to the brain and immune systems. We studied pharmacologic management of GIV patients treated with medications affecting the brain, immune, and other systems for associations with age and survival. Electronic records for 1,208 GIV patients were collected at a single institution from 1/1/2000 to 9/8/2021 and used to construct a multivariable time-dependent cox model that examined intra-cohort hazards associated with 14 different categories of outpatient medications. Date of diagnosis was identified as the start-time and subject death was used to define the event-time. Patients who were alive or lost to follow-up were censored to last recorded date. Survival time was calculated as the difference between diagnosis date and event time or censorship. For positive control, we confirmed that younger and older patients experienced an improved hazard with temozolomide. Strikingly, younger, but not older patients, experienced increased hazard with a steroid prescription (1.46 [1.12,1.90], **). In contrast, older, but not younger patients experienced improvements in hazard with prescription of neurological medications including benzodiazepines (0.47 [0.27, 0.80] **) and anti-convulsants (0.28 [0.08, 0.90] *), in addition to heart failure treatments including angiotensin II inhibitors (0.48 [0.27, 0.85] *) and diuretics (0.30 [0.15, 0.87] *). Patients prescribed antidepressants didn’t show improved hazard across younger and older GIV patients, which consistent with our previously published work. These data indicate novel associative benefits due to prescription of select pharmacologic agents that vary with patient age. This work may be used to better understand therapeutic approaches that selectively benefit younger vs. older adults with GIV.