A retrospective survival analysis of Glioblastoma patients treated with selective serotonin reuptake inhibitors

Abstract

Glioblastoma (GBM) is the most common and aggressive form of malignant glioma in adults with a median overall survival (OS) time of 16-18 months and a median age of diagnosis at 64 years old. Recent work has suggested that depression and psychosocial distress are associated with worse outcomes in patients with GBM. We therefore hypothesized that the targeted neutralization of psychosocial distress with selective serotonin reuptake inhibitor (SSRI) antidepressant treatment would be associated with a longer OS among patients with GBM. To address this hypothesis, we retrospectively studied the association between adjuvant SSRI usage and OS in GBM patients treated by Northwestern Medicine-affiliated providers. The medical records of 497 GBM patients were analyzed after extraction from the Northwestern Medicine Enterprise Data Warehouse. Data were retrospectively studied using a multivariable Cox model with SSRI use defined as a time-dependent variable for estimating the association with OS. Of the 497 patients, 315 individuals died, while 182 were censored due to the loss of follow-up or were alive at the end of our study. Of the 497 patients, 151 had a recorded use of SSRI treatment during the disease course. Unexpectedly, SSRI usage was not associated with an OS effect in both naïve (HR = 0.81, 95% CI = 0.64-1.03) and adjusted time-dependent (HR = 1.26, 95% CI = 0.97-1.63) Cox models. Ultimately, we failed to find an association between SSRI treatment and an improved OS of patients with GBM. Additional work is necessary for understanding the potential therapeutic effects of SSRIs when combined with other treatment approaches, and immunotherapies in particular, for subjects with GBM.

Keywords: Antidepressant; Biobehavioral; Depression; Glioma; Immunosuppression; Psychosocial.

Fig. 1

Inclusion and exclusion criteria for the analysis of subjects diagnosed with glioblastoma. Patient exclusions are noted and the division of patients who used SSRIs versus those who did not are summarized. Censored patients had no recorded death date and were censored at the date of their last appointment or, if they were still alive, at study end. TMZ, temozolomide; SSRI, selective serotonin reuptake inhibitor; NM, Northwestern Medicine.

Fig. 2

Distribution of SSRI use in patients. Patients were sorted by length of time to censor or death. For each patient: death is indicated by a black dot; time on SSRI is indicated by an orange line; time not on SSRI is indicated by a blue line. No clear pattern is visible between length of time on SSRI and length of follow-up.

Fig. 3

Trends in landmark analysis hazard ratios. The analyses presented in Table 1 were performed at the 1st quartile (202 days), median (395 days), and 3rd quartile (704 days) of follow-up. The boxes show follow-up times for all patients with the subset of the sample used in each of the final analyses colored red and blue, indicating SSRI use up to the landmark time.