6A-04 - Evaluating the real-world effect of CYP2C9-inhibiting selective serotonin reuptake inhibitors on the incidence of serious hypoglycemia among Medicare-insured sulfonylurea users

Background: Diabetes and depression are both highly prevalent among older Americans. Certain selective serotonin reuptake inhibitors (SSRIs) inhibit activity of the CYP2C9 enzyme responsible for the metabolism of sulfonylureas (SU), potentially increasing their blood concentrations and the risk of hypoglycemia. However, little is published on the real-world clinical impact of this interaction.

Objectives: To estimate the effect of concomitant use of CYP2C9-inhibiting SSRIs on the incidence of serious hypoglycemia, compared to use of non-CYP2C9 inhibiting SSRIs, among SU users.

Methods: Using a 20% national random sample of fee-for-service Medicare claims (2007-2019), we identified prevalent SU users with at least one year (baseline period) of continuous enrollment in Parts A, B, and D prior to initiating an SSRI. SSRIs inhibiting CYP2C9 (exposed) included fluoxetine, fluvoxamine, sertraline, and paroxetine, while non-inhibitors (unexposed) included escitalopram and citalopram. We followed patients from SSRI initiation until the outcome, disenrollment, or discontinuation of either SU or SSRI. We used a validated claims-based algorithm to identify serious hypoglycemia and a more sensitive definition to assess a secondary outcome of any hypoglycemic event. We conducted a subgroup analysis among patients with evidence of hypoglycemia during baseline, a strong risk factor. We used stabilized inverse of probability weighting (IPTW) to balance baseline covariates. We estimated incidence rates and hazard ratios with 95% confidence intervals (CI), treating death as a competing risk using the Aalen-Johansen estimator.

Results: We identified 95,874 concomitant users of SU and SSRI (50% exposed). Baseline characteristics appeared well-balanced prior to IPTW (average absolute standardized mean difference < 0.01). The unadjusted incidence of serious hypoglycemia was 9.5 (exposed) and 10.7 (unexposed) per 1,000 person-years. We did not observe an increased hazard of serious (adjusted HR, 0.93; 95%CI, 0.79-1.09) or any (aHR, 0.96; 95%CI, 0.91-1.02) hypoglycemia among CYP2C9 inhibiting SSRI users compared to users of non-CYP2C9 inhibiting SSRIs. Among patients with baseline hypoglycemia, unadjusted incidence of serious hypoglycemia was 20.9 (exposed) and 18.8 (unexposed) per 1,000 person-years. Exposure was associated with an increased, but less precise, hazard of serious hypoglycemia (aHR, 1.10; 95%CI, 0.77-1.56).

Conclusions: Overall, serious hypoglycemia was rare in concomitant users of SU and SSRI and not increased by use of CYP2C9 inhibiting SSRIs. However, among patients with prior hypoglycemia, the incidence of serious hypoglycemia was higher and patients initiating CYP2C9 inhibiting SSRIs had a slightly higher incidence of hypoglycemia.