Abstracts

Rationale: Outcomes data, especially comparative data, derived from pragmatic studies can inform clinician and payer decisions. Medical charts are particularly rich data sources for examining treatment response. Drug adherence and persistence can be inferred from analyses of pharmacy claims data. Together, these two genres of research may provide a more complete picture of a drug’s clinical utility. Once-daily Trokendi XR® (topiramate extended-release capsules, Supernus Pharmaceuticals) was shown to be bioequivalent to q12h immediate-release topiramate (TPM-IR, Topamax®). A modest but significant difference in neuropsychometric change scores suggested the potential for greater cognitive tolerability with Trokendi XR. We conducted both types of studies to retrospectively assess outcomes– eg, tolerability, adherence, retention - vs TPM-IR. Methods: Multisite chart review: HIPAA-compliant, IRB-approved study. Trained nurses abstracted data from medical records of randomly selected patients. Inclusion criteria: Trokendi XR initiated between 08/2013 and 06/2014; age ≥6 yrs; record of Trokendi XR-initiating visit; ≥1 post-initiation visit. Subset analysis of patients previously treated with TPM-IR. Claims analysis: Parallel comparison of Trokendi XR and TPM-IR. National longitudinal medical/pharmacy claims data from commercial health plans. Inclusion criteria: Date of first pharmacy claim (index date) for Trokendi XR or TPM-IR (Topamax or generic TPM-IR) during 8/2013 and 10/2014; age ≥6 yrs old; continuous plan enrollment ≥12 months pre- and ≥6 months post-index. Key outcomes: Medication Possession Ratio (MPR); adherence (MPR ≥80%); treatment persistence. Results: Multisite chart review: 83/485 (17%) epilepsy patients treated with Trokendi XR; 41/83 (49%) previously treated with TPM-IR. 14/83 (17%) reported side effects due to Trokendi XR (moodiness/irritability, 5%; somnolence, 5%; cognitive effect, 2%; fatigue, 2%). Compared with previous TPM-IR treatment, Trokendi XR associated with significantly fewer side effects (22% vs 40%, p Claims analysis (epilepsy cohort): TPM-IR, n=1364; Trokendi XR, n=99. MPR (65% vs 50%, p < 0.001) and adherence (44% vs 28%, p < 0.001) significantly greater in Trokendi XR cohort. Persistence significantly (p < 0.005) greater in Trokendi XR cohort; difference emerged early (≤2 months) and increased over time. Conclusions: Different retrospective study designs provided complementary findings for Trokendi XR vs. TPM-IR utility. Consistent with a signal of potential differences observed in healthy volunteers, Trokendi XR was associated with significantly greater tolerability, particularly cognitive tolerability, when compared with previous TPM-IR treatment (chart review). Greater tolerability may account, in part,  for the finding of greater persistence with Trokendi XR that emerged during treatment initiation (claims data). The axis of tolerability-adherence-persistence suggests potential for greater overall effectiveness with Trokendi XR and warrants further study. Funding: Studies sponsored by Supernus Pharmaceuticals, Inc.