(141) Assessing Exposure to Recently Approved Medications to Inform Feasibility for Signal Identification Analyses in the FDA’s Sentinel System

Background: At the time of approval, new molecular entity drugs and therapeutic biologics (i.e., new medications) have limited safety data. The U.S. Food and Drug Administration (FDA) primarily depends on spontaneous reporting systems and other passive information sources for identifying new safety signals after approval. Recently, the FDA began implementing an active surveillance approach by applying tree-based scan statistics (TBSS) to the Sentinel Distributed Database (SDD) to enhance post-marketing safety surveillance activities. However, the effectiveness of TBSS relies heavily on the size of the exposed population.

Objectives: To identify and evaluate utilization for medications approved by the FDA from 2017-2021 with adequate population exposure for effective post-marketing surveillance using TBSS.

Methods: For each of the 260 medications approved during the study period (48 in 2017, 60 in 2018, 48 in 2019, 53 in 2020, and 51 in 2021) individual cohorts, by year of approval, were created to analyze utilization in the SDD partners. The analyses were conducted from January of the year of approval to the most recent available data. An index exposure was defined as the first qualifying administration or dispensing of a given medication. We required individuals to have both medical and drug coverage for at least 183 days prior to their index exposure to allow for their characterization.

Results: There were seven medications with ≥100,000 users overall (3%): two approved in 2017 (semaglutide and netarsudil); four in 2018 (bictegravir/emtricitabine/tenofovir alafenamide, erenumab-aooe, galcanezumab-gnlm, and baloxavir marboxil); and one in 2019 (ubrogepant). A total of 41 medications (16%) had 10,000 to < 100,000 users (16 approved in 2017, 10 in 2018, 7 in 2019, 4 in 2020, and 4 in 2021). The number of medications with 5,000 to < 10,000 users were 19 (7%) (four approved in 2017, five in 2018, five in 2019, three in 2020, and two in 2021). The majority of the evaluated medications (193, 74%) had < 5,000 users.

Conclusions: Few of the analyzed medications accrued enough new users since approval to be feasible candidates for further consideration for signal identification analysis using TBSS. A majority of the identified medications with the highest utilization were approved on or before 2019. While these findings imply that most medications may not be optimal candidates for signal identification analyses in the initial years post-approval, several new medications have recently undergone successful TBSS analyses in the SDD. To facilitate the application of active surveillance strategies for products with lower initial uptake, periodic exposure assessments will be needed.