(067) Exposure Patterns of Immune Checkpoint Inhibitors Among Older Medicare Beneficiaries with Non-Small Cell Lung Cancer

Publication Number: 730

Background: Immune checkpoint inhibitors (ICI) are a class of immunotherapies that have dramatically improved the prognosis of patients with cancer. Using real world data to assess effectiveness and safety of ICIs require careful assessment of exposure patterns due to variable dosing schedule and high incidence of non-life threatening immune related adverse events that can alter the use patterns.

Objectives: To assess ICI exposure based on multiple definitions of ICI use among older Medicare beneficiaries with non-small cell lung cancer (NSCLC).

Methods: Using Surveillance, Epidemiology and End Results Program (SEER) cancer registry (2011-2017) linked with US Medicare data (2010-2019), we identified all adults ≥ 66 years with new diagnosis of NSCLC stage I-IV without any other cancer diagnosis who received ICI. We described patient demographics and stage at diagnosis. To describe the exposure patterns of ICIs, we assessed four treatment patterns allowing no more than 4, 6, 8, or 10 weeks between ICI doses. We described the time from diagnosis to the first ICI, time from the first dose to the last dose, number of total doses per the treatment course, and time between each dose in the overall cohort and subgroups by cancer stage.

Results: Among 73,412 adults with stage I-IV NSCLC, 7,954 (11%) received at least one dose of ICI after diagnosis. Among the 7,954 ICI users (median age 73, 51% male, 83% White, 11% stage I, 5% stage II, 25% stage III, 59% stage IV), the median time from diagnosis to ICI was 199 days (IQR 62-501) in overall cohort, 701 days in Stage I, 484 in Stage II, 325 days in Stage III and 96 days in Stage IV. In assessing on-ICI as no more than 4, 6, 8, and 10 weeks gaps between treatments, the median time on-ICI was 29 days (IQR 0-87), 63 days (IQR 21-154), 75 days (21-184), and 84 days (24-204). The median number of ICI doses received using treatment gaps of no more than 4, 6, 8, and 10 weeks were 3 (IQR 2-7), 5, (IQR 2-10), 5 (IQR 2-12), and 6 (IQR 2-12), respectively. Median number of doses and time between absolute first and last ICI (disregarding time gaps between ICI) was 6 doses (IQR 3-15) and 107 days (IQR 29-317). There was no difference in median length of ICI treatment or number of ICI doses received when stratified by stage.

Conclusions: We demonstrated variability in time to ICI and time exposed to ICI based on different ICI treatment patterns. ICI exposures should be defined based on observed use patterns, the population characteristics, and biologically plausible exposure window for specific outcomes under investigation. Further studies are needed to evaluate varying time-exposure windows and the impact of these time windows on estimating ICI safety and effectiveness.