Time to Treatment Discontinuation and Cost Effectiveness of Third-Line Therapies in Advanced Colorectal Cancer: Real-World Evidence from the NIH All of Us Research Program.

[BACKGROUND] Treatment options for patients with advanced colorectal cancer who progress after standard therapies remain limited, and real-world evidence on treatment durability and economic outcomes in later lines of therapy is needed to inform clinical and policy decision making. This study evaluated real-world outcomes, healthcare utilization, and cost effectiveness of immune checkpoint inhibitors (ICI) compared with regorafenib or trifluridine/tipiracil (Rego/Tri) in third-line (3L) advanced colorectal cancer (CRC), leveraging the nationally distributed NIH All of Us Research Program.

[METHODS] We conducted a retrospective cohort cost-utility analysis using All of Us Version 8 Curated Data Repository. Eligible adults (≥ 18 years) with advanced CRC initiated ICI (pembrolizumab, nivolumab ± ipilimumab) or Rego/Tri during 2017-2023. Outcomes included time to treatment discontinuation (TTD), healthcare utilization, costs, and quality-adjusted life years (QALYs; base-case utility = 0.75). Costs were estimated from CMS 2023 schedules (Part B ASP for infused agents, Part D unit costs for orals; inpatient/outpatient fee schedules for non-drug services), applied per cycle and proportionally for partial cycles. Analyses were conducted from the US Medicare payer perspective, with the time horizon limited to observed TTD and no discounting applied (< 1 year). Incremental cost-effectiveness ratios (ICERs), incremental net monetary benefit (INMB), and cost-effectiveness acceptability curves (CEACs) were estimated using nonparametric bootstrapping (1000 replicates). Sensitivity analyses varied utility weights (0.6-1.0).

[RESULTS] Among 82 patients (ICI n = 50; Rego/Tri n = 32), median TTD was significantly longer with ICI (8.0 vs 1.4 months; HR 0.49, 95% CI 0.30-0.79; p = 0.0037). Non-drug utilization costs were not significantly different, but ICI drug costs were higher (US$207,558 vs US$119,805; p = 0.001). Mean total costs were US$476,729 for ICI versus US$264,225 for Rego/Tri (p = 0.018). QALYs were higher for ICI (0.84 vs 0.43; p < 0.001). The ICER was US$518,000 per QALY gained. CEAC analysis demonstrated ~ 0% probability of cost effectiveness at US$150,000/QALY, ~ 50% at US$500,000/QALY, and ~ 80% at ≥ US$1,000,000/QALY. Sensitivity analyses yielded consistent findings.

[CONCLUSIONS] In this exploratory analysis using the novel NIH All of Us Research Program, ICI therapy was associated with longer treatment duration and greater QALYs but substantially higher costs compared with Rego/Tri. ICI was unlikely to be cost effective at conventional US thresholds but may be favorable at higher willingness-to-pay values. These findings highlight both the promise of leveraging the All of Us Research Program for nationally distributed, demographically diverse pharmacoeconomic research and the challenges of interpreting real-world cost effectiveness in small cohorts. As the dataset matures and sample sizes increase, broader application of All of Us may provide valuable insights to inform precision oncology policy and value-based pricing strategies.