Clinical Benefit From Molecularly Guided Cancer Care: A Matching-Adjusted Indirect Comparison of Larotrectinib Versus Standard of Care.

Abstract

Purpose: Widespread adoption of the tropomyosin receptor kinase (TRK) inhibitor larotrectinib has been hampered by limited comparisons against non-TRK inhibitor standard of care (SoC) regimens because of the rarity of TRK fusions. Cross-trial comparisons may be facilitated using matching-adjusted indirect comparison (MAIC), a validated methodology that balances baseline population characteristics.

Materials and methods: MAIC was conducted using individual patient-level data from three larotrectinib trials, compared with aggregate real-world data (RWD) from adult patients with locally advanced/metastatic TRK fusion-positive cancer in the Hartwig Medical Foundation database. Patient populations were matched on available clinical characteristics. Estimates of overall survival (OS; defined as the time from start of first postbiopsy treatment [or larotrectinib] to death) were compared between larotrectinib and non-TRK inhibitor SoC.

Results: The analysis included 24 patients receiving nontargeted TRK inhibitor SoC and 120 receiving larotrectinib. After matching, median OS was 50.3 months (IQR, 23.3-not estimable) for larotrectinib versus 13 months (IQR, 6.4-18.3) for SoC; larotrectinib was associated with an 84% risk reduction of death (adjusted hazard ratio, 0.16 [95% CI, 0.07 to 0.36]). To account for the longer follow-up time in the larotrectinib group than in the SoC group (median: 56.7 v 23.2 months), a restricted mean survival analysis was conducted up to 26.2 months (largest observed event time in SoC arm), which showed median survival of 22.6 months for larotrectinib and 12.8 months for SoC (mean difference: 9.8 months [95% CI, 5.6 to 14.0]).

Conclusion: In patients with TRK fusion-positive cancer, larotrectinib was associated with a mean survival advantage of 9.8 months compared with non-TRK inhibitor SoC, corroborating favorable findings for larotrectinib from a previous analysis that used a different set of RWD.