Background: Selective RET inhibitors (RETi) have changed the paradigm for treatment of RET altered cancers. Current evidence on their efficacy comes from single-arm basket trials. In this study, we performed intrapatient efficacy comparison between approved selective RETi (selpercatinib and pralsetinib) and prior and/or subsequent therapies. We used growth modulation index (GMI) as a tool to compare efficacy, where a value of >1.33 is presumed a cutoff to establish superior efficacy. Methods: We included patients with RET-positive tumors who were treated at our institution as part of RETi clinical trials. We excluded patients with no other systemic therapy received before or after the RETi of interest. We also excluded patients who received both drugs to avoid overlapping efficacy bias, and patients who were treated for ≤ 30 days or had non-evaluable disease. Since some patients have been treated beyond progression or discontinued RETi due to intolerance, we used ratios for both time to treatment takeoff (TTT) and time to progression (TTP) to compare efficacy of RETi (at the RETi treatment line (n)) to prior therapy (n-1) and to subsequent therapy (n+1). GMI was defined as the ratio between TTT/TTT (n±1) or TTP to TTP (n±1). Results: We included 66 patients who received RETi and met our inclusion criteria [39 received selpercatinib and 27 received pralsetinib]. Most patients had GMI>1.33 using either TTT or TTP (61% (n=40) and 58% (n=38), respectively). The median GMI based on pre-RETi therapy was 2.1 and 1.6 (using TTT and TTP, respectively); while the median GMI based on post-RETi therapy was 4.9 and 4.4 (using TTT and TTP, respectively). Patients with GMI>1.33 were more likely to have PR as best response compared to patients with GMI<1.33 (85% vs 53%, p=0.005 using TTT; and 87% vs 54%, p=0.003 using TTP). GMI using TTT (n/n-1) was higher in patients with RET fusions compared to patients with RET mutations (3.7 vs 1.4, p=0.048). GMI using TTT and GMI using TTP(n/n+1) were lower in patients with WBCs <8 at baseline (2.6 vs 13.8, p=0.014; 2.6 vs 11.1, p=0.014; for TTT and TTP) and GI cancer diagnosis (0.4 vs 9.3, p=0.042; 0.4 vs 10.4, p=0.042; for TTT and TTP). Conclusions: Intrapatient efficacy comparisons are feasible using GMI calculations and provide a proof of concept on the favorability of selective RETi compared to other systemic therapies.

* Compared to whichever line had better TTT and TTP.