A pooled analysis of response to selective RET inhibitors among patients with medullary thyroid cancer with M918T versus non-M918T RET mutations.

6078 Background: Medullary thyroid cancer (MTC) accounts for 1 to 2% of thyroid cancers in the United States; RET alterations occur in >95% of hereditary and 50% of sporadic forms. Up to 80% of patients with sporadic MTC have somatic M918T RET mutations, which is associated with poor prognosis (1). The tyrosine kinase inhibitors (TKIs) cabozantinib and vandetanib are approved to treat patients with MTC regardless of RET status; however, retrospective analyses have suggested that there may be greater benefit in patients with M918T mutations (1,2). Newly approved therapies selpercatinib and pralsetinib, developed for patients with RET mutations, have demonstrated higher response rates than previous first line therapies. In this analysis, we examine the differences in overall response rate (ORR) between patients with MTC with RET M918T non- RET M918T mutations. Methods: An analysis of ORR in patients with MTC with RET M918T mutations with non-M918T mutations was conducted using the efficacy populations used to support the approvals of pralsetinib and selpercatinib using the following groups: Patients who received prior cabozantinib or vandetanib (referred to as “previously treated”). Patients with no prior cabozantinib or vandetanib (“TKI naïve”). All patients regardless of prior therapy. Results: Exploratory analysis of ORR of pooled population of Selpercatinib and Pralsetinib in patients with MTC with RET M918T mutations and non-M918T mutations. 1 Prior vandetanib or cabozantinib. 2 No prior vandetanib or cabozantinib. Two groups of patients were analyzed ( RET M918T mutation and RET non-M918T mutation), with subgroups with respect to prior treatment. Among all patients regardless of prior therapy, the ORR was similar between M918T non-M918T groups. Among previously treated patients, the ORR was lower in the M918T group vs. the non-M918T group, while in the TKI naïve group the ORR was higher in the M918T groups vs the non-M918T group although the 95% CIs overlap in both comparisons. Conclusions: There were no major differences in ORR among mutational subtypes in patients with MTC treated with RET inhibitors, regardless of prior therapy. ORR was similar between patients with M918T and non-M918T mutations. Additional experience in ongoing clinical studies may provide additional data regarding responses across specific mutation types. References: 1.Sherman SI et al “Correlative analyses of RET and RAS mutations in a phase 3 trial of cabozantinib...” Cancer. 2016;122(24):3856-3864. 2.Wells SA Jr et al “Vandetanib in patients with locally advanced or metastatic medullary thyroid cancer...” J Clin Oncol. 2012;30(2):134-41.[Table: see text]

Prognostic factors in RET dependent cancers treateded with RET inhibitors in early phase clinical trials.

3117 Background: Activation of the RET proto-oncogene has been identified in multiple cancer types, for example, gene rearrangements in non-small cell lung cancer (NSCLC) and papillary thyroid cancer (PTC) and activating mutations in medullary thyroid cancer (MTC), amongst others. The recent FDA approval of two highly selective RET inhibitors, selpercatinib and pralsetinib has changed the treatment paradigm of RET-driven cancers, but the significance of historical prognostic factors is unknown. Herein, we analyzed the outcomes of patients with RET-altered cancers enrolled in phase I trials and assess the utility of prognostic scores. Methods: A retrospective analysis was performed of patients treated with selective RET inhibitors at the MD Anderson Cancer Center (MDACC). Baseline clinical factors and survival data were assessed. Overall and progression free survival (OS and PFS) were estimated using the Kaplan-Meier method and multivariable Cox models were constructed. For all a p-value of < 0.05 was consider significant. Results: Among 126 patients, median age was 58 years (range, 15-82), most with ECOG 0-1 (n = 124). RET-mutant MTC was most frequent (n = 81), followed by RET fusion-positive NSCLC (n = 30) and RET fusion positive thyroid cancer (n = 9). RET fusion partners were KIF5B (n = 17), CCD6 (n = 12) and NCOA4 (n = 7). RET M918T mutation was the most frequent (n = 50, 63%). Most patients were treated in the relapsed/refractory (R/R) setting (n = 85) and received a median of 1 prior line of therapy (range, 0-11). Median follow up was 20 months (range, 1-43). The estimated median PFS and OS were 24 and 35 months, respectively. Overall objective response rate was 64% (81/126), 2 complete response, 79 partial response, 32 had stable disease (25%) and 13 had progressive disease (PD). The following were associated with shorter PFS and OS: age ≥50 years (p < 0.05), albumin < 4 g/dL (p < 0.01), brain metastases (p < 0.0001), hemoglobin < 12 g/dL ( < 0.05), LDH > normal (p < 0.05), WBC ≥8 (p < 0.01), PD (p < 0.0001) and NSCLC (p < 0.01). The M918T mutation and ECOG > 0 were associated with shorter OS but not PFS (p < 0.05). > 3 metastatic sites and R/R disease were associated with inferior PFS (p = 0.04 and p = 0.01, respectively) but not OS. The Royal Marsden Hospital (RMH) and MDACC prognostic scores were significantly associated with PFS and OS (p < 0.01). In multivariable models including all variables significantly associated with PFS and OS (excluding LDH as this was only tested in 58 patients) albumin < 4 (HR 6.10, p = 0.013), brain metastases (HR 4.90, p = 0.027) and WBC ≥8 (HR 4.67, p = 0.031) were associated with inferior OS. NSCLC was significantly associated with inferior PFS (HR 5.45, p = 0.02). Conclusions: The RMH and MDACC prognostic scores predict OS in RET-aberrant cancers treated on early phase trials. Low albumin, WBC > 8 and brain metastases are significantly associated with inferior survival.