Association between clinical and disease characteristics and detectable or undetectable baseline ctDNA in patients with metastatic uveal melanoma.

Abstract

9537 Background: Tebentafusp, a bispecific soluble TCR specific for a gp100 peptide, is licensed for the treatment of metastatic uveal melanoma (mUM) in HLA-A*02:01+ individuals. 82% of patients (pts) with mUM in the IMCgp100-102 (previously treated; 2L+) and 61% in IMCgp100-202 (frontline; 1L) studies, respectively, had detectable baseline ctDNA. There is a strong relationship between decrease in ctDNA by 9 weeks and overall survival (OS) [Carvajal 2022; Hassel & Piperno-Neumann 2023]. Patients with undetectable baseline ctDNA had improved survival despite having macroscopic disease. In this study, we report the characteristics of baseline ctDNA undetectable vs detectable disease. Methods: Disease characteristics of 202 1L (Study 202) and 117 2L+ (Study 102) pts who had baseline undetectable or detectable ctDNA (unDNA vs detDNA) were compared in a post-hoc unplanned analysis. ctDNA detection was analysed at baseline and up to week 9 on tebentafusp using targeted mPCR-NGS assay for mutations in 15 genes including GNAQ, GNA11, SF3B1, CYSLTR2, PLCB4 and EIF1AX. Largest lesion size groups (≤ 3 cm, 3-8 cm, > 8 cm) were defined by American Joint Committee on Cancer 7th Ed. Presence of oncogenic mutations in GNAQ, GNA11, SF3B1 and BAP1 were assessed by WES on metastatic tissue biopsies. Results: Baseline ctDNA levels increased in association with size of largest lesion regardless of number of prior therapies. However, there was an association between disease burden (defined by largest lesion size) and unDNA vs detDNA in both studies, with all pts with largest liver lesion > 8 cm having detDNA. In Study 202, % of the population with unDNA vs detDNA was 50% vs 50% for ≤ 3cm and 27% vs 73% for 3.1-8 cm. While OS was similar for unDNA & detDNA in pts with ≤ 3cm lesions, OS was longer for pts with unDNA vs detDNA who had baseline tumor size 3.1-8 cm in both studies (Table). Within this 3.1-8 cm group, pts with unDNA tended to have lower LDH, AST, and ALP. Initial analysis showed no difference in the frequency of GNAQ, GNA11, SF3B1 or BAP1 mutations between unDNA and detDNA groups. Further data on somatic mutation defined groups and RNAseq gene expression in unDNA and detDNA groups will be presented. Conclusions: Patients with unDNA were more likely to have smaller disease burden than those with detDNA; however, many pts with unDNA had a significant burden of metastatic disease. Baseline unDNA was associated with improved OS outcome vs detDNA in pts with 3.1-8 cm baseline metastatic deposits, but not in pts with ≤ 3cm disease. No association was seen between oncogenic background of GNAQ/GNA11/SF3B1/BAP1 and unDNA vs detDNA. Clinical trial information: NCT02570308 ; NCT03070392 . [Table: see text]