Treatment with tebentafusp beyond radiographic progressive disease (PD) in metastatic uveal melanoma (mUM)

9585 Background: Tebentafusp (tebe) is the first T cell receptor therapeutic to demonstrate overall survival (OS) benefit in a randomized Phase 3 study vs investigator’s choice (IC) [ NCT03070392 ]. OS benefit was also observed in patients (pts) with best objective response (BOR) PD (HR 0.43), and in pts who had tumor growth ≥20% as best change in tumor size (HR 0.41), suggesting tebe-treated pts may exhibit atypical radiological responses and could benefit from treatment beyond radiographic progression (TBP), a well-established concept in immuno-oncology. Here we analyzed tumor kinetics and clinical benefit in pts treated with tebe beyond initial radiographic progression (TBP). Methods: 378 mUM pts were randomized 2:1 to tebe vs. IC. BOR was assessed by investigators using RECIST v1.1. TBP was permitted until: 1) additional ≥20% increase in tumor burden with absolute increase of ≥5 mm, or 2) unequivocal PD of non-target lesions; or 3) new non-measurable lesions. A Cox model adjusted for baseline covariates and for covariates at time of progression for TBP-eligible pts was used to compare survival post progression between those who did (TBP) and did not receive TBP (non-TBP). Stepwise selection of covariates (using p < 0.1 as the entry and staying criterion) was applied. Analysis performed on data cut-off 13Oct2020. Results: 183 tebe pts were eligible for TBP per protocol; 60% (109/183) received TBP with median duration of 8 wks. 21% of all tebe doses were administered as TBP. The proportion of pts with new lesions at initial progression (44% vs 57%) and median time to initial progression (2.9 mo vs 2.9 mo) were similar between TBP and non-TBP pts. Pts receiving TBP were more likely to have favorable key prognostic factors at baseline or at time of progression. After adjusting for these differences, a numerical benefit in post-progression OS favoring TBP was observed (HR 0.67, 95% CI [0.38,1.19]). Serial review of radiographic time points identified initial progression of sum of target lesions followed by stabilization for > 3 months after initial progression in some TBP pts. Safety profile during TBP was consistent with that expected for pts established on tebe and no pts experienced an AE leading to treatment discontinuation. Conclusions: An OS benefit observed for tebentafusp among mUM patients who have initial radiographic progression demonstrates that RECIST assessment underestimates benefit. In a post-hoc analysis of OS following initial radiographic progression, continued treatment with tebentafusp was associated with numerically longer OS after adjusting for key prognostic variables. Tebentafusp treatment beyond progression was tolerated without new safety signals and, in some patients, was associated with radiological stabilization of sum of target lesions for > 3 months following the initial progression. Clinical trial information: NCT03070392.