European Society of Medical Oncology (ESMO) 10Sept2022 – Paris ACIS Abstracts of Interest
Saturday 10Sept Note: There were no specific sessions or major presentations on Ocular Melanoma at ESMO 2022
Poster Session 3:
842P – A phase II trial of AU-011, an investigational, virus-like drug conjugate (VDC) for the treatment of primary indeterminate lesions and small choroidal melanoma (IL/CM) using suprachoroidal administration
Hakan Demirci (Univ. Michigan), C. Rich (Cadmus)
In dose escalation, adult subjects received up to 3 cycles of 3 weekly AU-011 treatments via SC administration with a maximum dose of 80μg with 2 laser applications. Preliminary safety results include 17 subjects in the dose escalation phase. The most common adverse events (AE) related to drug or laser were anterior chamber inflammation in 4 subjects, conjunctival hyperemia, eye pain, and punctate keratitis in 2 subjects each. There were no dose-limiting toxicities, treatment-related serious or grade 3/4 AEs, vitritis or serious AEs of vision loss. Two subjects developed 5 serious AEs unrelated to treatment.
831P – Outcomes of immune checkpoint inhibitors (ICI) in patients with metastatic uveal melanoma
(MUM) treated with tebentafusp Erika Koch (Princess Margaret Cancer Centre, Toronto)
Single center retrospective cohort study, included patients with MUM treated with tebentafusp and ICI (anti PD1 +/-
anti CTLA-4) between 2016-2021. Twenty patients were identified: 10 patients who received ICI following tebentafusp (T-I) and 10 patients treated with ICI prior to tebentafusp (I-T). Disease characteristics were similar amongst both groups. Response rates for the T-I group were: partial response (PR): 10% (1), stable disease (SD): 30% (3) and progressive disease (PD): 60% (6). For the I-T group the corresponding rates were: SD: 20% (2) and PD: 80% (8), p = 0.63 between groups. Median PFS from ICI initiation was 2.86 months (95%CI 2.37-NA) for T-I and 2.35 months (95% CI 0.92-NA) for I-T, (p=0.046). Median OS from start of first treatment was 21.47 months (95% CI 15.78-NA) for T-I, and 16.88 months (95% CI 8.52-NA) for I-T (p=0.38). In this small series, PD- 1 +/- CTLA-4 blocking antibodies may show better clinical efficacy in patients with MUM when administered after tebentafusp rather than before. A larger study is needed to confirm these data.
843P – Long-term survivors on tebentafusp in phase II trial of previously treated patients with metastatic uveal melanoma Takami Sato (Kimmel Cancer Center, Jefferson Univ, Philadelphia)
127 pts with 2L+ mUM received tebe during the expansion phase of Study 102, where treatment beyond initial RECIST progression (TBP) was allowed. The association between OS and baseline (BL) covariates, on-treatment RECIST response, baseline tumor biopsy and ctDNA changes were assessed.
With median follow-up of 29.9 months, 33% (42/127) of pts had OS ≥ 2 yrs, 79% of whom received TBP. 5 of 6 pts with PR were alive ≥ 2 yrs. 41% (25/61) of pts with best response of PD had OS ≥ 1 yr, including 10 pts with OS ≥ 2 yrs. In a multivariate analysis of BL covariates, LDH ≤ ULN and ALP ≤ ULN were most strongly associated with OS ≥ 2 yrs. Lower BL serum ctDNA levels and a low tumor M2 macrophage:CD3 T cell ratio were strongly associated with OS ≥ 2 yrs. BL gp100 expression, prior IO therapy (including anti-PD1 +/- ipilimumab), time from initial diagnosis or metastatic disease, or size of largest liver lesion were not significant predictors of OS ≥ 2 yrs. 20% (14/69) of pts with largest liver lesion of ≥ 3 cm had OS ≥ 2 years. ctDNA reduction (≥ 0.5 log) or clearance by Wk 9 were associated with long OS. Of 25 pts with OS ≥ 2 yrs and evaluable ctDNA, 52% (13/25) had ≥ 0.5 log ctDNA reduction, of whom 8 had ctDNA clearance. Most of the 13 pts had best RECIST response of PD or SD.
The strongest baseline predictors of OS ≥ 2 yrs on tebe were low serum LDH, ALP and ctDNA and, in tumors,
a low M2 macrophage:CD3 T cell ratio. Other BL covariates, e.g., size of largest liver lesion, were not significant predictors. ctDNA reduction/clearance may be a better early surrogate of long OS than radiographic response.
821P – Safety and efficacy of infrequent tebentafusp treatment omissions in patients with metastatic uveal melanoma Max Schlaak (University Hospital LMU Munich)
245 patients received tebe; median 23 doses. A total of 104 pts had omissions with 92/245 pts (38%) having an omission after the initial 3 wks. 56/92 pts (61%) had 1 omission; 14 pts had > 3 omissions. Most omissions were due to elective/other reasons (71%) or AEs (29%). 72% of omissions were ≤ 2 wks; 7% of omissions were > 3 wks.
Upon restarting, majority of pts did not have G3 TRAE (91%), G2 CRS (93%) or G2+ rash (93%) within 14 days. However, 6 pts had G2 cytokine-releae syndrome (CRS) within 14 days of restart and all had prior G2 CRS. 1 or 2 omissions did not have a significant impact on OS when controlling for immortal time bias. The small numbers of pts with omissions > 3 wks duration limit the ability to evaluate impact on OS.
After reaching 68 mcg, patients receiving tebentafusp can have 1-2 omissions of ≤ 2 weeks duration with minimal impact on safety and efficacy. Treatment restart was typically outpatient (95%), without dose modification from most recent dose (98%) or steroid premedication (98%). G2 CRS was uncommon at restart, and occurred mostly in patients with preceding G2 CR