Abstract
Purpose:
Several reports from the developers of the gene expression profile (GEP) test of uveal melanomas (Washington University, St. Louis [WUSTL] group) currently licensed for commercial use in the United States (DecisionDX UM test, Castle Biosciences, Inc., Pheonix, AZ) have suggested that patients with a GEP Class 2 tumor have a median survival time until death from melanoma metastasis of approximately 3 years and almost no chance of long-term survival. Our group has submitted FNAB specimens for GEP testing of patients with a clinically diagnosed posterior uveal melanoma prior to treatment since 9/2007. We analyzed our series of patients (9/2007 - 5/2012) to determine if our experience mirrored or differed from the published and presented survival results.
Methods:
Retrospective analysis of tumor size and GEP class in 273 primary posterior uveal melanomas evaluated by FNAB prior to or at the time of intraocular tumor treatment and actuarial assessment of cumulative probability of death from metastatic uveal melanoma in this group of patients.
Results:
Tumor size in our series ranged from 2.5 - 22.0 mm in largest basal diameter (mean 11.8 mm) and from 1.0 to 18.0 mm in thickness (mean 5.6 mm). 103 patients (37.3%) had a tumor ≤ 10 mm in largest linear tumor dimension (LTD) while 123 (45.1%) had a tumor > 10 but ≤ 15 mm in LTD and only 47 (17.2%) had a tumor > 15 mm in LTD. In contrast, 50.0% of tumors in the WUSTL group's validation series had LTD > 15 mm and only 10.4% had LTD ≤ 10 mm). 193 of the tumors in our series (70.7%) were categorized as GEP class 1 while only 77 (28.2%) were categorized as GEP class 2. In contrast, 51.2% of the tumors in the WUSTL validation series had a class 2 tumor. Three-year melanoma-specific cumulative actuarial mortality in patients with a GEP class 2 tumor in our series was 32.8%; in contrast, 3-year mortality in the class 2 cases in the WUSTL series was 44.3%.
Conclusions:
In our series, patients with a GEP class 2 tumor did not experience a short-term mortality rate from uveal melanoma metastasis as unfavorable as that reported by the WUSTL in their validation study. This observation suggests that (1) tumor size smaller than medium to large is associated with a longer metastasis-free latent interval, (2) GEP class 2 is not as predictably linked to short-term development of metastasis as indicated by the developers of this test, or (3) both.
Keywords: 589 melanoma •
533 gene/expression •
462 clinical (human) or epidemiologic studies: outcomes/complications