Abstract
Purpose :
Canine uveal melanoma (UM) is a tumor that rarely metastasizes. Recent studies have shown that these tumors in dogs have differences compared to human tumors and they do not require further treatment following enucleation. HMB 45, Melan A and COX 2 are common markers of UM in humans, but their expression in dogs has never been compared to humans. The aim of this study is to evaluate the expression of these markers in classic and melanocytoid-type UM in dogs, and compare this expression with humans.
Methods :
In total, 66 eyes from dogs with UM were evaluated using HMB-45, Melan A and COX 2 monoclonal antibodies. The canine tumors were classified as either classic or melanocytoid-type. Slides were evaluated based on immunostaining distribution (0=absence of staining; 1=1%–10% positive cells; 2=11%–50%; 3=51%–80%; and 4=81%–100%) and intensity (0=negative; 1=weak; 2=moderate; and 3=strong staining). Data were then converted to the numeric German Immunoreactive Score (IRS) by multiplying the distribution and intensity scores, resulting in a range from 0 to 12. Data pertaining to human staining were obtained from the most recently published studies from our laboratory. Differences between human and canine staining were determined using the Chi-squared test; significance was set at P<0.05. We compared expression between classic and melanocytoid-type tumors using the Student’s t-test.
Results :
In canine samples, HMB 45 was present in 30%, Melan A in 53%, and COX 2 in 86% of tumors. COX 2 expression was significantly greater in dogs compared to human tumors (58%; P<0.0001). Moreover, COX 2 IRS in classic tumors was significantly greater than melanocytoid-type tumors (P=0.00002). In 36% of the dog tumors, HMB 45 and Melan A were not expressed, and the overall expression of each marker individually was significantly less than in human tumors (P<0.0001).
Conclusions :
Melan A and HMB 45 are inadequate as diagnostic markers of canine UM, which is in stark contrast to human UM tumors. COX 2, on the other hand, showed very high expression in the vast majority of canine tumors, and the expression was significantly greater than in human tumors. Moreover, COX 2 was more highly expressed in the more aggressive classic canine variant. Therefore, research investigating anti-COX 2 treatment in canine UM is warranted. Additional UM markers, such as Tyrosinase, should be tested to improve diagnostic accuracy in dogs.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.