Like a sweaty, moonlit adventure in the Temple of Doom with justice-hunter Indiana Jones, who innocently stumbles on a dusty clue; examines it with his fingerprinted, partially cracked monocular lens; and combines this information with a rudimentary profile from a half-torn book published in Liverpool, England, an idea comes forth. With his self-assured grin, a clue is recognized, the criminal identified, and the puzzle solved. The criminal is uveal melanoma and the puzzle is its molecular and genetic profile providing clarification of its developmental pathway. We have trapped melanoma in the Temple of Doom and there is concerted effort to characterize it.

Uveal melanoma is a deadly tumor. Based on long-term follow-up, over 50% of patients succumb to this malignancy, and greater tumor thickness guarantees greater metastatic doom. Investigators worldwide are searching for genetic clues in molecular pathways leading to uveal melanoma development and metastatic spread. Teams have focused on delineating basic alterations in melanoma genome at the level of DNA or gene expression using measurements on RNA. Others have looked at techniques for harvesting adequate microscopic tissue for evaluation using needle biopsy without the need for enucleation. Still others have looked at more clinically relevant biologic signaling pathways. Let’s summarize what we have learned.

Regarding genome content (DNA) evaluation, the original detective work is credited to Prescher and associates in 1996, in which they demonstrated that uveal melanoma in enucleated eyes often showed chromosome 3 monosomy and this finding imparted poor prognosis. Subsequent larger studies by Scholes and associates and Damato and associates confirmed this finding and found correlation of monosomy 3 with epithelioid cell type, microvascular loops, basal tumor diameter, ciliary body involvement, and metastasis-related death. Damato and associates recently profiled 452 choroidal melanomas for DNA using multiplex ligation-dependent probe amplification (MLPA) and found high predictive value with 10-year melanoma-related mortality of 0% for those with disomy 3 compared to 55% for those with monosomy 3 and 71% for those with monosomy 3 plus 8q gain. They undoubtedly portrayed the predictive power of DNA evaluation using MLPA for melanoma prognosis.

Regarding RNA evaluation, gene expression profiling (GEP) has been employed. This elegant technique is a measure of genetic material within melanoma using messenger RNA expression from multiple genes. In 2003, GEP was explored by Tschentscher and associates in 20 eyes with uveal melanoma using 12 500 probes and found 2 groups of melanoma that correlated with monosomy 3 and disomy 3 tumors. In 2004, Onken and associates used GEP and confirmed the presence of 2 classes of melanoma in which class 1 (low grade) showed 95% survival and class 2 (high grade) showed only 31% survival at 8 years. Further refinements of this test have allowed for higher predictive value.

Regarding harvesting tissue for genetic evaluation, acquisition of melanoma using fine needle aspiration biopsy has been successful. The trans–pars plana approach with needle placement into tumor apex and aspiration of cells allowed for DNA detection in 97% of cases compared to 75% yield by transscleral approach into tumor base. With appropriate needle biopsy technique, tumor seeding was not seen in any case.

Regarding biologic signaling pathways, this could be the most rewarding and clinically applicable to practical management of melanoma, with hope for agents that target pathway abnormalities and lead to improved patient survival. In 2010, Van Raamsdonk and associates evaluated 186 uveal melanomas and found that 83% had somatic mutations in GNAQ or GNA11 . They concluded that the pathway involving these 2 genes was a major contributor to the development of uveal melanoma. Furthermore, mutations in GNA11 induced metastasis in mouse model and activated the mitogen-activated protein (MAP) kinase pathway. Harbour and associates discovered that breast cancer 1–associated protein (BAP1), implicated in numerous cancers including lung carcinoma, breast carcinoma, and renal cell carcinoma, was also found in uveal melanoma metastasis in 84% of cases. BAP1, located on chromosome 3, coordinates assembly of multiprotein complexes of transcription factors and cofactors for cellular processes. Depletion of BAP1 results in altered expression of key genes in cell cycle progression, DNA replication and repair, and cell metabolism and survival. BAP1 has tumor suppressor activity in vivo and in vitro and mutation could lead to tumorigenesis.

What does the clinician do with this treasure of information?

This information is critical to our understanding of the pathogenesis of uveal melanoma and might be relevant to targeted therapies in the near future, sooner than you think. No longer should we “shoot in the dark” with chemotherapy for uveal melanoma. Now we target the criminal based on pathway abnormalities. In 2011, Patel and associates provided an instructive overview of potential agents available for targeted therapy of uveal melanoma. They indicated, for example, that a patient with a documented pathway defect in MAP kinase pathway abnormality and elevated phosphorylated extracellular signal–related kinase (pERK) might be treated with targeted therapy of 17-allylamino-17-demethoxygeldanamycin (17-AAG). Identification of pathway abnormalities and institution of neoadjuvant therapy could be key in control of micrometastasis.

Our team in Philadelphia is currently investigating pathway defects in a mouse model of uveal melanoma. Identification of particular genetic alterations in this mouse model will provide an opportunity to test therapies preclinically. Spagnolo and associates recently described 7 targeted clinical trials now open to uveal melanoma patients. We know that 57% of uveal melanoma metastases carry mutations in codon 209 of GNA11 , while 22% carry mutations in codon 209 of GNAQ . For these patients, using inhibitors of the constitutively active MAP kinase pathway is rational. Alternatively, if BAP1 is altered in uveal melanoma metastases, poly-adenosine diphosphate-ribose polymerase (PARP) inhibitors may become the preferred treatment.

In this issue of the Journal , Chappell and associates performed GEP (RNA analysis) on 126 patients with uveal melanoma using tissue harvested by needle biopsy. They identified 64% class 1 (good prognosis) and 36% class 2 (poor prognosis) melanomas. Unlike previous similar publications, these authors found that tumor class did not predict rate of tumor regression following radiotherapy. Previous studies from Philadelphia, Boston, Jerusalem, and Los Angeles have indicated that high-grade melanoma at greater risk for metastasis shows more rapid regression following radiotherapy. In this current study, class 2 melanomas, regarded as high-grade tumors, did not show more rapid regression. This finding, which the authors admit was “surprising” to them, is contradictory to previous publications. The authors did not enumerate speculation to this contrary finding. We speculate that the disparity could represent an inadequate cohort size for reliable statistical analysis and we doubt that these experienced clinicians had inaccuracies in treatment parameters or genetic analysis that could have contributed to the lack of correlation. Despite these findings, this does not detract from the importance of genetic evaluation.

Now is an exciting time for uveal melanoma investigators. Major strides in our understanding of the development of melanoma come from teams around the world. We are gradually unmasking the long-hidden secrets in the pathogenesis of uveal melanoma. We have trapped melanoma in the Temple of Doom and we unwind its molecular profile, piece by piece.