Skip to main content
  • Uveal Melanoma: Managing Both Primary and Metastatic Disease

    By Marianne Doran, Contributing Writer

    This article is from September 2008 and may contain outdated material.

    Researchers are chipping away at this notoriously resistant neoplasm. Their tools include genetic profiling, preemptive chemotherapy, targeted drugs and novel immunotherapies to control metastasis.

    A discouraging fact: Nearly half of all patients diagnosed with melanoma of the uveal tract will develop metastases, usually to the liver, and the survival rate for metastatic disease has remained virtually unchanged for decades. Despite improvements in the diagnosis and local control of uveal melanoma, established metastases have proved remarkably refractory to conventional chemotherapy and are associated with a median survival time of less than six months after discovery. But recent advances in cytogenetics are enabling researchers to pinpoint more accurately which melanomas are most likely to spread, and allow preemptive treatment with agents targeted to the specific genetic abnormalities of a patient’s tumor. Preclinical studies have also suggested a number of proliferative mechanisms for metastases as well as antiproliferative agents thought to interrupt them.

    Predicting Metastatis

    In the last 15 years, the task of predicting which melanomas will spread has evolved from a reliance on clinical and histologic factors to more sophisticated predictions based on chromosomes and genes. Traditionally, a melanoma’s propensity to metastasize was assessed by relatively primitive measures such as tumor location (ciliary body vs. iris or choroid), size (large vs. small) and histology (primarily epithelial cells vs. spindle cells) or a high mitotic index. But a new understanding of genetic variability promises to shine a brighter light on prognostic factors.

    Chromosomal culprits. Uveal melanomas with certain chromosomal abnormalities—such as monosomy 3 or duplication of chromosome 8—are more likely to metastasize, and this knowledge has improved the identification of high-risk tumors, according to Arun D. Singh, MD, director of ophthalmic oncology at Cleveland Clinic.

    Dig deeper: gene expression. Recently, however, researchers have also been able to zero in on single genes. Through gene expression profiling, investigators have compared the expression patterns of more than 40,000 cancer-related genes in tumors of patients who had either survived melanoma or died of their disease. By determining which genes were consistently over- or underexpressed, they have identified a handful of genes that confer a high risk of metastasis. “Looking at the expression of many genes simultaneously, we have found a signature of gene expression that is very closely associated with metastasis,” said J. William Harbour, MD, professor of ophthalmology and visual sciences and director of the ocular oncology service at Washington University in St. Louis.

    This genetic signature, which currently comprises 12 genes, is being used to classify uveal melanomas as either class 1 (unlikely to metastasize) or class 2 (very likely to metastasize) based on the biopsy findings. “Class 2 tumors have a primitive stem cell-like signature, whereas class 1 tumors have a differentiated signature that is similar to that of normal melanocytes,” Dr. Harbour said. “We think the class 2 tumors develop the ability to metastasize by behaving more like primitive embryonic cells.”

    New tool to guide follow-up. Dr. Harbour’s group has developed a quick and inexpensive test to detect the 12 metastasis-associated genes. The research so far suggests that patients found to have class 1 tumors need little surveillance, whereas those with class 2 melanomas require more follow-up to detect metastases. “We are conducting a prospective multicenter study to validate the test’s accuracy and to determine which genes we can eliminate and still have an accurate test,” Dr. Harbour said. “We have a good indication that we’ll need even fewer than 12 genes—probably three or four.”

    Research conducted by Dr. Singh’s group supports the belief that only a small number of genes will be required to establish a patient’s prognosis. In fact, Cleveland Clinic researchers published a paper last year suggesting that underexpression of a single gene known as autotaxin predicts a poor prognosis in patients with uveal melanoma.1 The researchers are also launching a trial in which genetic testing will be used to establish each of their uveal melanoma patients’ risk of developing metastatic disease. Patients with a poor prognosis will be given the option of participating in a trial of preemptive treatment with a combination of experimental drugs. Those with a good prognosis will receive standard follow-up care (see “Share the Prognosis With Your Patient?”).

    Chemotherapies Old and New

    Uveal melanoma has a dismal track record of responding to conventional chemotherapy. Agents used to treat metastases from cutaneous melanoma, such as dacarbazine (DTIC-Dome) and temozolomide (Temodar), have not produced durable responses in uveal melanoma. “No conventional chemotherapy agent has ever been proven to significantly extend survival once the patient has widespread metastatic disease,” Dr. Harbour said. “So we think the best option is to treat patients preventatively when the primary tumor is diagnosed and the amount of undetectable metastatic disease is small, rather than waiting for it to grow to a much larger, detectable amount. We have a fairly good idea of which molecular pathways to look at, and we are testing agents right now. Ideally, we will find an agent that is already being used for another condition and is known to have a low toxicity because we would like to put patients on that drug for an extended period of time.”

    Dr. Harbour added that he hopes to be able to put his high-risk patients on some type of preemptive treatment within the next year.

    For patients beyond preemption . . . work to make conventional chemotherapy more effective continues. Dr. Harbour noted that for patients whose liver metastases are diagnosed early, hepatic arterial chemoembolization can significantly shrink the tumors. He added that although more patients need to be studied, his impression is that the procedure is extending survival.

    In addition, researchers at the National Cancer Institute are conducting a phase 3 trial of a procedure called percutaneous isolated hepatic arterial perfusion (see “Isolated Perfusion for Liver Metastases”). In a previous study, half of the participants with metastatic uveal melanoma responded to the treatment, including two complete responses.

    The ability to classify the metastatic capacity of uveal melanomas more accurately is also allowing more patient- specific follow-up. “If patients have a class 2 tumor, I monitor them much more closely,” Dr. Harbour said. “I get liver function studies every three months and image their liver every six months because if the metastatic disease can be detected early, they have a better chance of responding to chemoembolization.”

    Target the drug. The future of chemotherapy generally may lie in isolating so-called targeted agents. The tyrosine kinase inhibitor imatinib mesylate (Gleevec) stands as a shining example of the potential power of a targeted cancer therapy. Better tolerated and significantly more effective than conventional chemotherapy, the drug has revolutionized the treatment of chronic myeloid leukemia and gastrointestinal stromal tumors. For researchers in all areas of cancer therapeutics, finding the next Gleevec would be a home run.

    Blocking metastagenesis. In a review article published this year,2 Dr. Singh and colleagues in oncology and genomic medicine list several molecular pathways related to the metastasis of uveal melanomas, as well as agents that might block these pathways:

    • Apoptosis and cell proliferation: Bcl-2 (oblimersen), Rb (vorinostat, depsipeptide), p53 (bortezomib, vorinostat, depsipeptide), MAPK (imatinib mesylate, sorafenib, sunitinib), P13K/AKT (temsirolimus, everolimus, imatinib mesylate, sorafenib, sunitinib, nepafenac) and RTK (imatinib mesylate, sorafenib, sunitinib, cetuximab, erlotinib, gefitinib)
    • Invasion and metastasis: Adhesion molecules (vitaxin, volociximab) and MMP (marimastat)
    • Angiogenesis: bFGF (lenalidomide) and VEGF/VEGR (bevacizumab, VEGF Trap, AZD2171, sorafenib, sunitinib, lenalidomide)

    Among these agents, bortezomib, vorinostat, imatinib mesylate, sorafenib, sunitinib, VEGF Trap, AZD2171 and lenalidomide are all in phase 2 clinical trials for patients with uveal melanoma. These trials can be searched for at

    The Immunotherapy Model

    Monoclonal antibodies. Bradley R. Straatsma, MD, JD, professor and director emeritus of ophthalmology at the University of California, Los Angeles, has evaluated the ocular safety of a human monoclonal antibody called CP-675206. Also known as tremelimumab, CP-675206 targets cytotoxic T lymphocyte–associated antigen 4, an endogenous suppressor of the immune system. “We evaluated a series of patients in phase 1 and 2 clinical trials for metastatic melanoma arising from skin, mucosa, eye or paranasal sinuses,” Dr. Straatsma said. “Although the majority of patients did not respond, a subset of patients had extraordinary responses—and the responses appeared to be sustained.” The researchers detected no adverse effects from the monoclonal antibody. “This was a selected series, so we can’t draw any conclusions from it, but 40 percent of the patients in this group of 20 had either a partial or complete response,” Dr. Straatsma continued. “These results do suggest that ocular melanoma may be amenable to immunotherapy. The literature indicates that melanoma is the most antigenic of all cancers, so it’s a logical place to start.” Dr. Straatsma’s group has requested permission from the developer, Pfizer, to conduct a phase 1/2 study of ocular melanoma patients with proven metastatic disease.

    Antitumor vaccine. Melanoma vaccines are also under study. Under the direction of Jedd Wolchok, MD, researchers at Memorial Sloan-Kettering Cancer Center in New York serendipitously discovered an effective melanoma vaccine for dogs while testing vaccines intended for humans. That vaccine is now FDA-approved to treat canine melanoma. Memorial Sloan-Kettering has a large vaccine development program for ocular melanomas, according to David H. Abramson, MD, chief of the ophthalmic oncology service there. The experimental vaccines have been given to patients in an adjuvant form prior to the detection of metastases.

    Isolated Perfusion for Liver Metastases

    To treat metastatic melanoma, James F. Pingpank, MD, and his colleagues at the National Cancer Institute are investigating the efficacy of melphalan administered to the liver through a minimally invasive procedure known as percutaneous isolated hepatic arterial perfusion. “Melphalan is somewhat of an orphan drug these days,” said Dr. Pingpank, head of the surgical metabolism section in NCI’s surgery branch. “It’s pretty toxic but very effective in a variety of tumors if you can get it where it needs to go. The drug’s main toxicity is on bone marrow, so white blood cells, red cells and platelets get hit hard if you don’t find a way to limit their exposure.”

    LIMIT IT TO THE LIVER. To reduce melphalan’s systemic toxicity, the NCI researchers infuse the drug into the liver through a catheter inserted percutaneously into the hepatic artery. The melphalan perfuses the liver and exits through the hepatic veins. Hepatic venous outflow is then collected through a double balloon catheter in the retrohepatic inferior vena cava and filtered through two activated charcoal filters before it is returned to the patient’s circulation.1 The filtering procedure is between 75 percent and 80 percent effective, Dr. Pingpank said. As a result, all patients get some degree of bone marrow suppression, similar to what occurs with standard chemotherapy. “But in the overwhelming majority of patients this is treated with simple outpatient bone marrow support with a drug or two or maybe a transfusion,” Dr. Pingpank noted. “Patients are ready for a second treatment in 28 to 35 days. Most patients get four treatments, but some have received more.”

    FAST-TRACKING A PROMISING APPLICATION. Dr. Pingpank said the results of the phase 1 trial were so promising—about a 50 percent response rate—that the FDA allowed them to leapfrog to a phase 3 study. In that trial, which is about a third of the way to completion, patients were randomized to either the standard of care—usually chemotherapy—or the experimental treatment. If in either arm of the trial the treatment is ineffective and the tumors are progressing, patients can cross over to the other arm.

    Dr. Pingpank noted that so far, he has seen nothing in the phase 3 trial to suggest that the outcome will differ from the positive results seen in the earlier study.

    1 Pingpank, J. F. et al. J Clin Oncol 2005;23:3465–3474.

    Ophthalmology and Beyond

    The recent approaches to novel therapies underscore the need for accurate, archived biopsies and the importance of considering investigational therapies when managing high-risk patients. “All patients with metastatic uveal melanoma should be enrolled in a clinical trial,” Dr. Abramson said. “In the absence of any current treatment that works, patients will at least receive a scientific approach to treatment in a clinical trial, and the knowledge gained will be useful not only for them but for future patients.”

    Buddy-up with the oncologist. Ophthalmologists will increasingly be called upon to partner with their oncology colleagues in order to provide the best possible care for their patients with metastases. “I would encourage anyone with metastatic disease to be actively managed at a cancer hospital,” Dr. Abramson said. “These patients need the resources of people who are on the cutting edge of research and are willing to try different things.” He added that patients potentially have something to gain from being treated at a cancer center, and science has something to gain if patients are managed in an organized way that can be studied and reported. This approach also helps ensure that other patients are not receiving treatments that are shown not to work. “That’s a tragedy, and we see it all the time,” he said.

    Let’s give them all we’ve got. “We really can’t influence these patients’ lifespans with adjuvant therapy yet,” Dr. Abramson continued. “But we are all trying. And during their remaining life, patients need the resources of an institution that is dedicated to caring for them. They don’t need to be turned away and told that there is nothing that can be done because there is. Pain management is important. Focal radiation for metastases is important. We’re talking about quality of life, and that’s what life is all about, much more so than how long you live.”


    1 Singh, A. D. et al. Br J Ophthalmol 2007;91:1385–1392.

    2 Triozzi, P. L. et al. Cancer Treat Rev 2008;32(3):247–258.

    Share the Prognosis With Your Patient?

    The ability to screen for genetic risk factors for potentially fatal diseases poses dilemmas for the treating physician, especially in the absence of effective therapies. Ophthalmologists who treat patients with uveal melanoma differ in their approaches to at-risk patients.

    Some believe that sharing a troubling prognosis without being able to offer a reasonably promising treatment inflicts undue emotional stress on the patient. Other physicians, like Dr. Straatsma, maintain that it is a physician’s duty to discuss the prognosis with the patient. “Prognosis, favorable or unfavorable, has been part of the practice of medicine for thousands of years. It is something we recognize as one of the obligations of the medical profession.”

    Dr. Straatsma noted that his group is working with a doctoral candidate in psychology who is studying whether patients really want to know their prognosis. “When given the chance, about 90 percent of patients with ocular melanoma say they do want to know the prognosis information, good or bad. Patients want to be involved in their care; they want to be part of the total process.”

    Meet the Experts

    Chief of the ophthalmic oncology service at Memorial Sloan-Kettering Cancer Center in New York City.

    Financial disclosure: None.

    Professor of ophthalmology and visual sciences and director of the ocular oncology service at Washington University in St. Louis.

    Financial disclosure: None.

    Head of the surgical metabolism section at the National Cancer Institute in Bethesda.

    Financial disclosure: Research discussed in this article is supported by an agreement between Delcath Systems and the NCI, along with NCI/NIH intramural research funds.

    Director of ophthalmic oncology at the Cole Eye Institute, Cleveland Clinic.

    Financial disclosure: None.

    Professor and director emeritus of ophthalmology at the Jules Stein Eye Institute at the University of California, Los Angeles.

    Financial disclosure: None.