NEW YORK (GenomeWeb News) – Researchers from the Melbourne, Australia, branch of the Ludwig Institute for Cancer Research and New Zealand firm Pacific Edge Biotechnology reported today that they have developed a gene expression-based test for melanoma progression.
The team measured gene expression in samples taken from dozens of melanoma patients and looked for an expression signature linked to progression from stage III to stage IV melanoma. The research, published in today’s issue of Clinical Cancer Research, suggests that gauging the expression of just a handful of genes can help forecast clinical outcomes. Researchers say the test, which was 85 percent to 90 percent accurate in the small groups tested, may eventually help to guide treatments — and test new ones.
Some melanoma patients respond well to biological or cytotoxic chemotherapies, though most show little response to the treatment. It is also difficult to predict disease progression because of variability in the disease. For instance, many — but not all — patients with stage III melanoma, in which the cancer has spread to the lymph nodes, progress to stage IV melanoma very quickly. Consequently, the average five-year survival rate for those with stage III melanoma is less than 30 percent.
Being able to classify melanoma patients into molecular subtypes with different outcomes could help determine the most appropriate treatment as well as the biology driving cancer progression, the researchers noted.
“With current algorithms, it is not possible to predict which patients will achieve longer-term survival,” senior author Jonathan Cebon, director of the Joint Austin/Ludwig Institute for Cancer Research Medical Oncology Unit at Melbourne’s Austin Hospital, and his co-workers wrote. “We hypothesized that molecular profiling could be used to identify prognostic groups within patients with stage III melanoma while also providing a greater understanding of the biological programs underpinning these differences.”
Using microarrays, the researchers compared the expression of more than 30,000 genes in lymph node samples from 13 stage III melanoma patients with good outcomes and 16 with poor outcomes.
The approach turned up 2,140 genes that were differentially expressed between the two groups — including genes involved in apoptosis, signaling pathways, and anticancer immune responses.
Of these, the researchers verified the 21 most significant genes using quantitative PCR and eventually developed a predictive algorithm based on these genes and another based on the five most strongly significant genes.
They then tested the algorithms in prospective experiments, measuring gene expression in two different validation groups of ten stage III samples and 14 patients with published gene expression profiles. Using the two algorithms, the researchers were able to predict clinical outcome with 90 percent and 85 percent accuracy, respectively, in these validation groups.
Although they noted that the algorithms need to be tested in larger patient groups, the researchers expressed enthusiasm about the possibility of using their microarray and qPCR-derived test to predict clinical outcomes in “otherwise indistinguishable” stage III melanoma patients.
That, in turn, may help doctors target melanoma treatments more effectively and avoid giving unnecessary — and toxic — therapies to some. In addition, the researchers suggested, the results may also have implications for clinical trials of new melanoma treatments, allowing researchers to design trials and assess their outcomes more accurately.
“Although patients might all have the same type of cancers, there can be big differences in their survival simply because their cancers behave differently — and this may have nothing to do with the treatment,” Cebon said in a statement.
“It’s partly because we can’t clinically identify subtypes of patients that we have to do very large and very expensive trials. And, of course, this increases the time it takes to test the clinical benefit of potential new therapies.”