The findings, described in an advance on-line publication of the December issue of the Journal of Clinical Oncology, represent the first time that this type of genetic test has proven useful as a prognostic tool for ovarian cancer, which accounts for approximately 26,000 new cases and 16,000 deaths in the United States each year.

"Ovarian cancer is widely recognized as being extremely difficult to treat," explains Stephen A. Cannistra, M.D., Director of Gynecologic Medical Oncology at BIDMC and Associate Professor of Medicine at Harvard Medical School. "Because symptoms often do not appear until the disease has already spread to the upper abdomen, this malignancy is usually not diagnosed until it has reached an advanced stage." At that point, he adds, doctors typically use clinical data “ such as the amount of residual disease remaining following surgery “ to assess a patient's prognosis and determine their course of therapy, a method that Cannistra notes is admittedly imperfect.

Knowing that the behavior of cancers is partly dependent upon which genes are turned on and off in tumor cells, researchers have long suspected that a better understanding of the genetic profile of the tumors of individual patients could help in making a more accurate prognosis.

"With the advent of microarray analysis -- in which genes expressed by the cancer cells are labeled with a probe and then applied to a glass slide that contains embedded sequences of thousands of known human genes “ this type of genetic information has become much more accessible," explains Cannistra. "[Through this process] genes that are present in the tumor cell bind to their counterpart sequences on the glass slide, thereby permitting their identification with the aid of computer analysis."

Using tumor tissue from 68 ovarian cancer patients undergoing initial surgery at either BIDMC in Boston or Memorial Sloan-Kettering Cancer Center in New York, Cannistra and his co-authors employed microarray analysis to develop a "genetic snapshot" of ovarian cancer.

"We were ultimately able to identify 115 genes, which we refer to collectively as the Ovarian Cancer Prognostic Profile [OCPP]," Cannistra notes. "Simply knowing the expression pattern of these genes from the original tumor sample [i.e., whether genes were 'turned on' or 'turned off'] provided us with important information about prognosis that could not be gleaned from standard clinical features, such as tumor grade or residual disease status."

"Molecular profiling of epithelial ovarian cancer holds promise that goes beyond identifying the most aggressive tumors," according to an editorial published in the on-line version of the Journal of Clinical Oncology. "Gene expression signatures may also be a cornerstone to understanding the root causes of ovarian cancer, and to designing pathway-specific targeted therapies," the commentary concludes. According to Cannistra, future research will further evaluate this technology through prospective studies of patients with both advanced ovarian cancer, as well as early stage disease.

"The use of a powerful prognostic tool such as the OCPP may someday enable clinicians to identify those patients most appropriate for clinical trials [of investigative therapies]," writes Cannistra. "It may also provide insights into why tumors frequently develop resistance to chemotherapy, and may eventually permit individualized use of targeted therapies that are chosen on the basis of a given tumor's genetic profile," he notes.

bidmc.harvard/