Gene Signature Improves Prediction of Multi-Drug Resistant Ovarian Cancer Survival

This graph represents a four-level risk prediction analysis of patients’ overall-survival according to clinical factors and level of gene expression. For example, “high/high” indicates a high risk for shorter survival time plus a high gene signature expression. Patients with low gene signature expression had longer survival time than those with high expression, regardless of their clinical factors.

This graph represents a four-level risk prediction analysis of patients’ overall-survival according to clinical factors and level of gene expression. For example, “high/high” indicates a high risk for shorter survival time plus a high gene signature expression. Patients with low gene signature expression had longer survival time than those with high expression, regardless of their clinical factors.

CCR scientists found an 11-gene signature that improves prediction of overall survival of ovarian cancer patients whose disease becomes resistant to standard chemotherapy drugs.

Michael M. Gottesman, M.D., and Jean-Pierre Gillet, Ph.D., of the CCR Laboratory of Cell Biology, assessed gene expression patterns to determine whether a gene expression signature could improve current ovarian cancer prognostic methods. Ovarian cancer has an overall five-year survival rate of just 31 percent, it is usually diagnosed at late-stage, and many patients develop drug-resistant tumors. Therefore, identifying which patients are more likely to respond to the standard combination chemotherapy of carboplatin and paclitaxel, and which might benefit from an alternative treatment, could improve survival.

The researchers initially selected 380 genes that were associated with either intrinsic or acquired drug resistance in previous published findings. Then they analyzed the gene expression profiles of tumor samples of 80 ovarian cancer patients. After eliminating the genes that were expressed in fewer than 50 percent of the samples, they performed detailed analysis on the remaining 356 genes.

The researchers compared prediction for overall and progression free survival using various statistical models. They correlated the clinical variables—age at diagnosis, stage of disease, blood levels of CA-125 (a substance or antigen associated with ovarian and other cancers), and surgical debulking (the removal of cancerous tissue from ovaries and surrounding tissues and organs)—with the patients’ overall survival and divided the patients into high and low risk groups.  When they added to their analysis the gene expression profiles of the tissue samples, they found an 11-gene signature that gave a more precise prognosis for overall survival for patients who received standard chemotherapy. Patients who had a low expression for this gene signature had better survival outcomes than those with a high expression.

The gene signature did not predict progression free survival, however, and the small number of progressive disease samples in the study may have impeded the discovery of a predictive gene signature for progression free survival.

The researchers hypothesize that the 11-gene signature could be used to predict overall survival in patients with newly diagnosed disease. Additional studies on this gene signature could lead to new targeted therapies that improve clinical outcomes for ovarian cancer patients.

Summary Posted: 05/2012

Reference

Gillet JP, Calcagno AM, Varma S, Davidson B, Bunkholt-Elstrand M, Ganapathi R, Kamat A, Sook AK, Ambudkar SV, Seiden M, Rueda BR, Gottesman, MM. Multidrug Resistance-Linked Gene Signature Predicts Overall Survival of Patients With Primary Ovarian Serous Carcinoma. Clinical Cancer Research. 2012 April 5. PubMed Link