Identification of genetic alterations related to chemoresistance in epithelial ovarian cancer

Objective. After the completion of primary chemotherapy, the majority of advanced ovarian cancer patients have persistent, chemoresistant disease. Comparative genomic hybridization (CGH) has been used to study genetic alterations that may be responsible for chemoresistance in ovarian cancer. CGH is a useful, genomewide screen but resolution is limited to 5-10 Mb. Recently, quantitative microsatellite analysis (QuMA), a TaqMan-based quantitative PCR technology, has been used for higher resolution of DNA copy number abnormalities. Our goal is to identify specific chromosomal aberrations correlated with platinum resistance. Methods. Snap-frozen ovarian tissue samples taken from 22 patients with ovarian cancer between 1994 and 1998 were analyzed. Patients whose ovarian cancer actually demonstrated growth during platinum-combination treatment or no objective evidence of regression following four to six cycles of therapy were considered to have clinically defined platinum-resistant disease. QuMA was carried out at the following loci using the ABI Prism 7700 (TaqMan) instrument with a microsatellite repeat probe: D3S1553, D3S1617, D5S464, D5S630, D6S1581, D6S446, D8S557, D19S208, D20S196, DXS1068. Fisher's exact test, exact logistic regression, and the Cochran-Armitage trend test were used. Because of multiple hypothesis testing, the P values were adjusted with the Bonferroni procedure to limit the familywise error rate to at most 5%. Results. Of the 22 patients, 12 (54.5%) were platinum-sensitive and 10 (45.5%) were platinum-resistant. When comparing sensitive and resistant patients, no statistically significant difference was noted among stage, grade, histology, and age (P = 0.1292, P = 1.0000, P = 1.0000, P = 1.0000, respectively). In the QuMA analysis, 10 of the 14 (71.4%) patients who had a low copy number of D6S1581 were platinum-resistant, while none of the patients with a normal or high copy number of D6S1581 were platinum-resistant. This was statistically significant when the marker data were treated as either a continuous or a categorical variable (P = 0.0410 and P = 0.0170, respectively). No other loci correlated significantly with platinum resistance. Conclusions. D6S1581 was the only genetic marker, of those examined, significantly related to chemoresistance. Patients with a loss of D6S1581 are more likely to be platinum-resistant. Identification of genetic alterations associated with platinum resistance detected by QuMA may contribute to a better understanding of clinical behavior and chemotherapy treatment options for patients.