Clinicopathologic variables, operative characteristics, and DNA ploidy in predicting outcome in ovarian epithelial...

To test the hypothesis that DNA content can predict operative morbidity and survival in patients with ovarian carcinoma.

Subjects included patients diagnosed with invasive epithelial ovarian carcinoma at Brigham and Women's Hospital between July 1987 and November 1989. Fifty-nine patients were included in this analysis. In all cases, flow cytometry was performed on fresh tissue to evaluate DNA content. The medical records were reviewed in all patients for estimated blood loss, hospitalization days, intensive care unit days, operating room time, presence and size of residual disease, grade and type of tumor, stage, size of primary tumor, lymph node status, disease status, date of last examination, and number of months of follow-up.

Predictors for death included increasing age (P = .01), advanced stage (P = .007), the presence of malignant ascites (P = .03), residual tumor at completion of operation (P < .001), increased estimated blood loss (P < .001), increased hospitalization days (P < .001), and increased operating room hours (P < .001). When we controlled for age and stage, only estimated blood loss and residual tumor predicted poor outcome. Deoxyribonucleic acid ploidy, whether stratified as diploid or aneuploid or with DNA index cutoffs, did not predict tumor recurrence or survival rates.

Deoxyribonucleic acid ploidy has not yet been proven to be of independent prognostic importance for identifying groups of patients at high risk of dying from invasive epithelial ovarian carcinoma.