PubMedDate not supplied
It is increasingly common for cytology laboratories to receive ovarian, adnexal and pelvic cyst fluids obtained via sonographically directed aspiration and laparoscopic techniques, especially from women who are desirous of preserving fertility or who are undergoing in vitro fertilization (IVF). Accurate characterization of such cysts is a worthwhile goal, given the superior prognosis for ovarian carcinomas that are diagnosed at an early stage. In an effort to improve upon the false-negative diagnosis rate associated with cytology, we evaluated DNA ploidy as a possible adjunctive criterion. We examined 55 benign, 3 borderline and 6 malignant aspirates received by our cytopathology laboratory; 35 were aspirated directly from the patient from clinically and ultrasonographically benign cysts, and 29 were aspirated from surgically removed benign (20) and malignant (9) cysts. Adjunctive DNA ploidy and cell cycle analysis was performed using the Cell Analysis Systems CAS-200 on Feulgen-stained cytologic smears of the 64 cyst fluids. Adequate material for DNA analysis was obtained from 33/35 in situ aspirated cysts and from 19/29 surgical specimen cysts. Forty-seven of 52 cytologically benign cysts were diploid. Of the 5 nondiploid benign cysts, 3 were follicle cysts (2 from hormonally stimulated IVF patients and 1 from a postpartum patient), and 1 was a benign cystic teratoma. Their nondiploid DNA pattern or tetraploidy may be due to a high proliferative index. The fifth nondiploid benign aspirate was from a resected benign epithelium-lined cyst; its DNA histogram contained a conspicuous tetraploid population. All 9 malignant cysts were cytologically malignant. Of the 3 borderline cysts, 1 was nondiploid, and 2 were diploid. All 6 fully malignant cysts were nondiploid; 2 of them were tetraploid. Based on our results, we conclude that DNA ploidy analysis of cells derived from ovarian and adnexal cyst aspirates is feasible (in 95% of cases) and relatively specific (90%) and has a relatively high negative predictive value (92%). The results are not sufficiently predictive of the histology of the lesion to warrant therapeutic intervention based on ploidy alone (sensitivity of nondiploid results, 78%; positive predictive value of nondiploid results, 58%). Nondiploidy should suggest consideration, but is not conclusive, of a malignancy diagnosis. There may even be prognostic implications to the ploidy pattern, particularly in borderline tumors, in which nondiploidy portends a poor prognosis.