Retrograde seeding of malignant cells during hysteroscopy in presumed early endometrial cancer

Objective. The aim of this study was to determine the effect that preoperat ive hysteroscopy has on the frequency of positive cytology at the time of d efinitive surgical management in endometrial cancer. Methods. Charts of 222 patients with endometrial cancer were reviewed. Pati ents were divided according to whether (n = 64) or not (n = 158) they had h ysteroscopy with saline infusion. Each group was stratified into low or hig h risk for positive peritoneal cytology. Logistic regression analysis was u sed to compare the prevalence of positive peritoneal cytology with and with out hysteroscopy, before and after the stratification, adjusting for the co nfounding risk factors. Results. After adjusting for confounding variables there was a statistical difference in the frequency of positive peritoneal cytology in those who ha d hysteroscopy versus those who did not (odds ratio (OR) = 2.6, 95% confide nce interval (CI) = 1.02-6.63, P = 0.05). Even after stratifying patients i nto a low-risk group (OR = 2.12, 95% CI = 0.13-35.9, P = 0.6) and a high-ri sk group (OR 3.46, 95% CI = 1.3-9.12, P = 0.01) the difference in the high- risk group was statistically significant. Conclusion. Hysteroscopy seems to affect the prevalence of positive periton eal cytology, especially in those patients with highrisk cell types. Its us e in patients with suspicion of endometrial cancer should be reconsidered, (C) 2000 academic Press.