Hysteroscopic dissemination of endometrial carcinoma using carbon dioxide and normal saline: a retrospective study.

OBJECTIVE: The aim of this study was to compare the likelihood of disseminating endometrial carcinoma cells into the peritoneal cavity by hysteroscopic examination using carbon dioxide (CO(2)) or normal saline (NS) as the distension medium.

METHODS: A retrospective study of 162 consecutive patients with endometrial carcinoma treated at a university teaching hospital from 1994 to 1999 was undertaken. All patients had a hysteroscopic examination, using either CO(2) or NS as the distension medium, as part of the investigation for abnormal uterine bleeding or in determining whether the uterine cervix was invaded by tumor. Peritoneal fluid for cytology was collected immediately upon entry into the abdominal cavity. Positive peritoneal cytology was considered the primary statistical endpoint.

RESULTS: Among 162 patients, 39 cases were excluded from the study because of macroscopic intraperitoneal diseases (n = 32) or pathology other than endometrioid adenocarcinoma (n = 7). Another 3 cases were excluded because both distension mediums had been used in the hysteroscopy. Analysis was therefore based on the data of 120 patients. There was no statistically significant difference between the two groups of patients undergoing hysteroscopy using either CO(2) (n = 70) or NS (n = 50) with regard to age, pathologic stage, International Federation of Gynecology and Obstetrics grading, myometrial invasion, tumor size, cervical involvement, nodal involvement, and 2-year progression-free survival. However, there was a mean of 13.0 plus minus 5.0 days (range 3-21 days) time gap between laparotomy for definitive surgery and CO(2) hysteroscopy compared to immediate laparotomy after NS hysteroscopy (P < 0.001). Positive peritoneal cytology was noticed in 8 (6.7%) patients of which 7 were in the NS group and 1 was in the CO(2) group. Positive cytology was significantly more common among patients after hysteroscopy using NS than CO(2) (14.0% versus 1.4%, odds ratio = 11.2, 95% confidence interval = 1.3-94.5, P = 0.009). The presence of positive peritoneal cytology was not associated with age, tumor grade, tumor size, myometrial invasion, cervical involvement, or nodal metastasis. All 8 patients with positive cytology received no additional treatment and are disease free at 12 to 34 months of follow-up.

CONCLUSIONS: Our data suggested that endometrial malignant cells were introduced into the peritoneal cavity during hysteroscopy and might be more likely after the use of NS rather than CO(2). This report emphasizes the need for prospective evaluation for further clarification of this hypothesis. The clinical significance of the dissemination awaits the long-term follow-up of these patients.