Does Hysterectomy without Adnexectomy in Patients with Prior Tubal Interruption Increase the Risk of Subsequent Hydrosalpinx?
Drs. A. Morse, R. Hammer, A. Walter, P. Magtibay
University of Massachusetts Medical Center

Worcester, Massachusetts

Objective : To examine the hypothesis that hysterectomy without adnexectomy following tubal interruption increases the risk for subsequent development of hydrosalpinx.
Materials and Methods : This was a retrospective(case-control) study. The "disease" was a pathologic diagnosis of hydrosalpinx in any patient who underwent adnexectomy at the Mayo Clinic from 1995-2000(cases).  Patients with a pelvic malignancy, or who had not had a prior hysterectomy were excluded. The "exposure" was a history of tubal interruption followed by hysterectomy without adnexectomy at a later date. The exposure rate in patients with hydrosalpinx was compared to a all patients without hydrosalpinx who underwent adnexectomy at the Mayo Clinic from 1997-2000(controls). Patients with a pelvic malignancy and those without prior hysterectomy were again excluded. Fisher's exact test was used to

compare the two proportions.
Results : There was a statistically significant association between the development of hydrosalpinx and a history of hysterectomy without adnexectomy following tubal interruption. In patients for whom adnexal pathology was the primary indication for surgery (e.g. mass, pain), a total of 38 cases and 45 controls were eligible for the study. 9/38 cases and 2/45 controls were found to have a history of tubal interruption (OR=6.67,

95% CI 1.23-66.40, p=0.019). If patients whose adnexectomy was not the primary reason for surgery were included (e.g. the indication for surgery was incontinence or prloapse), then 9/43 cases and 4/97 controls were found to have a history of tubal interruption (OR=5.07, 95% CI 1.57-28.75, p=0.003).
Conclusions : In 1981, M.G. Gregory (J. Tennesee Med. Assoc, Oct, 1981) first proposed that patients undergoing hysterectomy without adnexectomy who had had a previous tubal interruption might be at risk for the development of hydrosalpinx. He hypothesized that this combination of procedures resulted in a segment of tube that was blocked at one end by the sterilization procedure and at the other by the suture securing the utero-ovarian ligament at hysterectomy. Our case-control study is the first we are aware of demonstrating a statistically significant association between hydrosalpinx and a history of hysterectomy without adnexectomy following tubal interruption. While the association is a moderately strong one, this case-control study does not confirm a causal relationship. If further study bears out Dr. Gregory's theory, then consideration should be given to performing salpingectomy at the time of hysterectomy even if the ovaries are being left behind.