Treatment of cervical carcinoma in situ in HIV positive women.

OBJECTIVES

To evaluate the safety and effectiveness of hysterectomy vs. cone biopsy in HIV seropositive patients with carcinoma in situ of the cervix (CIS).

METHODS

We performed a retrospective case-control study of all HIV seropositive patients diagnosed with carcinoma in situ of the cervix from 1989 to 1995. A control group of HIV(-) women with CIS was also ascertained matched for date of diagnosis of CIS, race and age.

RESULTS

There were 439 patients with CIS, of which 45 were HIV seropositive (10.3%). Nine were treated by hysterectomy, 30 by cone biopsy, and six remained untreated. Overall, 63% of HIV(+) patients did not receive any follow-up Pap smear (44% of hysterectomy patients, 50% of cone biopsy patients, and 83% of untreated patients; chi(2) P=0.41). According to Pap smear results, 67% (10/15) cone biopsy patients and 60% (3/5) hysterectomy patients had an abnormal Pap smear after treatment (P=0.9). Median time to recurrence was 12 months in hysterectomy patients vs. 14 months in cone biopsy patients. Deaths occurred in 22% of hysterectomy patients, 17% of cone biopsy patients, and 50% of untreated patients, none due to cervical cancer. Median time to death from presentation was 27.5 months for hysterectomy patients, 11 months for cone biopsy patients, and 7 months for untreated patients (P<0.05). There were no complications in the hysterectomy group, however, two patients were readmitted after cone biopsy for bleeding. When compared to HIV(-) women with CIS, HIV(+) patients were more likely to be treated by hysterectomy (chi(2) P=0.01).

CONCLUSION

All patients diagnosed with CIS should be counseled regarding HIV prevention and testing because of a significant seropositive rate. Compliance with gynecologic follow-up is very poor in this patient population. Special efforts should be made to enhance compliance. Cone biopsy and hysterectomy appear to be equally safe and effective in the treatment of CIS. CIS in HIV patients is a poor prognostic indicator for death from any cause.