OBJECTIVE

Radical hysterectomy for cervical cancer is associated with increased morbidity over an extrafascial hysterectomy. The goal of this study was to determine incidence of and risk factors for parametrial involvement (PI) based on conization specimen (CS) and to potentially identify candidates for less radical surgery.

METHODS

Patients with FIGO IA2-IIA cervical cancer treated with radical hysterectomy and pelvic lymph node dissection (RH) from 2000 to 2010 were retrospectively identified. Data was extracted from operative and pathology reports. Statistical analyses were performed using Fisher's exact test, t-test, and asymptotic logistic regression.

RESULTS

Of 267 RH patients identified, 118 (44%) had conization prior to RH. The incidence of PI was 15.7% overall and 7.5% in patients treated with conization prior to RH. There was no association between PI and histology, stage, grade, or tumor size. Conization patients with PI were more likely to have LVSI on CS (77.8% vs. 29.4%) and positive lymph nodes (LNP) (66.7% vs. 8.3%). Of patients with positive endocervical curettage, a modest 12% had PI, which was not statistically significant. Tumor size, depth of invasion, and margin status on CS were not statistically associated with PI. In logistic regression analysis, LNP alone or LNP+LVSI were predictive of PI.

CONCLUSIONS

The incidence of PI in early-stage cervical cancer is significant. Only LVSI on CS and LNP were predictors of PI in the current study. While there may be select patients with early stage cervical cancer who can be spared parametrectomy, additional research is warranted.