Utility of conization with frozen section for intraoperative triage prior to definitive hysterectomy

Objective: To review our experience with conization with intraoperative frozen section analysis and to compare results from our tertiary cancer center with those from 2 community hospitals. Methods: The records of all women who underwent conization with intraoperative frozen section analysis from January 1, 1997, through April 30, 2011, at The University of Texas MD Anderson Cancer Center and 2 community hospitals - The Woman's Hospital of Texas and St. Luke's Episcopal Hospital - were reviewed. Findings on pathologic analysis of frozen sections, permanent loop electrosurgical excisional procedure/conization specimens, and hysterectomy specimens were compared for each patient, and the results from the cancer center were compared to those from the community hospitals. Results: One hundred fifty-three patients met the inclusion criteria. Rates of accuracy of conization with frozen section analysis in predicting definitive pathologic findings were as follows: cervix with no residual disease after prior extirpative procedure, 96.5% (95% CI 86.9-100%); cervical squamous carcinoma in situ, 95.4% (95% CI 84.5-100%); cervical adenocarcinoma in situ, 98.7% (95% CI 92.7-100%); microinvasive carcinoma, 97.4% (95% CI 90.1-100%); and invasive carcinoma ≥ 3 mm, 100%. Most importantly, conization with frozen section analysis was 100% accurate for triaging patients to simple or radical hysterectomy. Finally, this approach performed equally well in the cancer center with subspecialized pathologists and the 2 community hospitals with general pathologists. Conclusion: Conization with frozen section analysis is an effective technique for intraoperative triage of patients to immediate simple or radical hysterectomy and can be accurately performed in both academic and community hospitals.