Predictors of the Residual Disease of High-Grade Lesions and Microinvasive Squamous Cell Carcinoma of the Cervix Following Conization
Nungrutai Saeaib MD*, Sathana Boonyapipat MD*, Kobkul Tungsinmunkong MD**, Tippawan Liabsuetrakul MD, PhD***
Affiliation : * Department of Obstetrics and Gynecology, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand ** Department of Pathology, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand *** Epidemiology Unit, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand
Objective : To determine the predictors of residual disease of high-grade lesion (HGL) and microinvasive
squamous cell carcinoma of the cervix (MICA) in subsequent hysterectomy following conization.
Material and Method: The medical records of women who underwent any conizations diagnosed of HGL and
MICA and followed by subsequent hysterectomy within 6 months were retrospectively reviewed. A case and
control was defined as whether or not a residual disease of HGL or more was detected in cervical tissue from
hysterectomy after conization. Demographic characteristics and pathological features of cases and controls
were recorded independently and blindly. Univariate and multivariate analysis were used. The Receiver
Operating Characteristics curve of predictors was created using the fitting value obtained from a logistic
regression model.
Results : A total of 185 women were diagnosed during January 1, 1997 and July 31, 2008 including 102
women without a residual disease and 83 with residual disease at cervical tissue from hysterectomy. The
multivariate analysis showed that postmenopausal status (OR = 3.5, 95% CI = 1.8-6.7), number of quadrant
involvement (OR = 3.8, 95% CI = 1.8-8.3), internal margin involvement (OR = 3.8, 95% CI = 1.7-8.2), severe
nuclear atypia (OR = 2.0, 95% CI = 1.1-3.8) and high mitotic activity (OR = 2.1, 95% CI = 1.1-3.7) were the
predictors of residual disease in hysterectomy specimens after conization. Three or more predictors involved
predicted the detection of residual disease.
Conclusion : The presence of postmenopause, three or four quadrants involved, positive internal margin,
severe nuclear atypia and high mitotic activity could be used to predict residual lesions after conization.
Keywords : Cervical dysplasia, Conization, Subsequent hysterectomy, Predictor
