Risk Factors For Prediction Of Persistence Or Recurrence After LEEP For Cervical Intraepithelial Neoplasia2-3

BackgroundCervical intraepithelial neoplasia (CIN) is a series of precancerous lesions which are associated with cervical cancer. A guideline of American Society for Colposcopy and Cervical Pathology (ASCCP) advised that both excision and ablation are acceptable treatment modalities for women with a histological diagnosis of CIN2-CIN3and satisfactory colposcopy. Loop electrosurgical excision procedure (LEEP) has been widely used as an effective approach for diagnosing and treating CIN2-CIN3, but cases of persistence/recurrence occurred in follow-up. The rate of persistence/recurrence varied from4.2%-51.8%for the difference among operators. Incidence of cervical cancer will decrease significantly in women with CIN2-CIN3after LEEP, but it still higher than the rate in population. Persistent/recurrent disease after LEEP for CIN2-CIN3was focused.To search for the risk factors of the persistent/recurrent of CIN2-CIN3, scholars analyzed these cases in whom the situation occurred. They find that high risk human papillomavirus load, grade of CIN, endocervical curettage diagnosis, surgical margins, lesions range, conizasion range, age, menopause, gravity, parity and immunosuppression were all connected to the persistence/recurrence of CIN2-CIN3. These researches were important and meaningful for the assessment and management of patients with CIN2-CIN3, but the results conflicted.In this study, we attempted to examine the demographic and pathologic parameters of patients with CIN2-CIN3and searches for the risk factors predicting persistent/recurrent disease after LEEP.Method1. Study subjectsPatients who received LEEP in Department of Gynecology and Obstetrics of Guangdong General Hospital from January2002to December2009and were diagnosed as CIN2or severe by histopathology were enrolled in our study. They also should fulfill the Inclusion criteria below:1. follow-up for at least12momths,2. had at lease one follow-up data of TCT or histology result. Conditions for study exclusion included hystoectomy in one year for diseases other than cervical, radical hystoectomy for invasive cervical cancer. Persistence was defined as histologically diagnosed CIN1-3with cytology abnormal in6months after LEEP. Recurrence was defined as histologically diagnosed CIN1-3after6months after LEEP.A total of1018women accepted LEEP during the period.129cases were diagnosed as CIN1, condyloma and polyp.9women were diagnosed as glandular epithelial lesions.880women were diagnosed as CIN2or severe, in whom103withdrew,21underwent hysterectomy within one year after LEEP for diseases other than cervical disease,19cases were diagnosed as invasive cervical caner and had further treatment, and the rest738women were enrolled in our study. In total,43cases of histologically conformed persistence/recurrence were found during follow-up (38cases of persistence and5cases of recurrence).Study essentials and assignment:1. Age:<40years (0),>40years (1) 2. Gravity:<5times (0),≥5times (1)3. Parity:<2times (0),≥2times (1)4. Pre-HPV-DNA load:≤100pg/ml (0),100-1000pg/ml (1),>1000pg/ml (2)5. Pre-pathological diagnosis:<CIN2(0), CIN2-3(1), CIN3(2)6. Post-pathological diagnosis1) External-cervix pathological diagnosis:<CIN2(0), CIN2-3(1), CIN3(2)2) Endocervical pathological diagnosis:external-cervix pathological diagnosis:<CIN2(0), CIN2-3(1), CIN3(2)7. Extend of CIN:group0(0)(patients who was diagnosed as CIN2or severe before LEEP, but proved to be<CIN2after LEEP),1quadrant (1),2-3quadrants (2),4quadrants (3)8. Gland involvement:negative (0), positive (1)9. Ecto-and endo-margin status of external-cervix:negative (0), positive (1)10. Margin status of endocervical:negative (0), positive (1).2. LEEP procedureLEEP procedure was performed in operating room of Department of Gynecology and Obstetrics of Guangdong General Hospital. Another colposcopic assessment was conduced by the operator before LEEP. After exposing the cervix with an adapted speculum, Lugol’s iodine was used to delineate the area of abnormality. Lesion area of external-cervix was excised under the power of50w, by a loop which was selected according to the size of the area. A second endocervical sweep was routinely performed in the same way. The base of the resulting crater was then coagulated and cauterized using a30w-power electrode. The specimens were supposed to be excised as a whole. Pins were used for orientation, and then specimens were placed in closed containers with right amount of10%formalin for pathology examination. 3. Management of specimensWhen sent to the department of pathology, specimens named "external-cervix" and "endocervical canal" were selected respectively according to the pin, delivered to12pieces averagely, and named as "1O’clock" to "12O’clock" clockwise. After slice-making and HE staining, the slices were sent to pathologist for pathological diagnosis.4. Post-LEEP follow-upApproaches of follow-up include TCT, HR-HPV test, colposcopic examination, and biopsy. A follow-up at the2nd week after LEEP was suggested to observe the cover of the cervical. It was generally suggested that these patients may accepted cervical cytology at the3rd-6th month and HPV test at the6th-12th month, and then go on their follow-up at the12th,18th and24th month after LEEP, and then once a year if the results are all normal. Women with abnormal TCT results will get advices below:a. For a TCT result with ASCUS, it was suggested to have a test of HR-HPV or go to colposcopic clinic directly. If HPV test is positive(≥1pg/ml), patients should go for colposcopic examination; if it was negative, patients can go on the follow-up previously described. b. For a TCT result with LSIL, it was suggested to repeat a cytological test after6-12months, or go for a colposcopic examination. For an unsatisfactory colposcopy, patients should have a cytological test after6-12months. c. For a TCT result with HSIL, patients should undergo a colposcopic examination as soon as possible. It was suggested to have a diagnostic LEEP with an unsatisfactory colposcopic exam.5. Statistic analysisSPSS13.0was used for the statistic analysis of our study. R×C crosstabs χ2test was used for univariate analysis. Multivariable logistic regression analysis was used to test for the value of parameters in predicting persistent/recurrent disease. The calculations of sensitivity, specificity were performed manually. A P value of<0.05was regarded as statistically significant.ResultsA total of1018women accepted LEEP during the period.129cases were diagnosed as CIN1, condyloma and polyp.9women were diagnosed as glandular epithelial lesions.880women were diagnosed as CIN2or severe, in whom103withdrew,21underwent hysterectomy within one year after LEEP for diseases other than cervical disease,19cases were diagnosed as invasive cervical caner and had further treatment, and the rest738patients were enrolled in our study. The mean age of the study group was36.48±7.208(19-74) years. The average time of follow-up was26.00±20.915(3-99) months. In total,43cases of histologically conformed persistence/recurrence were found during follow-up (38cases of persistence and5cases of recurrence). Rate of conformed persistence/recurrence was5.83%.1. Univariate analysisTo investigate the relationship between each elements and disease persistence/recurrence, R×C crosstabs test was used for univariate analysis. It was found that age, pre-HPV-DNA load, pathological diagnosis after LEEP, surgical margins, extend of lesion and gland involvement are statistically significant.2. Multivariable logistic regression analysisWe conducted multivariable logistic regression analysis to analyse the value of parameters which were significant in univariate analysis in predicting persistent/recurrent disease. Endocervical pathological diagnosis (CIN2-3: OR=14.690, P=0.000; CIN3:OR=30.597, P=0.000), ecto-margin status of external-cervix (OR=5.312,P=0.006) and margin status of endocervical (OR=2.574, P=0.035).3. Sensitivity, specificity of the risk factors To compare the value of elements with statistical significance in ultivariable logistic regression analysis in predicting persistent/recurrent disease, calculation of sensitivity and specificity were conducted. The pathological diagnosis of external-cervix was divided into3grades:<CIN2, CIN2-3, CIN3. When considered<CIN2as negative,≥CIN2as positive, the sensitivity, specificity were93.02%,28.06%respectively; when considered<CIN3as negative, CIN3as positive, the sensitivity, specificity were34.88%,86.62%respectively. The pathological diagnosis of endocervical was divided into3grades:<CIN2, CIN2-3, CIN3. When considered <CIN2as negative,≥CIN2as positive, the sensitivity, specificity were81.40%,80.58%respectively; when considered<CIN3as negative, CIN3as positive, the sensitivity, specificity were20.93%,98.27%respectively. The extent of CIN was divided into4grades:Oquadrant,1quadrant,2-3quadrants, and4quadrants. When considered0group as negative,≥1quadrant as positive, the sensitivity, specificity were95.35%,26.47%respectively; when considered<2quadrant as negative,≥2quadrant as positive, the sensitivity, specificity were81.40%,60.43%respectively; when considered<4quadrant as negative,4quadrant as positive, the sensitivity, specificity were27.91%,94.10%respectively. Sensitivity, specificity of ecto-margin status of external-cervix were23.25%,98.13%respectively. Sensitivity, specificity of endo-margin status of external-cervix were55.81%,84.17%respectively. Sensitivity, specificity of margin status of endocervical canal were39.53%,95.25%respectively.ConclusionIn patients with CIN2-CIN3who accept "cow-boy-hat" method of LEEP, elements that affect persistence/recurrence of disease are endocervical pathological diagnosis, endo-margin status of external-cervix and margin status of endocervical. The most predictive element is endocervical pathological diagnosis with a cut off point as CIN2. It should be sematic if the endocervical pathological diagnosis proved to be CIN3, or endocervical margin was involved, or CIN2-CIN3extended to4quadrants.