| Objective: To explore and screen the independent influencing factors of recurrence within two years after laparoscopic conservative operation in patients with Ovarian Endometriosis(OEM)without pharmacological management,and a predictive model with good discrimination,calibration and clinical validity was constructed to predict the risk of recurrence after surgery in the form of a nomogram.Meanwhile,our nomogram was compared with an existing model which evaluating the severity of OEM,namely revised American Society for Reproductive Medicine classification of endometriosis(r ASRM),to compare the discrimination and clinical validity of this model with the r ASRM score and r ASRM staging,so as to provide a better clinical guidance for screening patients requiring long-term pharmacological management after OEM removal and implementing individualized treatment for postoperative patients.Methods:A total of 747 patients who underwent laparoscopic ovarian cystectomy with postoperative pathological confirmation of OEM were selected from January 2010 to December 2019 in the First Affiliated Hospital of Jinan University.Their clinical information such as preoperative general condition,menstrual history,maternity history,imaging index,laboratory examination index,endometriosis(EMs)related condition,gynecological complications,history of chronic medical conditions,intraoperative condition and postoperative follow-up within 2 years were collected.The patients were randomly divided into training group(n = 522)and validation group(n = 225)in the ratio of 7:rASThe training group was used for screening variables and constructing predictive model,while the validation group was used for testing the performance of model.The nomogram was constructed by the independent risk factors for postoperative recurrence of OEM laparoscopic conservative operation,which were determined by univariate and multivariate Logistic regression analysis.Besides,the nomogram was internally verified by Bootstrap method.The discrimination of the predictive model was evaluated by Receiver Operating Characteristic(ROC)Curve and Area Under the ROC Curve(AUC),Net Reclassification Index(NRI)and Comprehensive Discriminant Improvement Index(IDI).The calibration of the model was tested by calibration curve and Hosmer-Lemeshow test.The clinical validity of the model was evaluated according to clinical Decision Curve Analysis(DCA).Results:1.The recurrence rate of 747 OEM patients who underwent expectant therapy after surgery within two years was 28.92%(216/747)in our research.2.Univariate analysis showed that the statistically significant variables of postoperative recurrence were age at surgery,dysmenorrhea,laterality of OEM,Carbohydrate Antigen 125(CA125),Neutrophil-to-Lymphocyte Ratio(NLR),history of EMs-related medication,EMs-related surgery and combined adenomyosis.Eight variables selected by univariate analysis were statistically correlated with postoperative recurrence were included in multivariate logistic regression analysis,which showed that patients who underwent operation in older age(OR=0.149,95%CI:0.043-0.472 for age >31 years old compared to ≤22 years old)was a protective factor for postoperative recurrence in OEM patients(P <0.05);the independent risk factors(all P <0.05)are:(i)preoperative dysmenorrhea(OR=2.152,95%CI:1.303-rAS614),(ii)high levels of CA125(OR=2.153,95%CI:1.040-4.677 for CA125 of 15.85-39.8 U/ml compared to CA125 ≤15.85 U/ml;OR=5.933,95%CI:2.974-12.601 for CA125 > 39.8U/ml compared to CA125 ≤15.85U/ml),(iii)NLR >2.6(OR=4.155,95%CI:2.565-6.795),(iv)history of medication related to EMs(OR=2.261,95%CI:1.186-4.309),(v)history of EMs-related surgery(OR=5.137,95% CI:2.213-12.360).rASSix clinicopathological factors such as age at surgery,dysmenorrhea,CA125,NLR,history of EMs-related medication and history of EMs-related surgery,were applied to construct nomogram and internally validated with the following results:(1)Discrimination: The AUC of nomogram was 0.825,significantly higher than the AUC of r ASRM score(AUC=0.576)and r ASRM staging(AUC=0.585)(P <0.001),the optimal Probability Threshold(Pt)of nomogram was 0.336.The nomogram had NRI and IDI of 1.017(95%CI: 0.669-1.345)and 0.343(P <0.001)respectively compared to r ASRM scores and 1.041(95%CI:0.742-1.316)and 0.338(P <0.001)respectively compared to r ASRM staging.(2)Calibration: The deviation between the fitting line of the nomogram from the standard line on the calibration plot was subtle,and the P-value of Hosmer-Lemeshow test was 0.103,which showed good prediction accuracy of the nomogram.(3)Clinical validity: The DCA curves showed that the NB of the nomogram was higher than the NB of the r ASRM score,r ASRM staging and the extreme curves(Curve None and Curve All)when the decision Pt was in the range of 0.2-0.9,suggesting that the use of this nomogram increases the net benefit of the OEM patient after conservative surgery.Conclusion:1.The recurrence rate of OEM patients within 2 years after conservative surgery in this research was 28.92%.The independent risk factors were age at surgery,dysmenorrhea,level of CA125,level of NLR,history of EMS-related medication and history of EMS-related surgery.2.A nomogram for postoperative recurrence in OEM patients was constructed based on six clinicopathological characteristics,including age at surgery,dysmenorrhea,CA125 level,NLR level,history of EMs-related medication and history of EMs-related surgery.Patients with nomogram prediction probability ≤0.336 were classified as patients with low risk of recurrence and those >0.336 were judged as patients with high risk of recurrence.Its predictive ability and clinical application value were superior to r ASRM score and r ASRM staging. |
PDF Full Text Request