P-748 Fertility and anatomical outcome following hysteroscopic adhesiolysis of intrauterine adhesions classified according to symptoms, imaging findings and hysteroscopic appearance of the uterine cavity

Abstract

Abstract Study question To classify patients with dense intrauterine adhesions based on clinical characteristics as well as, ultrasound, hysterosalpingography (HSG) and hysteroscopy findings into different prognostic groups. Summary answer We have defined a prognostic model to classify patients with dense intrauterine adhesions which could easily be implemented in patient counseling and management. What is known already Intrauterine adhesions are a major cause of hypo-amenorrhea and failure to conceive. A universally agreed classification to categorize intrauterine adhesions is paramount to provide individualized counseling and care. Previous efforts for categorization suffer from limitations such as variable assessment of anatomical outcomes, application of different techniques, pooling of results of different operators and lack of long-term follow-up. Many experts have voiced a call for a prognosis-oriented classification system. An ideal classification model should consider clinical characteristics, findings from imaging techniques (including ultrasound and HSG) together with hysteroscopic appearance of the uterine cavity and demonstrate a high predictive value. Study design, size, duration This is a retrospective analysis of 281 patients treated for intrauterine adhesions by a single operator (B.U.) between 2010 and 2021. Lysis of adhesions was affected in 479 office hysteroscopy procedures using the Versapoint bipolar cutting electrode under transabdominal ultrasound guidance. 227 patients were followed for at least 15 months after the last surgical procedure. Patients were classified into five categories (Class I to V) according to their symptoms, ultrasound, HSG and hysteroscopy findings. Participants/materials, setting, methods Clinical data and operative findings were reviewed from patient files and video recordings. The number of hysteroscopic interventions needed to restore the cavity and the reproductive outcome in women who were desirous for pregnancy were recorded. Predictive power of the model was assessed using the live birth rate as the primary and rate of cavity restoration and number of interventions as the second outcome parameters. Groups were compared using ANOVA, ROC and regression analyses. Main results and the role of chance Adhesions were classified as class I in 43 (15.3%), class II in 72 (25.6%), class III in 57 (20.3%), class IV in 82 (29.2%) and class V in 27 (9.6%) patients. They were due to previous curettages of pregnancies (79,7%) or retained products of gestation (3.9%), prior uterine surgery (6.8%), prior hysteroscopy of inappropriate technique (6.8%) and tuberculosis (2.8%). The cavity was septate in 12 and unicornuate in 2 patients. The mean age of the study group was 29.8 ± 3.7 (20-40). Age was not related with the severity of adhesions (p = 0.335). While the majority of patients with curettage-related adhesions were classified as Class II, uterine surgery, iatrogenic and tuberculosis related adhesions were higher in severity. The number of hysteroscopic adhesiolysis procedures (from 1.0 ± 02. to 2.3 ± 0.5) needed for optimal restoration of the cavity was directly related and the rate of full restoration (from 100% - 18.5%) was indirectly related with the severity of adhesions according to the proposed classification (p = 0.0001 for both). The live birth rates were 54.3%, 45%, 31.7%, 21% and 12.5% for patients in Class I to V, respectively (p = 0.0001). The proposed classification was fairly predictive (AUC: 0.654, 95%CI: 0.582-0.727) for live birth. Limitations, reasons for caution This is a retrospective analysis of consecutive patients with intrauterine adhesions in routine practice. The follow-up for reproductive outcome is limited. The study is hospital-based and single-center. Thus, the predictive value of the proposed classification needs to be validated in an external data set preferably in a prospective series. Wider implications of the findings In patients with intrauterine adhesions, a classification system based on patient symptoms, imaging findings and hysteroscopic appearance of the uterine cavity reliably predicts the postoperative outcome in terms of the extent of anatomical restoration of the uterine cavity and pregnancy and live birth rates. Trial registration number Not applicable