Preoperative Three-Dimensional Magnetic Resonance Angiography-Based Anatomical Risk Stratification in Transvaginal Sacrospinous Ligament Fixation: A Preliminary Predictive Analysis of Intraoperative Hemorrhage and Postoperative Complications

Background: To establish a preliminary three-tier anatomical risk classification for sacrospinous ligament fixation (SSLF) using preoperative three-dimensional (3D) magnetic resonance angiography (MRA), evaluate its correlation with intraoperative hemorrhage and postoperative complications, as well as compare perioperative outcomes between MRA-guided and conventional SSLF while accounting for potential confounding between risk stratification and tailored surgical approaches in the MRA cohort. Methods: This single-center, parallel-group randomized controlled trial was conducted at a tertiary gynecological center in China from January 2024 through December 2025. A total of 140 patients with pelvic organ prolapse quantification (POP-Q) stage III–IV uterine prolapse scheduled for uterus-preserving transvaginal SSLF were randomly assigned in a 1:1 ratio to either a 3D MRA-guided group (n = 70), which underwent preoperative MRA with 3D reconstruction and individualized surgical planning, or a conventional palpation-guided group (n = 70), which underwent standard tactile-guided surgery without preoperative MRA. In the MRA-guided group, bilateral sacrospinous ligament (SSL) morphometry and the spatial relationship with the internal pudendal artery (IPA) were quantified using 3D reconstructions (Mimics version 19.0). Patients were classified into three anatomical risk categories: Type I (standard risk: IPA distance ≥20 mm and thickness ≥3.5 mm), Type II (moderate risk: IPA distance 15–19 mm or thickness 2.5–3.4 mm), and Type III (high risk: IPA distance <15 mm or thickness <2.5 mm). Primary outcomes comprised operative time, estimated blood loss (EBL), and length of hospital stay. Prespecified secondary outcomes included 6-month overall postoperative complication rates and 1-month Pelvic Floor Distress Inventory-20 (PFDI-20) scores. Risk-stratified subgroup analyses, univariate receiver operating characteristic (ROC) curve analyses, and Cochran–Armitage trend testing were designated as post hoc exploratory analyses, with no formal adjustment for multiple comparisons applied. Multivariable logistic regression analysis was performed solely in the full study cohort; subgroup regression analysis within the MRA-guided group was omitted due to severe overfitting risk from limited complication events. Results: Baseline demographic characteristics demonstrated acceptable intergroup balance, with all standardized mean differences (SMDs) <0.2, except for BMI (SMD = 0.448) showing moderate imbalance; BMI was included as a covariate in multivariable regression to account for this imbalance, rather than relying on p‑value comparisons to confirm randomization success. The MRA-guided group exhibited a significantly shorter operative time (56.2 ± 9.8 vs. 74.8 ± 11.6 min, p < 0.001), reduced median EBL (39 [32–51] vs. 64 [53–79] mL, p < 0.001), and shorter hospitalization (3.2 ± 0.5 vs. 4.5 ± 1.1 d, p < 0.001). Median EBL was reported alongside mean ± standard deviation (SD) to account for four extreme hemorrhage outliers (210, 225, 240, and 270 mL) in the conventional arm, which distorted the distribution of raw EBL values. The overall complication rates were significantly lower in the MRA-guided group than in the conventional group (7.1% [5/70] vs. 24.3% [17/70], χ2 = 8.14, p = 0.004). Within the MRA-guided group, risk-type distribution was: Type I 44.3% (31/70), Type II 37.1% (26/70), and Type III 18.6% (13/70). Complication rates increased descriptively across risk types (3.2%, 7.7%, and 15.4% for Types I–III, respectively), although the Cochran–Armitage trend test did not reach statistical significance (p-value = 0.157). All four intraoperative vascular injuries occurred in the non-MRA group; no neurovascular injuries were observed in the MRA-guided cohort. On multivariable logistic regression including all 140 patients, assignment to the non-MRA group emerged as an independent predictor of postoperative complications (adjusted odds ratio [OR] = 4.1; 95% confidence interval [CI]: 1.3–12.8, p = 0.015). Separate univariate ROC curve analyses were conducted in distinct study populations. Among all participants, binary group allocation yielded an area under the curve (AUC) of 0.66 (95% CI: 0.57–0.75); within the MRA cohort, the three-tier risk classification achieved an AUC of 0.67 (95% CI: 0.54–0.80), outperforming IPA distance alone (AUC = 0.63; 95% CI: 0.50–0.76) and SSL thickness alone (AUC = 0.60; 95% CI: 0.47–0.73). A combined predictive model integrating group assignment and MRA risk type was excluded from analysis, as anatomical risk stratification data were unavailable for all patients in the conventional arm, thereby precluding construction of a unified model. Direct DeLong comparisons between ROC curves derived from non-overlapping patient subsets were not performed in accordance with methodologic guidance. Conclusions: The 3D MRA-derived SSL–pudendal vascular risk classification provides a preliminary framework for preoperative anatomical stratification in patients undergoing SSLF, with incremental descriptive associations with intraoperative hemorrhage and postoperative adverse events. Type III anatomical features were associated with a numerically higher complication risk despite the use of image-adapted surgical technique. Clinical Trial Registration: The trial was registered with the Chinese Clinical Trial Registry (ChiCTR) at https://www.chictr.org.cn (registration number ChiCTR2600121272).