Accuracy of the Deep Pelvic Endometriosis Index

Accuracy of the Deep Pelvic Endometriosis Index | Image Credit: © Pixel-Shot - © Pixel-Shot - stock.adobe.com.

The Deep Pelvic Endometriosis Index (dPEI) may be able to predicthospital stay, operating time, postoperative complications, and de novo postoperative voiding dysfunction, according to a recent study published in JAMA Network Open.

Endometriosis is a common gynecologic disorder that causes severe chronic pelvic pain and infertility, leading to massive detriments in patients’ quality of life. First-line hormonal treatment followed by surgery if unsuccessful is recommended in patients with endometriosis by the European Society of Human Reproduction and Embryology.

Surgery for deep pelvic endometriosis (DPE) leads to major risks of intraoperative and postoperative complications, making preoperative evaluation and inclusion of the patient in the decision-making process vital. Surgery should be performed by a gynecologic surgeon who is an expert in endometriosis, with additional experts on the surgery team.

In 2020, the dPEI was published with the goal of using imaging to standardize pretherapeutic staging. It was the first score to include lateral locations of DPE and uses a compartment-based approach.

To evaluate the efficacy of the dPEI score in predicting postoperative complications in women with DPE, investigators conducted the ENDOVALIRM study. Participants were identified through queries to 7 French endometriosis centers.

Women who received surgery and preoperative magnetic resonance imaging (MRI) for DPE from January 1, 2019, to December 31, 2020, were included in the analysis. Women were excluded if they were aged under 18 years, had menopausal status, pregnancy, or history of DPE surgery, were missing clinical, pathologic, surgical, or imaging data, or had gone more than 12 months between MRI and surgery.

Data on parity, gravidity, infertility, presurgical hormonal treatment, and clinical symptoms was evaluated. Clinical symptoms included dyspareunia, dysmenorrhea, pain on defecation or dyschezia, catamenial diarrhea, chronic pelvic pain, rectal bleeding, and preoperative dysuria.

A phased array coil was used to collect MRI sequences at 1.5T or 3T. The MR images were reviewed independently by 7 radiologists with 5 or more years’ experience in pelvic MRI. Radiologists were unaware of histologic results, clinical history, and surgical examination, but knew about DPE presence.

Surgical procedures included partial colpectomy, hysterectomy, unilateral or bilateral uterosacral ligaments resection, unilateral or bilateral parametrectomy, unilateral or bilateral ovarian cystectomy, discoid resection, partial bladder resection, segmental digestive resection, rectal shaving, unilateral or bilateral ureterolysis, and ureteral reimplantation.

Operating times were evaluated during the study. Postoperative data included postoperative complications, duration of hospital stay, presence of de novo voiding dysfunction, and late postoperatice complications.

There were 605 women in the final cohort, aged a mean 33.3 years. Of this cohort, 63.8% were nulliparous women, 88.8% of which had never been pregnant. Also, 18.5% were uniparous and 17.7% were multiparous. Hormonal treatment before surgery was seen in 63.6% of patients.

An average 159.3 days between MRI and surgery was reported, with an average operating time of 152.2 minutes and an average hospital stay of 3.9 days. Mild disease was reported in 61.2% of women, moderate in 25.8%, and severe in 13.1%.

Women with mild disease had observed adnexal locations, superficial locations, and central endometriosis in 58.9%, 11.9%, and 93.2% of cases respectively. Central endometriosis was observed in all patients with moderate or severe disease, while adnexal locations were observed in 74.4% of women with moderate disease and 73.4% with severe disease.

Superficial locations were found in 12.2% of women with moderate disease and 32.9% with severe disease. Lateral endometriosis was found in 31.2% of the cohort, 6.7% of women with mild disease, 48.7% with moderate disease, and 98.7% with severe disease. The absence of intraoperative findings was accurately predicted through MRI.

Longer operating times and hospital stays were reported in women with DPE compared to those without DPE. The risk of severe postoperative complications was greater in women with DPE in the posterocentral, mediolateral, and extrapelvic compartments. De novo voiding dysfunction was also associated with the posterocentral, posterolateral, mediolateral, and anterocentral compartments.

A significant impact on operating time was made by dPEI score. Severe DPE was associated with longer median times than moderate DPE, while moderate DPE was associated with longer times than mild DPE. Similar associations were found between dPEI score and hospital stay.

Of patients with severe DPE, 11.4% had severe postoperative complications, compared to 5.1% with moderate DPE and 2.7% with mild DPE. When comparing junior and senior readers for DPEI scores, strong interobserver agreement was found, indicating reproducibility.

Accurate predictions were found for operating time, hospital stay, and postoperative complications using the dPEI score. This may help clinicians prepare patients for surgery.

Reference

Thomassin-Naggara I, Monroc M, Chauveau B, et al. Multicenter external validation of the deep pelvic endometriosis index magnetic resonance imaging score. JAMA Netw Open. 2023;6(5):e2311686. doi:10.1001/jamanetworkopen.2023.11686