Initial Results of FAPI PET/MRI to Assess the Extent of Endometriosis

Endometriosis is a common gynecologic disorder often associated with chronic pelvic pain, infertility, and reduced quality of life. Conventional imaging modalities, such as transvaginal ultrasound and MRI, may underestimate the extent of disease, particularly in deep infiltrating endometriosis. Fibroblast activation protein (FAP) is highly expressed in remodeling endometrium and has been implicated in endometriotic lesions. This study investigates the feasibility of integrating FAP-targeted PET/MRI to improve endometriosis detection. Methods: This retrospective study included 18 premenopausal women. The test cohort consisted of 9 patients with suspected or confirmed endometriosis who had undergone FAP-targeted PET/MRI or PET/CT with the FAP inhibitor (FAPI) [68Ga]BED003 (previously known as [68Ga]Ga-OncoFAP-DOTAGA). The reference cohort consisted of 9 premenopausal women without known endometriosis who had undergone FAPI PET/MRI or PET/CT with either [68Ga]BED003 or [68Ga]Ga-FAPI-46 for initial staging of breast cancer. Examinations were not scheduled with regard to the patients’ menstrual cycles. Lesions were assessed using a segment-based analysis derived from the #Enzian classification. Focal radiopharmaceutical uptake was semiquantitatively assessed using SUVmax and SUVmean. Clinical assessments and laparoscopic findings, when available, served as reference standards. Results: In the test cohort with available PET/MRI data (7 patients and 91 segments), MRI alone identified endometriosis-suspicious findings in 19% of segments. In contrast, combined PET/MRI identified findings in 49% of segments. Among all PET datasets (108 segments per cohort), focal extrauterine FAP uptake was observed in 41% of the test cohort’s segments compared with 13% of the reference cohort’s segments (P < 0.01). The mean number of PET-positive segments per patient was higher in the test cohort than in the reference cohort (4.9 ± 2.4 vs. 1.7 ± 1.3, P = 0.009), with the greatest difference observed in peritoneal, ovarian, and tubal segments. Uterine SUVmax did not differ significantly between the 2 cohorts (P = 0.489). Where available, PET/MRI findings showed substantial concordance with laparoscopic results. Conclusions: FAPI PET/MRI increased the detection of lesions suggestive of endometriosis compared with MRI alone. These findings support further prospective evaluation of FAPI PET/MRI as an adjunct imaging modality for preoperative assessment of endometriosis.