ORF22-01 - Spectrum of Gonadal Neoplasia in Molecularly Confirmed Complete Androgen Insensitivity Syndrome: MRI Criteria to Define Benignity vs. Malignancy

Background: In complete androgen insensitivity syndrome (CAIS), retained gonads may harbor a spectrum of proliferative lesions from benign Leydig cell hyperplasia (LCH) to invasive neoplasia. As management increasingly shifts toward gonadal retention and active surveillance, the inability of standard imaging to confidently distinguish benign mimics from true neoplasia remains a major barrier. We performed a radiologic–pathologic correlation study to define quantitative imaging criteria of benignity in adult CAIS.

Methods: We analyzed 32 gonads from 16 postpubertal patients with molecularly confirmed CAIS (pathogenic AR variants) who underwent gonadectomy. Preoperative multiparametric MRI including diffusion-weighted imaging (DWI) was correlated with gonadal volume and comprehensive histopathology with immunohistochemistry (e.g., OCT4, SALL4, inhibin). The diagnostic performance of MRI features was assessed, with an emphasis on volumetric thresholds for discriminating benign hyperplastic nodules from malignancy.

Results: Macroscopic neoplasia occurred in 3/16 patients (18.7%), including seminoma, large-cell calcifying Sertoli cell tumor, and serous borderline tumor. Non-neoplastic gonads frequently exhibited benign mimics, including dense LCH and hamartomas. DWI restriction was highly sensitive but showed limited specificity (84.6%) due to a consistent false-positive pattern (“Leydig cell effect”), in which hyperplastic LCH nodules mimicked restricted diffusion. Quantitative morphometry revealed a marked volumetric separation: all benign lesions (including DWI false-positives) measured < 22 cm³, whereas all neoplastic gonads exceeded 116 cm³ (p < 0.001), supporting a proposed volumetric safety threshold < 30 cm³ to characterize benignity. In this cohort, absence of diffusion restriction showed 100% negative predictive value for macroscopic neoplasia, supporting benignity in homogeneous gonads.

Conclusions: In molecularly confirmed CAIS, benign LCH is common and can mimic neoplasia on functional MRI, particularly via restricted diffusion. We propose a pragmatic multimodal surveillance framework: gonads without diffusion restriction are highly reassuring, whereas gonads with restricted diffusion but low volume ( < 30 cm³) likely represent benign LCH rather than neoplasia. These quantitative criteria help define imaging “benignity,” support safe active surveillance, and may reduce unnecessary anxiety and potentially avoidable surgery.