Accuracy of MRI without intracavernosal prostaglandin E1 injection in staging, preoperative evaluation, and operative planning of penile cancer.

Abstract

To evaluate the performance of non-erectile MRI in staging and preoperative evaluation of penile carcinomas, compared to postoperative histopathology. In this retrospective study, MRI scansof patients who had undergone surgery for penile carcinoma (n = 54) between January 2012 and April 2018 were read by two radiologists; anddisagreement was solved in the presence of a third experienced radiologist. Data necessary for preoperative evaluation and staging were collected and compared with final postoperative histology and the type ofsurgery performed. All MRI had been performed without intracavernosal injection of prostaglandin E1 and with IV Gadolinium, as per local protocol. 54 patients were included in the study (mean age 57.52 ± 12.78). The number of patients with T1, T2, and T3 staging in histopathology were 32, 14, and 8. Moderate interobserver agreement was found for staging, disease-free penile length, and all subsites except urethra, which had weak agreement. Strong agreement of consensus MRI with final histopathological staging was found (49/54, weighted κ = 0.85), with high sensitivity and specificity. Sensitivity and specificity for involvement of corpus spongiosum, corpora cavernosa, and urethra were 95.5% and 93.8%, 87.5% and 97.8%, and 90.9% and 86.1%, respectively. Sensitivity (89.6%) and specificity (100%) of MRI for predicting adequate disease-free penile length were high. There were acceptable interobserver agreement and good diagnostic performance of MRI for staging and preoperative assessment without intracavernosal injection, especially for higher stages and higher degrees of invasion which require more extensive surgery.