Unenhanced MRI of the abdomen and pelvis for surveillance of patients with stage 1 testicular cancer post-radical orchiectomy.

Abstract

To evaluate the ability of unenhanced magnetic resonance imaging (MRI) to detect metastatic lymphadenopathy in men with stage 1 testicular cancer on surveillance. With IRB approval, we identified 23 consecutive men with stage 1 testicular cancer (diagnosed after orchiectomy) undergoing surveillance with unenhanced MRI of the abdomen and pelvis. Three blinded radiologists (2 inexperienced R1/R2 and 1 experienced R3) independently assessed MRI for: presence, location and size of abnormal lymph nodes and degree of confidence (5-point Likert scale) in diagnosis. Diagnostic accuracy was tabulated and compared between groups using ROC. Inter-observer agreement was assessed using Cohen's kappa statistic. 17.4% (4/23) men developed 6 metastatic lymph nodes (reference standard: interval development from baseline, size > 1.0cm short axis). R1 and R2 detected 75% (3/4) patients with abnormal lymph nodes, compared to R3 who detected all four cases. False positive interpretations occurred in: 5.2% (1/19) R1, 10.5% (2/19) R2 and no patients for R3. Sensitivity, specificity and area under the ROC curve (AUC) were: R1: 75% (95% CI 19.4-99.4%), 94.7% (74.0-99.9%) and 0.85 (0.59-1.00), R2: 75% (19.4-99.4%), 89.5% (66.9-98.7%) and 0.82 (0.57-1.00) and, R3: 100% (95% 39.8-100.0%), 100% (82.4-100%) and 1.00 (1.00-1.00) with no difference in AUC between readers (p = 0.383). Comparison in accuracy between readers is limited due to the small sample size. Inter-observer agreement was substantial (K = 0.62). Median (range) degree of confidence scores were rated: R1 5 (5-5), R2 4 (3-5) and, R3 5 (5-5). In this study, unenhanced MRI was adequate for surveillance of stage 1 testicular cancer; however, radiologist inexperience may lead to errors.