RETURN TO ABSTRACT LISTING


Objective:
Clinical presentations of the acute scrotum may be confounding, leading to abdomino-pelvic computed tomography (CT) imaging in lieu of initial sonographic evaluation. Though the scrotum is not routinely included at CT, radiologist assessment of the spermatic cord may aid in diagnosis. The purpose of this study was to evaluate qualitative and quantitative spermatic cord CT findings in the presence or absence of symptomatic scrotal pathology (SSP) at ultrasound.

Materials and Methods:
This retrospective study included 112 male patients (mean age 49.5 years) undergoing contrast-enhanced (CECT, n = 88) or noncontrast (NECT, n = 24) CT and scrotal ultrasound within a 24-hour time period. For CECT cases, attenuation of the spermatic cord was determined in Hounsfield units and differential spermatic cord enhancement calculated. Three fellowship trained body radiologists, blinded to the diagnosis, also independently assessed qualitative spermatic cord CT abnormalities for both CECT and NECT examinations. Both qualitative and quantitative findings were compared with the presence or absence of SSP determined at contemporaneous scrotal ultrasound. Reader performance, interobserver agreement and reader confidence were assessed for all CTs, and were compared between NECT and CECT examinations. Laterality and sidedness were evaluated with binomial tests and independent samples t-tests were used to examine differences in confidence. The Intraclass Correlation Coefficient (ICC) was used to evaluate interobserver agreement and ROC curves with area under the curve (AUC) used to characterize quantitative and qualitative evaluations and identify optimal cutoff points.

Results:
SSP was present in 34 out of 112 patients (30.3%). Positive cases were unilateral in 28 (82.4%) and bilateral in 6 (17.6%), p =.001. For unilateral cases, there was no difference in the sidedness of positive results (35.7% left; 64.3% right), p = .19. Among CECT positive cases, a 25% cutoff value for differential cord enhancement had the highest diagnostic accuracy (87.5%), with 85.7% positive predictive value (PPV) and 95.2% specificity. For qualitative image interpretation, CECT reader performance was excellent (aggregate AUC = 0.86; p < .001), with sensitivity and specificity of 71.4 % and 90.3% , respectively. NECT performance was nondiscriminatory (aggregate AUC = 0.61; p = .18). Readers had significantly higher confidence levels when evaluating CECT (p < .001). Inter-observer agreement for qualitative reads was excellent for both CECT (ICC = .98) and NECT (ICC = .96). For CECT reads, a 25% differential enhancement threshold would have decreased false negatives from 17% to 9% (mean across readers).

Conclusion:
Spermatic cord differential enhancement and qualitative abnormalities at CECT are both accurate predictors of SSP. A differential enhancement threshold of 25% may improve radiologist performance. However, NECT was not predictive of SSP.