Finding Vessels – Accuracy of CT for Identifying Anomalous Feeding Vessels in Patients with Congenital Cystic Lung Lesions

Abstract

Contrast computer tomography (CT) scanning is the investigation of choice for the further assessment of suspected cystic congenital lung lesions (CCLL). Its use to identify the presence of anomalous feeding vessels supplying the lesion is well documented, but data regarding its accuracy is limited. This study compares CT results to operative and pathological findings to determine the accuracy of CT in identifying these anomalous vessels. 51 consecutive cases of cystic congenital lung lesions managed in one hospital by a single consultant were reviewed. All cases had contrast CT scans performed preoperatively, as standard practice in this institution. We compared the results of these CT scans to the macroscopic appearance at surgery and histological findings postoperatively. We also compared the results of 2 CT protocols used in our institution between 1999-2007 and 2007-2009, respectively. Anomalous vessels were reported on CT in 9 cases. All but 1 had concordant operative and pathological findings. In the remaining 42 cases, no anomalous vessels were seen on CT. Of these, 9 cases were found to have an anomalous blood supply at surgery, 6 of which were hybrid lesions and 3 isolated sequestrations. The specificity of CT in identifying feeding vessels in the study was 97% (95% CI: 0.83-0.99) and the sensitivity was 47% (95% CI: 0.23-0.71). The positive predictive value was 89% (95% CI: 0.50-0.99) and negative predictive value 79% (95% CI: 0.62-0.89). The most recent protocol yielded an improved sensitivity of 75% (95% CI: 0.22-0.98) and a specificity of 100% (95% CI: 0.46-1.0) with a 100% (95% CI: 0.31-1.0) positive and 83% (95% CI: 0.36-0.99) negative predictive value. CT is a specific investigation for identifying anomalous vessels in CCLL but lacks sensitivity, leading to a relatively low negative predictive value. This emphasises the need in every case to look for anomalous vessels at surgery to avoid morbidity and potential mortality. An improved protocol for CT scans leads to improved specificity and sensitivity and predictive values.