OBJECTIVE. The purposeof our studywas to assessthe potentialof thin-sectionmul tiphasichelicalCT in diagnosisandstagingof hilar cholangiocarcinomas.
SUBJECTS AND METHODS. IdenticallycollimatedhelicalCT studieswereperformedbefore and during the hepatic artery dominant phase and during the portal vein dominant phase of contrastenhancementin 29 consecutivepatientswith proven hilar cholangiocarcinomas.Dif ferencesin attenuation betweenthetumorandtheliverwerecalculatedin eachcaseby subtract ing the average attenuation of the tumor from that of the liver. A four-point scale termed a â€oe¿ lesion conspicuityscore― was usedto determineratesof tumordetection.CT findingswere correlated with surgically assessed extent of tumor, histologic findings, or both in all cases. RESULTS. Ten (34%) of the 29 hilar cholangiocarcinomas were detectedon unenhanced images.All hilar cholangiocarcinomas (100%) were seenon hepaticartery dominantphase scans,and 25 (86%) of 29 hilar cholangiocarcinomas were seen on portal vein dominant phase scans, regardless of the morphologic appearance. An infiltrating stenotic lesion was found in 17 (59%) of 29 patients, an exophytic hilar lesion was found in 11 patients (38%), and one patient (3%) had an intraluminal polypoid lesion. Mean differences in enhancement between infiltrating stenotic lesions and the liver were significantly greater on hepatic artery dominant phase scans (28 ±10 H) than on portal vein dominant phase scans (10 ±8 H), whereas the mean difference in enhancement between the exophytic lesions and the liver was statisticallygreaterduringtheportalveindominantphase(p < .01). Two of the hilar cholang iocarcinomas were resectableat surgery, and I 8 were not.The overallaccuracyof helicalCT for assessing resectability was 60%. In 10 (56%) of 18 patients, unresectable disease was cor rectlydiagnosedwith helicalCT (sensitivity, 56%). Eight (44%) of I 8 patientsconsideredto have resectable tumors with helical CT had unresectable tumors at surgery. A resectable Ut mor was correctly diagnosed in two patients with helical CT.
CONCLUSION.MultiphasichelicalCT canbe usedto detectandclassifyhilar cholang iocarcinomas. However,theexactproximaltumorextentalongbile ductstendsto be underes timated with helical CT; therefore, helical CT is inaccurate for determining resectability.H ilar cholangiocarcinomas are typi cally small, slow-growing, locally invasive tumors that have a dismal prognosis ifleft untreated, witha meansurvival of approximately 3 months after initial presen tation [1â€"7]. The anatomic location of hilar cholangiocarcinoma makesresectiondifficult [7]. Surgical exploration of these patients should be undertaken only when preoperative examination has shown a potential for curative resection because the risks of palliative surgery for malignant obstructive jaundice are high, with surgical mortality rates of 20â€"30% [8,9]. Accordingly, preoperative assessment of resec tability of hilar cholangiocarcinoma has in creased in importance in recent years because percutaneous and endoscopic p...