Two-phase helical CT for detection of early gastric carcinoma: Importance of the mucosal phase for analysis of the abnormal mucosal layer

Purpose: To evaluate the efficacy of two-phase dynamic helical computed tom ography (CT), including the gastric mucosal phase, for detection of early g astric carcinoma with typical hyperattenuating and atypical nonhyperattenua ting enhancement patterns. Method: Two-phase helical CT scanning was performed using the water-filling method as negative oral contrast material for 212 patients with proven ade nocarcinoma on endoscopic biopsy. Two gastrointestinal radiologists prospec tively analyzed the focal alterations of the inner hyperattenuating mucosal layer and the outer hypoattenuating layer before the information obtained at barium study and pathologic examination was available. The first, so-cal led mucosal phase was obtained at 38-45 seconds after the start of intraven ous injection of contrast material for a total of 150 ml/sec at a rate of 4 ml/sec to obtain maximum enhancement of the inner mucosal layer, The secon d delayed phase was obtained at 3 minutes. Results: Fifty-four cases of early gastric cancer were suspected on two-pha se helical CT preoperatively. Postoperatively, 49 cases of early gastric ca ncer were pathologically confirmed. The detection rate for the typical hype rattenuating early gastric cancer, that is the type 1 enhancement pattern d efined as the localized thickening of the inner hyperattenuating layer, usi ng two-phase helical CT was 18% (9/49 patients). The type 2 enhancement pat tern, defined as the focal interruption of the inner hyperattenuating mucos al layer without abnormal enhancement of the outer hypoattenuating layer on the mucosal phase, was seen in 15 cases. These were pathologically confirm ed as early gastric cancer IIb + IIc (three patients), IIc (four patients), IIc + IIa (one patient), IIc + III (three patients), IIb + IIc (one patien t), and advanced cancer (T2) lesions (three patients). The type 3 enhanceme nt pattern was defined as the focal polypoid protrusion of the inner hypera ttenuating layer without abnormal enhancement of the outer thickened hypoat tenuating layer on the mucosal phase, and was seen in seven patients who we re pathologically confirmed with early gastric cancer IIb + IIc (three pati ents), IIc + IIa (one patient), and IIc + IIb (three patients). The lesions became less distinct on the delayed phase. Five T2 advanced gastric cancer s were falsely interpreted as early gastric cancer. The detection rate for early gastric cancer after considering type 2 and 3 atypical enhancement pa tterns was increased to 57% (28/49 patients). Conclusion: Helical CT with two-phase scan including the mucosal phase was efficient for identifying the enhancement patterns of early gastric cancer.