Objective: To determinate in a retrospective clinical study the benefit of helical-CT for the diagnosis and staging of pancreatic carcinoma compared with endoscopic ultrasonography.
Method: Results of helical-CT and endoscopic ultrasonographic investigations relative to 24 cases of histologically proved adenocarcinoma were reviewed. The criteria studied were: a) existence of a pancreatic mass, b) diagnosis of malignancy, c) existence of vascular involvement, d) lymph node involvement. Statistical analysis (binomial test) was performed in 24 patients for the assessment of mass and malignancy and in 10 patients for the assessment of vascular involvement (9 patients who underwent surgery and one patient with a positive arteriography).
Results: For diagnosis of pancreatic mass both methods were equivalent: 91.5% (22 cases out of 24) for helical-CT and 87.5% (21 cases out of 24) for endoscopic ultrasonography. Two pancreatic masses were seen only with endoscopic ultrasonography. Their size was under 3 cm. Two of the 3 masses not seen with endoscopic ultrasonography were infiltrative lesions measuring more than 3 cm. Helical-CT was significantly superior to endoscopic ultrasonography for the diagnosis of malignancy (96% vs 71%) (P = 0.035). The accuracy of helical-CT for vascular involvement was significantly superior to endoscopic ultrasonography (90% vs 40%) (P = 0.031). Endoscopic ultrasonography underestimated vascular involvement in 5 cases: mesenterico-portal confluence (3 cases), superior mesenteric artery (2 cases). In the 6 cases with histologically proved lymph node involvement, endoscopic ultrasonography (6 correct assessments) was superior to helical-CT (3 correct assessments).
Conclusion: This study confirms the complementarity of helical-CT and endoscopic ultrasonography for the diagnosis of pancreatic carcinoma. It suggests that helical-CT is superior to endoscopic ultrasonography for diagnosis of malignancy and assessment of vascular involvement and endoscopic ultrasonography is superior to helical-CT for diagnosis of lymph node involvement. These preliminary results have to be confirmed by a prospective study including a large number of patients with surgical correlation.