OBJECTIVE. The purpose of this study was to determine the accuracy of
helical CT scanning in predicting the stage of carcinoma of the exocri
ne pancreas using TNM staging, guidelines and in predicting resectabil
ity. MATERIALS AND METHODS. Twenty-six patients with proven adenocarci
noma of the pancreas underwent uniphasic or biphasic helical CT scanni
ng. Two observers unaware of the patient's surgical stage evaluated th
e CT examinations using the TNM system (with specific assessment and d
escription of disease sites), In addition, the two observers rated con
fidence of nonresectability using a 5-point scale (ranging from 1, def
initely resectable, to 5, definitely not resectable). Observer results
and preoperative interpretations were compared with surgical findings
. RESULTS. Nineteen of 26 patients had nonresectable disease. The comb
ined observer scores showed correct determination of T stage in 77% of
patients, of N stage in 58%, and of M stage in 79%. The overall accur
acy in determining lack resectability was 96% and 84% for the two obse
rvers. All errors in determining resectable versus nonresectable disea
se occurred when the observer was not maximally confident of his or he
r diagnosis. CONCLUSION. Helical CT is an effective screening techniqu
e for assessing T and M stages of pancreatic carcinoma. However, helic
al CT is poor at detecting regional lymph node involvement. In patient
s with equivocal T-stage findings (such as questionable venous involve
ment), other studies such as endoscopic sonography may be of value.