RATIONALE AND OBJECTIVES: Little has been reported on the ability of endoluminal ultrasound (EUS) to identify a normal pancreas after an abnormal axial computed tomogram (CT). Many clinicians still use axial technology, as opposed to helical or spiral CT, which differ in scanning speed. Spiral CT and EUS are considered equal in their ability to diagnose pancreatic tumors. Although this is not the case with axial CT, the complementary role by EUS has not been defined. This study reports on the ability of EUS to identify the "true-negative" pancreas deemed abnormal by axial CT. METHODS: Sixty-five consecutive patients suspected of having a small pancreatic lesion were studied by comparing axial CT and EUS examinations, using each patient as his or her own control. Identification of a normal pancreas was reviewed using surgery, biopsy, and long-term clinical follow-up as the standard of truth. RESULTS: Thirty-three patients were documented as having small pancreatic lesions; the remaining 32 were normal. The sensitivity and specificity, respectively, were 91% and 41% for axial CT and 88% and 88% for EUS. The positive and negative predictive values, respectively, were 61% and 82% for CT and 88% and 88% for EUS. The statistical differences between axial CT and EUS were significant. CONCLUSION: An axial CT positive for a small pancreatic mass requires confirmation with additional imaging before invasive management. The specificity of EUS--twice that of CT--is strong evidence that EUS can fulfill this role. Review of the literature supports the conclusion that EUS should be required in the workup of small pancreatic lesions identified at axial CT.
RATIONALE AND OBJECTIVES: Little has been reported on the ability of endoluminal ultrasound (EUS) to identify a normal pancreas after an abnormal axial computed tomogram (CT). Many clinicians still use axial technology, as opposed to helical or spiral CT, which differ in scanning speed. Spiral CT and EUS are considered equal in their ability to diagnose pancreatic tumors. Although this is not the case with axial CT, the complementary role by EUS has not been defined. This study reports on the ability of EUS to identify the "true-negative" pancreas deemed abnormal by axial CT. METHODS: Sixty-five consecutive patients suspected of having a small pancreatic lesion were studied by comparing axial CT and EUS examinations, using each patient as his or her own control. Identification of a normal pancreas was reviewed using surgery, biopsy, and long-term clinical follow-up as the standard of truth. RESULTS: Thirty-three patients were documented as having small pancreatic lesions; the remaining 32 were normal. The sensitivity and specificity, respectively, were 91% and 41% for axial CT and 88% and 88% for EUS. The positive and negative predictive values, respectively, were 61% and 82% for CT and 88% and 88% for EUS. The statistical differences between axial CT and EUS were significant. CONCLUSION: An axial CT positive for a small pancreatic mass requires confirmation with additional imaging before invasive management. The specificity of EUS--twice that of CT--is strong evidence that EUS can fulfill this role. Review of the literature supports the conclusion that EUS should be required in the workup of small pancreatic lesions identified at axial CT.