Clinical and laboratory data or imaging results cannot provide a positive diagnosis of septic complications of pancreatic and peripancreatic necrosis in patients with acute pancreatitis. Confirmation can be obtained only after percutaneous computed tomography (CT)-guided aspiration of the necrotic tissues or fluid collection; although the important role of 99Tc(m)-HMPAO-labeled granulocyte scintigraphy has been recently emphasized. The aim of this study was to determine the sensitivity and specificity of 99m-technetium-hexamethylpropyleneamine oxime (99Tc(m)-HMPAO)- or 111In-oxine-labeled granulocyte scintigraphy for the diagnosis of infection in pancreatic or peripancreatic necrosis to define the ideal label for diagnosis. Thirty-six scintigraphic examinations were performed in 34 consecutive patients (mean age, 58 +/- 2 years) 20 +/- 2 days after the onset of acute pancreatitis (Balthazar classes A-C, n = 7; classes D and E, n = 29). The scintigraphic study included scintigraphic tomography and static acquisition 1 and 3 h, respectively, after reinjection of the autologous 99Tc(m)-labeled granulocytes and static images 3-4 and 24 h after the simultaneous reinjection of 111In-oxine-labeled autologous granulocytes. The diagnosis of infected pancreatic or peripancreatic necrosis was confirmed with percutaneous CT-guided aspiration (14 positive aspirates among 20 performed) and sterile necrosis after negative aspiration (6 negative aspirates) or after a 6 +/- 1-month follow-up free of clinical or biological signs of ongoing sepsis. The sensitivity and specificity were 86 and 73%, respectively, for scintigraphic tomography, 100 and 55% for 3-h 111In images, 93 and 68% for 3-4-h 111In images, and 100 and 64% for 24-h 111In images. The fall in splenic activity between the 3-4 and the 24-h 111In images was 26 +/- 3% in patients with septic pancreatic and peripancreatic necrosis (n = 14) and 16 +/- 3% in those with sterile necrosis (n = 22) (p < 0.01). Labeled granulocyte scintigraphy was thus shown to be an effective tool for the diagnosis of infection in pancreatic and/or peripancreatic necrosis due to acute pancreatitis, especially when the scintiscans are performed early after injection of 99Tc(m) or when the fall in splenic activity over the 24 h following reinjection of 111In is measured in particularly difficult cases. These promising preliminary results should be confirmed by a prospective study.
Read full abstract