Portal streamlining as a cause of nonuniform hepatic distribution of sodium pertechnetate during per-rectal portal scintigraphy in the dog

The purpose of this study was to evaluate nonuniform patterns of vascular distribution of pertechnetate in the dog during per-rectal portal scintigraphy. Ninety-two studies were reviewed retrospectively to document the patterns of radionuclide distribution. Forty-five studies were classified as normal and 47 were classified as diagnostic for a macrovascular portosystemic shunt. In these dogs, shunt fractions were calculated and compared using a t-test. In dogs with sufficient liver radioactivity for evaluation, the study was classified as having uniform, dorsal, central, or ventral radiopharmaceutical distributions. There were 51 animals (45 normal and six dogs with low-magnitude portosystemic shunts) with sufficient liver activity to assess the radionuclide distribution within the liver. A one-way ANOVA was used to compare shunt fractions between each of the distribution patterns. Two dogs were anesthetized and selective portovenograms were performed. Portovenograms were compared with the scintigraphic images to correlate the vascular distribution of the right, central, and left divisional branches of the portal vein. The shunt fraction in the 45 normal dogs was significantly lower than in the dogs with portosystemic shunts (5.7% +/- 4.8% vs. 78.6% +/- 20.0% (mean +/- SD), P < 0.001). Of the 51 dogs with sufficient liver activity to classify the pattern of distribution, there were 15/51 (31.4%) with uniform radionuclide distribution, 10/51 (19.6%) with focal dorsal distribution, 15/51 (29.4%) with focal ventral distribution, and 10/51 (19.6%) with focal central distribution. There was no significant difference in the shunt fractions between the groups. There were six dogs diagnosed with low-magnitude portosystemic shunt with sufficient liver radioactivity to categorize the vascular distribution of the radionuclide within the liver. Of these six dogs, two had focal dorsal distribution, one had focal central, one had focal ventral and two had uniform distribution. Focal dorsal distribution could result from streamlining of the radionuclide into the right divisional branch of the portal vein. Focal ventral distribution could result from streamlining the radionuclide into the left divisional branch of the portal vein. Focal central distribution could result from streamlining the radionuclide into the central divisional branch of the portal vein.     

VALIDATION OF DECONVOLUTIONAL ANALYSIS FOR THE MEASUREMENT OF HEPATIC FUNCTION IN DOGS WITH TOXIC-INDUCED LIVER DISEASE

The extraction of the hepatobiliary radiopharmaceutical ^99mTc-mebrofenin ( Choletec ) by the liver can be used to evaluate the severity of hepatocellular disease. The hepatic parenchymal cells extract mebrofenin from the blood by the same active transport mechanism as bilirubin. The ability of the liver to extract ^99mTc-membrofenin is a measure of hepatic parenchymal cell function. In this study, we induced hepatocellular disease by administration of a hepatotoxic drug and compared a direct method of determining the hepatic extraction of ^99mTc-membrofenin to hepatic extraction fraction derived from deconvolutional analysis. We also compared both methods of calculating the hepatic extraction of ^99mTc-membrofenin to liver histopathology. Hepatic extraction fraction derived from deconvolutional analysis correlated very well to the direct measurement technique (R=0.922, p<0.001). Both methods of determining hepatic extraction correlated well to quantitative histopathology, having the same correlation coefficient and p values. (R=-0.833, p=0.003). As the hepatic extraction ^99mTc-membrofenin decreased, the severity of the histopathologic lesions of the liver increased in a linear fashion. There was a significant correlation of the hepatic excretion T_1/2 to quantititative histopathology (R=0.949, p<0.001). The hepatic excretion T_1/2 increased as the severity of the histopathologic lesions of the liver increased. Hepatic extraction (HEF) and excretion of ^99mTc-membrofenin are good predictors of the severity of hepatocellular damage in toxic induced liver disease. This study helps validate the premise that HEF derived from deconvolutional analysis ois a good predictor of the actual first pass hepatic extraction of ^99mTc-membrofenin.     

EVALUATION OF THE FELINE PANCREAS USING COMPUTED TOMOGRAPHY AND RADIOLABELED LEUKOCYTES

This study was designed to test the feasibility and utility of computed tomography and radiolabeled granulocytes in evaluating the feline pancreas in six normal cats. Autologous granulocytes were labeled with 99mTc-hexamethylpropyleneamine oxime (HMPAO) and injected into each cat. Whole body scintigraphic images were acquired at 1, 5, 15, and 30 minutes, and 1, 2, and 4 hours following injection. The following day, each cat was anesthetized and computed tomographic images of the abdomen were acquired both pre- and post-contrast. Following CT, a surgical pancreatic biopsy was collected. Feline granulocytes were successfully labeled with 99mTc-HMPAO with a labeling efficiency of 15-42% (average of 27%). An average of 5.42 x 10(7) cells in a 2 mL volume were injected into each cat. Less than 1 minute was required to acquire 500,000 count images. Granulocytes distributed predominantly to the lung, spleen and liver in order of decreasing activity. Only background activity was identified in the region of the pancreas. The pancreas was easily identified on CT images of the abdomen. The pancreas was hypoattenuating relative to both the spleen and liver. The pancreas enhanced with the administration of contrast medium, peaking immediately, then gradually clearing over the 30-minute test period. Following contrast medium administration the pancreas remained hypoattenuating relative to the spleen. All biopsies confirmed the absence of pancreatic inflammation in the study cats and no adverse effects were recognized as a result of pancreatic biopsy. Both computed tomography and radiolabeled granulocytes appear to hold promise as imaging procedures for the detection of feline pancreatitis. We predict that these described normal parameters may be altered in the face of inflammation, facilitating detection of feline pancreatitis. Data from cases of suspect feline pancreatitis are needed to evaluate these methods for clinical utility.