Hepatobiliary scintigraphy and the string test in the evaluation of neonatal cholestasis.

We evaluated [99mTc]diisopropylphenyl-carbamoylmethylimidodiacetic acid ([99mTc]DISIDA) cholescintigraphy with measurement of duodenal fluid radioactivity collected by the string test in patients with neonatal cholestasis. Twenty-six infants with prolonged jaundice and acholic stools were studied prospectively. Twelve patients had neonatal hepatitis, 12 biliary atresia, and one each Alagille syndrome and alpha 1-antitrypsin deficiency liver disease. All infants except the biliary atresia patients and four of the neonatal hepatitis patients revealed bowel activity on scan 6 h after tracer administration. At 24 h, three of these latter patients with neonatal hepatitis and two of the patients with biliary atresia revealed bowel activity. String radioactive counts for neonatal hepatitis ranged from 99,574 to 967,205 cpm (374,504 +/- 232,210 cpm; mean +/- SD) and for biliary atresia from 8,342 to 370,346 cpm (117,149 +/- 98,698 cpm; mean +/- SD). While neither test alone was capable of correctly differentiating among all patients, those patients with biliary atresia had either a negative hepatobiliary scan at 24 h or string radioactive count below 197,007 cpm. Disparity between the hepatobiliary scan and the string radioactive counts mandates further diagnostic investigation. These data suggest that simultaneous administration of the string test with hepatobiliary scintigraphy is advantageous in the evaluation of infants with cholestatic jaundice.