Acute haemorrhagic pancreatitis in HIV-positive patients.

Abstract

Acute pancreatitis in HIV-positive patients is less common than alcohol-related pancreatitis. Haemorrhagic pancreatitis is very rare. A 23-year-old HIV-positive woman presented with vague upper abdominal pain of 3 days duration pain radiating to the back and vomiting. There was no history of trauma alcohol use recent viral illness or drug treatment. She had a tender epigastrium on deep palpation but no rebound mass or ascites. A chest radiograph was normal and an abdominal radiograph showed dilated loops of small bowel with air in the rectum. Laboratory results were as follows: full blood count: white cell count 10.7×109/l (neutrophils 85.5%) haemoglobin concentration 12.9 g/dl platelet count 146×109/l; urea and electrolytes: sodium 130 mmol/l potassium 3.0 mmol/l chloride 110 mmol/l urea 13.9 mmol/l creatinine 103 µmol/l; liver function tests: all normal; serum amylase 391 U/l. She was admitted for observation and treated symptomatically. Her condition deteriorated with generalised peritonitis a very high base excess (-12.5 mmol/l) on arterial blood gas measurement and a serum amylase level of 341 U/l. An exploratory laparotomy showed free haemorrhagic peritoneal fluid (Fig. 1) (amylase level 3 281 U/l) retroperitoneal blood tracking from the transverse colon to the caecum (Figs 1 and 2) and a grossly inflamed pancreas (Fig. 3). A baseline computed tomography (CT) scan of the abdomen was done on postoperative day 1 (Fig. 4). The CD4 count on postoperative day 3 was 34 cells/µl. The patient improved steadily and was discharged via an antiretroviral (ARV) clinic for follow-up with no sequelae of acute pancreatitis. A follow-up CT scan of the abdomen on day 20 was normal. What triggered the attack remains obscure. Discussion We found only one citation on acute haemorrhagic pancreatitis in HIV-positive patients1 with 1 of the 2 patients reported to have survived. CT scan facilities would help in the preoperative diagnosis. If the serum and urine amylase remain normal diagnosis is difficult. We did not have the facility to measure the serum lipase level which might have been helpful. The APACHE II score has been suggested as a more reliable means of monitoring these patients in the ICU setting. (full-text)