Hyperammonemic Encephalopathy in a Four-Year-Old Child With Prune Belly Syndrome

RE is a 4-year-old Hispanic boy with prune belly syndrome who presented to the Emergency Room with altered mental status. At home he was lethargic and was unable to recognize his parents. There was a history of tactile fever for 1 day for which he received no medication. There was one episode of nonbilious, nonbloody vomiting. He also refused to eat. There was no history of trauma. Three days before this Emergency Room visit he was seen at the General Pediatrics Clinic for “not feeling well.” A urine dip stick test was done in light of a history of multiple urinary tract infections. The test was positive for nitrite along with 2+ occult blood and 3+ leukocyte esterase. Therapeutic doses of trimethoprim-sulfamethoxazole were substituted for his nitofurantoin prophylaxis. Urine specimens were also sent to the laboratory for urinalysis and cultures. The deficient abdominal musculature in this patient led to the requirement of intermittent catheterization, which was being done by the patient's mother once every 12 h. Past history was significant for recurrent urinary tract infections and pneumonia, which often required hospitalization. He had received prolonged antibiotic prophylaxis for recurrent urinary tract infections. In the Emergency Room the patient had a temperature of 36°C, heart rate of 166 beats/min, respiratory rate of 24/min and a blood pressure of 118/88 mm Hg. He was mentally obtunded but responsive to pain. He had very sluggish pupillary reaction to light bilaterally. He was moderately dehydrated. His neck was supple with full range of motion. The cardiopulmonary examination was normal. The abdomen was distended but this subsided upon catheterization of the urinary bladder, which yielded 1300 ml of urine. No hepatosplenomegaly was present. The capillary refill time was 3 s. Neurologic examination was significant for unresponsiveness except to pain, sustained ankle clonus, Babinski reflexes and diminished deep tendon reflexes, bilaterally. There were a few laboratory test results available in the Emergency Room before the patient's transfer to the ward. The complete blood count showed 12 900/mm3 white blood cells with a differential of 70% segmented forms, 12% band forms, 3% lymphocytes and 5% monocytes and a hemoglobin of 11.9 g/dl. The platelet count was within normal limits. Random serum glucose was noted to be 168 mg/dl. Serum electrolytes were sodium 141 meq/l, potassium 4.5 meq/l, chloride 106 meq/l and bicarbonate 17 meq/l. Serum values of aspartate aminotransferase and alanine aminotransferase were both 28 units/l, alkaline phosphatase was 156 units/l and lactate dehydrogenase was 728 units/l. Other results on the serum chemistry panel were normal. Serum ammonia was 217 μmol/l (reference range, 10 to 50 μmol/l). An arterial blood gas showed a pH of 7.49, pCOs of 24 mm Hg, pO2 of 119 mm Hg in room air. Urine pH was 8.5 on urinalysis. A urine toxicology screening test was normal. Cerebrospinal fluid analysis showed no white blood cells, one red blood cell and normal glucose and protein values. Urine and blood specimens were sent for culture before the patient's transfer to the ward. Serum amino acids and urine organic acids done at a later time were normal. What then was the reason for this patient's hyperammonemic encephalopathy? Aninda Das, M.D.; Diane Henderson, M.D. Division of General Pediatrics Children's Hospital of Los Angeles Los Angeles, CA For denouement see p. 925. DENOUEMENT—continued from p. 923 This patient had hyperammonemic encephalopathy from a Proteus mirabilis urinary tract infection (UTI). The urine culture yielded > 100 000 colonies/ml of P. mirabilis and 11 000 to 25 000 colonies/ml, respectively, of Klebsiella and Enterobacter species. This patient's hyperammonemia resulted from an increased production in the urinary bladder by the bacteria and its subsequent absorption through the vesicoureteric venous system. The association between hyperammonemic encephalopathy and prune belly syndrome patients with a P. mirabilis UTI has been documented in the past.1-4 Although hyperammonemic encephalopathy is rare the clinician should be alert to the possibility of this acutely evolving condition, particularly in patients with prune belly syndrome. The management of this condition is essentially medical. Ammonia (NH3), the toxic product of nitrogen metabolism, is converted to urea via the Krebs-Henseleit cycle and then excreted through the kidneys. The various sources of NH3 production include amino acid metabolism, bacterial deamination of unabsorbed amino acids and protein derived from the diet in the lower gastrointestinal tract, and the deamination of glutamine in the kidneys. In advanced liver disease there is a decrease in hepatic urea production leading to accumulation of NH3 in the blood.5 An excessive quantity of NH3 in blood escaping the hepatic detoxification mechanism contributes to the hyperammonemia and the resulting encephalopathy. Therefore patients with portacaval shunts are at increased risk because NH3 is carried directly from the lower gastrointestinal tract to the systemic circulation. The buffer system governing the existence of ammonia in solution1 is NH3 + H+ [right harpoon over left harpoon] NH+. The pKa of this reaction is 9.15. This means that for a pH of 7.4, approximately 1 of 100 molecules is a free base NH3 in solution in equilibrium with NH+. This increases to about 50% at a pH of 9. Therefore the proportion of free base NH3 increases in an alkaline environment. Free base NH3 is lipid-soluble in contrast to water-soluble NH4+. Therefore the former permeates cell membranes much more easily than the latter. The gradient of partial pressures of the free base across the cell membrane determines the net direction of diffusion. This again is a function of the pH and the total NH3 concentration on each side of the membrane. Once across into the bloodstream it is converted to the poorly permeable NH4+ by a process termed “diffusion trapping.”1 P. mirabilis is a urease-producing bacterium.6 The hydrolysis of the urea to ammonia by the enzyme and the resultant increase in the pH of the urine are characteristic of P. mirabilis UTI. The elevated urine pH also promotes struvite stone formation, which obstructs urinary outflow. The stones perpetuate the infection to chronicity by serving as foci of infection. Prune belly syndrome7 is a congenital condition in which there is an absence of the abdominal musculature. Massive dilations of the upper urinary tracts and the ureters along with an enlarged urinary bladder are the commonly associated urinary tract abnormalities with this syndrome. Undescended testes are also a characteristic finding in males. These patients are at increased risk for recurrent UTI. This patient was managed with intravenous hydration. His urinary bladder was continuously drained with an indwelling catheter. Based on the results of the urine culture from the General Pediatrics Clinic (growth of > 100 000 colonies/ml of both Proteus species and Pseudomonas aeruginosa), imipenem therapy was started. His mental status improved dramatically and the serum NH3 was 91 μmol/l within 12 h of admission and decreased further to 12 μmol/l on the third hospital day. A renal ultrasound examination showed bilateral hydronephrosis and hydroureters, and a voiding cystourethrogram was significant for vesicoureteric reflux on the left side. A residual function of 50% in the left kidney was seen on a dimercaptosuccinic acid scan. The urologists opted to do bilateral ureteral stent placement. The patient was discharged home after 9 days. Our patient had a hugely dilated bladder containing alkaline urine. This was conducive to the production of free base NH3 as well as its absorption through the increased surface area presented by the dilated urinary tracts. Moreover the venous drainage of the vesicoureteral system bypasses the portal system. Thus the large quantity of NH3 absorbed from the urinary tract directly entered the systemic circulation to cause the encephalopathy in this patient.