Mark Paller Minneapolis, DESCRIPTION OF THE NEPHROLOGY TRAINING PROGRAM AT EMORY UNIVERSITY SCHOOL OF MEDICINE

Transcription

1 DITORIAL COMMITI Tomas Ben, ditor Denver, CO William Dallas, Henrich TX Mark Paller Minneapolis, MN Fred Silva Oklahoma City, OK DSCRIPTION OF TH NPHROLOGY TRAINING PROGRAM AT MORY UNIVRSITY SCHOOL OF MDICIN The Nephrology Fellowship Training Program at mory is designed to prepare the candidate for a career in academic nephrology (3 yr) or clinical practice (2 yr). The first year of training involves extensive instruction in all aspects of clinical nephrology using large outpatient and inpatient services, a transplantation service (2/yr). and three diverse hospitals. Research training is based on grants from the Federal Government to eight faculty. plus an NIH. George M. O Brien Kidney Research Center grant and an NIH Training Grant, Training for clinical practice includes 18 months of patient care at three separate. contrasting institutions. The Atlanta VA Medical Center is a 524-bed referral center with both ambulatory and inpatient dialysis. consultative nephrology. and a large clinical outpatient service, Grady Memorial Hospital is an 856-bed hospital serving Atlanta and oftering ambulatory and inpatient dialysis programs, a large continuous ambulatory peritoneal dialysis program. and an active consultative nephrology service. mory University Hospital is a 4-bed tertiary-care hospital with kidney transplantation and inpatient dialysis plus consultative nephrology. Fellows develop skills in all aspects of dialysis. renal biopsy. and the medical management of renal transplant recipients. Fellows play important roles in teaching medical students and housestaft and regularly participate in clinical nephrology conferences. and journal clubs. The fellows are trained to perform and interpret ultrasound of the kidney and to place cufted. permanent vascular dialysis-access catheters and Tenckhoft peritoneal catheters percutaneously. Fellows in the research program complete 12 months of inpatient and outpatient care. They then begin a 2-yr research period designed to provide training in basic or clinical research, Fellows design and personally complete a research project while faculty preceptors assist in the design of the projects and teach techniques needed for the research project. The training program is designed so that all of the fellow s time may be focused on the research project. except for one afternoon per week in the renal clinic, This intensive period should optimize the fellow s exposure to basic research and will provide him/her with the skills necessary to conduct independent research, Unexpected ncephalopathy in Chronic Renal Failure: Hyperammonemia Complicating Acute Peritonitis1 J. Luis Pimentel, Jr., Saul W. Brusilow, and William. Mitch2 J.L. Pimentel, W. Mitch, Renal Division, mory University School of Medicine. Atlanta. GA S.w. Brusilow. Pediatrics Department, Johns Hopkins University School of Medicine, Baltimore, MD (J. Am. Soc. Nephrol. 1994; 5:lO#{243}&-173) ABSTRACT A woman with mild chronic renal insufficiency was being treated with glucocorticoids for a presumed chronic inflammatory disease. She developed peritonitis arising from a pelvic abscess, which was drained without complications. Unexpectedly, she became Received July 3, Accepted March Correspondence to Dr. W. MItch, Renal Division, mory UniversIty 5chool of Medicine, 1364 Clifton Road, N, Atlanta, GA &6673/ $3./ Journal of the American Society of Nephrology Copyright 1994 by the American society of Nephrology 166 obtunded, and eventually, the neurologic dysfunction was linked to hyperammonemia in spite of normal liver function tests. Hyperammonemia was only transiently controlled in spite of protein restriction, repeated hemodialysis, and the use of biochemical means to reduce ammonia. A recurrent pelvic abscess was drained, and hyperammonemia disappeaned. A review of ammonia and nitrogen metabolism indicates that bypassing the liver with shunting of ammonia into the systemic circulation should be added to the causes of symptomatic hyperammonemia. Treatment requires the elimination of the bacterio. Key Words: Hyperammonemia, chronic renal failure, acute peritonitis, hemodialysis I n adults, the most common cause of hyperammonemia is liver disease ( 1 ). Other less common causes include inherited abnormalities in urea cycle enzymes (2), in dibasic amino acid metabolism (3,4), Volume 5 Number

2 Pimentel et al and In organic acidemias (5), as well as acquired disorders such as portal-systemic shunts ( 1,6), Reye s syndrome (7), high-dose cytotoxic chemotherapy (8), valproate toxicity (9), and urinary tract infections by urea-splitting bacteria ( 1). An increase in blood ammonla has also been described in chronic renal failure (CRF) ( 1 1 ) and in patients undergoing hemodialysis ( 1 2). Unfortunately, the cause of hyperammonemia is not always easily determined and therapeutic strategies are comphex. We describe the clinical course ofa patient with CRF but no evidence of impaired hepatic function, who developed hyperammonemia and encephabopathy presumably from acute peritonitis associated with perforation ofa sigmold diverticulum. Medical therapy was unsuccessful, and hemodialysis had only a transient effect, but the ammonia level returned to normal values after the surgical drainage of a pelvic abscess. Patients with unexplained hyperammonemia should prompt a search for an occult intra-abdominal infection. CAS RPORT A 47-yr-old woman was admitted to mory University Hospital because of fever and abdominal pain. In 1969, after a thyroldectomy for Graves disease, she developed hypoparathyroidism and was treated with vitamin D plus calcium supplements. Over the ensuing years, she developed nephrocalcinosis and CRY. Five weeks before admission, she was given prednisone for necrotlzing pannicuhitis; evaluations for systemic diseases were negative. Two weeks before admission, she developed abdominal pain and distention, frequent bowel movements, and a low-grade fever. She had no history of alcohol abuse or liver disease. At the time of admission, she was febrile and had a distended abdomen with bilateral lower quadrant and suprapubic pain to palpation without rebound tenderness. Pelvic and rectal examinations were normal. Stool was negative for occult blood. A neurobogic examination was normal. Laboratory evaluation showed a hematocrit of 34%; a white blood cell count of 16,6/mm3; platelet count and coagulation times were normal. Liver tests (SGOT, SGPT, alkaline phosphatase, and bihirubin) were within normal limits. Blood urea nitrogen was 64 mg/dl, and serum creatinine 3.7 mg/dl. Urinalysis showed only trace protein; there was no pyuria. Abdominal computed tomography displayed a heft adnexal mass and thickened small boweh hoops. An exploratory laparotomy revealed generalized peritonitis (a total of 6 ml of pus was drained), and a site of purulent, fecal-appearing material was seen between the sigmoid colon and the left posterior leaf of the broad ligament. A diverting, transverse loop cobstomy was created. Blood, urine, and peritoneal fluid cultures showed no growth, but the hatter contained vegetable matter. These findings were felt to be most consistent with a ruptured sigmoid diverticula. On the sixth postoperative day, she became confused and lethargic and exhibited generalized hyperreflexia, bilateral Babinski signs. rigidity, and decorticate posturing. The hemogram and blood chemistries, including serum levels of hepatobihiary enzymes, were unchanged. Arterial blood gases while breathing room air were: ph. 7.52; Po2, 1 1 mm Hg; and Pco2, 24 mm Hg. A brain computed tomographic scan and spinal tap were normal except for a cerebrospinal fluid opening pressure of 1 85 cm H2. A venous blood ammonia level was 478.tmol/L (normal, 1 1 to 35). To evaluate whether the hyperammonemia could be caused by a defect in the urea cycle (2,4), plasma amino acids were obtained. Citrulhine and arginine values were normal, as were ahanine and glycine values (Table 1 ). Orotic acid was not detected in the urine (2) (Tabhe 2). The urea nitrogen appearance rate was estimated with a volume of distribution of urea equal to 6% of body weight, daily determinations of blood urea nitrogen, and the average rate of urea nitrogen excretion (13). Treatment consisted of restricting nitrogen intake and Infusing 1% dextrose; lactulose and neomycin were given orally. When this was unsuccessful, hemodialysis was instituted with a 1.8-m2 surface-area polysulfone membrane dialyzer (initial blood flow rate of 3 ml/min, then 4 ml/min during the second hour of treatment, and a dialysate flow rate of 8 ml/min). After dialysis, the arterial blood ph and TABL 1. Plasma amino acid palfern before the iv infusion of anginine, benzoate, and phenylacetatea.. Amino Acid Plasma Amino Acid Values (,mol/l) Normal (mol/l) Taurine 41 L Aspartic Acid 19 H 2-13 Threonine 86 N Serine 5L Glutamic Acid 31 N Glutamine 711 H Proline 19L Glycine 21 7 L Alanine 75L Citrulline 8 N 6-25 Valine 121 L Cystine 81 H 9-43 Methionine 22N Isoleucine 48 N 4-58 Leucine 68L Tyrosine 5 N Phenylalanine 81 H 5-69 Ornithine 75L Lysine 11OL Histidine 51 L Arginine SON a Plasma amino acid values (second column) varied widely, high (H), normal (N), or low (L), but were not characteristic of a urea cycle enzyme deficiency (see Text). Journal of the American Society of Nephrology 167

3 Hyperammonemia in Chronic Renal Failure TABL 2. Urinary excretion of organic acidsa xcretion Normal Range Organic Acid (mg/mg of (mg/mg of Creatinine) Creatinine) A modiaiysis p-hydroxyphenylacetic p-hydroxyphenyllactic p-hydroxyphenylpyruvic Glyoxylic Orotic Acid none detected.-.1 a These values are not consistent with an organic acidemia; in par- S C. 3 2 ticular, there was no evidence of excess orotate production. 1 ammonia levels decreased sharply, but 3 h later, ammonia returned to the pretreatment level (443 moh/l) (Figure la). The calculated dialyzer clearance of ammonium at the start of dialysis was 1 42 ml/min (clearance = blood flow (arterial - venous concentration I arterial concentration). Because dialysis only transiently corrected the hyperammonemia, we used another treatment on the basis of increasing the activity of the amino acid acylation pathway to reduce ammonia accumulation (14-16). Arginine hydrochloride (5 g). sodium benzoate ( 1 g), and sodium phenylacetate ( 1 g) were infused iv over 2 h, followed by 1 g of sodium phenylacetate over 1 8 h. Nonetheless, the blood ammonia continued to rise (Figure ib), and a second hemodialysis was performed. Again, the blood ammonia dropped sharply but subsequently rose, and the patient developed rigidity, decorticate posturing. and seizures, requiring endotracheal intubation and respiratory support. Cystoscopy to exclude a bladder infection or fistula was normal ( 1), but paracentesis unveiled peritonitis and an exploratory laparotomy disclosed a puruhent cavity in the retrouterine cul-desac, consistent with an abscess. Immediately after surgery, the blood ammonium began to fall and was within the normal range after 36 h (Figure ic). Unfortunately, the patient remained comatose and an electroencephalogram was consistent with severe, generalized cerebral injury. After a prolonged course in the intensive care unit, she died. A limited autopsy revealed marantic endocarditis of the mitral valve, bilateral pneumonia, splenomegaly with organizing fibropuruhent perisplenitis, colonic diverticubosis, dense adhesions throughout the abdomen, papillary necrosis, nephrocalcinosis, and an organizing thrombus in the pelvis. There was no histologic evidence of liver disease. B. C S C hours iv infusion P hours 4-. $ u r g. ry AMMONIA Ammonia MTABOLISM Production Ammonia is formed mainly in the liver, but skeletal muscle, brain, and kidney also produce it (renal yenous ammonia is always higher than the arterial value). Amino acids released from the degradation of dietary or cellular proteins are deaminated by a days Figure 1. The effect of therapeutic interventions on blood ammonia level. The normal venous ammonia level is 11 to 35 M. (A) The transient change in blood ammonia after hemodialysis. (B) The infusion of a mixture of arginine, phenylacetate, and benzoate changed the blood level of ammonia minimally. (C) The surgical drainage of a pelvic abscess led to a sustained decrease in the blood ammonia concentration. 168 Volume 5 Number

4 Pimentel et al transaminatlon reaction to yield L-glutamate, which undergoes oxidative deamination in a reaction catalyzed by L-glutamate dehydrogenase (Figure 2). In addition to the ammonia arising from protein catabolism, considerable ammonia is generated from the degradation of peptides and urea In the colon by deaminases and ureases present in bacteria (17). Ammonia is also formed by the nonoxidative deamination of amino acids and adenosine plus the oxidation of physiologic amines, such as epinephrine and dopamine. Ammonia (NH3) diffuses freely across lipid membranes, but at physiologic ph values, 98.5% of ammonia Is protonated and ammonium (NH4) does not readily penetrate membrane barriers. Alkalemia enhances its diffusion into the brain (and other tissues) by maintaining a higher nonlonized fraction of ammonia. When alkalosis is complicated by hypokalemia, renal ammonlagenesis is stimulated, providing a mechanism by which hyperammonemla coexisting with alkalemia may precipitate encephalopathy in patients with liver disease. Removal of Ammonia Ammonia in the blood is removed by several organs, including the liver, where It is used In the synthesis of urea (Figure 3). TIssues such as muscle and brain use alternative pathways to detoxify ammonia. First, ammonia is used by glutamine synthetase to form L- glutamine from L-glutamate. Second, ammonia can be used in the reversal of the glutamate dehydrogenase a-amino acid ct-keto acid a-ketoglutarate NH3 URA URA L-Glutamate CYCL CO2 DANAON Figure 2. Nitrogen flow in amino acid catabolism. reaction to form L-glutamate from a-ketoglutarate. Under normal resting conditions, there Is a small uptake of ammonia by nonhepatic tissues, but in hyperammonemia, skeletal muscle ammonia extraction may exceed 5% of the arterial level ( 18). Finally. the kidneys extract glutamine from plasma to form L-glutamate and ammonia and then a-ketoglutarate and ammonia by the actions of gbutaminase and glutamate dehydrogenase, respectively. The adult human kidneys excrete about 3 to 5 mmol of ammonia each day. The Urea Cycle The major source of nitrogen for urea synthesis in the urea cycle Is glutamine. In the mitochondria of hepatocytes, ammonium from glutamine is used by carbamylphosphate synthetase to form carbamylphosphate from ATP and bicarbonate. Subsequently, an additional nitrogen is removed from aspartate to synthesize urea. The urea cycle requires 4 mol of ATP per mole of urea synthesized, so the reactions are virtually irreversible. The liver has an enormous capacity, removing 8% of portal venous ammonia in a single passage (see reference 2). Ammonia Measurement Methods for measuring ammonia include diffusion into an ammonlum electrode and enzymatic reactions coupled to NADH with or without separation techniques, including ion-exchange resins. The enzymatic method uses glutamate dehydrogenase to convert ammonia plus a-ketoglutarate and NADH to L-glutamate plus NAD. The increase in NAD Is measured by ultraviolet spectrophotometry at 34 nm. It is important to maintain the blood sample (collected into lithium heparmn or DTA so that plasma can be obtained) on ice, separate the plasma Immediately, and run it within 1 h to suppress the enzymatic formation of ammonia from amides such as glutamine; the ammonia content in blood rises at the rate of.3 p.g/ml per minute If the sample Is kept at room temperature. The difference between arterial and venous NH levels is not significant, so venous blood can be used. Blood ammonia levels only indirectly reflect tissue levels because ammonia rapidly enters cells and is metabolized to urea, glutamine, etc. The normal range of plasma ammonia concentration is 12 to 55 mol/l. TH HYPRAMMONMIC SYNDROMS The causes of hyperammonemia are summarized in Table 3. Inherited deficiencies of each of the five enzymes of the urea cycle and of N-acetyl glutamate synthetase have been described. xcept for arginase and heterozygous ornithine transcarbamylase (HOTC) deficiency. these enzymopathles are clinically manifested in the neonatal period. Arginase deficiency becomes clinically evident as a progressive neurologic impairment later in childhood. HOTC deficiency, an X-llnked dominant disorder, generally causes episodes of vomiting. lethargy, and coma early in child- Journal of the American Society of Nephrology 169

5 Hyperammonemia in Chronic Renal Failure ATP HCO3 ADP N o Nacetyl glutamate Acetyh Co A glutamate CARBAMYL PHOSPHAT ORNITHIN a -keto e-aminocaproic acid lysine URA * N CITRULLIN ARGININ C=O * ASPARTAT N ARGINOSUCCINAT Oxaloace o Carbamyl phosphate synthetase (CPS) O Omithine transcarbamylase (OTC), FUMARAT Arginosuccinate synthetase (AS) \.,. Arginosuccmase Arginase (AL) succinate N-acetyl glutamate synthetase (NAGS) methylmalonylcoa Propionyl CoA carboxylase Methylmalonyl CoA mutase k&o L-lysine dehydrogenase propionyl CoA 11 a -ketoglutarate Figure 3. The urea cycle. Numbered steps 1 to 9 are catalyzed by the enzymes shown in the left lower portion of the diagram; reactions 1 to 5 constitute the urea cycle. Steps 6 to 9 represent other sites of known primary enzymes defects. Carbamylphosphate synthetase and ornithine transcarbamylase are located in the mitochondrial matrix; the other urea cycle reactions occur in the cytosol. The urea nitrogens are labeled to show their origin. Open circles denote an inhibitory effect. CoA, coenzyme A. hood in affected females, but the disorder may not be recognized until adulthood ( 19). Acquired acute and chronic hyperammonemia primarily occur in patients with severely impaired hepatic function. Because the normal liver has a tremendous capacity to remove ammonia, it is rare for patients with normal hepatic function to develop hyperammonemia, even when endogenous ammonia generation is high or ammonium salts are administered (2). Thus, when hyperammonemia develops In patients without evidence of liver disease, as In this patient. it is likely that the liver is bypassed, as in portacaval shunts. Differential Diagnosis of Adult-Onset Hyperammonemia In the absence of hepatocellular damage, hyperammonemia has been reported to occur in patients with CRY, In rare subjects given valproate, in urinary tract infection with urea-splitting bacteria, and in HOTC deficiency in women. Watson et al. also uncovered a syndrome characterized by transient hyperammonemia in three adult patients treated for acute leukemia (2 1). Of these possible explanations for hyperammonemia in this case, only the HOTC deficiency could not be excluded, because we did not measure hepatic ornithine transcarbamylase activity nor were cytogenetic studies performed. Nevertheless, certain clinical features make this diagnosis unlikely. The patient had no family or personal history of hyperammonemia (untreated patients with HOTC deficiency generally experience episodes ofvomiting, lethargy, and coma In the first year of life or later in childhood), urinary orotic acid was not increased, and the pattern of plasma amino acids (Tables 3 and 4) was not charac- 17 Volume 5. Number

6 Pimentel et al TABL 3. Causes of hyperammonemia Inherited (Ref. No.) Urea cycle enzyme defects (2) Dibasic aminoacidopathies (3,4) Organic acidemias (5) Congenital lactic acidosis (5) Carnitine deficiency (28) Acquired (Ref. No.) Intrinsic liver disease (1) Idiopathic transient hyperammonemia (21) Portacaval shunts (1,6) Reye s syndrome (7) Drug induced Valproate therapy (9) Hypoglycin A ingestion (Jamaican vomiting sickness) (1) Salicylate intoxication (1) Cytotoxic chemotherapy (8) Asparaginase therapy in leukemia (1) Urinary tract infection (1) Parenteral nutrition (1) Sorbent regenerative hemodialysis (ammonia breakthrough ) (12) Systemic herpes virus infection (1) Perinatal asphyxia (1) teristic because there was no decrease in citrulline and arginine nor was there an Increase in glutamine and alanine (2,4,22). In short, a urea cycle enzyme deficiency causing the hyperammonemia was unlikely. Treatment We tried several therapeutic strategies. Initially, treatment with oral hactubose and neomycin was unsuccessful, and even though hemodialysis rapidly reduced the blood ammonia level, the benefit was only transient (Figure 1 ). We also use alternative pathways of waste nitrogen excretion because they have been useful in treating inborn errors of urea synthesis in children and they effectively change nitrogen metabolism in patients with CRY ( 14, 15). As Illustrated In Figure 4. the synthesis and excretion of arginosuccinate and citrulline may be enhanced by the administration of arginine. Because arginosuccinate contains both nitrogen atoms that will be used In the synthesis of urea and its renal clearance is equal to the GFR, the excretion of this amino acid is equivalent to that of urea ( 14). In addition, stimulating amino acid acylation Is possible by administering benzoate and phenylacetate, substrates for this pathway. The acylation of benzoate by an equlmolar amount of glycine forms hippurate, whereas phenylacetate is conjugated to glutamine to form phenylacetylgbutamine. Hippurate and phenyhacetylglutamine are cleared by glomerular filtration and by the organic anion secretory pathway of the proximal tubule. Unless removed by dialysis, hippurate and phenybacetylgbutamine will accumulate in patients with severe CRY, but the administration of these compounds will still change nitrogen metabolism. For example, benzoate administration to azotemic patients was shown to reduce the rate of urea production rapidly because of nitrogen shunting from ammonia and urea formation into ghycine production ( 15). When renal function is normal, this strategy can remove as much as five times more nitrogen than does hemodialysis. The amino acid values we found after administering arginine, benzoate, and phenylacetate (Table 5) were lower than those measured in children receiving iv benzoate and phenylacetate and might explain why this therapy failed to decrease blood ammonium levels sharply. Perhaps high blood ammonium levels overwhelmed the capac- Ity of these pathways ( 1 6,2 1 ) because ammonia production must have been very high because the blood ammonium rapidly returned to predlalysis levels soon after Comments treatments. In considering pathophysiohogic mechanisms that might have been involved in the development of hyperammonemia in this patient. there are several questions: ( 1 ) How could severe hyperammonemia occur in an adult with normal liver function and intact ureagenesis? (2) Was ammonia production Increased? (3) Did hyperammonemia cause the neurohogic findings? and (4) How should patients with CRY and hyperammonemia be treated? The clinical course of patients with severe, transient, idiopathic, hyperammonemic coma has been described (8,2 1). ven though the patients described in those reports did not have clinical evidence of serious liver dysfunction, they seemed to exhibit an Impairment In urea synthesis In association with diseases causing protein catabolism. When hyperammonemla was first detected in our patient. she had a total nitrogen excretion of 7.2 g in 24 h (76% was urea nitrogen). This exceeded the amount of nitrogen she had been ingesting as protein (estimated from food tables) and suggested a catabolic condition. Later, her urea appearance rate was calculated to be 4.5 g of N/day ( 13). This rate was slightly less than nitrogen intake (5.5 g) and below the maximal rate of urea synthesis that normal individuals ( 1 3 to 1 34 g of N/day) or patients with liver cirrhosis can achieve (78 to 11 g of N/day) (23). On the basis of these calculations, we concluded that the patient was able to synthesize urea and should have had much more capacity to remove ammonia by converting it to urea. This conclusion led us to search for shunting of ammonia to bypass the liver. The patient had peritonitis from ruptured diverticula but did not develop clinical evidence of hyperammonemia until the Infection was localized as a pelvic abscess. The clinical course, therefore, might be explaned as follows: during the early, diffuse intraabdominal infection, the ammonia produced was absorbed by the peritoneum into the portal vein and was Journal of the American Society of Nephrology 171

7 Hyperammonemia in Chronic Renal Failure TABL 4. Urea cycle defects nzyme Defect Blood NH3 Urinary Orotate Plasma amino acids Cit Arg GIn Ala Asa CPS 111 I I - NAGS II I I - OTC III III I I - AS III I III I I - AL III I I I I I III Arginase I I - I I I a nzyme abbreviations are as in Figure 3. Cit. citrulline; Arg, arginine; GIn, glutamine; Ala. alanine; Asa, argininosuccinate. [Benzoate] N4+ Glutamate Glutamine [Phenylacetate Glycine \ a-ketog1utarate Hippurate Hco3. 4 Phenylacetylgiutamine Carbamyl phosphate Ornithine Citrulline t,-#{248} Aspartate URA Urea Cycle AS Argniase L :i,,osuccinate Fumarate Figure 4. Pathways of waste nitrogen synthesis and excretion. Regardless of the defect in urea synthesis, the formation and excretion of phenylacetylglutamine and hippurate are readily accomplished bythe acylation of glutamine and glycine by the administration of phenylacetate and benzoate, respectively. Arginine promotes the synthesis of citrulline In AS deficiency and of arginosuccinate in AL deficiency. Modified from Reference 1. Abbreviations are as in Figure 3. detoxified by the liver. Later, hyperammonemia developed when the Infection became localized and ammonia was delivered into the systemic circulation via the internal iliac vein into the Inferior vena cava (25). Thus, the ammonia produced would bypass the liver and cause encephalopathy ( 1). Because of antibiotic therapy, cultures of peritoneal and abscess fluids were negative, but urease-producing bacteria must have been present, because the blood ammonia level rapidly decreased after the second surgery, in which the pelvic abscess was successfully drained. Two additional factors could have played a robe in the severity of hyperammonemia: there was a boss of skeletal muscle associated with glucocorticoid therapy and infection and CRY was present. After the liver, skeletal muscle is quantitatively the most Important organ capable of removing ammonia (24), and a reduced muscle mass woubd limit the capacity to detox- 172 Volume 5. Number

8 Pimentel et al TABL 5. Plasma levels of specific amino acids and metabolites before and after an iv dose of arginine, benzoate, and phenylacetate Before Infusion (mmol/l) nd of 2 h After Infusion Infusion Hippurate Phenylacetylglutamine Phenylacetate Not done Benzoate Not done Ify ammonia. In addition, there is evidence that some CRY patients have high arterial blood ammonia levels (26) and increased ammonia extraction by the brain (27). On the other hand, the rapid normalization of blood ammonia after the successful drainage of the abscess (Figure ic) makes it unlikely that CRY alone, or in combination with loss of muscle mass, caused hyperammonemia. The effect of changes in brain ammonia metabolism that might have predisposed this patient to develop Irreversible brain damage could not be addressed. Regarding the third question about the cause of her coma, the rapid neurobogic deterioration together with raised blood ammonium levels points to a cause and effect relationship. We could find no other explanation for the rapid development of coma and Irreversible brain damage. In summary, hyperammonemia can occur even in patients with no evidence of hepatic disease If an Infection increases ammonia production and ammonia is shunted Into the systemic circulation. ffective therapy requires the drainage or eradication of the infection, and other therapies, including dialysis, may prove Ineffective. ACKNOWLDGMNTS We are indebted to Dr. Michael Gravanis. Department of Pathology. mory University School of Medicine. for his careful review of the liver histology. Supported In part by Grant HD (SW. Brusilow) from the NIH. RFRNCS 1. Batshaw ML: Hyperammonemia. Curr Probl Pediatr 1984; 14: Walser M: Urea cycle disorders and other hereditary hyperammonemic syndromes. In: Stanbury JB. Wyngaarden JB, Frederickson DS, et al., ds. The Metabolic Basis of Inherited Disease. 5th d. New York, McGraw- Hill; 1983: Simell, Perheentupa J, Rapola J, Visokorpi JK, skeun L: Lysinuric protein intolerance. Am J Med 1975; 59: Valbe D, Simeib : The hyperornithinemias. In: Stanbury JB, Wyngaarden JB, Frederickson DS. et at., ds. The Metabolic Basis of Inherited Disease. 5th d. New York, McGraw-Hill; 1983: Flannery DB, Hsia Y, Wolf B: Current states of the hyperammonemic syndromes. Hepatohogy 1982;2: Mendenhall CL, Rouster S. Marshall L, Wesner R: A new therapy for portah system encephahopathy. Am J Gastroenterol 1986;81 : Trauner DA: Reyes syndrome. Curr Probl Pediatr 1982: 12: Mitchell RB, Wagner J, Karp J. et al.: Syndrome of idiopathic hyperammonemia after high dose chemotherapy: Review of nine cases. Am J Med l988;85: Couhter DL, Allen RJ: Hyperammonemia with vaiproic acid therapy. J Pediatr 1981;99:317-3l9. 1. Drayna CJ, Titcomb CP, Varma RR, Soerceh KH: Hyperammonemic encephalopathy caused by Infection in a neurogenic bladder. N ngl J Med : Walser M: Determinants of ureagenesis, with particular reference to renal failure. Kidney Int 198; h7: Canzanelbo VJ, Rasmussen RT, McGobdrick MD: Hyperammonemic encephalopathy during hemodiabysis. Ann Intern Med 1983;99: Maroni BJ, Steinman TI, Mitch W: A method for estimating nitrogen intake of patients with chronic renal failure. Kidney mt 1 985;27: Batshaw ML, Thomas GH, Brusilow SW: New approaches to the diagnosis and treatment of inborn errors of urea synthesis. Pediatrics 198 1;68: Mitch W, Brusibow SW: Benzoate-lnduced changes in glycine and urea metabolism in patients with chronic renal failure. J Pharmacoh xp Ther 1982;222: Brusilow SW, Danney M, Waber U, et at.: Treatment of episodic hyperammoneznla in children with inborn errors of urea synthesis. N ngi J Med 1984;3 1: van Berlo CLH, van Leeuwen PAM, Soeters PB: Porcine intestinal ammonia liberation. Influence of food intake. hactulose and neomycin treatment. J Hepatol 1988;7: Bessman SP, Bradley J: Uptake of ammonia by muscle: Its implications in ammoniagenic coma. N ngl J Med 1955;253: Short M, Conn HO, Snodgrass PJ, Campbell AGM, Rosenberg L: vidence for X-hinked dominance inheritance of ornithine transcarbamyhase deficiency. N ngi J Med 1973;288: Conn HO, Lieberthab MM: The Hepatic Coma Syndromes and Lactulose. Baltimore, Williams & Wilkins, 1979: Watson AJ, Karp J. Walker WG, Chambers T, Risch VR, Brusibow SW: Transient idiopathic hyperammonemia in adults. Lancet 1985;7: Maestri N, McGowen KD, Brusiow SW: Plasma glutamine concentration: A guide in the management of urea cycle disorders. J Pediatr 1992; 12 1: Reopens B, Henderson JM, Fulenwider JT: A tracer method for measurement rate of urea synthesis in normab and cirrhotic subjects. Gastroenterology 198:78: Ruderman NB, Berger M: The formation of glutamine and alanine in skeletal muscle. J Biol Chem 1974;249: Williams PL, Warwick R, Dyson M, Bannister LH: Veins of the abdomen and pelvis. In: Gray s Anatomy. 37th d. New York: Churchill Livingstone; 1989: Tizianeblo A, Deferrari G, Garibotto G, Robaudo C: Amino acid metabolism and the liver in renal failure. Am J Chin Nutr 198;33: Bianchi PG, Malbo AT, Taghiabue M, Terno G: Matabolismo cerebrale deicarboidrate e dellammonio nelhinsufficienza renale cronica. Minerva Nefrol 1969; 16: Chapoy PR, Angelini C, Brown WJ, Stiff J, Shug AL, Cederbaum SD: Systemic carnitine deficiency-a treatable inherited lipid-storage disease presenting as Reye s syndrome. N ngl J Med 198;33: Journal of the American Society of Nephrology 173