European Journal of Clinical Nutrition (2015) 69, 279–281 © 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15 www.nature.com/ejcn
CLINICAL CASE REPORT
Refeeding encephalopathy in a patient with severe hypophosphataemia and hyperammonaemia S Becker, G Dam and CL Hvas The refeeding syndrome is a potentially fatal condition that affects multiple organ systems. It is the consequence of fluid and electrolyte shifts that may occur in a malnourished patient following the introduction of nutrition therapy. The most prominent characteristic is hypophosphataemia. Although hyperammonaemia is usually seen in decompensated liver cirrhosis or acute liver failure, it may occur in other settings. We report a clinical case of prolonged and severe encephalopathy accompanied by hypophosphataemia and hyperammonaemia in a 59-year-old woman with no preexisting liver disease, urea cycle defects or portosystemic shunting. We suggest that these biochemical abnormalities were caused by uncontrolled refeeding and that the clinical picture was consistent with refeeding encephalopathy. European Journal of Clinical Nutrition (2015) 69, 279–281; doi:10.1038/ejcn.2014.244; published online 12 November 2014
INTRODUCTION The refeeding syndrome (RFS) is a condition of severe fluid and electrolyte shifts that may occur when nutrition is reintroduced to a malnourished patient.1 RFS is characterised by hypophosphataemia, hypokalaemia, hypomagnesaemia, fluid imbalance and frequently thiamine deficiency. RFS may be caused by oral, enteral or parenteral nutrition.1 Hyperammonaemia is most commonly seen in decompensated liver cirrhosis but may occur in other conditions such as acute liver failure, portosystemic shunting or urea cycle defects.2 Here, we describe a case of RFS presenting with encephalopathy associated with both hypophosphataemia and hyperammonaemia in a patient without liver disease. Case report A 59-year-old woman with a 30-year history of perianal Crohn’s disease was admitted to the hospital. Two days earlier, she had been discharged from a surgical department after revision of sacral ulcers. Because of poor nutritional status and a low food intake, oral thiamine supplementation and nasogastric tube feeding had been established at discharge with an initial dose of 500 ml (625 kcal) per day. On admission, the patient was deeply comatose with no response to pain stimuli. She had myoclonus, spontaneous pronations of all extremities and an extensive plantar response. Physical examination revealed two spider angiomas but no palmar erythema or caput medusae. She was malnourished with a body weight of 43 kg and a body mass index of 16 kg/m2. Before admission, she had a daily alcohol consumption of 2–3 glasses of wine. Her daily medications included morphine, acetaminophene, magnesium, potassium and B vitamins. Blood tests showed elevated C-reactive protein 80 mg/l (o8 mg/l), leukocytosis 18.1 × 109/l (3.5–10 × 109/l), hypoalbuminaemia 13 g/l (36–45 g/l), normal liver function tests, sodium, potassium and magnesium levels, severe hypophosphataemia o 0.17 mmol/l (0.76–.41 mmol/l) and markedly increased arterial ammonia of
252 µmol/l ( o50 µmol/l). Urea excretion was 89–117 mmol/day (300–600 mmol/day) and the glomerular filtration rate was normal. Arterial blood gas revealed an alkalosis with a pH of 7.54 (7.37–7.45). Two cerebral computed tomography scans and one magnetic resonance imaging (MRI) revealed no pathology, particularly no pontine myelinolysis. Spinal fluid had normal cell counts, cerebrospinal fluid/serum glucose ratio and marginally elevated protein at 0.94 g/l (0.15–0.5 g/l). Neuroinfection was disproved by PCR, microscopy and culture. Electroencephalography was consistent with metabolic encephalopathy and ruled out epileptic state. An abdominal computed tomography showed no signs of cirrhosis, portal hypertension or portosystemic shunting. Over 10 days, the patient’s mental status did not improve despite intravenous antibiotics, phosphate 20–80 mmol per day and thiamine (Figure 1). Following normalisation of plasma phosphate levels, enteral tube feeding including lactulose was re-established. Haemodialysis was carried out to lower ammonia. On the 12th day of hospitalisation, the patient’s cerebral state slowly improved. During recovery, she complained of muscle pain and disorientation. She was discharged on day 31 on a stable tube feeding regimen. Although awake and oriented, the patient experienced cognitive deficits and muscle weakness. DISCUSSION Hypophosphataemia is a predominant feature of RFS. RFS is potentially fatal and can affect any organ system.1 Phosphate is essential for most cellular processes and is particularly important during the anabolic phase.1 In this patient, multiple risk factors predisposed to RFS––that is, alcohol consumption, inflammation, weight loss, low body mass index and low food intake before realimentation. The initial enteral feeding dose was low (15 kcal/kg/day and 0.7 g protein/kg/day) but sufficient to induce a severe and prolonged hypophosphataemia. Guidelines recommend that patients at a high risk of RFS should be started at 10 kcal/kg/day or even lower in extreme cases.1 Profound hypophosphataemia
Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus C, Denmark. Correspondence: Dr CL Hvas, Department of Hepatology and Gastroenterology, Aarhus University Hospital, Nørrebrogade 44, DK-8000, Aarhus C, Denmark. E-mail: [email protected] Received 12 July 2014; revised 22 September 2014; accepted 28 September 2014; published online 12 November 2014
Refeeding encephalopathy S Becker et al
280 may explain the coma, myoclonus, respiratory insufficiency, neurological deficits and muscle pain. Encephalopathy during refeeding may be caused by hypophosphataemia alone,3,4 but other factors such as thiamine deficiency
Figure 1. Plasma levels of ammonia (upper level) and phosphate (lower level) during hospitalisation for refeeding encephalopathy.
or centropontine myelinolysis5 should be considered. In this case, thiamine was provided orally at initiation of enteral tube feeding, and MRI excluded centropontine myelinolysis. Hyperammonaemia is usually seen in decompensated liver cirrhosis but may occur in acute liver failure, portosystemic shunting and urea cycle defects.2 Our patient had no pre-existing liver disease. No previous episodes of coma indicated primary urea cycle defects, and she received no medications that could interfere with the urea cycle—for example, valproic acid.2 In the absence of predisposing factors, we suspected a link between hyperammonaemia and hypophosphataemia. Normally, ammonia is detoxified through glutamine synthesis in liver, muscle, kidney and brain tissue or via the urea cycle in the liver. Both are ATP-demanding processes (Figure 2). In order not to underestimate plasma ammonia, arterial blood samples could therefore be preferred.6 Hyperammonaemic encephalopathy has been described in protein-deprived children7 and in adults hyperalimentated with essential amino acids.8 A suggested mechanism was limited substrate for the urea cycle.8 In protein-deprived rats,8,9 ammonia increased and urea excretion decreased during starvation, and the authors found reduced urea enzyme and glutamine synthetase (GS) activity. Although the urea cycle enzymes rapidly reversed upon realimentation, GS took 4 weeks to recover and was compensated by increased GS activity in muscle tissue.9 In the present case, we believe that the hyperammonaemia was caused by a combination of refeeding-induced nitrogen challenge and limited urea detoxification via both the urea cycle and GS. A low muscle mass may have contributed, and hypophosphataemia may have impaired the ATP-demanding steps in both processes. Hyperammonaemia was ultimately treated by haemodialysis in combination with lactulose. Other ammonia-lowering therapies such as rifaximine were omitted in the absence of chronic liver disease.6 Potentially, carnithine deficiency could contribute to hyperammonaemia, but this has mainly been associated with
Figure 2. Two essential pathways for ammonia detoxification: glutamine synthesis and the urea cycle. (a) Glutamine synthetase (GS) is present in liver, muscle, kidney and brain tissue. It catalyses the ATP-dependent condensation of glutamate with ammonia to yield glutamine. (b) The urea cycle is restricted to the liver and located in the mitochondria and cytoplasm of periportal hepatocytes. Urea is formed through five enzymatic steps, which require the cofactor N-acetylglutamate and four ATP equivalents.2 European Journal of Clinical Nutrition (2015) 279 – 281
© 2015 Macmillan Publishers Limited
Refeeding encephalopathy S Becker et al
long-term enteral feeding and anticonvulsant drug therapy and is unresponsive to haemodialysis.10 In conclusion, this case illustrates the prolonged encephalopathy induced by severe hypophosphataemia and hyperammonaemia following unmonitored enteral feeding. The case stresses the importance of adequate vitamin and electrolyte substitution in the malnourished patient during nutrition therapy.1 The substitution of electrolytes, particularly phosphate, may be needed for weeks in all types of nutrition therapy. Hyperammonaemia may explain coma also in non-cirrhotic patients, where acute liver failure, portosystemic shunting, urea cycle defects and possibly refeeding should be considered. We recommend patience if coma persists after correction of phosphate and ammonia levels and to consider haemodialysis in severe hyperammonaemia where other ammonia-lowering therapies are insufficient.
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS The authors thank Peter Ott for the valuable comments.
© 2015 Macmillan Publishers Limited
281
REFERENCES 1 Ormerod C, Farrer K, Harper L, Lal S. Refeeding syndrome: a clinical review. Br J Hosp Med (Lond) 2010; 71: 686–690. 2 Haberle J. Clinical and biochemical aspects of primary and secondary hyperammonemic disorders. Arch Biochem Biophys 2013; 536: 101–108. 3 Megarbane B, Guerrier G, Blancher A, Meas T, Guillausseau PJ, Baud FJ. A possible hypophosphatemia-induced, life-threatening encephalopathy in diabetic ketoacidosis: a case report. Am J Med Sci 2007; 333: 384–386. 4 Ardalan MR, Pourafkari L, Tubbs RS, Shoja MM. Hypophosphatemic encephalopathy in a CAPD patient. Am J Med Sci 2008; 335: 492–494. 5 Leroy S, Gout A, Husson B, de Tournemire R, Tardieu M. Centropontine myelinolysis related to refeeding syndrome in an adolescent suffering from anorexia nervosa. Neuropediatrics 2012; 43: 152–154. 6 Vilstrup H, Amodio P, Bajaj J, Cordoba J, Ferenci P, Mullen KD et al. Hepatic encephalopathy in chronic liver disease: Practice Guideline by the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver. Hepatology 2014; 60: 715–735. 7 Morsy MR, Madina H, Sharaf SA, Soliman AT, Elzalabany MM, Ramadan MA. Hyperammonemia in Marasmic children. J Trop Pediatr 1994; 40: 97–99. 8 Grazer RE, Sutton JM, Friedstrom S, McBarron FD. Hyperammonemic encephalopathy due to essential amino acid hyperalimentation. Arch Intern Med 1984; 144: 2278–2279. 9 Felipo V, Minana MD, Grisolia S. Control of urea synthesis and ammonia utilization in protein deprivation and refeeding. Arch Biochem Biophys 1991; 285: 351–356. 10 Limketkai BN, Zucker SD. Hyperammonemic encephalopathy caused by carnitine deficiency. J Gen Intern Med 2008; 23: 210–213.
European Journal of Clinical Nutrition (2015) 279 – 281
CLINICAL CASE REPORT
Refeeding encephalopathy in a patient with severe hypophosphataemia and hyperammonaemia S Becker, G Dam and CL Hvas The refeeding syndrome is a potentially fatal condition that affects multiple organ systems. It is the consequence of fluid and electrolyte shifts that may occur in a malnourished patient following the introduction of nutrition therapy. The most prominent characteristic is hypophosphataemia. Although hyperammonaemia is usually seen in decompensated liver cirrhosis or acute liver failure, it may occur in other settings. We report a clinical case of prolonged and severe encephalopathy accompanied by hypophosphataemia and hyperammonaemia in a 59-year-old woman with no preexisting liver disease, urea cycle defects or portosystemic shunting. We suggest that these biochemical abnormalities were caused by uncontrolled refeeding and that the clinical picture was consistent with refeeding encephalopathy. European Journal of Clinical Nutrition (2015) 69, 279–281; doi:10.1038/ejcn.2014.244; published online 12 November 2014
INTRODUCTION The refeeding syndrome (RFS) is a condition of severe fluid and electrolyte shifts that may occur when nutrition is reintroduced to a malnourished patient.1 RFS is characterised by hypophosphataemia, hypokalaemia, hypomagnesaemia, fluid imbalance and frequently thiamine deficiency. RFS may be caused by oral, enteral or parenteral nutrition.1 Hyperammonaemia is most commonly seen in decompensated liver cirrhosis but may occur in other conditions such as acute liver failure, portosystemic shunting or urea cycle defects.2 Here, we describe a case of RFS presenting with encephalopathy associated with both hypophosphataemia and hyperammonaemia in a patient without liver disease. Case report A 59-year-old woman with a 30-year history of perianal Crohn’s disease was admitted to the hospital. Two days earlier, she had been discharged from a surgical department after revision of sacral ulcers. Because of poor nutritional status and a low food intake, oral thiamine supplementation and nasogastric tube feeding had been established at discharge with an initial dose of 500 ml (625 kcal) per day. On admission, the patient was deeply comatose with no response to pain stimuli. She had myoclonus, spontaneous pronations of all extremities and an extensive plantar response. Physical examination revealed two spider angiomas but no palmar erythema or caput medusae. She was malnourished with a body weight of 43 kg and a body mass index of 16 kg/m2. Before admission, she had a daily alcohol consumption of 2–3 glasses of wine. Her daily medications included morphine, acetaminophene, magnesium, potassium and B vitamins. Blood tests showed elevated C-reactive protein 80 mg/l (o8 mg/l), leukocytosis 18.1 × 109/l (3.5–10 × 109/l), hypoalbuminaemia 13 g/l (36–45 g/l), normal liver function tests, sodium, potassium and magnesium levels, severe hypophosphataemia o 0.17 mmol/l (0.76–.41 mmol/l) and markedly increased arterial ammonia of
252 µmol/l ( o50 µmol/l). Urea excretion was 89–117 mmol/day (300–600 mmol/day) and the glomerular filtration rate was normal. Arterial blood gas revealed an alkalosis with a pH of 7.54 (7.37–7.45). Two cerebral computed tomography scans and one magnetic resonance imaging (MRI) revealed no pathology, particularly no pontine myelinolysis. Spinal fluid had normal cell counts, cerebrospinal fluid/serum glucose ratio and marginally elevated protein at 0.94 g/l (0.15–0.5 g/l). Neuroinfection was disproved by PCR, microscopy and culture. Electroencephalography was consistent with metabolic encephalopathy and ruled out epileptic state. An abdominal computed tomography showed no signs of cirrhosis, portal hypertension or portosystemic shunting. Over 10 days, the patient’s mental status did not improve despite intravenous antibiotics, phosphate 20–80 mmol per day and thiamine (Figure 1). Following normalisation of plasma phosphate levels, enteral tube feeding including lactulose was re-established. Haemodialysis was carried out to lower ammonia. On the 12th day of hospitalisation, the patient’s cerebral state slowly improved. During recovery, she complained of muscle pain and disorientation. She was discharged on day 31 on a stable tube feeding regimen. Although awake and oriented, the patient experienced cognitive deficits and muscle weakness. DISCUSSION Hypophosphataemia is a predominant feature of RFS. RFS is potentially fatal and can affect any organ system.1 Phosphate is essential for most cellular processes and is particularly important during the anabolic phase.1 In this patient, multiple risk factors predisposed to RFS––that is, alcohol consumption, inflammation, weight loss, low body mass index and low food intake before realimentation. The initial enteral feeding dose was low (15 kcal/kg/day and 0.7 g protein/kg/day) but sufficient to induce a severe and prolonged hypophosphataemia. Guidelines recommend that patients at a high risk of RFS should be started at 10 kcal/kg/day or even lower in extreme cases.1 Profound hypophosphataemia
Department of Hepatology and Gastroenterology, Aarhus University Hospital, Aarhus C, Denmark. Correspondence: Dr CL Hvas, Department of Hepatology and Gastroenterology, Aarhus University Hospital, Nørrebrogade 44, DK-8000, Aarhus C, Denmark. E-mail: [email protected] Received 12 July 2014; revised 22 September 2014; accepted 28 September 2014; published online 12 November 2014
Refeeding encephalopathy S Becker et al
280 may explain the coma, myoclonus, respiratory insufficiency, neurological deficits and muscle pain. Encephalopathy during refeeding may be caused by hypophosphataemia alone,3,4 but other factors such as thiamine deficiency
Figure 1. Plasma levels of ammonia (upper level) and phosphate (lower level) during hospitalisation for refeeding encephalopathy.
or centropontine myelinolysis5 should be considered. In this case, thiamine was provided orally at initiation of enteral tube feeding, and MRI excluded centropontine myelinolysis. Hyperammonaemia is usually seen in decompensated liver cirrhosis but may occur in acute liver failure, portosystemic shunting and urea cycle defects.2 Our patient had no pre-existing liver disease. No previous episodes of coma indicated primary urea cycle defects, and she received no medications that could interfere with the urea cycle—for example, valproic acid.2 In the absence of predisposing factors, we suspected a link between hyperammonaemia and hypophosphataemia. Normally, ammonia is detoxified through glutamine synthesis in liver, muscle, kidney and brain tissue or via the urea cycle in the liver. Both are ATP-demanding processes (Figure 2). In order not to underestimate plasma ammonia, arterial blood samples could therefore be preferred.6 Hyperammonaemic encephalopathy has been described in protein-deprived children7 and in adults hyperalimentated with essential amino acids.8 A suggested mechanism was limited substrate for the urea cycle.8 In protein-deprived rats,8,9 ammonia increased and urea excretion decreased during starvation, and the authors found reduced urea enzyme and glutamine synthetase (GS) activity. Although the urea cycle enzymes rapidly reversed upon realimentation, GS took 4 weeks to recover and was compensated by increased GS activity in muscle tissue.9 In the present case, we believe that the hyperammonaemia was caused by a combination of refeeding-induced nitrogen challenge and limited urea detoxification via both the urea cycle and GS. A low muscle mass may have contributed, and hypophosphataemia may have impaired the ATP-demanding steps in both processes. Hyperammonaemia was ultimately treated by haemodialysis in combination with lactulose. Other ammonia-lowering therapies such as rifaximine were omitted in the absence of chronic liver disease.6 Potentially, carnithine deficiency could contribute to hyperammonaemia, but this has mainly been associated with
Figure 2. Two essential pathways for ammonia detoxification: glutamine synthesis and the urea cycle. (a) Glutamine synthetase (GS) is present in liver, muscle, kidney and brain tissue. It catalyses the ATP-dependent condensation of glutamate with ammonia to yield glutamine. (b) The urea cycle is restricted to the liver and located in the mitochondria and cytoplasm of periportal hepatocytes. Urea is formed through five enzymatic steps, which require the cofactor N-acetylglutamate and four ATP equivalents.2 European Journal of Clinical Nutrition (2015) 279 – 281
© 2015 Macmillan Publishers Limited
Refeeding encephalopathy S Becker et al
long-term enteral feeding and anticonvulsant drug therapy and is unresponsive to haemodialysis.10 In conclusion, this case illustrates the prolonged encephalopathy induced by severe hypophosphataemia and hyperammonaemia following unmonitored enteral feeding. The case stresses the importance of adequate vitamin and electrolyte substitution in the malnourished patient during nutrition therapy.1 The substitution of electrolytes, particularly phosphate, may be needed for weeks in all types of nutrition therapy. Hyperammonaemia may explain coma also in non-cirrhotic patients, where acute liver failure, portosystemic shunting, urea cycle defects and possibly refeeding should be considered. We recommend patience if coma persists after correction of phosphate and ammonia levels and to consider haemodialysis in severe hyperammonaemia where other ammonia-lowering therapies are insufficient.
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS The authors thank Peter Ott for the valuable comments.
© 2015 Macmillan Publishers Limited
281
REFERENCES 1 Ormerod C, Farrer K, Harper L, Lal S. Refeeding syndrome: a clinical review. Br J Hosp Med (Lond) 2010; 71: 686–690. 2 Haberle J. Clinical and biochemical aspects of primary and secondary hyperammonemic disorders. Arch Biochem Biophys 2013; 536: 101–108. 3 Megarbane B, Guerrier G, Blancher A, Meas T, Guillausseau PJ, Baud FJ. A possible hypophosphatemia-induced, life-threatening encephalopathy in diabetic ketoacidosis: a case report. Am J Med Sci 2007; 333: 384–386. 4 Ardalan MR, Pourafkari L, Tubbs RS, Shoja MM. Hypophosphatemic encephalopathy in a CAPD patient. Am J Med Sci 2008; 335: 492–494. 5 Leroy S, Gout A, Husson B, de Tournemire R, Tardieu M. Centropontine myelinolysis related to refeeding syndrome in an adolescent suffering from anorexia nervosa. Neuropediatrics 2012; 43: 152–154. 6 Vilstrup H, Amodio P, Bajaj J, Cordoba J, Ferenci P, Mullen KD et al. Hepatic encephalopathy in chronic liver disease: Practice Guideline by the American Association for the Study of Liver Diseases and the European Association for the Study of the Liver. Hepatology 2014; 60: 715–735. 7 Morsy MR, Madina H, Sharaf SA, Soliman AT, Elzalabany MM, Ramadan MA. Hyperammonemia in Marasmic children. J Trop Pediatr 1994; 40: 97–99. 8 Grazer RE, Sutton JM, Friedstrom S, McBarron FD. Hyperammonemic encephalopathy due to essential amino acid hyperalimentation. Arch Intern Med 1984; 144: 2278–2279. 9 Felipo V, Minana MD, Grisolia S. Control of urea synthesis and ammonia utilization in protein deprivation and refeeding. Arch Biochem Biophys 1991; 285: 351–356. 10 Limketkai BN, Zucker SD. Hyperammonemic encephalopathy caused by carnitine deficiency. J Gen Intern Med 2008; 23: 210–213.
European Journal of Clinical Nutrition (2015) 279 – 281
