Secondary Logo

Journal Logo

Neuroscience in Anesthesiology and Perioperative Medicine: Case Report

Perioperative Exacerbation of Valproic Acid–Associated Hyperammonemia

A Clinical and Genetic Analysis

Bezinover, Dmitri MD, PhD*; Postula, Marek MD†,‡; Donahue, Kathleen DO*; Bentzen, Brian MD*; McInerney, James MD§; Janicki, Piotr K. MD, PhD*

Author Information
doi: 10.1213/ANE.0b013e318228a001

Sudden exacerbation of chronic hyperammonemia (HA) is a medical emergency requiring immediate treatment. During metabolic stress and/or surgery with general anesthesia, ammonia levels can increase dramatically and lead to undetected cerebral edema and subsequent herniation. Careful selection of medications, recognition of symptoms, and exploration of the cause of HA, as well as careful perioperative management, may prevent perioperative deterioration.

Cases of severe HA related to the antiepileptic valproic acid (VPA) have been published in the emergency and pediatric medicine literature.1 Reports of HA patients requiring surgery are rare and most often limited to the pediatric population.2 There are very few reports of anesthesia for adult patients with chronic HA, usually because of inborn urea-cycle metabolic errors.3 None of these cases is related to VPA treatment. We report a case of HA associated with VPA treatment, together with genetic analysis of the underlying mutation and the perioperative management of the associated complications.


Before the preparation of this report, IRB approval from the Penn State Milton S. Hershey Medical Center and Penn State College of Medicine and informed written consent from the patient were obtained.

A 35-year-old Caucasian man with a 13-year history of intractable grand mal seizures, despite a combination of antiepileptic medications, presented for vagal nerve stimulator battery replacement under general anesthesia. His surgical history was significant for the resection of a neuroepithelial tumor. After surgery, the patient did not have any residual focal deficits but the seizures did not subside. Phenytoin, used as an initial treatment, was replaced with carbamazepine to control epileptiform activity. The addition of lamotrigine and topiramate improved the patient's condition but did not completely abolish the seizures. Five years before the current admission, VPA was added and a vagal nerve stimulator was placed. This improved control of the seizures. The patient, however, experienced dramatically increased sleepiness and mental status deterioration. Extensive neurological evaluation failed to uncover the cause of this condition. Four years ago, the patient was admitted emergently to another hospital after he became unresponsive. Laboratory results revealed a significantly increased level of serum ammonia (240 μmol/L) with normal renal and hepatic function. Therapy with lactulose enema and L-carnitine was initiated. The ammonia level decreased to 58 μmol/L, and the patient was discharged with an L-carnitine prescription and recommendation for a low-protein diet. No follow-up evaluation was performed, and VPA was noted as a possible cause for the HA. Attempts to taper the VPA failed because of uncontrollable seizures. Several months before the current admission, seizure episodes increased in frequency. Evaluation revealed that vagal nerve stimulator battery replacement was necessary.

Perioperative Management

The patient was initially evaluated in the anesthesia preoperative clinic 7 days before the procedure. Because of the history of chronic HA, additional diagnostic tests were performed. Blood samples for evaluation of liver and kidney function (alanine aminotransferase, aspartate aminotransferase, γ-glutamyltransferase, bilirubin, and creatinine blood urea nitrogen) were obtained. Additionally, a blood amino acid profile (glycine, glutamine, lysine, alanine, arginine, citrulline, carnitine, and ornithine) and urine orotic acid levels were ordered. The serum ammonia level at this time was 68 μmol/L. Preoperative laboratory results demonstrated normal liver and kidney function. Blood levels of all amino acids and the urine orotic acid were within normal range.

After discussion with the patient and neurosurgeon, the patient was consented for moderate sedation and local anesthesia for battery replacement. Five days before the procedure, high-dose L-carnitine administration was initiated to reach the recommended 50 mg/kg/d (3500 mg/d instead of 1000 mg/d).4

The patient was admitted to the hospital on the day of surgery. The ammonia level evaluated on the day of surgery was 89 μmol/L. He received premedication with 2 mg midazolam IV before the procedure, and propofol (120 μg/kg/min) was used for sedation. The surgical field was infiltrated with 10 mL of 0.5% bupivacaine + 1% lidocaine by the neurosurgeon. An infusion of 10% dextrose was also begun. At the end of an uneventful surgery lasting 90 minutes, the propofol infusion was stopped and the patient regained consciousness. Neurological examination results were similar to baseline. Over the next 2 hours, the patient became increasingly somnolent. The serum ammonia level increased to 143 μmol/L. The patient received a lactulose enema and his status improved, with his serum ammonia level decreasing to 128 μmol/L. The patient further recovered uneventfully and was discharged on postoperative day 1 to the care of his primary care provider.

Genetic Analysis

One milliliter heparinized blood was obtained for analysis. The sequencing chromatogram (Fig. 1, Table 1) demonstrates that the patient was heterozygous for a single nucleotide polymorphism (SNP) in the carbamoyl phosphate synthetase 1 (CPS1) gene (CPS1 4217C>A, rs1047891, Thr1406Asn).

Figure 1
Figure 1:
Sequencing chromatogram illustrating the presence of the investigated mutation. The polymerase chain reaction template was sequenced on an ABI 3130XL Capillary Sequencer (Applied Biosystems, Life Technologies Corp., Carlsbad, CA) according to manufacturer recommendations. All sequence chromatograms were read and aligned to the reference sequence of the CPS1 gene with GenBank accession number NM_001875 (in both the sense and antisense directions) to determine the CPS1 4217C>A genotype using MacVector version 11.1 software (MacVector, Inc., Cary, NC). The left panel (A) illustrates the sequencing chromatogram obtained for reverse (−) strand, and the right panel (B) represents chromatograms obtained for the forward (+) strand of the polymerase chain reaction template. Arrows indicate the chromatographic peaks for nucleotides in the investigated position. Please note that for both the reverse and forward strand there are 2 nucleotides in this position (corresponding to the heterozygous character of the sample).
Table 1
Table 1:
Genotyping Strategy for Detection of CPS1 4217C>A Single Nucleotide Polymorphism Based on CPS1 Genomic Sequence with GenBank Accession Number NT_005403.17


We present a case of significant deterioration of chronic VPA-induced HA after a surgical procedure despite aggressive perioperative treatment. Preoperative evaluation revealed normal levels of key amino acids involved in the urea cycle. Patients with VPA-related HA might have increased levels of glutamine and alanine, and decreased levels of citrulline, arginine, and carnitine. The amino acids profile can, however, be within the normal range, especially in cases of a chronic condition.4 The results in this case, in combination with a negative urine orotic acid level and normal liver and kidney function, suggest that the HA was likely related to VPA.

The effects of VPA on urea metabolism are complex and occur primarily in the liver mitochondria. VPA preferentially inhibits the enzyme necessary for the first step in the urea cycle—CPS1. This may lead to a dose-independent increase in the concentration of its substrate, ammonia, in the blood (Fig. 2).4,5

Figure 2
Figure 2:
1, In the hepatic mitochondrial matrix, valproic acid (VPA) inhibits carbamoyl phosphate synthetase 1 (CPS1) directly and indirectly (through suppression of N-acetylglutamate, which is an activator of CPS1, and increases concentrations of mitochondrial pyruvate, which is an inhibitor of CPS1.18 Suppression of CPS1 leads to a decrease in ammonia utilization. 2, In the hepatic cytosol, VPA forms a complex with carnitine and enhances carnitine's renal excretion.5 Carnitine is essential for transportation of fatty acids into the mitochondria. Reduced carnitine concentration results in the reduction of mitochondrial fatty acids available for β-oxidation and a subsequent increase in protein utilization. Lack of energy supply from β-oxidation may also cause a decrease in ammonia metabolism in the urea cycle. 3, In the kidney, VPA enhances the transport of glutamine across the mitochondrial membrane with simultaneous activation of glutaminase. This process increases transformation of glutamine to glutamate with a subsequent increase in ammonia production.19 21

Carnitine is required for transport and the subsequent oxidation of fatty acids in the mitochondria. VPA inhibits carnitine transport, which leads to increased renal carnitine excretion.5 Lack of carnitine results in the reduction of fatty acids metabolism and the subsequent increase of protein utilization (Fig. 2).4 VPA also increases ammonia production in the kidney (Fig. 2).

A number of studies have demonstrated that HA occurs frequently during VPA treatment; however, investigators have found a variable prevalence of HA (ranging from 5.6%6 to 100%).7,8 In the majority of cases, HA was mild (<45 μmol/L) and patients were asymptomatic.9,10 The highest reported blood ammonia level after VPA treatment was 991 μmol/L.11 It is important to note, however, that a relatively low HA of 83 μmol/L has also been associated with severe encephalopathy and coma.12 Studies have found that the combination of VPA with another antiepileptic medication was usually associated with a higher blood ammonia level.13,14 A mechanistic rationale for the interaction between VPA and CPS1, with resulting HA, has been recently suggested by Yagi et al.13 This study demonstrated that 1 particular missense SNP in the coding part (exon 36) of the CPS1 gene (CPS1 4217C>A, rs1047891, Thr1406Asn) leads to substitution of asparagine with threonine and decreased enzymatic activity in the urea cycle.13 Genomic analysis revealed that our patient was heterozygote for a missense polymorphism in the carbamoyl phosphate synthase 1 (4217C>A, rs1047891, Thr1406Asn). The association of this mutation with VPA-induced HA was previously described only in the Asian population, and never in a Caucasian. The prevalence of this genetic polymorphism (based on the frequency of the minor A allele) is approximately 10% to 30% in the Caucasian population, which may explain the relatively high frequency of VPA-induced HA.a

The pharmacological mechanism of action of VPA is based on the inhibition of γ-aminobutyric acid transamination, with a subsequent increase in γ-aminobutyric acid concentration predominantly in the central nervous system. This mechanism of action is probably not itself responsible for HA because other medications with similar activity, such as gabaculine, phenelzine, and vigabatrin, are not associated with increased blood ammonia levels.

Patients with chronic HA associated with VPA treatment requiring anesthesia for surgical procedures are at risk for dramatic exacerbation because of underlying errors of metabolism. The goal of perioperative management is to minimize any type of stress, such as anxiety, dehydration, or pain. All of these factors can potentially activate protein metabolism with subsequent deterioration of HA.15 The initial perioperative management should include replacement of VPA with another antiepileptic drug, if possible. A high-caloric, low-protein diet for 36 to 72 hours before surgery can help to prevent increased production of nitrogen waste. The next essential step is preoperative administration of L-carnitine in the dose of 50 to 100 mg/kg/d.4 Premedication with an anxiolytic, such as midazolam, is recommended. Perioperative hydration with 10% dextrose can help to minimize protein metabolism. A combination of general and regional anesthesia, ensuring adequate analgesia, is favorable.15 Regional anesthesia with sedation is preferred because of the necessity of neurological assessment postoperatively. Routine monitoring of the ammonia level throughout the perioperative period is essential. Treatment with lactulose or neomycin is usually helpful if ammonia is increased or if there is subsequent deterioration of the patient's mental status. If HA is refractory to standard treatment, other measures can be undertaken such as IV arginine, sodium benzoate or sodium phenylacetate administration, or oral administration of sodium phenylbutyrate (for less-acute cases). These medications activate alternative pathways for removal of nitrogen waste.16,17 Finally, hemodialysis can be used to decrease blood ammonia levels.15

VPA has been frequently used for the treatment of epilepsy. Taking into consideration the relatively high prevalence of VPA-related HA,14 surgical procedures and anesthesia for these patients may be associated with the possibility of acute neurological deterioration. Routine evaluation of the blood ammonia level of patients receiving VPA treatment, scheduled for surgical procedures, could identify affected patients. If future investigations confirm the significance of the CPS1 4217C>A, rs1047891 mutation for VPA-induced HA, genomic screening could be considered for identification of patients at risk. The advisability of VPA treatment for this particular population should be reconsidered.


Name: Dmitri Bezinover, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Dmitri Bezinover approved the final manuscript.

Name: Marek Postula, MD.

Contribution: This author helped design the study, conduct the study, and analyze the data.

Attestation: Marek Postula approved the final manuscript.

Name: Kathleen Donahue, DO.

Contribution: This author helped conduct the study, analyze the data, and write the manuscript.

Attestation: Kathleen Donahue approved the final manuscript.

Name: Brian Bentzen, MD.

Contribution: This author helped conduct the study, analyze the data, and write the manuscript.

Attestation: Brian Bentzen approved the final manuscript.

Name: James McInerney, MD.

Contribution: This author helped conduct the study, analyze the data, and write the manuscript.

Attestation: James McInerney approved the final manuscript.

Name: Piotr K. Janicki, MD, PhD.

Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.

Attestation: Piotr K. Janicki approved the final manuscript.


The authors are grateful for the assistance of Victor Ruiz-Velasco, PhD, Associate Professor of Anesthesiology, Neural and Behavioral Science, and Pharmacology, in preparation of the manuscript.

a Entrez SNP: National Center for Biotechnology Information. Available at: Accessed May 13, 2011.
Cited Here


1. Dealberto MJ. Valproate-induced hyperammonaemic encephalopathy: review of 14 cases in the psychiatric setting. Int Clin Psychopharmacol 2007;22:330–7
2. Dutoit AP, Flick RR, Sprung J, Babovic-Vuksanovic D, Weingarten TN. Anesthetic implications of ornithine transcarbamylase deficiency. Pediatr Anesth 2010;20:666–73
3. Mühling J, Dehne MG, Fuchs M, Sablotzki A, Weiss S, Spatz J, Hempelmann G. Conscientious metabolic monitoring on a patient with hyperornithinemia-hyperammonemia-homocitrullinuria (HHH) syndrome undergoing anaesthesia. Amino Acids 2001;21:303–18
4. Wadzinski J, Franks R, Roane D, Bayard M. Valproate-associated hyperammonemic encephalopathy. J Am Board Fam Med 2007;20:499–502
5. Camina MF, Rozas I, Castro-Gago M, Paz JM, Alonso C, Rodriguez-Segade S. Alteration of renal carnitine metabolism by anticonvulsant treatment. Neurology 1991;41:1444–8
6. Altunbasak S, Baytok V, Tasouji M, Herguner O. Asymptomatic hyperammonemia in children treated with valproic acid. J Child Neurol 1997;12:461–3
7. Navarro-Quesada FJ, Lluch-Fernández MD, Vaquero-Abellán M, Marchante-Serrano C, Jiménez C. Evaluation of the effect of long term valproic acid treatment on plasma levels of carnitine, ammonia and amino acids related to the urea cycle in pediatric epileptic patients. Rev Neurol 1997;25:1037–44
8. Warter JM, Brandt C, Marescaux C, Rumbach L, Micheletti G, Chabrier G, Krieger J, Imler M. The renal origin of sodium valproate induced hyperammonemia in fasting humans. Neurology 1983;33:1136–40
9. Wyllie E, Wyllie R, Rothner AD, Erenberg G, Cruse RP. Valproate induced hyperammonemia in asymptomatic children. Cleve Clin Q 1983;50:275–7
10. Iinuma K, Hayasaka K, Narisawa K, Tada K, Hori K. Hyperamino-acidaemia and hyperammonaemia in epileptic children treated with valproic acid. Eur J Pediatr 1988;148:267–9
11. Nicolai J, Smith SJ, Keunen RW. Simultaneous side effects of both clozapine and valproate. Intensive Care Med 2001;27:943
12. Elgudin L, Hall Y, Schubert D. Ammonia induced encephalopathy from valproic acid in a bipolar patient: case report. Int J Psychiatry Med 2003;33:91–6
13. Yagi M, Nakamura T, Okizuka Y, Oyazato Y, Kawasaki Y, Tsuneishi S, Sakaeda T, Matsuo M, Okumura K, Okamura N. Effect of CPS1 4217C>A genotype on valproic acid-induced hyperammonemia. Pediatr Int 2010;52:744–8
14. Chicharro AV, de Marinis AJ, Kanner AM. The measurement of ammonia blood levels in patients taking valproic acid: looking for problems where they do not exist? Epilepsy Behav 2007;11:361–6
15. Schmidt J, Kroeber S, Irouschek A, Birkholz T, Schroth M, Albrecht S. Anesthetic management of patients with ornithine transcarbamylase deficiency. Paediatr Anaesth 2006;16:333–7
16. Kleppe S, Mian A, Lee B. Urea cycle disorders. Curr Treat Options Neurol 2003;5:309–19
17. Baum VC, O'Flaherty JE. Anesthesia for Genetic, Metabolic, and Dysmorphic Syndromes of Childhood. 2nd ed. Philadelphia: Lippincott Williams & Wilkins/Wolters Kluwer, 2007:280–1
18. Duarte J, Macias S, Coria F, Fernandez E, Claveria LE. Valproate-induced coma: case report and literature review. Ann Pharmacother 1993;27:582–3
19. Rumbach L, Cremel G, Marescaux C, Warter JM, Waksman A. Valproate induced hyperammonemia of renal origin: effects of valproate on glutamine transport in rat kidney mitochondria. Biochem Pharmacol 1989;38:3963–7
20. Mallet L, Babin S, Morais JA. Valproic acid-induced hyperammonemia and thrombocytopenia in an elderly woman. Ann Pharmacother 2004;38:1643–7
21. Verrotti A, Trotta D, Morgese G, Chiarelli F. Valproate-induced hyperammonemic encephalopathy. Metab Brain Dis 2002;17:367–73
© 2011 International Anesthesia Research Society