Skip to main content
Clinical Liver Disease logoLink to Clinical Liver Disease
. 2023 Jan 23;21(1):9–12. doi: 10.1097/CLD.0000000000000004

A curious case of confusion

Jian Chu 1,, Tavia Buysse 1, Justin Mitchell 2
PMCID: PMC9875787  PMID: 36937777

LEARNING OBJECTIVES

After participating in this activity, the learner will be better able to:

  1. Explain the need for assessment of nutritional, infectious, metabolic, and vascular causes in patients with non-cirrhotic hyperammonemia.

CASE REPORT

A 74-year-old woman was found acutely nonverbal and unresponsive at her rehabilitation facility and taken to the emergency department. Her medical history was significant for ischemic stroke, subclinical seizures on levetiracetam, and cholangiocarcinoma status post Whipple procedure 1 month before presentation.

She was afebrile and hemodynamically stable on admission. She did not respond to noxious stimuli nor visually track with a Glasgow Coma Scale of 10. Her basic metabolic panel, liver enzymes, complete blood count, vitamin B12, thyroid-stimulating hormone, urine drug screen, and arterial blood gas were unremarkable. Lactic acid was elevated to 3.0 mmol/L. Due to concern for postictal state from breakthrough seizures, levetiracetam was changed to valproic acid. MRI of the brain was unremarkable and electroencephalogram was negative for epileptiform activity. Microbial, autoimmune, and paraneoplastic studies from cerebrospinal fluid were also negative. A venous ammonia level was ordered for metabolic workup and found to be elevated to 218 µg/dL (reference: 58–94 µg/dL). She was empirically started on lactulose and rifaximin but repeat ammonia levels remained elevated with only modest improvement in mental status.

Question 1: What Additional Studies Would You Recommend?

  1. Serum valproic acid level

  2. Liver biopsy

  3. Triple phase CT/MRI of the abdomen

  4. Serum and urine protein electrophoresis

While valproic acid was not thought to be the cause of her hyperammonemia, it was changed to lacosamide and levocarnitine was started. She was maintained on a high-protein diet to prevent sarcopenia and received zinc supplementation for mild deficiency. A liver ultrasound showed normal parenchymal size and echogenicity and absence of vascular thromboses. A urine culture that had been obtained for infectious workup grew Proteus mirabilis and she was treated with i.v. piperacillin-tazobactam. Serum and urine amino acid and orotic acid levels were sent for workup of metabolic defects and were within normal limits.

A triple-phase CT abdomen/pelvis was obtained to investigate vascular causes for hyperammonemia and revealed an atretic main portal vein and a large superior mesenteric vein (SMV) to left common iliac vein portosystemic shunt (Figure 1). This shunt was identified as the primary etiology of her persistent hyperammonemia.

FIGURE 1.

FIGURE 1

Triple-phase CT abdomen/pelvis with i.v. and oral contrast demonstrating a large SMV-iliac portosystemic shunt (red arrowhead) in axial (A) and coronal (B) cross-sections. (C) Fluoroscopy during interventional radiology procedure demonstrating large caliber shunt. Abbreviation: SMV, superior mesenteric vein.

Question 2: What Is The Best Next Step In Management For This Patient?

  1. Shunt embolization

  2. Titration of lactulose and rifaximin

  3. Screening for esophageal and gastric varices

  4. Transjugular intrahepatic portosystemic shunt (TIPS) procedure

She underwent balloon-occluded retrograde transvenous obliteration of the shunt with marked improvement in mental status. Two months later, she had fully recovered without recurrence of encephalopathy. Repeat ammonia level was 62 µg/dL and imaging confirmed stable positioning of embolization devices (Figure 2).

FIGURE 2.

FIGURE 2

Repeat CT abdomen/pelvis with i.v. contrast demonstrating stable embolization devices (blue arrows) at the SMV-shunt (A) and shunt-left (B) iliac connections. Abbreviation: SMV, superior mesenteric vein.

DISCUSSION

Ammonia is an essential substrate for protein metabolism via the urea cycle but is also a potent neurotoxin that can lead to hyperammonemic encephalopathy, which is a rare but life-threatening cause of acute altered mental status. Although the majority of cases are due to intrinsic liver disease, noncirrhotic hyperammonemia can result from overproduction or impaired metabolism of ammonia in patients without liver disease.1 As illustrated in the presented case, noncirrhotic hyperammonemia is a diagnostic challenge, but early recognition can prevent severe neurological consequences.

Noncirrhotic hyperammonemia can result from overproduction of ammonia due to increased catabolism or microbial ammoniagenesis (Table 1). Increased catabolism occurs across a spectrum of clinical conditions, including seizure, starvation, or sepsis—all of which were important considerations in this case. Genitourinary infection by bacteria with urease activity, such as Proteus mirabilis, have been also implicated in ammonia overproduction.2 Bacteriuria-associated hyperammonemia is most frequently observed in patients with concurrent obstructive uropathy or neurogenic bladder, as increased intravesicular pressures promote venous ammonia absorption.

Table 1.

Schematic for differential diagnosis of noncirrhotic hyperammonemia

Differential for noncirrhotic hyperammonemia
Increased ammonia production Impaired ammonia metabolism
Increased catabolism  Steroids  Starvation  Burn wounds  Hematologic cancers Hyperalimentation  Total parenteral nutrition Microbial ammoniagenesis via urease  Proteus mirabilisEscherichia coliKlebsiella pneumoniae Severe gastrointestinal bleed Medications  Antiepileptic drugs (valproic acid, carbamazepine, topiramate)  Chemotherapy (5-fluorouracil)  Salicylate Errors of metabolism  Urea cycle disorders  Organ acidemias  Fatty acid oxidation defects Portosystemic shunts  Congenital vs. spontaneous  Transjugular intrahepatic portosystemic shunt Acute liver failure

Impaired ammonia metabolism via enzymatic dysfunction or nutritional deficiencies can also cause hyperammonemia. A thorough medication review is important as select medications—including valproic acid—directly inhibit the activity of key urea cycle enzymes. With long-term use, valproic acid can also deplete carnitine stores, which further impairs urea cycle metabolism.3 Intrinsic causes of impaired ammonia metabolism include nutritional deficiencies and inborn errors of metabolism. Trace metals like zinc are cofactors for urea cycle enzymes and deficiencies have been linked to hyperammonemia.4 Inborn errors of metabolism are rare causes of hyperammonemic encephalopathy. The most common late-onset urea cycle disorder is ornithine transcarbamylase deficiency, which is primarily inherited through an X-linked recessive pattern. Due to X-inactivation, hyperammonemia can occur in about 15%–20% of older adult female carriers during times of increased catabolic stress.5 The diagnosis is supported by increased levels of plasma glutamine or alanine with elevated urinary orotic acid levels and can be confirmed with DNA mutation testing.6

Our patient’s pre-Whipple imaging showed an atretic portal vein and large SMV-iliac shunt. This was most consistent with a congenital malformation as the SMV is rarely implicated in portosystemic shunts that result from cirrhotic portal hypertension.7 The shunt was also significantly larger in diameter than would be expected from spontaneous collateralization. In some clinical scenarios, ultrasonography can be used to measure abnormal shunt blood flow relative to native portal venous flow and the ratio reflects the severity of shunting.

Management of noncirrhotic hyperammonemia is typically focused on addressing the underlying cause. Nonabsorbable disaccharides, such as lactulose, decrease ammonia production and gut absorption. While lactulose is a first-line therapy for hepatic encephalopathy, no randomized controlled trials for its use in noncirrhotic hyperammonemia have been performed. In a study of patients with hyperammonemia of unknown etiology, lactulose did not improve ammonia level, intensive care unit length of stay, or mortality.8 While the risks of using lactulose are relatively low, its role in treatment of noncirrhotic hyperammonemia remains uncertain.

Due to the temporal correlation between onset of encephalopathy and surgery, we surmise that the Whipple procedure increased portal diversion through her congenital shunt. A Whipple involves resection, ligation, and re-anastomosis of structures near the portal-SMV-splenic confluence. In normal individuals, the portal venous system accounts for nearly two third of hepatic vascular supply.9 However, our patient’s portal vein was atretic and surgical manipulation of surrounding mesentery may have resulted in consequential portosystemic shunting. Our patient underwent comprehensive evaluation for noncirrhotic causes of hyperammonemia (Table 2), which revealed multiple contributory factors including Proteus bacteriuria, valproic acid use, increased postsurgical catabolism, and a portosystemic shunt. Her encephalopathy and hyperammonemia persisted despite antimicrobial treatment, dietary protein supplementation, withdrawal of valproic acid, and administration of lactulose and rifaximin. Therefore, her portosystemic shunt was identified as the primary driver of hyperammonemia after her Whipple procedure and the patient clinically improved following shunt embolization.

Table 2.

Diagnostic studies to be considered during workup of noncirrhotic hyperammonemia

Diagnostic studies for noncirrhotic hyperammonemia
Serum Urine Imaging
Arterial blood gas Lactate Chem-7 (bicarbonate, anion gap, glucose) Ketones Quantitative amino acids DNA mutation analysisa Urinalysis (pH, ketones)bUrinary orotic acid MRI brainc Triple-phase CT or MRI abdomen
a

Mutation analysis offers confirmatory testing for urea cycle disorders, of which OTC deficiency is the most common cause of late-onset hyperammonemia.

b

Measurement of ketone bodies, specifically beta-hydroxybutyrate, may be helpful in identifying starvation as a catabolic stressor.

c

In hyperammonemic encephalopathy, MRI brain findings range from reversible symmetric cortical diffusion restriction with FLAIR hyperintensities to cerebral edema.

Abbreviations: FLAIR, fluid-attenuated inversion recovery, OTC, ornithine transcarbamylase.

CME QUESTIONS

CME Question 1: What additional studies would you recommend?

  1. Serum valproic acid level

  2. Liver biopsy

  3. Triple phase CT/MRI of the abdomen

  4. Serum and urine protein electrophoresis

Answer:

Abdominal vessel imaging with CT or MRI can reveal vascular abnormalities that shunt blood from the portal system away from the liver. Valproic acid-induced hyperammonemia typically takes 1–2 weeks of therapy. In these cases, serum valproic acid levels may be normal and do not correlate with symptoms or level of hyperammonemia. Liver biopsy at this time would be an unnecessary procedure and would not aid in diagnosis unless underlying liver disease was suspected. While patients with multiple myeloma can rarely have hyperammonemic encephalopathy, this patient’s other workup was not consistent with a blood cell dyscrasia.

KEY POINTS

  1. A portosystemic shunt in the absence of intrinsic liver disease is a rare cause of hyperammonemia.

  2. Patients with noncirrhotic hyperammonemia should undergo investigation for nutritional, infectious, metabolic, as well as vascular causes.

  3. Hyperammonemia may be precipitated by gastropancreatic surgeries.

Acknowledgments

CONFLICT OF INTEREST

The authors have no financial or commercial interests to declare.

Contributor Information

Jian Chu, Email: jian_chu@rush.edu.

Tavia Buysse, Email: tavia_buysse@rush.edu.

Justin Mitchell, Email: Justin_Mitchell@rush.edu.

REFERENCES

  • 1.Kalra A, Norvell J. Cause for confusion: noncirrhotic hyperammonemic encephalopathy. Clin Liver Dis (Hoboken). 2020;15:223–227. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Kenzaka T, Kato K, Kitao A, Kosami K, Minami K, Yahata S, et al. Hyperammonemia in Urinary Tract Infections. PLoS One. 2015;10:e0136220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Baddour E, Tewksbury A, Stauner N. Valproic acid-induced hyperammonemia: Incidence, clinical significance, and treatment management. Ment Health Clin. 2018;8:73–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Katayama K. Zinc and protein metabolism in chronic liver diseases. Nutr Res. 2020;74:1–9. [DOI] [PubMed] [Google Scholar]
  • 5.Batshaw M, Msall M, Beaudet A, Trojak J. Risk of serious illness in heterozygotes for ornithine transcarbamylase deficiency. J Pediatr. 1986;108:236–241. [DOI] [PubMed] [Google Scholar]
  • 6.Redant S, Empain A, Mugisha A, Kamgang P, Attou R, Honoré P, et al. Management of late onset urea cycle disorders–a remaining challenge for the intensivist? Ann. Intensive Care. 2021;11:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nardelli S, Riggio O, Gioia S, Puzzono M, Pelle G, Ridola L. Spontaneous porto-systemic shunts in liver cirrhosis: Clinical and therapeutical aspects. World J Gastroenterol. 2020;26:1726–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Cutler N, Sadowski B, MacGregor D. Use of lactulose to treat hyperammonemia in ICU patients without chronic liver disease or significant hepatocellular injury. J Intensive Care Med. 2022;37:698–706. [DOI] [PubMed] [Google Scholar]
  • 9.Lautt W. Hepatic Circulation: Physiology and Pathophysiology. Morgan & Claypool Life Sciences; 2009. [PubMed] [Google Scholar]

Articles from Clinical Liver Disease are provided here courtesy of American Association for the Study of Liver Diseases

RESOURCES