Recurrent Pyroglutamic Acidosis in the Context of Undiagnosed Liver Cirrhosis—A Case Report

Abstract

Metabolic associated fatty liver disease, previously known as nonalcoholic fatty liver disease, is the most common cause of chronic liver disease across all ethnic groups; however, it remains enormously underestimated.^1,2 Sepsis, hepatotoxic medications and malnutrition in the acute settings on top of unknown cirrhosis can lead to decompensation and various metabolic complications. Pyroglutamic acidosis is a rarely recognised cause for unexplained high anion gap metabolic acidosis that is felt to be frequently underdiagnosed. Particular patients at risk include women, the elderly, those on regular paracetamol and those suffering with malnourishment or sepsis. Other risk factors include alcohol abuse and chronic liver disease (3). We present the case of a patient with recurrent episodes of pyroglutamic acidosis and encephalopathy in the context of undiagnosed nonalcoholic fatty liver disease with cirrhosis.

A 44-year-old gentleman was admitted to the hospital from an outpatient gastroenterology clinic with severe hypokalaemia (2.3 mmol/L). He was complaining of muscle cramping, paraesthesia and lethargy. He was being seen in clinic with unexplained weight loss, macrocytic anaemia and intermittent postprandial vomiting. Previous upper endoscopy was normal; and a CT thorax, abdomen and pelvis did not reveal a cause for his weight loss. The patient's medical history included depression with anxiety, macrocytic anaemia and folate deficiency. His Body Mass Index (BMI) was 32.6, triglyceride and cholesterol levels were 2.2 and 4, respectively, and he was not prediabetic. Regular medications included co-codamol 30/500 mg as needed, esomeprazole 20 mg OD and mirtazipine 30 mg at night. His alcohol consumption was negligible for the last 7 years, and he had not had alcohol in excess for any significant period of time.

On admission, he was alert, comfortable but lethargic. His examination was unremarkable. An initial venous blood gas (VBG) in the emergency department showed a potassium (K+) concentration of 1.65 mmol/L; other blood tests of note showed a stable haemoglobin level of 87 g/L, with raised mean corpuscular volume of 131.9 fL, a thrombocytopaenia of 65 × 10⁹/L, hypokalaemia at 2.3 mmol/L. a low urea of <1 mmol/L and a borderline low albumin (Alb) of 32 g/L and total bilirubin of 26 μmol/L. The SGPT level was normal, and C-reactive protein was mildly elevated at 56 mg/L. Other bloods were grossly normal. A blood borne virus screen for hepatitis A, B, C, E and human immunodeficiency virus was negative.

The patient was transferred to intensive care for a monitored bed to replace K+.

Episode 1

On day three, after step down to a gastroenterology ward, the patient demonstrated his first of many acute deteriorations, whereby he became progressively more unwell. The initial episode was characterised by tachycardia, tachypnoea, pyrexia and hypotension with sepsis. He was found to have a severe left leg cellulitis. VBG showed a lactic acidosis (pH 7.32) with a lactate of 10.3 mmol/L, along with a compensatory low partial pressure of carbon dioxide (pCO2 1.17 kPa) and a low bicarbonate (HCO3- 4.4 mmol/L) and base excess (BE -19.5 mmol) which is consistent with a high anion gap metabolic acidosis. The anion gap here was calculated to be 37.8. Otherwise, there was a normal partial pressure of oxygen (PO2 18.73 kPa) and a stable potassium concentration (K+ 3.6 mmol/L). The patient was transferred to the intensive care again for inotropic and fluid support.

The patient underwent Doppler ultrasound and was found to have an acute left common femoral deep vein thrombosis (DVT); this was treated with anticoagulation and deemed to be a provoked DVT. Of note, an echocardiogram showed good biventricular function with no vegetations or suggestion of endocarditis. Flucloxacillin antibiotic was commenced for a positive Staph. hominis leg wound culture.

Episode 2

After returning to the ward, the patient continued on treatment for cellulitis with flucloxacillin and hypokalaemia with oral and intravenous (IV) potassium replacement. He was intermittently confused and still demonstrated a stable anaemia and persistent unexplained thrombocytopaenia of 33 × 10⁹/L; this was felt to be a consumptive disorder especially with normal haematinics and fibrinogen levels. The patient again became acutely unwell with tachyhypnoea, tachycardia, a falling Glasgow Coma Score (GCS) of 13 (E3, V4, M6) and peripheral circulatory shutdown. An arterial blood gas (ABG) this time showed a pH of 7.194, a PCO2 of 1.4 kPa, a PO2 of 21 kPA (on air), a HCO3– of 4.0 mmol/L with a base excess of −22 mmol, a lactate of 9.78 mmol/L and an anion gap of 35.8. Shortly after this, he was again readmitted to intensive care unit (ITU) for management of his high anion gap acidosis.

Episode 3

The patient became drowsy with a grade 2 encephalopathy (confirmed on electroencephalography (EEG)) and displayed hyperammonaemia (ammonia at 169 μmol/L). He had pancytopaenia, hypoalbuminaemia and had a persistent lactic acidosis (lactate 3–5 mmol/L) with an associated severe cellulitis (still on flucloxacillin). The patient was behaving more similar to a patient with typical decompensated liver disease. An ultrasound showed a diffusely fatty liver without signs of cirrhosis. Other investigations showed a negative liver screen, normal CT head and echocardiogram. Renal function was normal with a normal Alb/creatinine ratio and a urinary pH of 5.0. An MRI lower limb showed bilateral acute myositis without osteomyelitis. He was treated with regular lactulose, rifaximin 550 mg twice daily orally and regular sodium bicarbonate at 2 g four times a day intravenously. A liver Fibroscan was performed and produced a score of 28.8 kPa – displaying established cirrhosis; this was felt to be secondary to nonalcoholic steatohepatitis (NASH). Treatment continued for this NASH cirrhosis, and flucloxacillin was stopped after 21 days of treatment. His 3rd deterioration begun gradually with increasing Kussmaul breathing, a falling GCS and again high anion gap metabolic acidosis with hypokalaemia (2.2 mmol/L). After extensive investigations, the metabolic workup for hyperammonaemia including zinc, urinary organic acid electrophoresis, urinary 5-oxoproline and d-lactate was sent.

At this point, the patient had been in ITU on 4 occasions and had received a total of 3 weeks of flucloxacillin. He was improving globally with a more stable GCS at 15 most of the time, his ammonia level was down at 69 μmol/L and his cellulitis was recovering. On detailed history, the patient admitted having been taking at least 10 tablets of 30mg/500 mg of co-codamol daily for chronic back pain for most of the last 4 years. After this, his paracetamol use was halved to 500 mg four times daily, and the sodium bicarbonate supplementation was slowly weaned over the course of 7 days. The metabolic results returned which showed a notable positive 5-oxoproline level in his urine. This signified high levels of pyroglutamic acid in the blood (at the time of testing) that would have certainly caused his persistent high anion gap metabolic acidosis and his recurrent clinical deteriorations requiring intravenous fluids, bicarbonate and ITU admissions. The patient continued to improve clinically and started to receive physiotherapy and rehabilitation before he was discharged.

Discussion

Pyroglutamic (or 5-oxoproline) acidosis is a cause of a high anion gap metabolic acidosis that is deemed relatively rare but felt to be underdiagnosed by lack of knowledge and of test availability.^3,4 The risk factors for the condition are shown in Table 1, and particular emphasis is placed on the chronic use of paracetamol at even therapeutic doses.^3,5,6

Table 1.

Risk Factors for the Development of Pyroglutamic Acidosis.

Paracetamol use	Other medication
Malnutrition/chronic alcohol use	-Flucloxacillin
Chronic liver disease	-Vigabatrin
Female sex/pregnancy	-Ciprofloxacin
Sepsis	-Netilmicin
Renal failure	Enzyme deficiency
Elderly	-Glutothione synthase
	-5-oxoprolinase

Understanding of the pathophysiology begins with the glutathione metabolism and the γ-glutamyl cycle – as shown in Figure 1. Glutathione is an antioxidant found in all our cells where it acts to neutralise endogenous waste and xenobiotics.⁷ Chronic paracetamol use saturates our normal glutathione storage and results in a net deficit in glutathione and cysteine.³ which is further confounded by nutrition deficits of cystine precursors and glycine seen in malnutrition and sepsis.^3,4 These factors cause a negative feedback loop, whereby γ-glutamyl cysteine levels rise and become a substrate for an alternative later reaction catalysed by γ-glutamyl cyclotransferase resulting in an increase in 5-oxoproline (pyroglutamic acid).⁵ In normal physiology, 5-oxoprolinase can metabolise 5-oxoproline; however, the enzyme can be deficient and is easily saturated and can notably be inhibited by flucloxacillin use.^4,5 This ultimately equates to a 5-oxoprolinaemia and acidosis, which is further complicated by poor renal clearance in cases of renal failure.

Figure 1.

The γ-glutamyl cycle shown highlights the relevant risk factors and their contribution to a pathological state of 5-oxoprolinaemia. With cysteine, glycine and glutathione depletion, the alternative pathway is favoured leading to the breakdown of γ-glutamyl cysteine into γ-glutamyl amino acid and a thus an accumulation of 5-oxoproline. This is further confounded by inhibition of its breakdown by 5-oxoprolinase and renal impairment. Ultimately, this results in the pyroglutamic acidosis as discussed in this case and the pathophysiology that occurs.

A 5-oxoproline metabolic acidosis, also known as pyroglutamic acidosis, presents as a high anion gap metabolic acidosis.^8,9 Clinically, this manifests with tachypnoeic compensation, tachycardia, altered consciousness and can include encephalopathy, haemolytic anaemias and neurological disorders.⁴ Diagnosis is by the presence of elevated levels of 5-oxoproline in the urine, signifying stress of the glutathione metabolism – such as in sepsis.^8,10 The levels of 5-oxoproline normally vary with normal metabolic stress such as menstruation but lie within normal ranges in healthy individuals.¹¹ Optimal treatment includes the cessation of paracetamol, utilisation of N-acetylcysteine, sodium bicarbonate and a reduction in the other risks factors and inhibitory medication.^3,4

In hindsight, in our patient, his pyroglutamic acidosis was complicated by coexisting decompensated NASH cirrhosis with a lactic acidosis (from sepsis) that clearly only partially accounted for this high anion gap and hyperammonaemia. His lactate persisted between 3 and 10 mmol/L whilst his anion gap was much higher at 35.8–37.8 mmol/L. Diabetic ketoacidosis, methanol, ethylene glycol, salicylate and renal failure were excluded as causes, and metabolic workup did not indicate any enzyme deficiency contributing to his lactic acidosis. In the meantime, focused history taking revealed that the patient had been taking excessive paracetamol for chronic back pain.

In the context, in our patient, it has previously been observed that malnutrition combined with low glutathione and hyperammonaemia can cause encephalopathy with minimally raised ammonia levels.¹² The increasing annual prevalence rate of metabolic associated fatty liver disease parallels that of obesity. In a meta-analysis conducted in 2016, it showed that the global prevalence of non-alcoholic fatty liver disease in particular is 25.24%, reflecting similarly the rates of type 2 diabetes and obesity in the United States.¹³

In many scenarios, metabolic associated fatty liver disease (MAFLD) remains a silent disease, underdiagnosed despite being associated with many comorbidities. NASH cirrhosis is currently the leading indication for liver transplantation in women and the second leading indication for men in the United States.¹⁴

On reflection, our patient's presentations were not typical. His underlying unexplained weight loss can be explained by multifactorial metabolic losses. His poor appetite was attributable to sepsis and intermittent drowsiness due to hepatic encephalopathy. His extensive leg cellulitis and resulting sepsis contributed to a state of malnutrition and moreover liver cirrhosis is known to induce a catabolic state with muscle wasting.¹⁵ Although his hepatic synthetic function tests were mildly elevated or normal, a clinical picture of hepatic dysfunction remained, consisting of encephalopathy, pancytopaenia, raised serum ammonia (in the absence of other causes) and mild abnormal LFT's - including high T bilirubin. We excluded other metabolic causes for encephalopathy such as respiratory or renal failure. Although his ultrasound scan and CT did not show radiologic evidence for liver cirrhosis, a fibroscan was performed based on his clinical features. Studies have shown that ultrasound has a sensitivity of 65–95% to detect chronic liver disease,¹⁶ and the limited spatial resolution of CT allows detection of only relatively thick fibrous septum which makes accurate texture analysis difficult.¹⁷

Moving forward with our patient, with the reduction of paracetamol dosing to a maximum of 3 g daily, cessation of flucloxacillin with resolving cellulitis and the use and careful downtitration of sodium bicarbonate, our patient slowly recovered. He became less encephalopathic, increased his nutritional intake and was later discharged. We are fortunate to be able to have found a cause for his recurrent deteriorations and have been able to provide him advice moving forward to use paracetamol more sparingly and to avoid long-term flucloxacillin use.

It is clear that the combination of the risk factors for developing pyroglutamic acidosis and that were met by our patient would seem to be extremely common in all hospitals¹, which adds to the consensus that this process is likely more common than recognised². The GOLD MARK acronym for the causes of high anion gap metabolic is seen in Table 2 and includes the aforementioned 5-oxoproline.¹⁸

Table 2.

GOLD MARK Acronym—Causes of a High Anion Gap Metabolic Acidosis.

Glycols - (ethylene/propylene)	Methanol
Oxoproline	Aspirin
L-Lactate	Renal failure (uraemia)
D-Lactate	Ketoacidosis

5-oxoprolinuria with pyroglutamic acidosis is a rarely diagnosed but common occurrence in the hospital setting. Particular emphasis should be placed on considering and excluding this as a cause of a high anion gap acidosis. This is especially pertinent in the presence of chronic paracetamol use and particularly in the context of chronic liver disease which often coexists with severe malnutrition, encephalopathy and a predisposition for sepsis. In addition, increased awareness and understanding of MAFLD cirrhosis is essential, especially with increasing prevalence of obesity and diabetes. Moreover, further consideration should be made when considering a diagnosis of liver cirrhosis if a clinical picture suggests it without ultrasonographic features.

CRediT authorship contribution statement

Bradley J.W. Allen: Conceptualization, Methodology, Writing - original draft, Visualization, Investigation. Ahmed A. Abu Shanab: Supervision, Writing - review & editing. Mark R. Anderson: Writing - review & editing. Edward N. Fogden: Writing - review & editing.

Conflicts of interest

All authors have none to declare.

Funding

None.