Hypoxic hepatitis and furosemide

Abstract

Hypoxic hepatitis is a rather common complication of heart, circulatory or respiratory failure. We present the case of a patient with hypoxic hepatitis in the setting of heart failure and dehydration from furosemide as a reminder of an important clinical lesson. The pathogenesis of hypoxia (especially in the case of heart failure) is explained by a two-hit mechanism in which the liver at risk of hypoxic injury by passive hepatic congestion (right heart failure) is subsequently exposed to systemic hypoperfusion, which leads to a marked and transient elevation of aminotransferases. In the case presented, the use of furosemide (at least partially) promoted the second hit because it helped to generate hypotension and splanchnic hypovolaemia and favoured hepatic hypoxia.

Background

In our opinion, this case is relevant because it shows the association between furosemide use and hypoxic hepatitis, which is an unusual one.

Case presentation

We present the case of a 60-year-old female patient who received doxorubicin-based chemotherapy for breast cancer which led to cardiomyopathy.

She was admitted to our hospital because of heart failure. An echocardiogram showed a dilated left ventricle, global hypokinesia and a left ventricular ejection fraction of 35%.

After a successful clinical progression, the patient was discharged with a furosemide dose of 40 mg orally two times a day. Beta blockers were started at low doses. Kidney function was not hindered, and a hepatic enzymogram showed no alterations at discharge.

Ten days after discharge from the hospital, she presented with syncope.

At the initial evaluation, the patient was found to be hypotensive with blood pressure (BP) of 60/40 mm Hg (mean arterial pressure -MAP- of 47 mm Hg estimated by formula) with a reperfusion time of less than 2 s. The heart rate was 60 beats/min. Mild jugular vein distention, perimalleolar oedema and bibasilar crackles were evident. Glasgow Coma Scale score was 15 and no alterations were observed in the neurological examination. Intravenous fluid therapy with normal saline was started.

Investigations

Laboratory test results on admission were: urea 125 mg/dL (12.8–42.8 mg/dL), creatinine 2.5 mg/dL (0.5–0.7 mg/dL), Na +126 mEq/L (136–145 mEq/L), K+3.6 mEq/L (3.5–5.1 mEq/L), total bilirubin 0.71 mg/dL (0.20–1.20 mg/dL), alanine aminotransferase (ALT) 343 UI/L (<33 UI/L), aspartate aminotransferase (AST) 343 UI/L (<32 UI/L), gamma glutamyl transferase 87 UI/L (5–36 UI/L), alkaline phosphatase 110 UI/L (35–105 UI/L), haemoglobin 11.4 g/dL (12–16 g/dL), white cell counts 8,74 ×10^9/L (4,5–11 ×10^9/L), platelets 118 ×10^9/L (140 –400 ×10^9/L) and an unaltered prothrombin time. The abdominal and urinary tract ultrasound appeared normal.

The patient was diagnosed with prerenal acute kidney injury, secondary to the use of diuretics in the context of congestive heart failure.

BP levels between 80/50 and 90/50 (MAP 60–63 mm Hg) were maintained during the first 24 hours of admission. She received an intravenous fluid therapy of 1000 cm³ of normal saline every 12 hours, her usual dose of beta blockers was halved and her diuretics dose was suspended.

During the first 24 hours of hospitalisation, the patient’s transaminase levels increased (ALT 952 UI/L and AST 1152 UI/L), without bilirubin alterations. Serology results for hepatitis B and C viruses, Epstein-Barr virus and cytomegalovirus were negative. Therefore, a diagnosis of hypoxic hepatitis was proposed.

Differential diagnosis

The hypoxic aetiology of the liver injury in this clinical setting is clear. However, when confronted with a patient with liver injury, infectious, drug induced, autoimmune and vascular (portal vein thrombosis or hepatic vein thrombosis)aetiologies should be considered. In this particular case, only infectious aetiology was ruled out (and it could be argued that it was not mandatory to do so, due to the clinical characteristics of this case).

Treatment

Following the diagnosis of hypoxic hepatitis, intravenous fluid therapy with normal saline was maintained and diuretics remained suspended. The patient tolerated the treatment well, and did not present with acute congestive failure.

Outcome and follow-up

A favourable clinical progression was observed with the established treatment. Intravenous fluid therapy was suspended after 4 days. Systolic BP of 100–120 mm Hg and diastolic BP of 60–70 mm Hg (MAP 73–87 mm Hg) were observed, no low cardiac output symptoms were noted, renal function improved and the transaminase levels normalised a week after admission. Additionally, no symptoms or signs of heart failure were noted on discharge.

Discussion

Hypoxic hepatitis is the name that is currently used for the entity previously called ischaemic hepatitis or shock liver, because the haemodynamic mechanism that leads to hypoxia is caused by low hepatic flow during hypotension. This leads to passive congestion of the liver, arterial hypoxaemia and dysoxia.¹ For example, in one study, only 51% of patients with hypoxic hepatitis had experienced a known hypotensive episode (defined as a sustained systolic BP of less than 80 mm Hg for at least 15 min).²

Hypoxic hepatitis occurs more frequently in critically ill patients. In a 2015 systematic review, 2% of intensive care unit admissions and 0.2% of hospital admissions had hypoxic hepatitis.³

The current diagnosis of hypoxic hepatitis is generally based on three criteria: (1) a clinical context resulting from decreased oxygen delivery or decreased oxygen use by the liver; (2) an increase in significant and commonly transient aminotransferase levels and (3) the exclusion of other potential causes of liver injury. These potential causes are viruses, bacteria (leptospirosis, anaplasmosis), protozoa (malaria, Babesia sp), fungi (Histoplasma), drugs, autoimmune diseases and vascular liver disease.

Acute heart failure, septic shock and respiratory failure are the most common conditions that predispose patients to hypoxic hepatitis.^{2 4–6} These three predisposing causes cover more than 90% of the cases, although other causes have been documented in the literature.¹ There is often more than one predisposing factor,⁶ but shock or maintained hypotension is not necessarily observed.^1–3 The level that transaminases must exceed varies according to different studies: between 2.5 times to 20 times the upper limit of the normal range. Therefore, an increase of 10 times the upper limit of transaminases could be a reasonable diagnostic value.^{3 7} When these three criteria are met, a diagnosis of hypoxic hepatitis can be assumed without performing a biopsy, but in cases of doubt, a biopsy can provide a definitive diagnosis.⁸ In the biopsy, the histological correlation is centrilobular hepatocyte necrosis in the absence of other abnormalities.⁷

Current knowledge of the pathophysiology of hypoxic hepatitis suggests that it occurs through a two-hit mechanism. The first hit occurs in a low flow state, typically in the context of right-sided heart failure that leads to elevated liver pressures. In this particular case, one could argue that the first hit occurred in the setting of mild right heart failure (patient only had perimalleolar oedema and mild jugular vein distention). This places the liver at risk of hypoxic injury. The second hit occurs during an acute episode of heart or circulatory failure that leads to systemic hypotension and hepatic ischaemia. Or it occurs during acute respiratory failure that leads to hypoxia and hepatic ischaemia. In heart failure or other cardiac pathologies, elevated central venous pressure is transmitted directly to the hepatic veins, which results in elevated hepatic venous pressures, thus reducing the gradient between the portal and venous pressures. The high venous pressures ultimately result in congestive liver disease with pathological findings of central venous congestion and hepatocytic injury in zone 3 of the hepatic acinus. This reduces the effectiveness of the liver’s buffering capacity to mitigate hypotension during an acute event.⁷

This case reminds us that although the first hit is the well-known right-sided heart failure that leads to venous liver congestion, the second hit can be explained, at least in part, by furosemide use. A furosemide dose of 80 mg on a daily basis could be a predisposing cause of hypovolaemia, hypotension and hypoxic hepatitis.⁹ In another published case report in which high doses of intravenous furosemide (250–500 mg) were administered, it was suggested that this drug may have contributed to splanchnic hypovolaemia in some patients, either by volume depletion or by volume transfer to another vascular department by selective vasoconstrictive effects.¹⁰

The pattern of liver function tests in patients with hypoxic hepatitis is characterised by a rapid increase in aminotransferase levels, with peak levels between 25 and 50 times the upper limit. Higher elevations of aminotransferase levels are more specific for ischaemic insult, with a narrower correlation with levels greater than 75 times the normal upper limit. It has been suggested that AST exists at higher levels in zone 3 of the hepatic acinus, which results in a greater increase in AST levels in hypoxic hepatitis compared with ALT levels. The highest levels of aminotransferases are seen at 24–48 hours after an ischaemic event, with the resolution starting immediately after the injury and reaching normal values within 7–10 days (depending on the correction of the underlying processes). Typically, there is a slight elevation in bilirubin (less than 3 mg/dL) and prothrombin time, whereas alkaline phosphatase levels remain normal. Bilirubin and the international normalised ratio may be elevated at disease onset and/or remain elevated.⁷ Our patient showed aminotransferases levels greater than 30 times the normal upper limit with normal levels of bilirubin and prothrombin time.

Hypoxic hepatitis occurs in the context of systemic hypoperfusion; therefore, there is a strong correlation with ischaemia in other organs. In a retrospective review of 182 patients with hypoxic hepatitis, 67% of the patients had renal injury, 41% had rhabdomyolysis and 25% had ischaemic pancreatitis.⁵ It can even lead to acute liver failure. In our case, prerenal acute kidney injury (according to the Kidney Disease: Improving Global Outcomes guidelines)¹¹ was observed. Treatment in each case is based on treating the underlying pathology.⁷ The mortality rates for patients with hypoxic hepatitis range from 25% to 51%.^{3 12} Our patient had a markedly benign course and recovered fully only with support treatment.

Learning points.

• The pathogenesis of hypoxic hepatitis (especially in the case of heart failure) is explained by a two-hit mechanism in which a liver that is at risk of hypoxic injury due to passive hepatic congestion (right heart failure) is subsequently exposed to systemic hypoperfusion, which leads to a marked and transient elevation in aminotransferases.

• The presence of shock or sustained evident hypotension is not exclusive conditions for the development of hypoxic hepatitis.

• Furosemide may predispose patients to hypoxic hepatitis by generating splanchnic hypovolaemia, either by volume depletion or by volume transfer to another vascular department by selective vasoconstrictive effects. This effect has been observed at an oral dose of 80 mg/day.