Article Text

Download PDFPDF
Hydralazine-induced liver injury: a review and discussion
  1. Meeta Bhardwaj,
  2. Nakul Jay Bhardwaj,
  3. Kendra Cueto and
  4. T Colin Killeen
  1. Department of Internal Medicine and Geriatrics, Cleveland Clinic Foundation, Cleveland, Ohio, USA
  1. Correspondence to Dr T Colin Killeen; killeet{at}ccf.org

Abstract

Hydralazine is a commonly prescribed antihypertensive agent. Some of its labelled adverse reactions include lupus-like syndrome, tachycardia, headache and fever. Despite its well-known side effects, little is known about hydralazine’s hepatotoxic effects. We report the case of a 54-year-old female patient who was started on hydralazine for hypertension management but later presented with hydralazine-induced liver injury. Her initial presentation consisted of non-specific symptoms and a hepatocellular injury pattern. Liver biopsy revealed hepatic steatosis. Three weeks after discontinuation of hydralazine, the patient’s liver enzymes normalised, and her symptoms resolved. Few studies have examined the incidence and mechanism by which hydralazine induces a liver injury pattern. With this case, we review the literature, the pathogenesis involved and the eventual management of hydralazine-induced liver injury. We propose close monitoring of liver enzymes for patients on hydralazine throughout their treatment course.

  • liver disease
  • drugs
  • gastroenterology
  • gastrointestinal system

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Hydralazine is a phthalazine derivative often prescribed in severe hypertension and heart failure.1 It vasodilates arteries and arterioles, leading to a reduction in afterload of blood on the heart. Biochemically, it increases the level of cyclic guanosine monophosphate, leading to smooth muscle relaxation.1 This relaxation and reduction of afterload function to decrease blood pressure. Well-known adverse effects reported with hydralazine use include headache, dizziness, fluid retention and lupus-like syndrome.1 Increasingly, however, it has been associated with liver injury.2

Among the cases of hydralazine-related liver injury, non-specific symptoms were mostly noted, including fever, fatigue and nausea.2 In these cases, liver injury developed over the interval of a few weeks to months after introduction of hydralazine.2 Most symptoms resolve following discontinuation of hydralazine and recur following rechallenge of the drug.2 Although previous cases of hydralazine-related hepatitis have been reported, its connection and categorisation under drug-induced liver injury (DILI) have not been fully discussed or characterised.

The following case presents a middle-aged Caucasian woman diagnosed with hydralazine-induced liver injury pattern and her constellation of symptoms throughout. The manifestations and mechanisms of hydralazine-related liver injury, diagnosis of this clinical disorder and resolution following discontinuation of the medication are further discussed. Finally, we end with a discussion on its categorisation under DILI and review the pathophysiology of this disorder.

Case presentation

A 54-year-old Caucasian woman with a medical history of hypertension, stage three chronic kidney disease, asthma, Crohn’s disease and cerebrovascular accident presented to the office for hypertension management. Due to insurance issues, intolerance of certain antihypertensives and suboptimal control of blood pressure, her medication regimen was adjusted to include hydralazine 50 mg three times daily with titration to 100 mg after 2 weeks on the medication. Her blood pressure regimen further included triamterene 200 mg daily and losartan 100 mg daily.

Three days after starting the 100 mg dose of hydralazine, the patient presented to the emergency department for influenza-like symptoms including fevers, chills, fatigue, tachycardia, non-bloody diarrhoea, nausea, headaches, rhinorrhoea, myalgias and a dry cough.

Investigations

Labs and infectious disease workup were mostly unrevealing apart from elevated transaminases, including an elevated aspartate transaminase (AST) at 170 U/L and alanine aminotransferase (ALT) at 155 U/L with normal bilirubin and alkaline phosphatase (ALP). Her complete blood count was unremarkable. An ECG revealed sinus tachycardia, which eventually resolved with intravenous fluid hydration. A COVID-19 rapid PCR test was negative. The patient was sent home with a diagnosis of a non-specific viral syndrome.

The next day, she had a follow-up with her primary care provider (PCP) to further discuss her symptoms, but no significant changes were made, and conservative treatments for symptoms were continued. Five days later, the patient returned to the emergency department with generalised malaise, fatigue, fever, myalgias, intermittent right upper quadrant abdominal pain, nausea, darkened urine and exertional dyspnoea. Although her vitals remained stable, fevers persisted despite taking acetaminophen every 8 hours with temperatures reaching 102°F. The patient was fully alert and oriented and, on examination, was only remarkable for dark urine output. There were no signs of jaundice or scleral icterus. The patient denied any personal or family history of chronic liver disease or any alcohol abuse. Labs were again unrevealing apart from elevated transaminases, with AST 249 U/L, ALT 519 U/L, ALP 282 U/L and unremarkable bilirubin. An acetaminophen level was undetectable. A hepatitis panel, including anti-hepatitis A virus (HAV) (IgM), hepatitis B core antibody (HBcAb) (IgM), hepatitis B surface antigen (HBsAg) and anti-hepatitis C virus (HCV), was all negative as was a second COVID-19 test. Abdominal ultrasound was unremarkable and did not reveal any gallstones, dilated bile ducts, wall thickening or biliary obstruction. Further immunological studies were unrevealing for antinuclear antibody, antimitochondrial antibody, anti-smooth muscle antibody, liver kidney microsome type 1 antibodies, antihistone antibody and antineutrophil cytoplasmic antibodies. IgG and IgA, ceruloplasmin and ferritin levels were normal. Additional virological testing including Epstein-Barr virus, cytomegalovirus and herpes simplex virus was negative. Coagulation panels remained within normal limits throughout the patient’s symptom course.

During her hospital stay, the patient’s hydralazine was eventually stopped 9 days after maximal dose increase, completing 3 weeks on hydralazine. Her liver function tests (LFTs) slowly began down trending with the exception of ALP. Her AST was 146 U/L, ALT 409 U/L and ALP 428 U/L. On this same day, a magnetic resonance cholangiopancreatography demonstrated a normal pancreas and biliary tract and incidental findings of small side-branch intraductal papillary mucosal neoplasms. Mild hepatic steatosis was detected, but no focal hepatic lesions were found. The patient’s LFTs dramatically reduced 6 days following discontinuation of hydralazine: AST 33 U/L, ALT 105 U/L and ALP 219 U/L. A month later, the patient’s liver enzymes returned closer to baseline with AST 29 U/L, ALT 41 U/L and ALP 70 U/L.

Two months following discontinuation of hydralazine, the patient presented for a liver biopsy due to a lack of normalisation of transaminases. The biopsy demonstrated mild inflammation of the portal tracts without significant interface activity, duct injury or ductular reaction. In addition, mild steatosis (grade 1 on the Brunt scheme) was seen without ballooned hepatocytes, Mallory hyaline or cholestasis. Hepatic parenchyma demonstrated scattered lobular inflammation without hepatocellular injury. Overall, there were no features of steatohepatitis and autoimmune hepatitis or signs of non-alcoholic steatohepatitis.

Outcome and follow-up

After further consultation with gastroenterology, it was concluded that her symptoms resulted from a hydralazine-induced liver injury. It was believed that her postbiopsy symptom course appeared to be more consistent with DILI based on present inflammation and lack of features consistent with non-alcoholic steatohepatitis. The patient continued her diltiazem 240 mg, triamterene 200 mg daily and losartan 100 mg. A month later, the patient followed up with her PCP, at which time she felt to be back to her baseline state of health.

Discussion

Liver injury, marked by irregular liver enzymes, can present with or without associated clinical symptoms and can have multiple aetiologies, including the direct and indirect effect of drugs.3 Despite being associated with multiple drugs, the incidence of DILI is unknown with reported incidences ranging from 14 to 19 cases per 100 000 persons.4 Variability of its incidence is due to its clinical presentation that often leads to a misdiagnosis.5 Clinically, patients may present with non-specific symptoms such as fatigue, nausea, abdominal pain or immunoallergic signs such as fever, rash and adenopathy.6 DILI has multiple mechanisms of liver injury including direct and indirect toxic effects of the drug and variability in dose-related toxicities, which are not yet fully understood.6 7

DILI can be designated into two forms of hepatotoxicity: direct/intrinsic and idiosyncratic. DILI by means of direct hepatotoxicity generally has a short latency period with symptom onset in 1–5 days following ingestion of supratherapeutic doses.4 Acetaminophen is the most common cause of direct dose-related hepatotoxicity by means of direct/intrinsic mechanism.4 The majority of DILI cases, however, tend to be classified as ‘idiosyncratic’ drug reactions, in which toxicity is not dose-dependent and instead arises from a mechanism different than the therapeutic mechanism of the medication.8 9 DILI by idiosyncratic mechanisms is difficult to reproduce in animal models and is variable in incidence, occurring after 1 in 2000 to 1 in 100 000 patient exposures.10

In patients with idiosyncratic DILI, patients with certain risk factors appear to be more susceptible to serious liver injury.11 These include advanced age, female sex, alcohol abuse and malnutrition.12 Such risk factors may negatively influence cytochrome P450 (CYP450) and interfere with metabolism of various drugs leading to DILI.13 Due to the CYP450’s numerous roles in the body, including steroid biosynthesis and overall growth regulation,8 any issue within this system can lead to deleterious effects.

Two clinical patterns of hydralazine-induced DILI have been described: one with a shorter latency period of 2–6 weeks and a second with longer latency period of 2 months to greater than 1 year.14 15 Symptoms noted during the shorter latency period include fever, rash and eosinophilia, along with a rapid recovery of symptoms following discontinuation of the drug.11 Although the symptomology remains similar between both clinical patterns, biopsy of the liver in the longer latency form shows chronic hepatitis and fibrosis14 and has been associated with liver microsomal antibody detection.2 The recovery during this time is often prolonged, and such cases have also been found to be more likely associated with lupus-like syndrome.14 16 Our patient’s liver injury resembled the first clinical pattern, involving a short latency period with rapid recovery of symptoms following discontinuation of the drug.

The pathogenesis of hydralazine-induced liver injury is believed to occur both directly and indirectly due to metabolism of the drug by the hepatic system. Hydralazine and dihydralazine, its structurally similar variant, are both metabolised by liver enzymes, N-acetyltransferase (NAT) and cytochrome P450-1A2, respectively.13 Variants in these two enzymes have been associated with hydralazine-induced hepatitis. Per Manns et al, individuals with variants in NAT leading to decreased enzyme activity or a description of ‘slow acetylator’ phenotype were found to develop hydralazine-induced hepatitis more often. These individuals relied on their intact P450-1A2 enzyme for proper metabolism, but such reliance on one path is believed to increase the potential for liver injury.2 These same ‘slow acetylators’ were also associated with a higher risk of developing anti-DNA antibodies and lupus-like syndrome.17 Finally, metabolism of the drug by P450-1A2 can also produce reactive metabolites that can further induce an immune response leading to DILI.13

Liver biopsy, though not required to diagnose most cases of DILI, was obtained in our case and demonstrated mild inflammation of the portal tracts, mild steatosis and scattered lobular inflammation without hepatocellular injury. These findings were not as distinct as those found in prior studies and help reflect the heterogeneity of biopsy findings associated with hydralazine. Past studies have demonstrated a wide range of histological findings, which can be associated with other disease processes, including drug-induced autoimmune hepatitis and even granuloma formation.18 In one study, liver biopsies from hydralazine-induced liver injury were associated with normal parenchyma but presence of epithelioid granulomas and few giant cells.12 This same study noted complete regression of the granulomas following resolution of the disorder. Conversely, other studies have shown liver biopsy revealing fibrosis and marked inflammation with interface and bridging necrosis.19

Diagnosis of DILI: a closer look

Diagnosing DILI can be challenging due to its multiple aetiologies, manifestations and lack of specific biomarkers that identify it.20 Although there are several biomarkers that are associated with liver injury, including components of the LFTs, autoantibodies, drug metabolites and serum detection of reactive adducts, these are not specific to a particular drug or type of injury.8 These can be taken into account when evaluating if a condition meets criteria, but their presence does not indicate diagnosis.8 Further derangements in LFTs can be a marker of the severity of disease and therefore were followed throughout this case. LiverTox, the National Institutes of Health-sponsored website on drug-induced hepatotoxicity, describes more than 1200 agents (ie, prescription drugs, herbal products, over-the-counter agents and nutritional supplements), which can lead to liver injury.4 Antimicrobials and antiepileptics are some of the most reported reasons for DILI, with antihypertensive and diabetic medications being less common.21–23 Furthermore, as highlighted in this case, many of the symptoms accompanied are non-specific, which can initially make it difficult to attribute them solely to a particular condition. For these reasons, DILI is generally a diagnosis of exclusion.24

Acute DILI can be considered with symptoms lasting less than 6 weeks; subacute, if lasting 6 weeks to 6 months; or chronic, if lasting more than 6 months.25 Further DILI classification is based on the type of injury pattern: hepatocellular, cholestatic or mixed, as described in figure 1.26–28 The pattern type is determined by calculating the R value, which is defined as serum ALT/upper limit of normal (ULN) divided by the serum ALP/ULN (ALT/ALT ULN ÷ ALP/ALP ULN).28 The R values for each DILI pattern are cholestatic injury R≤2, mixed injury R between 2 and 5 and hepatocellular injury R≥5.28 Following calculation of the R value, the Roussel Uclaf Causality Assessment Method (RUCAM) Score can be further used to determine the causality of a DILI.28

Figure 1

Liver injury is classified as cholestatic, mixed or hepatocellular based off the calculated R value. The R value is determined as the serum alanine aminotransferase (ALT)/upper limit of normal (ULN) divided by the serum ALP/ULN (ALT/ALT ULN ÷ ALP/ALP ULN).28 This figure lists the various differentials associated with each type of liver injury. AST, aspartate transaminase; CMV, cytomegalovirus; EBV, Epstein-Barr virus; HSV, herpes simplex virus.

A case of suspected DILI is taken through seven categories in which possible points are assigned. A greater sum of points is correlated with a higher likelihood of DILI.28 These categories are listed in figure 2 and include the onset, duration of injury, risk factors, drug and non-drug causes and a repeat liver injury following drug readministration.28 The total point score can classify DILI as excluded (<1 point), unlikely (1–2 points), possible (3–5 points), probable (6–8 points) or highly probable (>8 points).28 The R value is required for successful calculation of the RUCAM Score and will influence the point scores for the time of liver injury onset, time course and risk factors leading to the injury.28

Figure 2

The following is a stepwise depiction of the diagnosis of drug-induced liver injury. Abnormal liver function tests are often the first step to finding a liver abnormality. The R value is then calculated to determine the type of liver injury. Subsequently, this is used to generate a score in the Roussel Uclaf Causality Assessment Method (RUCAM) assessment, where the greater the score, the higher likelihood that this liver injury resulted from the drug in question.28 ALP, alkaline phosphatase; ALT, alanine aminotransferase; ULN, upper limit of normal.

When first suspecting DILI, an assessment of symptoms and liver enzymes aberrations for RUCAM scoring helps in the assessment for a potential aetiology. The R value and type of liver injury pattern can progress over time, with most acute DILI cases presenting as a hepatocellular pattern and chronic DILI presenting as mixed or cholestatic pattern.26 In our case, as shown in table 1, our patient initially had an R value of 5.5, indicating a hepatocellular injury. Four days later (3 weeks after hydralazine initiation), this liver injury progressed to a mixed hepatocellular and cholestatic injury with an R value of 2.9. Since we did not rechallenge our patient with hydralazine following improvement of her symptoms, our patient scored a 6 on the RUCAM Scale at both time intervals indicating a probable and likely cause of hydralazine-induced liver injury. To exclude other potential aetiologies of liver injury, our patient scored a 2 on the revised original score for autoimmune hepatitis, likely ruling this out.

Table 1

Transition of R value in case presentation

There is no specific treatment for DILI apart from supportive management.8 This is in part due to the numerous aetiologies of DILI. The extent of disease is varied based on the particular drug in question and its impact on the individual, making treatment individualised. In many cases, terminating the usage of the drug can lead to recovery, as seen in our patient.

As the incidence of DILI continues to increase, it is vital to gain more understanding on its aetiologies, pathogenesis, diagnoses and treatments. Developing a standard protocol for DILI remains elusive given the multitude of drug-related aetiologies. While diagnosis of DILI remains complex, multiple diagnostic tools exist and should be used. Like with any medication, it is important to be cognisant of any potential side effects that can arise. While hydralazine is commonly used for its antihypertensive effects, it is imperative that clinicians remain vigilant when prescribing this drug by checking routine LFTs and stopping the offending agent should symptoms or lab abnormalities occur.

Learning points

  • Previous cases of hydralazine-induced liver injury have been reported, but its connection and categorisation under drug-induced liver injury have not been fully discussed or characterised. Many of the symptoms are non-specific, further adding to this difficulty.

  • Hydralazine-induced liver injury has been found to be more prominent in individuals who have a variant in their liver enzymes, NAT and cytochrome P450-1A2. Individuals with a variant in their NAT led to decreased activity, as they relied on the cytochrome P450-1A2 pathway to metabolise hydralazine. It was found that these individuals were more likely to develop hydralazine-induced liver injury, as reliance on one path is believed to increase the chance for liver injury.

  • In order to diagnose drug-induced liver injury, the Roussel Uclaf Causality Assessment Method Scale is used. This involves first determining the kind of liver injury (cholestatic, hepatocellular or mixed) by calculating the R value, followed by a series of questions and point values associated with a diagnosis of drug-induced liver injury. This patient’s presentation, lab values and workup further categorised hydralazine as an instigator of liver injury.

  • There is no current treatment for drug-induced liver injury apart from supportive treatment and removal of the offending agent. It is crucial that clinicians remain vigilant when prescribing hydralazine by checking routine liver function tests and stopping the offending agent should symptoms or lab abnormalities occur.

Ethics statements

References

Footnotes

  • Contributors MB performed the literature review, analysed the case report, created tables/figures and was the major contributor in writing the manuscript. NJB assisted in analysis of the case report, writing of the manuscript and major editing of the manuscript. KC assisted in the analysis of the case report, creating tables/figures and writing of the manuscript. TCK assisted in the analysis of the case report, writing of the manuscript and major editing of the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.