Noninvasive Tests to Monitor Methotrexate‐Induced Liver Injury

Watch a video presentation of this article

Answer questions and earn CME

Abbreviations

ALT

alanine aminotransferase

AST

aspartate aminotransferase

BMI

body mass index

IBD

inflammatory bowel disease

MRE

magnetic resonance elastography

MTX

methotrexate

NAFLD

nonalcoholic fatty liver disease

P3NP

procollagen III peptide

PsA

psoriatic arthritis

PsO

psoriasis

RA

rheumatoid arthritis

SWE

shear wave elastography

TE

transient elastography

Case

You are performing a clinic consultation for a 35‐year‐old woman with psoriatic arthritis (PsA) receiving methotrexate (MTX) for 5 years and with elevated liver tests. What are the indications and noninvasive options for fibrosis testing in patients who are taking MTX?

Discussion

Initially developed as a chemotherapeutic agent in the 1950s, low‐dose weekly MTX, given either orally or subcutaneously, has become a mainstay in the treatment of a variety of autoimmune diseases including psoriasis (PsO), PsA, and rheumatoid arthritis (RA). Although MTX overall has good efficacy and safety, hepatotoxicity has been well described.

Although precise mechanisms of MTX‐induced hepatotoxicity are unknown, a variety of both folate‐dependent and ‐independent pathways have been theorized1, 2 (Fig. 1). There have been differing viewpoints in the literature about whether the type of rheumatic disease (i.e., PsO/PsA versus RA) confers differing risk for MTX hepatotoxicity, with some studies suggesting a higher risk for patients with PsO/PsA,3 and others suggesting similar risk among rheumatic diseases.4

Figure 1.

Theorized folate‐dependent and ‐independent pathways of MTX‐induced hepatotoxicity.

Patients at high risk for MTX hepatotoxicity are described in Table 1. Recommendations for monitoring have differed between major society guidelines (Table 2). A key difference in rheumatology and dermatology guidelines is the emphasis on liver biopsies for monitoring in the latter, particularly based on cumulative MTX dose, despite a lack of evidence that cumulative MTX dose correlates to the development of hepatic fibrosis. Serum transaminase testing has traditionally been used for screening and as an adjunct to liver biopsy for MTX hepatotoxicity monitoring, the latter particularly in high‐risk groups. Liver biopsy is complicated by both sampling error and procedural risk. Recently, there has been interest in noninvasive testing, including a new generation of serum testing and imaging, which is the focus of this review.

Table 1.

Risk Factors for MTX‐Induced Hepatotoxicity11

History of moderate‐to‐severe alcohol consumption

Persistently abnormal AST/ALT

History of liver disease (including hepatitis)

Family history of liver disease

Diabetes

Obesity

NAFLD

Hepatotoxic drugs (other than MTX)

Hyperlipidemia

Absence of folate supplementation during MTX administration

Table 2.

Key Differences in Hepatotoxicity Monitoring Between Dermatology and Rheumatology Guidelines

Dermatology Guidelines11, 12

Rheumatology Guidelines13, 14

Low‐Risk Patients: Perform routine blood sampling every 1‐3 months to measure serum AST/ALT and albumin. If 5 out of 9 months demonstrate aminotransferase elevation, a liver biopsy is recommended. Liver biopsy should also be considered every time MTX reaches a cumulative dose of 3.5‐4.0 g.* High‐Risk Patients: Pretreatment liver biopsy and baseline laboratory tests are recommended. Patients with increased BMI and risk for NAFLD should have pretreatment liver ultrasound to confirm. For continued surveillance, liver biopsy is recommended again after 2‐6 months of treatment. Liver biopsy should also be considered every time the MTX reaches a cumulative dose of 1‐1.5 g. In high‐risk patients who require continued MTX therapy, a hepatologist should be consulted for comanagement.

Both Low‐ and High‐Risk Patients: During the first 2 months of therapy, blood samples should be drawn at 4‐ to 8‐week intervals to measure aminotransferases (AST/ALT) and serum albumin. At 3‐6 months of MTX therapy, serum AST/ALT and albumin should be measured at 8‐ to 12‐week intervals. At greater than 6 months of therapy, serum AST/ALT and albumin levels should be measured at 12‐week intervals. If an abnormal value is obtained, the physician should decrease the MTX dose and repeat the blood tests in 2‐4 weeks. If repeat samples show persistent elevation of AST/ALT greater than twice the upper limit, MTX should be held for 1‐2 weeks. If transaminase levels continue to be elevated, the clinician should seek additional cause for hepatotoxicity. If the patient is not at high risk and AST/ALT elevations persist, then MTX should be discontinued. High‐Risk Patients: Pretreatment liver biopsy and baseline laboratory tests are recommended. Patients with increased BMI and risk for NAFLD should have pretreatment liver ultrasound to confirm. For continued surveillance, the above protocol is still recommended for high‐risk patients; however, the clinician should monitor the patient’s AST/ALT and albumin levels at shorter intervals. Liver biopsy is not recommended if the above protocol is followed.

^*Notably, before the 2009 American Academy of Dermatology guidelines, liver biopsy was recommended for all patients with 1.5 g cumulative MTX dose.

Serum testing of procollagen III peptide (P3NP) continues to be used in Europe as an adjunct to routine transaminase testing, and serial testing is recommended in a number of European dermatology guidelines for MTX monitoring in PsO. As a marker of extracellular matrix turnover, levels of P3NP can be elevated in hepatic fibrosis. A lack of commercial availability in the United States and a lack of specificity because levels can be falsely elevated in other disease states including PsA limit its use as a sole marker for routine MTX monitoring, and it is not supported by expert opinion.5

Further serum testing can be accomplished by a panel of five distinct markers (α₂‐macroglobulin, haptoglobin, apolipoprotein A‐I, γ‐glutamyltranspeptidase, and total bilirubin) entered into an algorithm with patient factors including age and gender; this is available commercially in the United States as FibroSURE and in Europe as FibroTest. Because FibroSURE testing has been validated as a predictor of fibrosis in other disease states, notably hepatitis C and nonalcoholic fatty liver disease (NAFLD), a variety of studies have investigated its use for MTX monitoring in autoimmune diseases including PsO and PsA.

Finally, advanced imaging techniques such as transient elastography (TE), shear wave elastography (SWE), and magnetic resonance elastography (MRE) have been developed as adjunctive markers for hepatic fibrosis.6 TE has been the most extensively studied in the MTX population. TE uses an ultrasound‐based technique that measures propagation of a shear wave throughout the liver as a surrogate of liver stiffness. TE, however, can be less reliable in patients with obesity, a known comorbid condition in patients with PsO and PsA. MRE can be more reliable in obese patients, but is significantly more expensive. SWE is a newer technique that also uses ultrasound but with higher frequencies, and perhaps higher performance, as compared with TE, but it has less published data in the MTX population. A summary of studies of imaging and blood markers in MTX‐treated patients is provided in Table 3.

Table 3.

Summary of Key Selected Studies

Assay	Key Illustrative Studies	Patient Population	Study Details	No. of Patients	Conclusions
TE	Barbero et al.15	IBD on MTX	Cross‐sectional observational; no liver biopsy	46	METAVIR ≥F2 by TE was rare Neither IBD type nor MTX cumulative dose was associated with TE fibrosis score
TE and FibroSURE	Berends et al.7	PsO on MTX	Cross‐sectional observational, liver biopsy within 18 months of TE and FibroScan	24	For the detection of METAVIR ≥F2 fibrosis on biopsy: FibroSURE was more sensitive 83% (specificity 61%) TE was more specific at 88% (sensitivity 50%)
	Laharie et al.16	Autoimmune disorders (RA, PsO, and others) on MTX	Case‐control; liver biopsy in minority (n = 13) of patients	390 cases (versus 128 controls)	Neither TE nor FibroSURE was elevated in cases as compared with controls Only BMI and alcohol consumption correlated with severe liver fibrosis on TE on multivariate analysis
TE, FibroSURE, and P3NP	Lynch et al.9	PsO on MTX	Cross‐sectional observational, liver biopsy in minority (n = 5) of patients with elevated P3NP	77	Obesity correlated with invalid TE results Of five liver biopsies with elevated P3NP levels, three did not have fibrosis (TE and FibroTest were normal in these three patients) Authors suggest liver biopsy if two out of three abnormal tests between P3NP, TE, and FibroSURE as strategy for further validation in prospective trials
SWE	Kim et al.17	RA on MTX; compared with healthy controls	Case‐control; no liver biopsy	185 (versus 81 healthy controls)	Fibrosis scores for cases were significantly higher than healthy controls BMI, but not cumulative MTX dose, correlated with fibrosis scores

Many studies have examined serum and imaging markers, sometimes in combination, for the detection of liver fibrosis. In this review, we focus on two key studies in the MTX‐treated PsO population. Berends et al.7 performed a study involving 24 patients with PsO with routine liver biopsies based on cumulative MTX dose who also had both TE and FibroSURE blood testing within 18 months of biopsy. This study found complementary roles for testing, with the FibroSURE blood testing being highly sensitive at 83% (specificity 61%), and TE being highly specific at 88% (sensitivity 50%) for the detection of METAVIR ≥F2 fibrosis on biopsy. Obesity was noted to obscure results of TE testing. Of additional interest was that only one patient in the study had blood aspartate aminotransferase (AST) or alanine aminotransferase (ALT) elevated beyond twice the upper limit of normal, as well as the poor correlation of cumulative MTX dose with liver fibrosis on biopsy. This study does contradict a systematic review of FibroSURE blood testing in hepatitis C–related fibrosis, where specificity was noted to be 90% and sensitivity 47%.8 The main limitation of Berends et al.’s7 study was its small sample size.

A second study by Lynch et al.9 investigated 77 PsO patients with TE, as well as FibroSURE and P3NP blood testing in the majority of these patients, with five liver biopsies done in the setting of elevated P3NP levels. All five of the liver biopsies demonstrated steatosis (two of five at least moderate), and two of the biopsies demonstrated some degree of fibrosis. Specific conclusions were difficult to reach in this descriptive study due to a lack of uniformity in patients receiving all of the testing modalities, but of note for the three liver biopsy patients without fibrosis, both TE and FibroSURE testing were normal. The authors suggested that liver biopsy can be considered only in patients with two out of three abnormal tests between TE, FibroSURE, and P3NP, with a need to evaluate this strategy in prospective studies.

Although both dermatologists and rheumatologists use MTX for the treatment of PsA and PsO, at least one study from the Netherlands suggested differences in practice patterns between types of clinician, with dermatologists using more frequent laboratory monitoring and more resultant drug withdrawal as compared with rheumatologists. There was no difference in serious adverse events or death between the two practice patterns.10

MTX remains an important tool in the management of autoimmune diseases. Even though compelling studies to date have been conducted in the utility of noninvasive monitoring of MTX liver toxicity, clearly more research in larger cohorts with a prospective approach is needed.