Should the theory of methotrexate-induced liver toxicity be abandoned?

Table 1 Published cohorts evaluating liver fibrosis in methotrexate-treated patients

Ref.	Number of patients	Findings	Risk of fibrosis
Aithal et al[1], 2004	69	Fibrosis risk: 0% at 1500 mg, up to 32% at > 10000 mg; no correlation between cumulative dose and fibrosis progression	No; no correlation between cumulative MTX dose and fibrosis progression (r = 0.052; P = 0.65)
Zachariae et al[2], 1980	328 patients, 764 biopsies	Cirrhosis in 6.4%; 25.6% in patients treated > 5 years	Yes; strong dose-response observed; fibrosis at doses < 2 g in some cases
Roenigk et al, 1982[3]	Approximately 800 psoriasis patients (summary of multiple biopsy series)	24%-34% any fibrosis; approximately 3% cirrhosis overall (0%-6% across series); higher rates only at > 4 g	Moderate fibrosis in historical high-dose cohorts; advanced fibrosis rare and concentrated in high-dose + risk-factor patients
Chalmers et al[4], 2005	253	11 biopsied due to elevated PIIINP; 3 mild fibrosis, similar rate to routine biopsy controls	Low overall risk with once-weekly MTX regimens under PIIINP monitoring
Berends et al[5], 2007	24	Fibrosis: F0 (21%), F1 (54%), F2 (17%), F3-F4 (8%); FibroTest and FibroScan effectively identify fibrosis; no clear MTX dose association	No; no clear association between cumulative MTX dose and presence of fibrosis
Bray et al[6], 2012	21	9.5% liver fibrosis (F3); 90.5% had NAFLD/NASH	It was difficult to determine whether the fibrosis was due to MTX itself or a combination with NAFLD/metabolic syndrome
Kim et al[7], 2015	185	4.9% had liver stiffness > 8.6 kPa; no significant correlation with MTX dose, but high BMI correlated	No; only a high body mass index but not the cumulative MTX dose was associated with substantial liver fibrosis
Atallah et al[8], 2023	999	Elevated liver stiffness in 15.3%; elevated ELF in 29.4%. No significant association with cumulative MTX dose or duration. Fibrosis strongly linked to metabolic factors (BMI, diabetes), not MTX	Low (MTX risk likely overestimated)
Gelfand et al[9], 2021	40237 (PsO 5687; PsA 6520; RA 28030)	PsO/PsA MTX users had higher rates of liver disease and cirrhosis than RA MTX users; differences persisted after adjustment for MTX dose and comorbidities; RA had highest cumulative dose but lowest liver disease	Higher liver disease/cirrhosis risk in PsO/PsA vs RA, independent of MTX dose; supports role of underlying disease/metabolic factors rather than intrinsic MTX fibrogenicity
European Association for the Study of the Liver et al[12], 2016	Not applicable (clinical guideline)	MTX is not identified as a primary hepatotoxic agent in NAFLD	MTX not considered a relevant risk factor for fibrosis in this guidance
Chalasani et al[10], 2018	Not applicable (clinical guideline)	Methotrexate is not identified as a primary cause of fibrosis in NAFLD	MTX not listed among hepatotoxic agents relevant to NAFLD progression
Lertnawapan et al[13], 2019	108	Mild/moderate fibrosis in 10.3%; safe short-term MTX use	Yes; the study shows that MTX itself is an independent risk factor for liver fibrosis, but that the risks are amplified by concurrent metabolic influence (obesity, fatty liver)
Pongpit et al[14], 2016	165	Significant fibrosis in 10.9%; no correlation with cumulative MTX dose; associated with metabolic factors	No
Martyn-Simmons et al[15], 2014	27	Mild fibrosis: 63%, moderate to severe: 11%, cumulative dose (mean): 7530 mg	Methotrexate-induced liver fibrosis remains a complication that can limit the use of methotrexate for psoriasis
Maybury et al[16], 2014	429 patients (from 8 biopsy-based studies)	Pooled RD: 22% for any fibrosis, 9% for significant fibrosis, 4% for cirrhosis	Yes - increased fibrosis risk; no clear dose-effect; limited by study heterogeneity
Weinblatt et al[17], 1992	26	0% fibrosis at 24 months; mild fibrosis in 1 patient at 48 months and 72 months	No, low; “there have been no cases of either moderate fibrosis or cirrhosis in this cohort”. “Results from the liver biopsies are reassuring”
Arias et al[18], 1993	16	Fibrosis in 9.1% biopsies; 1 mild, 1 moderate-severe	No, “those patients with the shortest duration of therapy were actually those with the most severe liver abnormalities on biopsy, although this inverse relationship was not statistically significant”
Boffa et al[19], 1996	87	21% fibrosis, 4% cirrhosis, 30% inflammation at initial biopsy	Low, “the risk of significant liver damage from low-dose, once-weekly MTX therapy for psoriasis is low”
Richard et al[20], 2000	57	33.8% mild fibrosis by Roenigk; 94.6% fibrosis (48.6% mild, 41.8% moderate, 4% severe) by SSS. No progression at 2 g cumulative MTX dose	No, “neither the Roenigk score nor the SSS demonstrated any progression of hepatic fibrosis in patients having received 2 g of MTX, or showing abnormal levels of transaminases”
Lémann et al[21], 2000	49	11 biopsied; 3 normal, 5 mild steatosis, 1 mild portal fibrosis, no advanced fibrosis over median 18 months	No, MTX-related liver fibrosis was rare and mild in this cohort. No progressive or advanced fibrosis was reported, despite MTX use over a median of 18 months (range 7-59 months)
Maurice et al[22], 2005	38	Fibrosis in 13% patients; progression in 16% paired biopsies at median dose of 5960 mg	Low overall fibrosis risk under modern weekly low-dose regimens: Median cumulative MTX dose at time of fibrotic biopsy: 5960 mg; whole cohort median: 2740 mg; advanced fibrosis was more likely in patients with: Higher cumulative MTX dose
Laharie et al[23], 2006	62	1 cirrhosis, 3 mild fibrosis; no difference by MTX dose	No, “there was no significant difference between patients in group 1 (high dose) and group 2 (naive) and the two groups were comparable according to the CDAI”
Halonen et al[24], 2006	16	19% mild fibrosis; no dose-response relationship with MTX	No, “no minimum cumulative MTX dose as a potential indicator for liver fibrosis was detected”. “MTX or its metabolites did not predict histological liver disease”
Carneiro et al[25], 2008	13	Minimal fibrosis < 2 g cumulative MTX; increased fibrosis ≥ 3 g cumulative dose	Yes, low risk up to 2 g cumulative MTX dose - no clear signs of new or advanced fibrosis. Increased risk at ≥ 3 g - several patients developed portal fibrosis, fibrous septa, and regenerative nodules
Arena et al[26], 2012	100	Liver stiffness strongly correlated with cumulative MTX dose; fibrosis observed with doses > 4000 mg and LS > 9 kPa	Yes, increased liver stiffness was clearly correlated with higher cumulative MTX dose, even after adjustment for other factors. Fibrosis was only detected in patients with LS > 9 kPa and high MTX dose (> 4000 mg)
Barbero-Villares et al[27], 2012	49	Advanced fibrosis in 7.5%; no association with MTX duration or cumulative dose	No, “regarding LF development, MTX therapy is safe”
Dubey et al[28], 2016	204	Biopsy-confirmed fibrosis in 2 patients (1%); no correlation with cumulative MTX dose up to 8 g	No, no statistical association between MTX dose and liver damage, even at doses up to approximately 8 g
Bordbar et al[29], 2018	78	MTX use not associated with increased fibrosis risk; no difference with chronic hepatitis	No
Cervoni et al[30], 2020	131	Higher cumulative MTX dose independently associated with fibrosis; amplified risk with obesity/fatty liver	Yes
Rivera et al[31], 2022	457	Hepatic steatosis in 64.1%; advanced fibrosis risk in 26.2%-37.2%, significantly correlated with longer MTX exposure, metabolic syndrome, and alcohol use	Moderate (increases with treatment duration)
Wong et al[32], 2024	228	Significant fibrosis in 26.5%; correlated with obesity and diabetes, not MTX dose/duration	No, “liver fibrosis was not correlated with the total cumulative dose of MTX or duration of MTX use, but was significantly correlated with obesity and diabetes status”

“Number of patients” refers to the analysed cohort size. “Findings” summarize the main liver assessment results. “Risk of fibrosis” is reported as defined in each study (biopsy stage or non-invasive thresholds). MTX: Methotrexate; PIIINP: Procollagen III aminoterminal propeptide; NAFLD: Non-alcoholic fatty liver disease; NASH: Nonalcoholic steatohepatitis; BMI: Body mass index; ELF: Enhanced liver fibrosis; PsO: Psoriasis; PsA: Psoriatic arthritis; RA: Rheumatoid arthritis; CI: Confidence interval; RD: Risk difference; SSS: Scheuer scoring system; CDAI: Clinical disease activity index; LS: Liver stiffness; LF: Liver fibrosis.

Table 2 Non-invasive monitoring: How to deal with common pitfalls

Tool/scenario	Common pitfalls	What to do (mitigation)	When to escalate
Transient elastography/shearwave elastography	False-highs with acute inflammation, cholestasis, or postprandial state; obesity/central adiposity lowers reliability; device/probe cut-off variability; suboptimal IQR/median	Repeat in fasting state; use correct probe (XL if BMI/skin-capsule distance warrants); ensure quality metrics (e.g., IQR/median ≤ 30%, ≥ 10 valid shots); standardize device/protocol; interpret with ALT, platelets, clinical context	If kPa remains high on repeat under proper conditions and discordant with labs/clinical picture - consider turnover panel or specialist review
Serum scores (FIB-4, APRI, ELF, PIIINP)	Age-dependence; low PPV for moderate fibrosis in low-prevalence settings; systemic inflammation elevates components and may over-call risk	Use rule-out/rule-in bands; re-test when disease activity is quiescent; combine with elastography for concordance	Concordant high risk across two modalities or rising trend despite risk optimization, escalate work-up or therapy discussion
Discordance management	Enzymes high but TE low; TE high with normal labs	Consider MASLD activity, extrahepatic drivers, transient flares. Confirm fasting/probe; repeat TE; consider turnover markers before biopsy	Persisting discordance after optimized repeats, and where biopsy would change management
Follow-up cadence/interpretation	Over-reliance on single thresholds during changing clinical status	Prefer trendbased interpretation (bi-directional movement over time) with standardized timing/conditions	Escalate if sustained upward trends in TE and/or serum scores occur despite riskfactor optimization

Quality criteria for transient elastography commonly include interquartile range/median ≤ 30% and ≥ 10 valid measurements; thresholds vary by device/probe and clinical context. IQR: Interquartile range; BMI: Body mass index; ALT: Alanine aminotransferase; FIB-4: Fibrosis-4 index; APRI: Aspartate aminotransferase/platelet ratio index; ELF: Enhanced liver fibrosis test; PIIINP: Procollagen III aminoterminal propeptide; PPV: Positive predictive value; TE: Transient elastography; MASLD: Metabolic dysfunction-associated steatotic liver disease.

Table 3 Stepwise algorithm for liver fibrosis risk stratification and monitoring in methotrexate-treated patients

Step	Criteria and interpretation (key thresholds)	Actions and followup
Baseline risk	Alcohol intake; metabolic risk: BMI/waist, diabetes, lipids, blood pressure. Review co-medications (polypharmacy); check ALT/AST, platelets, eGFR. Known chronic liver disease (viral hepatitis, cholestatic/autoimmune), significant alcohol use, or CKD start in a higher-risk lane	Place the patient in the appropriate risk lane before NITs
Primary triage (FIB-4)	Calculate FIB-4 for steatosis risk or unexplained enzyme elevation. Interpretation: FIB-4 < 1.3 (low risk). If age > 65: < 2.0 (low risk). FIB-4 ≥ 1.3 (or ≥ 2.0 if > 65 years) proceed to secondary assessment	Low risk: Continue MTX and risk-factor optimization; repeat FIB-4 every 2-3 years (every 1-2 years with diabetes or > 2 metabolic risk factors)
Secondary assessment	Use VCTE first (fast ≥ 3 hours; correct probe; document IQR/median). VCTE: < 8 kPa (low risk); 8-12 kPa (indeterminate); > 12 kPa (high likelihood of advanced fibrosis). Alternatives/complements: ELF (rule-out < 9.8; consider cirrhosis risk if ≥ 11.3), MRE (≥ 3.63 kPa suggests advanced fibrosis)	Repeat if suboptimal conditions; if uncertainty persists, consider hepatology consultation
Follow-up - low risk	FIB-4 < 1.3 and VCTE < 8 kPa	Continue MTX; optimize metabolic risk. Recheck FIB-4 in 1-3 years; repeat VCTE if clinical status changes
Follow-up - intermediate risk	FIB-4 ≥ 1.3 or VCTE 8-12 kPa or ELF 9.8-11.2	Continue MTX; intensify metabolic management. Repeat NITs in 6-12 months
Follow-up - high risk	VCTE > 12 kPa, or ELF ≥ 11.3, or MRE ≥ 3.63 kPa, or falling platelets	Discuss with hepatology. Consider dose reduction/alternative therapy; evaluate for portal hypertension. Initiate HCC surveillance if cirrhosis is suspected
Switch therapy despite mild fibrosis	Consider switching from MTX (or reducing dose) if either of the following occurs despite optimized metabolic care: ≥ 20% increase in VCTE/MRE with the final value in or approaching the indeterminate band (e.g., 6.5-8.0 kPa), confirmed on repeat testing in strict fasting; persistent ALT/AST elevations ≥ 2 × ULN on ≥ 2 tests ≥ 4 weeks apart without another cause	Switch or reduce MTX after risk-benefit discussion
Biopsy and referral triggers	Non-invasive tests discordant (e.g., FIB-4 high but VCTE low). Confirmed VCTE ≥ 12 kPa or ELF ≥ 11.3. New cytopenias or synthetic dysfunction; persistent clinical suspicion despite equivocal NITs	Refer to hepatology; consider liver biopsy

Vibration-controlled transient elastography quality: Interquartile range/median < 30%, ≥ 10 valid shots; correct probe (XL for high body mass index). Prefer fasting when possible; standardize timing and conditions between measurements. Interpret results with clinical context (alanine aminotransferase, platelets). BMI: Body mass index; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; eGFR: Estimated glomerular filtration rate; CKD: Chronic kidney disease; NIT: Non-invasive test; FIB-4: Fibrosis-4 index; MTX: Methotrexate; VCTE: Vibration-controlled transient elastography; IQR: Interquartile range; ELF: Enhanced liver fibrosis; MRE: Magnetic resonance elastography; HCC: Hepatocellular carcinoma; ULN: Upper limit of normal.