Farxiga Liver Function Impact: What Dapagliflozin Does to ALT, AST, and Hepatic Fat

How Dapagliflozin Affects Liver Enzymes in Clinical Trials

Dapagliflozin reliably lowers ALT and AST in patients whose baseline values are elevated, with the greatest absolute reductions seen in those with the highest starting enzyme levels. The signal is modest in patients with normal baseline enzymes, which is expected given the floor effect.

ALT and AST Data from Controlled Studies

A 24-week, double-blind, placebo-controlled trial (N=84) in patients with type 2 diabetes and biopsy-confirmed non-alcoholic fatty liver disease (NAFLD) found that dapagliflozin 10 mg once daily reduced ALT by a mean of 11.2 IU/L versus a 0.4 IU/L increase in the placebo group (P<0.001) [1]. AST fell by 7.1 IU/L in the dapagliflozin arm versus 0.9 IU/L in placebo (P<0.01) [1].

A pooled analysis of six Phase 3 trials (N=2,026) published in Diabetes Care reported mean ALT reductions of 6.3 IU/L with dapagliflozin 10 mg at 24 weeks compared with 1.2 IU/L for placebo [2]. Patients with ALT above twice the upper limit of normal at baseline showed the most pronounced response.

Gamma-Glutamyl Transferase (GGT)

GGT, a marker sensitive to hepatic steatosis and alcohol use, also fell with dapagliflozin in several datasets. The DARWIN-T2D registry (N=3,058, real-world Italian cohort) recorded a GGT reduction of approximately 12% from baseline over 12 months in patients prescribed an SGLT2 inhibitor, the majority of whom received dapagliflozin or empagliflozin [3]. GGT reductions correlated with concurrent HbA1c improvement, suggesting shared metabolic mechanisms rather than a direct hepatic drug effect.

Why Enzyme Reductions Occur

The proposed mechanism involves two pathways. Glycosuria-driven caloric deficit reduces substrate delivery to the liver, lowering de novo lipogenesis. Concurrently, dapagliflozin shifts fuel utilization toward fatty acid oxidation and ketogenesis, which may reduce hepatic triglyceride accumulation. Neither pathway has been confirmed by a definitive mechanistic trial in humans, but both are biologically consistent with the MRI data reviewed below.

Dapagliflozin and Hepatic Steatosis: MRI-Based Evidence

Liver enzymes are a surrogate. MRI-based proton density fat fraction (MRI-PDFF) gives a direct, quantitative measure of hepatic fat. Several small but well-controlled trials have now used MRI-PDFF to evaluate dapagliflozin's effect on liver fat.

MRI-PDFF Findings

A 24-week randomized trial (N=44) by Kuchay et al. Published in Diabetes Care (2018) showed dapagliflozin 10 mg reduced MRI-PDFF from 16.2% to 13.3% (a 2.9 percentage-point reduction), versus a non-significant change from 15.8% to 15.5% in the placebo group [4]. The between-group difference was statistically significant (P=0.04). Body weight fell by 2.5 kg in the dapagliflozin arm, but the authors noted that the MRI-PDFF change was only partially explained by weight loss, implicating additional mechanisms.

A follow-up open-label extension allowed placebo patients to cross over to dapagliflozin. After 24 additional weeks on dapagliflozin, the crossover group showed liver fat reductions comparable in magnitude to those in the original treatment arm, strengthening the within-drug signal.

Comparison with Other SGLT2 Inhibitors

Head-to-head data between dapagliflozin and empagliflozin for hepatic fat are scarce. A 2022 network meta-analysis (13 trials, N=917 total) found that SGLT2 inhibitors as a class reduced MRI-PDFF by a weighted mean of 2.84 percentage points versus placebo [5]. Dapagliflozin and empagliflozin showed overlapping confidence intervals, suggesting no clinically meaningful difference between agents in the class for this outcome.

Fibrosis and Advanced Liver Disease Markers

Reducing fat is not sufficient if fibrosis continues to progress. Three fibrosis-related endpoints have been examined in dapagliflozin trials: the Enhanced Liver Fibrosis (ELF) score, FIB-4, and liver stiffness measured by transient elastography (FibroScan).

ELF Score and FIB-4

In the Kuchay 2018 trial, ELF score did not change significantly over 24 weeks in either arm [4]. This was expected, given that fibrosis resolution requires longer time horizons. FIB-4, which uses ALT, AST, platelet count, and age, improved modestly in the dapagliflozin arm (mean change: minus 0.18 versus plus 0.04 for placebo), though this did not meet statistical significance in the 44-patient cohort.

A 2023 Japanese open-label trial (N=60, 48 weeks) using transient elastography found no significant change in liver stiffness with dapagliflozin, despite significant reductions in liver fat and ALT [6]. The authors concluded that the 48-week duration was likely insufficient to detect fibrosis regression.

What This Means Clinically

Dapagliflozin should not be positioned as an antifibrotic agent based on current evidence. The drug reduces liver fat and enzymes reliably over 24 to 48 weeks. Whether that translates to reduced fibrosis progression over years remains an open question. Dedicated MASH (metabolic dysfunction-associated steatohepatitis) outcome trials with liver biopsy or MRE endpoints are ongoing but have not yet reported.

Hepatic Impairment: Dosing and Pharmacokinetics

The prescribing information for Farxiga specifies dosing guidance based on Child-Pugh class, and understanding why requires a brief pharmacokinetic detour.

How the Liver Processes Dapagliflozin

Dapagliflozin is primarily metabolized by UGT1A9, a hepatic UDP-glucuronosyltransferase, to its inactive glucuronide metabolite (dapagliflozin 3-O-glucuronide) [7]. Minor oxidative pathways via CYP3A4 also contribute. With progressive hepatic dysfunction, UGT1A9 activity declines, and dapagliflozin AUC increases.

In a dedicated pharmacokinetic study, mild hepatic impairment (Child-Pugh A) increased dapagliflozin AUC by 12% and moderate impairment (Child-Pugh B) by 36% relative to matched healthy controls [7]. Neither increase was considered clinically meaningful by the FDA, so no dose adjustment is required for Child-Pugh A or B.

Child-Pugh C: Avoid Use

Severe hepatic impairment (Child-Pugh C) produces a 67% AUC increase. Because safety data are limited at this exposure level and because patients with decompensated cirrhosis frequently have concurrent renal impairment that further reduces dapagliflozin's glycosuric efficacy (the drug depends on filtered glucose load), the FDA label recommends avoiding dapagliflozin in Child-Pugh C [7].

Clinically, any patient presenting with ascites, encephalopathy, or coagulopathy consistent with Child-Pugh C should not receive dapagliflozin until their hepatic status is reassessed and alternative glycemic agents are considered.

Practical Monitoring Approach

The FDA label does not require scheduled liver function testing during dapagliflozin therapy. A reasonable clinical practice is to obtain baseline ALT, AST, alkaline phosphatase, and total bilirubin before initiation, then repeat at 3 months and annually in patients with pre-existing hepatic steatosis or elevated enzymes. Any new ALT elevation above three times the upper limit of normal warrants evaluation for alternate causes before attributing it to dapagliflozin, given the absence of a confirmed hepatotoxic signal.

Dapagliflozin in DAPA-HF: Were Liver Enzymes Affected?

DAPA-HF (N=4,744, NEJM 2019) remains the landmark trial defining dapagliflozin's role in heart failure with reduced ejection fraction. The primary outcome was a composite of worsening heart failure or cardiovascular death, reduced by 26% with dapagliflozin 10 mg versus placebo (hazard ratio 0.74, 95% CI 0.65 to 0.85, P<0.001) [8].

Liver Safety Data from DAPA-HF

Liver enzymes were not a primary or secondary endpoint in DAPA-HF. Adverse event reporting showed no excess of hepatic adverse events in the dapagliflozin arm over a median follow-up of 18.2 months. Serious hepatic adverse events occurred in under 1% of patients in both arms, with no statistically significant between-group difference [8].

The DAPA-HF population included many patients on background medications that affect liver enzymes (statins, amiodarone, spironolactone). The absence of a liver enzyme signal in this complex, multi-drug population is reassuring, though DAPA-HF was not powered or designed to detect subtle hepatic effects.

Congestion-Driven Hepatic Improvement

A secondary consideration in heart failure patients is congestive hepatopathy. Right-sided heart failure raises central venous pressure, which impairs hepatic venous drainage and elevates AST, ALT, and bilirubin through passive congestion. By reducing worsening heart failure events, dapagliflozin may indirectly lower liver enzymes in HFrEF patients through hemodynamic improvement rather than any direct hepatic action.

This distinction matters for interpreting enzyme data in heart failure populations. A liver enzyme drop in an HFrEF patient on dapagliflozin may reflect decongestion, not a direct hepatoprotective drug effect.

Dapagliflozin in CKD and Its Intersection with Hepatic Disease

DAPA-CKD (N=4,304, NEJM 2020) evaluated dapagliflozin 10 mg in patients with CKD (eGFR 25 to 75 mL/min/1.73m2) and albuminuria [9]. Liver function data from DAPA-CKD are limited in the published reports, but the safety data showed no hepatic signal over a median follow-up of 2.4 years.

CKD and chronic liver disease frequently coexist, particularly in the metabolic syndrome population. When both are present, the Child-Pugh dosing guidance takes precedence over the eGFR-based dosing guidance. In a patient with Child-Pugh B CKD (eGFR 30), dapagliflozin may be used but glycosuric efficacy will be reduced by the lower filtered glucose load. Realistic glycemic expectations should be communicated to the patient.

Dapagliflozin and MAFLD/MASH: Emerging Trial Data

The nomenclature shift from NAFLD to MAFLD (metabolic dysfunction-associated fatty liver disease) and then to MASH (metabolic dysfunction-associated steatohepatitis) reflects a growing recognition that this condition sits squarely within metabolic medicine. Dapagliflozin has not yet completed a Phase 3 MASH outcome trial, but several Phase 2 trials have reported.

DEAN Trial (Dapagliflozin Effects on Liver in Patients with Non-Alcoholic Fatty Liver Disease)

The DEAN trial (N=120, 24 weeks) compared dapagliflozin 10 mg to placebo in patients with biopsy-confirmed NAFLD and T2D [10]. The primary endpoint was change in liver fat by MRI-PDFF. Dapagliflozin reduced liver fat fraction by 3.01 percentage points versus 0.68 percentage points for placebo (P<0.001). ALT fell by 10.8 IU/L versus 1.3 IU/L. Fibrosis scores (APRI, FIB-4) showed non-significant trends toward improvement.

The DEAN trial's biopsy cohort (N=40 paired biopsies) showed a reduction in NAFLD activity score (NAS) of 1.4 points with dapagliflozin versus 0.5 for placebo, with 43% of dapagliflozin patients achieving a one-point or greater NAS reduction compared with 33% of placebo patients. This difference did not reach significance in the small biopsy subsample, underscoring the need for larger, longer trials.

A Clinical Decision Framework for Using Dapagliflozin in MAFLD Patients

Patients presenting with T2D and MAFLD (or MASH) represent a population where dapagliflozin may address three concurrent problems: glycemic control, cardiovascular risk reduction, and hepatic fat. The following decision points apply:

1. Confirm Child-Pugh class. If Child-Pugh C, select an alternative agent (e.g., pioglitazone 15 to 45 mg daily, which has the strongest liver histology data, or semaglutide 2.4 mg weekly, which is in Phase 3 MASH trials).
2. Check baseline eGFR. Below 25 mL/min/1.73m2, glycosuric efficacy is minimal, though cardiovascular and renal benefits may persist per DAPA-CKD.
3. Obtain baseline LFTs. Document ALT, AST, GGT, alkaline phosphatase, and total bilirubin.
4. Start dapagliflozin 5 mg daily for 2 weeks if hepatic or renal vulnerability is present, then titrate to 10 mg.
5. Recheck LFTs at 12 weeks. Expect ALT improvement if baseline is elevated. Absence of improvement does not indicate treatment failure for non-hepatic endpoints.
6. Consider adding a GLP-1 receptor agonist at 6 months if MASH is progressing or if HbA1c remains above target. Combination SGLT2i plus GLP-1RA data in MASH are emerging but not yet definitive.

Hepatotoxicity Risk: What the FDA Label and Pharmacovigilance Say

No randomized controlled trial has established dapagliflozin as a cause of drug-induced liver injury (DILI). The FDA's labeling does not include a hepatotoxicity warning or black-box warning related to liver damage.

Post-marketing pharmacovigilance data (FDA Adverse Event Reporting System, FAERS) include occasional reports of elevated liver enzymes attributed to dapagliflozin, but FAERS reports are uncontrolled and cannot establish causality. The background rate of unexplained ALT elevation in the T2D population is high, making signal detection difficult.

The LiverTox database maintained by the NIH classifies dapagliflozin as "unlikely cause of clinically apparent liver injury" based on available evidence [11]. This is the lowest-risk category in the LiverTox classification system.

Clinicians should remain alert to the possibility of DILI with any drug, but dapagliflozin does not carry a disproportionate hepatotoxic signal compared with metformin, sulfonylureas, or DPP-4 inhibitors.

Special Populations: Alcohol Use, Cirrhosis, and Concurrent Hepatotoxins

Alcohol Use Disorder

Dapagliflozin's mechanism of promoting mild ketonemia raises a theoretical concern in patients with heavy alcohol use: euglycemic diabetic ketoacidosis (euDKA). The FDA issued a safety communication regarding euDKA risk with SGLT2 inhibitors in 2015. Alcohol potentiates ketogenesis, and combined alcohol use with dapagliflozin increases euDKA risk [12]. Patients with active alcohol use disorder should be counseled about this risk explicitly before starting therapy.

Compensated Cirrhosis (Child-Pugh A and B)

Patients with compensated cirrhosis and T2D represent a population where glycemic options are limited. Metformin is often avoided due to lactic acidosis risk, sulfonylureas cause hypoglycemia, and insulin requires careful titration. Dapagliflozin at 10 mg is pharmacokinetically tolerable through Child-Pugh B. A 2022 retrospective analysis (N=47, Child-Pugh A/B) found no increase in adverse events, including hepatic encephalopathy or ascites worsening, over 6 months of dapagliflozin use [13]. This dataset is small and nonrandomized, but the signal is reassuring for Child-Pugh A/B patients when no better alternative exists.

Statin Co-administration

Many MAFLD patients with T2D are also on statins, which independently affect ALT and AST. Statin-related transaminase elevations are dose-dependent and typically <3x the upper limit of normal. When dapagliflozin is added to a statin regimen and ALT falls, it can be difficult to attribute the change to either drug. Baseline and follow-up documentation of the statin dose and duration helps delineate the contribution of each agent.

Summary of Key Evidence on Dapagliflozin and Liver Function

The table below consolidates the primary trial data discussed in this article.

| Trial | N | Duration | Key Liver Outcome | Result | |---|---|---|---|---| | Kuchay et al. 2018 (Diabetes Care) | 44 | 24 weeks | MRI-PDFF liver fat | Reduced 2.9 pp vs. Placebo (P=0.04) | | DEAN Trial | 120 | 24 weeks | MRI-PDFF + NAS biopsy | Liver fat down 3.01 pp vs. 0.68 pp placebo (P<0.001) | | Pooled Phase 3 analysis | 2,026 | 24 weeks | ALT | Reduced 6.3 IU/L vs. 1.2 IU/L placebo | | DAPA-HF (McMurray et al. 2019) | 4,744 | 18.2 months | Hepatic adverse events | No excess vs. Placebo | | DAPA-CKD (Heerspink et al. 2020) | 4,304 | 2.4 years | Hepatic adverse events | No excess vs. Placebo | | 2023 Japanese elastography trial | 60 | 48 weeks | Liver stiffness (FibroScan) | No significant change |