Recurrent Drug-Induced Autoimmune-like Hepatitis due to Turmeric Supplementation

Abstract

Drug-induced autoimmune-like hepatitis (DI-ALH) is a rare form of drug-induced liver injury (DILI) that mimics autoimmune hepatitis (AIH). Commonly implicated agents include minocycline, nitrofurantoin, and infliximab, though a wide range of other drugs has been reported. We report a 65-year-old woman with a history of suspected turmeric-induced liver injury six years prior who presented with nausea and painless jaundice two weeks after restarting the same turmeric supplement containing black pepper extract. She denied alcohol or hepatotoxic medication use, and family and personal history of autoimmune or liver disease were unremarkable. Laboratory evaluation revealed marked transaminase elevation (AST 4310 U/L, ALT >2500 U/L), total bilirubin 18.3 mg/dL, and INR 1.1. Autoimmune testing revealed positive smooth muscle antibodies and elevated IgG, while viral, metabolic, and genetic studies were negative. Imaging demonstrated periportal edema. Despite supplement discontinuation, liver enzymes worsened, prompting a liver biopsy, which demonstrated acute hepatitis with portal and lobular inflammation, plasma cell predominance, and canalicular cholestasis without fibrosis consistent with DI-ALH. Corticosteroid therapy led to biochemical improvement over three weeks. The temporal association with exposure, autoimmune serologies, characteristic histology, and steroid responsiveness supported the diagnosis of recurrent turmeric-induced DI-ALH, with more severe injury on re-exposure suggesting immune sensitization. Piperine, included to enhance curcumin bioavailability, may potentiate hepatotoxicity. DI-ALH should be considered in patients with liver injury and recent supplement use, as immunosuppressive therapy may be beneficial when injury persists after drug withdrawal.

Introduction

Drug-induced liver injury (DILI) is a common cause of acute liver failure and an increasingly recognized consequence of both prescribed medications and herbal and dietary supplements. ¹ Among these, immune-mediated phenotypes that closely mimic autoimmune hepatitis (AIH) represent a diagnostic challenge and are described as drug-induced autoimmune-like hepatitis (DI-ALH). ² These entities share overlapping clinical, serological, and histopathological features with AIH, often making differentiation difficult in routine clinical practice. ²

Turmeric-containing supplements, frequently co-formulated with piperine to enhance bioavailability, have been increasingly implicated in hepatotoxicity. ³ While most reported cases of turmeric-associated liver injury present as hepatocellular injury with variable latency, immune-mediated presentations remain rare. ³ Emerging evidence suggests genetic susceptibility factors, including specific human leukocyte antigen (HLA) alleles, may predispose to turmeric-associated liver injury. ⁴ Furthermore, recurrence following re-exposure and steroid responsiveness may support an immune-mediated mechanism, although standardized diagnostic criteria to distinguish DI-ALH from AIH are lacking.^2,5 Distinguishing these entities is clinically relevant, as DI-ALH may not require long-term immunosuppression. ²

We report a case of severe acute hepatitis with autoimmune features following re-exposure to a turmeric supplement containing black pepper extract, with histologic evidence of interface hepatitis and rapid improvement with corticosteroid therapy. This case highlights the diagnostic challenges in distinguishing DI-ALH from AIH and underscores the importance of recognizing herbal supplements as potential triggers of immune-mediated liver injury.

Case Presentation

A 65-year-old woman with a history of presumed supplement-induced DILI six years prior presented with one week of nausea and painless jaundice two weeks after restarting the same turmeric supplement containing black pepper extract. She was taking one tablespoon of the supplement daily. She denied alcohol use, hepatotoxic medication exposure, or recreational drug use, and family history was unremarkable. Physical examination was notable for scleral icterus without stigmata of chronic liver disease.

Laboratory evaluation demonstrated sodium 132 mmol/L, total bilirubin 18.3 mg/dL, direct bilirubin 9.62 mg/dL, alkaline phosphatase 153 IU/L, aspartate aminotransferase (AST) 4310 U/L, alanine aminotransferase (ALT) >2500 U/L, international normalized ratio (INR) 1.1, and platelet count 276 K/uL. HLA typing revealed HLA-A01, HLA-A11, HLA-B08, and HLA-B35. Compared with her prior episode, she demonstrated a more severe biochemical presentation (peak total bilirubin 18.3 mg/dL vs 7.3 mg/dL, Figure 1). Autoimmune evaluation revealed positive smooth muscle antibodies (77 U) and elevated immunoglobulin G (IgG) (1798 mg/dL). Cross-sectional imaging demonstrated periportal edema without biliary obstruction. Viral, metabolic, and genetic evaluation was negative. Despite supplement discontinuation, liver enzymes continued to rise, prompting a liver biopsy on hospital day 4, which demonstrated acute hepatitis with lobular and portal inflammation, numerous plasma cells, and canalicular cholestasis without significant fibrosis. Given concern for DI-ALH and worsening biochemistries, corticosteroid therapy with intravenous methylprednisolone 60 mg daily was initiated before liver biopsy was obtained (on hospital day 4) with transition to oral prednisone starting with prednisone 40 mg daily with taper by 10 mg every 2 weeks. Progressive biochemical improvement over three weeks was observed (Figure 2). She was to have follow-up with gastroenterology closely for monitoring of liver biochemistries.

Figure 1.

Total bilirubin trend from this hospitalization and follow up 20 days after discharge

Figure 2.

AST and ALT trend from this hospitalization and follow up 20 days after discharge

Discussion

AIH typically follows a chronic and progressive course that can lead to fibrosis, liver failure, and death if not treated. ⁶ The global annual incidence and prevalence of AIH is estimated to be 1.37 and 17.4 per 100,000 individuals, respectively, according to a recent meta-analysis. ⁷ DILI is a leading cause of acute liver injury and an increasingly recognized consequence of herbal and dietary supplement use. ¹ In prospective studies, the incidence of DILI has been reported at approximately 14–19 cases per 100,000 people per year. In one cohort of 261 patients with AIH, 24 cases (9.2%) were later reclassified as DI-ALH.^8,9 Conversely, among patients with DILI, about 3–8.8% meet criteria for DI-ALH. ⁵ Reported associations with DI-ALH include agents such as minocycline, nitrofurantoin, methyldopa, hydralazine, statins, ¹⁰ immunotherapies. ¹¹ Herbal and dietary supplements have rarely been implicated in DI-ALH.

Distinguishing DI-ALH from AIH remains challenging, as both conditions share similar clinical, serological, and histological features. ² Diagnostic features supporting DI-ALH include an acute onset, hypersensitivity features (fever, rash, eosinophilia), temporal relationship with drug exposure, improvement after drug withdrawal with or without corticosteroid therapy, absence of cirrhosis, and lack of personal or family history of autoimmune disease. ² Latency period ranges from 1 week up to even exceeding 12 months. ¹² Differentiation of these two conditions is clinically relevant, as immunosuppressants may not be required in DI-ALH and, when used, can often be safely discontinued without recurrence once the offending agent is withdrawn. ⁵ Liver biopsy should be considered if the diagnosis is unclear, if there is evidence of severe injury from laboratory values, or if corticosteroids are being considered for treatment. ² Interface hepatitis with portal and periportal infiltrates of lymphocytes, lobular hepatitis, plasma cells, and eosinophils can be observed in AIH and DI-ALH. ² However, advanced fibrosis or cirrhosis are typically absent in DI-ALH, in contrast to AIH. ¹³ Notably, the patient’s HLA-B*35 positivity may be of potential relevance. Prior studies have reported involvement of HLA-B*35:01 in turmeric-associated liver injury, as well as in other medication-related hepatotoxicity, ⁴ including green tea extract ¹⁴ and Polygonum multiflorum, ¹⁵ suggesting a possible immunogenetic association that warrants further investigation.

The Revised Electronic Causality Assessment Method (RECAM) is an electronic adaptation of the Roussel Uclaf Causality Assessment Method (RUCAM) used to assess causality in DILI, categorizing likelihood from excluded (≤−3) to highly probable (≥8). ¹⁶ Conversely, the Simplified Autoimmune Hepatitis (AIH) score incorporates autoantibodies, serum IgG levels, histology, and exclusion of viral hepatitis, with a score ≥7 favoring definite AIH. ¹⁷ In this case, the RECAM score was 14, favoring highly probable DILI, while the Simplified AIH score was 7 supporting definite AIH. However, both scoring systems should be interpreted as supportive rather than definitive, as no validated diagnostic criteria exist for DI-ALH, and neither was designed to distinguish immune-mediated DILI phenotypes. Furthermore, early glucocorticoid administration may have influenced biochemical evolution and potentially confounded causality assessment. Overall, the clinical, serological, histological, and temporal features in this case are most consistent with an immune-mediated DILI with autoimmune features, DI-ALH, rather than AIH.

Treatment of DI-ALH requires cessation of the offending agent with close monitoring until clinical symptoms and laboratory abnormalities resolve and remain sustained. ¹⁸ Additional management relies on retrospective studies and individualization to the patient. Glucocorticoids can be considered if symptoms persist or if laboratory values worsen or fail to improve despite discontinuation of the implicated agent. ² Hy’s law defines significant hepatocellular injury as aminotransferases >3× upper limit of normal (ULN) with concomitant bilirubin >2× ULN, which is associated with approximately a 9%–12% risk of death or liver transplantation in DILI cohorts. ¹⁹ Glucocorticoid therapy may be warranted in patients who fulfill Hy’s law, ¹⁸ as in this patient. Longer glucocorticoid therapy for 1-2 months can be used in patients with laboratory abnormalities, however total duration of therapy is based on clinical response. ² Dosing is also based on clinical experience of the provider. ²⁰ Faster clinical improvement is observed with DI-ALH (2 months) with treatment with glucocorticoids, similar to this patient with resolution within 3 weeks, compared to AIH (16.8 months). ²¹ However, extended clinical monitoring is required to confirm sustained remission without immunosuppression and to distinguish this entity from AIH, which commonly requires long-term therapy.^2,18

Turmeric has gained widespread popularity as a natural anti-inflammatory supplement and is increasingly recognized among the leading causes of herb- and dietary supplement-associated DILI. ³ Piperine, a component of black pepper commonly added to turmeric supplements to enhance curcumin bioavailability by up to 20-fold, may contribute to hepatotoxicity.^3,22 Turmeric-induced liver injury typically presents after a latency of weeks to months (most often 1-4 months). ²³ Based on our literature review, two cases of DI-ALH were linked to turmeric use, with recovery after discontinuing turmeric and without need for glucocorticoids.^9,24 In our case, re-exposure to turmeric resulted in a more severe recurrence of liver injury compared with the initial episode, providing highly supportive evidence. Rapid biochemical improvement following glucocorticoid therapy further supports an immune mediation. Positive rechallenge is among the most specific indicators of DILI and is rarely observed due to ethical constraints, likely reflecting immune sensitization with an amplified inflammatory response upon re-exposure. ²⁵

Furthermore, classification of this case as DI-ALH is supported by the temporal relationship to turmeric exposure, recurrence upon re-exposure, and autoimmune features on serologic and histologic evaluation. However, the distinction between DILI and DI-ALH remains poorly defined, with substantial overlap in clinical, serological, and histopathological features. Although scoring systems such as the RECAM and Simplified AIH score may aid in assessment, ⁵ no standardized diagnostic criteria exist to reliably differentiate these entities, ² and patients may fulfill criteria for both. ²⁶

This case highlights several important clinical considerations. Herbal and dietary supplements should not be presumed benign and should be included in the differential diagnosis of acute hepatocellular injury. In cases with incomplete biochemical recovery after withdrawal of the suspected agent, liver biopsy may be warranted to evaluate for autoimmune features; however, it is often deferred in typical DILI, potentially contributing to underrecognition of DI-ALH. Recognition of DI-ALH is clinically relevant, as selected patients may benefit from glucocorticoid therapy.

Conclusion

DI-ALH remains a diagnostically challenging entity due to its overlap with AIH in clinical, serological, and histological features. This case illustrates a severe immune-mediated liver injury following turmeric re-exposure in a patient with HLA-B35 positivity, supported by autoimmune serologies, liver biopsy findings, and a robust response to glucocorticoid therapy. To our knowledge, this is among the few reported cases of biopsy-confirmed recurrent turmeric-associated immune-mediated liver injury with documented rechallenge. Herbal and dietary supplements, including widely perceived “natural” products such as turmeric, should be considered in the differential diagnosis of acute hepatocellular injury. Re-exposure may result in more severe immune-mediated injury, underscoring the importance of thorough medication and supplement histories. Early recognition of this phenotype is clinically relevant, as selected patients may benefit from glucocorticoid therapy and avoidance of re-exposure is critical to prevent recurrence. Extended follow-up is warranted to confirm sustained remission off immunosuppression.