Consider steroids in anti-TNF liver injury

Patients with liver injury secondary to anti–tumor necrosis factor–alpha agents often exhibit histological changes similar to those seen in spontaneous autoimmune hepatitis, wrote Dr. Marwan Ghabril and his colleagues. The report was published in the May issue of Clinical Gastroenterology and Hepatology.

The finding comes from a review seeking to characterize the presentation, injury pattern, course of illness, and treatment of anti-TNF agent–induced liver injury, which Dr. Ghabril, of Indiana University in Indianapolis, said "may be severe and prolonged."

To that end, he and his colleagues studied six patients from the U.S. Drug-Induced Liver Injury Network (DILIN) database who had liver damage that was likely secondary to anti-TNFs, according to DILIN criteria.

Additionally, 28 cases were culled from a literature review of the PubMed database using the search terms "hepatotoxicity," "liver injury," "tumor necrosis factor," and the generic names of all TNF-alpha antagonists.

Among all patients, infliximab was most commonly the offending agent, followed by etanercept and adalimumab; no cases were attributed to natalizumab, golimumab, or certolizumab.

Diseases for which the anti-TNFs were prescribed included ankylosing spondylitis, Crohn’s disease, chronic ulcerative colitis, juvenile inflammatory arthritis, psoriatic arthritis, psoriasis, and rheumatoid arthritis.

First, Dr. Ghabril looked at the six patients from the DILIN database. Among these patients, the median duration of drug use before onset of liver injury was 16 weeks (range, 2-52 weeks).

"At presentation, half had jaundice, half had nausea, but only one had fever and none had immuno-allergic features of skin rash or eosinophilia," wrote the authors.

Only one of these patients developed significantly impaired coagulation, with an INR of 3.5, and none of the DILIN patients developed ascites or other signs of hepatic failure, they added.

The peak ALT ranged from 384 to 1,687 U/L, and the peak bilirubin from 1.5 to 27.7 mg/dL.

"Five of the six patients were treated with corticosteroids. One patient had a protracted illness, but all ultimately recovered and could be withdrawn from corticosteroid therapy without recurrence," they wrote.

Next, the authors turned their attention to the PubMed cases.

"Peak serum ALT ranged from 140 to 2,250 U/L and bilirubin from normal to 27.7 mg/dL," the authors reported.

Most patients (n = 22) had autoimmune serological markers and/or histological features at some point during the clinical course, they found.

Twelve patients stopped the drug and initiated corticosteroid therapy, oral or parenteral; all recovered. The rest improved after discontinuation of the anti-TNF, without steroid treatment, except for one patient who had underlying cirrhosis and required liver transplantation.

Notably, several patients were able to continue anti-TNF treatment with another agent without further incident, the authors wrote. Indeed, "three tolerated treatment with etanercept without recurrence of liver injury after cessation of infliximab or adalimumab. Two did well with adalimumab after [drug-induced liver injury] associated with infliximab, and one was successfully switched from adalimumab to abatacept."

Finally, the investigators contrasted patients with serological or histological autoimmune features with patients who lacked any autoimmune characteristics.

The autoimmune patients tended to have a longer latency period. The median period after the drug was initiated but before liver injury was noted was 16 weeks for the autoimmune patients versus 10 weeks for the nonautoimmune patients (P = .17). The autoimmune patients also had a higher peak ALT (median 784 vs. 528; P = .03).

"The mechanism by which the TNF-alpha antagonists lead to drug-induced liver injury is unknown," the authors wrote.

"Because the injury can occur after only one infusion, dose-dependent toxicity is unlikely. Unpredictable, idiosyncratic drug-induced liver injury seems most likely, as in this series no patients had clinical evidence of a rash or eosinophilia, and only one presented with fever."

"Further studies are needed to ascertain whether genetic or other markers of the hepatotoxicity associated with TNF-alpha antagonists can be identified," they said.

The DILIN database is supported by the National Institute of Diabetes and Digestive and Kidney Diseases. The authors reported having no conflicts of interest.

Patients with liver injury secondary to anti–tumor necrosis factor–alpha agents often exhibit histological changes similar to those seen in spontaneous autoimmune hepatitis, wrote Dr. Marwan Ghabril and his colleagues. The report was published in the May issue of Clinical Gastroenterology and Hepatology.

The finding comes from a review seeking to characterize the presentation, injury pattern, course of illness, and treatment of anti-TNF agent–induced liver injury, which Dr. Ghabril, of Indiana University in Indianapolis, said "may be severe and prolonged."

To that end, he and his colleagues studied six patients from the U.S. Drug-Induced Liver Injury Network (DILIN) database who had liver damage that was likely secondary to anti-TNFs, according to DILIN criteria.

Additionally, 28 cases were culled from a literature review of the PubMed database using the search terms "hepatotoxicity," "liver injury," "tumor necrosis factor," and the generic names of all TNF-alpha antagonists.

Among all patients, infliximab was most commonly the offending agent, followed by etanercept and adalimumab; no cases were attributed to natalizumab, golimumab, or certolizumab.

Diseases for which the anti-TNFs were prescribed included ankylosing spondylitis, Crohn’s disease, chronic ulcerative colitis, juvenile inflammatory arthritis, psoriatic arthritis, psoriasis, and rheumatoid arthritis.

First, Dr. Ghabril looked at the six patients from the DILIN database. Among these patients, the median duration of drug use before onset of liver injury was 16 weeks (range, 2-52 weeks).

"At presentation, half had jaundice, half had nausea, but only one had fever and none had immuno-allergic features of skin rash or eosinophilia," wrote the authors.

Only one of these patients developed significantly impaired coagulation, with an INR of 3.5, and none of the DILIN patients developed ascites or other signs of hepatic failure, they added.

The peak ALT ranged from 384 to 1,687 U/L, and the peak bilirubin from 1.5 to 27.7 mg/dL.

"Five of the six patients were treated with corticosteroids. One patient had a protracted illness, but all ultimately recovered and could be withdrawn from corticosteroid therapy without recurrence," they wrote.

Next, the authors turned their attention to the PubMed cases.

"Peak serum ALT ranged from 140 to 2,250 U/L and bilirubin from normal to 27.7 mg/dL," the authors reported.

Most patients (n = 22) had autoimmune serological markers and/or histological features at some point during the clinical course, they found.

Twelve patients stopped the drug and initiated corticosteroid therapy, oral or parenteral; all recovered. The rest improved after discontinuation of the anti-TNF, without steroid treatment, except for one patient who had underlying cirrhosis and required liver transplantation.

Notably, several patients were able to continue anti-TNF treatment with another agent without further incident, the authors wrote. Indeed, "three tolerated treatment with etanercept without recurrence of liver injury after cessation of infliximab or adalimumab. Two did well with adalimumab after [drug-induced liver injury] associated with infliximab, and one was successfully switched from adalimumab to abatacept."

Finally, the investigators contrasted patients with serological or histological autoimmune features with patients who lacked any autoimmune characteristics.

The autoimmune patients tended to have a longer latency period. The median period after the drug was initiated but before liver injury was noted was 16 weeks for the autoimmune patients versus 10 weeks for the nonautoimmune patients (P = .17). The autoimmune patients also had a higher peak ALT (median 784 vs. 528; P = .03).

"The mechanism by which the TNF-alpha antagonists lead to drug-induced liver injury is unknown," the authors wrote.

"Because the injury can occur after only one infusion, dose-dependent toxicity is unlikely. Unpredictable, idiosyncratic drug-induced liver injury seems most likely, as in this series no patients had clinical evidence of a rash or eosinophilia, and only one presented with fever."

"Further studies are needed to ascertain whether genetic or other markers of the hepatotoxicity associated with TNF-alpha antagonists can be identified," they said.

The DILIN database is supported by the National Institute of Diabetes and Digestive and Kidney Diseases. The authors reported having no conflicts of interest.

Patients with liver injury secondary to anti–tumor necrosis factor–alpha agents often exhibit histological changes similar to those seen in spontaneous autoimmune hepatitis, wrote Dr. Marwan Ghabril and his colleagues. The report was published in the May issue of Clinical Gastroenterology and Hepatology.

The finding comes from a review seeking to characterize the presentation, injury pattern, course of illness, and treatment of anti-TNF agent–induced liver injury, which Dr. Ghabril, of Indiana University in Indianapolis, said "may be severe and prolonged."

To that end, he and his colleagues studied six patients from the U.S. Drug-Induced Liver Injury Network (DILIN) database who had liver damage that was likely secondary to anti-TNFs, according to DILIN criteria.

Additionally, 28 cases were culled from a literature review of the PubMed database using the search terms "hepatotoxicity," "liver injury," "tumor necrosis factor," and the generic names of all TNF-alpha antagonists.

Among all patients, infliximab was most commonly the offending agent, followed by etanercept and adalimumab; no cases were attributed to natalizumab, golimumab, or certolizumab.

Diseases for which the anti-TNFs were prescribed included ankylosing spondylitis, Crohn’s disease, chronic ulcerative colitis, juvenile inflammatory arthritis, psoriatic arthritis, psoriasis, and rheumatoid arthritis.

First, Dr. Ghabril looked at the six patients from the DILIN database. Among these patients, the median duration of drug use before onset of liver injury was 16 weeks (range, 2-52 weeks).

"At presentation, half had jaundice, half had nausea, but only one had fever and none had immuno-allergic features of skin rash or eosinophilia," wrote the authors.

Only one of these patients developed significantly impaired coagulation, with an INR of 3.5, and none of the DILIN patients developed ascites or other signs of hepatic failure, they added.

The peak ALT ranged from 384 to 1,687 U/L, and the peak bilirubin from 1.5 to 27.7 mg/dL.

"Five of the six patients were treated with corticosteroids. One patient had a protracted illness, but all ultimately recovered and could be withdrawn from corticosteroid therapy without recurrence," they wrote.

Next, the authors turned their attention to the PubMed cases.

"Peak serum ALT ranged from 140 to 2,250 U/L and bilirubin from normal to 27.7 mg/dL," the authors reported.

Most patients (n = 22) had autoimmune serological markers and/or histological features at some point during the clinical course, they found.

Twelve patients stopped the drug and initiated corticosteroid therapy, oral or parenteral; all recovered. The rest improved after discontinuation of the anti-TNF, without steroid treatment, except for one patient who had underlying cirrhosis and required liver transplantation.

Notably, several patients were able to continue anti-TNF treatment with another agent without further incident, the authors wrote. Indeed, "three tolerated treatment with etanercept without recurrence of liver injury after cessation of infliximab or adalimumab. Two did well with adalimumab after [drug-induced liver injury] associated with infliximab, and one was successfully switched from adalimumab to abatacept."

Finally, the investigators contrasted patients with serological or histological autoimmune features with patients who lacked any autoimmune characteristics.

The autoimmune patients tended to have a longer latency period. The median period after the drug was initiated but before liver injury was noted was 16 weeks for the autoimmune patients versus 10 weeks for the nonautoimmune patients (P = .17). The autoimmune patients also had a higher peak ALT (median 784 vs. 528; P = .03).

"The mechanism by which the TNF-alpha antagonists lead to drug-induced liver injury is unknown," the authors wrote.

"Because the injury can occur after only one infusion, dose-dependent toxicity is unlikely. Unpredictable, idiosyncratic drug-induced liver injury seems most likely, as in this series no patients had clinical evidence of a rash or eosinophilia, and only one presented with fever."

"Further studies are needed to ascertain whether genetic or other markers of the hepatotoxicity associated with TNF-alpha antagonists can be identified," they said.

The DILIN database is supported by the National Institute of Diabetes and Digestive and Kidney Diseases. The authors reported having no conflicts of interest.