Registry Data Shows Liver Risk With Avacopan Save

A new study may add fuel to the fire already burning around avacopan (Tavneos), approved in 2021 to treat anti-neutrophil cytoplasmic autoantibody (ANCA)-associated vasculitis (AAV), which the FDA wants drugmaker Amgen to pull from the U.S. market.

Among 238 patients in a U.S.-based rheumatology registry who received avacopan for AAV, a total of 14 instances of clinically significant elevations in liver enzymes were recorded, and two patients experienced drug-induced liver injury (DILI) that was highly likely to stem from avacopan, according to Eric T. Roberts, PhD, MPH, of the University of California San Francisco, and colleagues.

One of those DILI cases was severe, the researchers noted in their report appearing Monday in Arthritis & Rheumatology. That patient and the other with less severe DILI recovered within about 2 months of stopping the drug, and neither developed vanishing bile duct syndrome (VBDS).

Avacopan is an oral C5a complement inhibitor and has sold well, with an estimated $459 million in global sales for Amgen last year. But it is now mired in controversy. In February 2026, the FDA said it had identified liver risks that went beyond the warning of potential hepatotoxicity already listed in the drug's label. It also said questions had arisen about data integrity in the principal randomized trial supporting the drug's approval, called ADVOCATE, and it told Amgen that it should remove avacopan from the market. The next month, the FDA announced that eight people had died from liver injury tied to avacopan, along with dozens of nonfatal DILI cases.

Another shoe dropped in June, when the New England Journal of Medicine retracted the paper that reported ADVOCATE's results. The paper's two academic authors had requested the move when they discovered some data had been changed after unblinding, apparently by authors affiliated with ChemoCentryx, the company that had developed avacopan before selling its U.S. rights to Amgen.

Amgen, however, has thus far refused to take avacopan off the U.S. market, even after the European Union revoked its authorization earlier this month. As of yet, the FDA has not issued a mandatory withdrawal order of its own.

The new report is likely to bolster concerns about hepatotoxicity, insofar as it involves patients receiving the drug in routine practice, based on relatively systematic data collected in the American College of Rheumatology's RISE registry. More than 1,000 U.S. rheumatology practices contribute to RISE -- about 25% of the total -- covering an estimated 3.5 million patients. Roberts and colleagues searched RISE for patients with at least one recorded prescription for avacopan from October 2021 to March 2025, along with an AAV diagnosis as indicated by ICD-9/10 codes, coming up with 238.

Liver enzyme elevations were identified mostly using the FDA's standard definitions: three times the upper limit of normal for aspartate and alanine aminotransferase and twice the upper limit of normal for alkaline phosphatase. The FDA also considers elevations in bilirubin as an important marker, but has not specified a threshold; therefore, the study authors defined it as >2 mg/dL, which is twice the upper limit of normal at most clinical labs, they said.

Patients having at least two enzymes elevated were then evaluated for DILI, which involved contacting their treating clinicians to obtain more detailed information. The clinicians were also asked to complete the so-called RUCAM scoring form that sets the probability that a specific drug is responsible.

The two DILI cases found this way both involved middle-aged women with the granulomatosis-with-polyangiitis subtype of AAV, who developed hepatotoxicity after 9 weeks on avacopan. Their RUCAM scores were 9, indicating high likelihood that avacopan was the cause. Fortunately, neither progressed to VBDS, a life-threatening complication in which bile ducts are permanently lost, which had been noted in some adverse event reports submitted to regulators.

Roberts's group noted that the rates of enzyme elevations were comparable to those seen with methotrexate. But overt DILI is much rarer with that drug.

Another important finding reflects back on the participating rheumatologists: Roberts and colleagues found that liver testing was performed a median of 32 days (IQR 9-68) after starting avacopan, with a median interval for subsequent tests of 125 days (IQR 76-208). In other words, one-quarter of rheumatologists treating these patients waited more than 2 months for the first test, and one-quarter let the interval for subsequent tests to stretch past 6 months. "Our data suggest U.S. clinicians are not screening as frequently on average as recommended by the FDA," the researchers observed; the agency has said liver panels should be administered every 2 weeks during the first month and then monthly through month 6.

In concluding, Roberts and colleagues stopped short of calling for avacopan's withdrawal. "Avacopan represents an important therapeutic option for patients with AAV," they wrote, "however these risks should be thoroughly discussed with patients and routine monitoring timelines, particularly in the first 3 months, should be scrupulously adhered to."

John Gever was Managing Editor from 2014 to 2021; he is now a regular contributor.

Disclosures

The study was funded by the Agency for Healthcare Research and Quality and the National Institute of Arthritis and Musculoskeletal and Skin Diseases.

One co-author reported relationships with Amgen (avacopan's manufacturer), AstraZeneca, Gilead, and Aurinia. Other authors including Roberts declared they had no relevant relationships with commercial entities.

Primary Source

Arthritis & Rheumatology

Source Reference: Roberts ET, et al "Hepatotoxicity of avacopan: real-world incidence and confirmed cases of severe drug-induced liver injury in a US national rheumatology registry" Arthritis Rheumatol 2026; DOI: 10.1002/art.70304.