This article is not a discussion of treatment decisions or whether avacopan should be prescribed. It is a synthesis of publicly available information drawn from Kissei Pharmaceutical's May 15, 2026 announcement, alongside regulatory communications from the US FDA, the European EMA, and reporting from major news outlets. Individual treatment decisions should always be made in consultation with a qualified physician.

⚠️ Twenty patients in Japan have died after taking a drug for a rare autoimmune disease. The drug — Tavneos, generic name avacopan — was hailed five years ago as the first new treatment in decades for ANCA-associated vasculitis. Causality has not been established in every case, but the Japanese marketer of the drug has urged doctors to stop starting new patients on it, the FDA has moved to pull it from the US market, and Europe's EMA is reviewing the trial data that got it approved.

Here is what's actually known, what's still unclear, and why this story stretches from a pharmaceutical office in Matsumoto, Nagano, all the way to Thousand Oaks, California.

What Kissei announced on May 15

On Friday, May 15, 2026, Kissei Pharmaceutical Co., Ltd. — a mid-size drugmaker based in Matsumoto, in the Japanese Alps — disclosed that 20 patients had died after taking Tavneos in Japan since the drug went on sale in 2022. The company holds the exclusive license to sell avacopan in Japan.

A few numbers anchor the announcement, and they're worth keeping straight.

By April 27, 2026, an estimated 8,500 patients in Japan had been treated with the drug since launch. Among those patients, 22 cases of vanishing bile duct syndrome (VBDS) — a rare and severe form of liver injury in which the bile ducts inside the liver are progressively destroyed — were reported. Thirteen of the 20 deaths fell within that VBDS group. Most of the severe liver injury cases appeared within three months of starting treatment, with a particularly high concentration between days 29 and 56.

Kissei emphasised — and this is important — that the deaths include cases where a direct causal link to Tavneos has not been established. Patients with severe ANCA vasculitis are already seriously ill, often with kidney involvement, and many take immunosuppressants and high-dose steroids on top of avacopan. Separating drug effect from disease effect is genuinely difficult in postmarketing data.

But the pattern was concerning enough that Kissei asked physicians not to start new patients on the drug, and to carefully reassess whether to continue treatment for existing patients. On May 1, 2026, the company added vanishing bile duct syndrome to the "serious adverse reactions" section of the Japanese label, with frequency listed as "unknown." The drug's marketing approval in Japan remains in effect for now, and Kissei said it is in discussions with Japan's Ministry of Health, Labour and Welfare while gathering information from US and European regulators.

The drug and the disease it treats

To make sense of why this matters, you need to know what Tavneos was supposed to do.

ANCA-associated vasculitis (AAV) is a group of rare autoimmune diseases — granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA) being the two main forms — in which the immune system attacks small and medium-sized blood vessels. The damage can hit the kidneys, lungs, nerves, skin, and upper respiratory tract. Japan's Ministry of Health classifies AAV as an "intractable disease" (指定難病), with patient registries exceeding 10,000 people. Worldwide, it remains a serious, often life-threatening condition.

The historical treatment is a heavy regimen: rituximab or cyclophosphamide to suppress the immune system, plus high-dose glucocorticoids — prednisone, essentially — to calm the inflammation. The steroids work, but extended high-dose steroid exposure brings its own roster of harms: bone loss, infections, diabetes, weight gain, mood changes. Doctors have wanted a steroid-sparing option for decades.

Avacopan was that option. It's a selective complement C5a receptor (C5aR) antagonist, given orally as 30 mg twice daily. The mechanism is reasonably elegant: C5a is an inflammatory signal that primes neutrophils to attack blood vessels, and blocking its receptor on those neutrophils interrupts the cycle without broad immune suppression. In the pivotal Phase 3 ADVOCATE trial (331 patients), avacopan plus standard immunosuppression was at least as good as the high-dose steroid arm at inducing remission, and looked better at sustaining it at 52 weeks.

Japan approved Tavneos in September 2021 — the world's first regulatory green light. The US FDA followed in October 2021, and the European Union in January 2022.

How the safety picture changed

The label always warned about liver injury. In the ADVOCATE trial, serious hepatotoxicity was observed, and "hepatotoxicity" sits in the Warnings and Precautions section of the US prescribing information.

What changed after the drug went on the market — particularly in Japan, which had the largest patient population — was that doctors started reporting a specific pattern: not just elevated liver enzymes, but bile ducts being destroyed. Vanishing bile duct syndrome is rare and bad. Patients develop jaundice, intense itching, fatigue, dark urine. The damage can be permanent. In 2024, Amgen — the US company that by then owned the drug — proactively submitted a request to the FDA to add VBDS to the Tavneos label. That request was still pending when the bigger crisis hit.

On March 31, 2026, the FDA issued a Drug Safety Communication. The agency said it had identified 76 cases of drug-induced liver injury (DILI) globally with "reasonable evidence of a causal association" with avacopan, based on data through October 9, 2024. Of those, 74 were serious: 54 hospitalisations and eight deaths. Seven cases were biopsy-confirmed VBDS, three of them fatal. The vast majority of cases — 66 out of 76 — were reported in Japan. Median time to onset was 46 days from the first dose.

Kissei's May 15 announcement updated those numbers with fresher Japanese data: 22 VBDS cases, 20 deaths, roughly a year and a half further along than the FDA's cutoff.

The bigger fight in the US

The liver issue is only half the story. The other half is a regulatory standoff that started before the latest fatalities surfaced.

On January 16, 2026, the FDA quietly asked ChemoCentryx — the original developer of avacopan, which Amgen had bought in October 2022 for around 3.7 billion dollars (roughly 590 billion yen at current rates) — to voluntarily withdraw Tavneos from the US market. The reason wasn't safety. It was data integrity: the FDA had concerns about how ChemoCentryx re-adjudicated the primary endpoint for 9 of the 331 patients in the ADVOCATE trial.

Twelve days later, Amgen formally refused. The company said it remained confident in the drug's benefit-risk profile and would not pull it.

Then the safety communication landed in March. Then, on April 27, 2026, the FDA escalated. In a formal letter, the Center for Drug Evaluation and Research stated it "can no longer conclude that there is, or has ever been, a valid demonstration of substantial evidence of effectiveness for TAVNEOS" — and proposed withdrawing the approval outright, citing both the data integrity questions and the worsening liver-safety picture. Withdrawing a drug over an applicant's objection is rare and procedurally slow; Tavneos remains on the US market for now, but the agency has made its position explicit.

Tavneos is one of Amgen's faster-growing products: 459 million dollars (about 73 billion yen) in 2025 sales, up 62% from the previous year, with peak forecasts that had climbed past 1 billion dollars by 2031. Amgen shares fell 2.1% on Friday after Kissei's announcement.

Europe is also looking

The European Medicines Agency started its own review at the European Commission's request in late January 2026, after the same data integrity concerns about ADVOCATE that triggered the FDA's request. The CHMP, EMA's human medicines committee, will decide whether to maintain, amend, suspend, or revoke Tavneos's EU marketing authorisation. As of mid-May, that review is ongoing. The UK's MHRA has not announced a parallel review.

It's worth noting what European patient groups have emphasised: the EMA review is about data, not new safety signals. The Drug Safety Communication from the FDA, and now Kissei's Japanese update, change that framing. The data fight and the safety fight have now merged.

Why so many of the cases came from Japan

One question that hangs over this story: why does Japan, with a much smaller patient population than the US, account for the overwhelming majority of reported cases?

A few plausible factors, none of which has been formally established:

Japan was first to approve Tavneos and built up the largest post-approval patient base — 8,500 patients by April 2026 is a meaningful denominator. Japan also has unusually robust postmarketing surveillance for designated intractable diseases, and clinicians routinely report adverse events to the Pharmaceuticals and Medical Devices Agency (PMDA). It is genuinely possible that the same rate of liver injury was happening in other countries but being underreported.

There is also a clinical-research literature on East Asian populations and idiosyncratic drug-induced liver injury that some hepatologists will point to — genetic variations in drug metabolism, in HLA haplotypes associated with DILI, in immune background. None of that has been formally invoked here, and reading too much into it would be speculation. What can be said is that the regulatory paths in Japan, the US, and Europe have all converged on the same drug, from different directions, within the past four months.

What patients are being told

For new patients in Japan: Kissei is asking doctors not to start them on Tavneos at this time. The drug remains approved, so individual clinical judgment still matters, but the company's safety letter is unambiguous.

For existing patients in Japan: don't stop on your own. Kissei's letter asks physicians to reassess whether continued treatment is appropriate for each patient, particularly looking at liver enzyme trends and time on therapy. Stopping a vasculitis treatment abruptly carries its own risks — disease flare can be severe — and the trade-off is one a physician needs to weigh.

For patients in the US and Europe: the labelling and clinical-use guidance hasn't formally changed in either jurisdiction yet, though the FDA's March 31 safety communication advised patients to discuss safety risks with their healthcare professional and consider alternatives.

The signs of liver trouble to watch for, repeated in every regulator's communication: unusual fatigue, itching, nausea, vomiting, light-coloured stools, yellowing of skin or eyes, dark urine, swelling or pain in the upper right abdomen.

A drug story that isn't over

Step back, and the Tavneos situation captures several things at once. A genuine clinical need — ANCA vasculitis is a brutal disease with limited treatment options. A regulatory approval based on a single pivotal trial that is now being interrogated for data integrity. A foreign acquisition story (ChemoCentryx → Amgen) layered on top of a licensing deal (Vifor → Kissei). And a postmarketing safety signal that took several years to fully crystallise.

What's particularly Japanese about this story isn't the drug, which is American. It's that Japan was the first regulator to approve it, has the largest patient base, has done some of the most assertive postmarketing surveillance, and is now the country where the safety alarm is loudest.

How is rare-disease drug oversight handled where you live? Do regulators in your country move faster than the FDA on safety concerns, slower, or about the same — and how confident are you in the postmarketing reporting system catching problems like this one?

References