Tavneos FDA Withdrawal: Why the Decision Is Complicated
The FDA says Tavneos approval relied on compromised trial data. See what the evidence shows, why patients still value it, and how withdrawal could unfold.
Tavneos, the FDA and the Hard Question of Drug Withdrawal
Drug approval is supposed to take research we're not fully sure about and turn it into a decision patients and clinicians can actually use. Tavneos - an option for a rare type of vasculitis - shows what can happen when the evidence behind that decision later gets questioned. The FDA says the pivotal trial may have involved undisclosed changes in how patient outcomes were assessed, and it has proposed withdrawing the drug's US approval. Amgen, Tavneos's manufacturer, argues the treatment still belongs in clinical care.
For patients, this isn't a technical fight over study design. Tavneos was meant to help reduce reliance on steroids, which can cause serious side effects. But the drug has also been linked to reports of severe liver injury, and whether it truly works as intended is now disputed. The FDA's formal withdrawal process only adds to the uncertainty - and the decision could shape how other drug approvals get handled, too.
This article lays out what the FDA says happened, where clinicians see it differently from Amgen, what the company argues, and how patients and doctors can think through the uncertainty while the regulatory process plays out.
Why Tavneos matters to patients
Tavneos is the brand name for avacopan. It was approved in the United States in October 2021 for adults with ANCA-associated vasculitis (AAV). AAV refers to a group of uncommon autoimmune diseases where the immune system attacks small blood vessels. When inflammation hits vital organs - especially the kidneys and lungs - it can become life-threatening without treatment.
Most people start with high-dose steroids to control inflammation quickly, along with other medications aimed at the underlying disease. Steroids can work, but staying on them for long stretches can take a heavy toll. Chloe Cusano, who was diagnosed with AAV at 20, described gaining 40 pounds within weeks of starting steroids, along with mood changes, sleep problems, and nightmares.
Tavneos was supposed to offer another approach - one that might lower steroid exposure. That matters, because "alternative" doesn't always mean "nothing." Sometimes it means switching to other medicines that come with their own risks. When you're dealing with a rare, serious illness, even imperfect evidence can feel like the difference between acceptable options and fewer choices.
What the original trial showed - and did not show
The FDA reviewed a clinical trial sponsored by ChemoCentryx, the California biotechnology company that developed Tavneos. In the trial, participants receiving the drug appeared to have higher rates of symptom remission after one year. But FDA reviewers questioned how confidently those results could be interpreted, including concerns about the study's design and what treatment was given to the comparison group.
In 2021, the FDA convened an advisory committee. Members split on whether the evidence was strong enough for approval. Some thought Tavneos offered a real benefit and another treatment option worth having; others said the trial didn't clearly prove the benefit. The committee narrowly recommended approval, and the FDA approved the drug.
Even then, questions remained. One committee member later said his vote assumed the sponsor followed the trial rules. That assumption is now at the center of the dispute: it's not only about whether the drug might help, but whether the original evidence can be relied on to justify the decision.
The trade-off is not simply drug versus no drug
Tavneos is expensive. The source article cited a price of about $220,000 per year, and it reported sales nearing half a billion dollars in fiscal year 2025. Those figures raise obvious questions about value, but they don't answer the clinical question of who benefits and what level of risk is acceptable.
Patients and clinicians also have to weigh Tavneos's potential harms against the harms of alternatives, including prolonged steroid use. Cusano credits the drug with helping her taper steroids faster. That kind of individual experience matters for understanding why people want access - but one person's response can't replace broader, dependable clinical evidence.
What the FDA says went wrong with the trial
The FDA's most serious concern involves a later review of the pivotal trial data. In an investor lawsuit, a report alleged the company's original analysis didn't show better one-year remission rates for Tavneos patients compared with the comparison treatment.
The FDA says that after that, company employees reviewed unblinded trial data - meaning they could see which participants received the drug and which received a placebo. After that, the FDA says a selected set of patients had their remission outcomes reassessed. According to the FDA, this readjudication violated the study protocol and involved selecting patients in a way that favored the company's desired result. The FDA also alleges ChemoCentryx did not disclose the original analysis or the reassessment to the agency.
These are allegations from the regulator, not a final court finding based on the source material. Amgen disputes the FDA's characterization and says it wasn't data manipulation. Still, the procedural issue matters: if outcome rules get revisited after trial results are visible, can the revised analysis genuinely act like an unbiased test of the treatment?
Why changing the analysis can change confidence
A trial protocol lays out the rules in advance for how outcomes will be measured and how results will be analyzed. Predefining those rules reduces the chance that researchers will make choices after seeing which results look better. If a sponsor changes how outcomes are assessed once it knows who received the drug, the concern isn't automatically that anyone altered underlying patient records. The concern is that choices about which records count - or how they're interpreted - could shape the conclusion.
Amgen says no clinical data were changed. It describes the reassessment as a way to correct a small number of errors by applying the trial's prespecified rules consistently, and it says the reassessment followed a prespecified procedure. In the FDA's view, as described in its withdrawal proposal, the process wasn't properly disclosed and the selection of cases was problematic.
Those positions lead to different ways of judging the episode. Amgen focuses on whether raw data were altered and whether the reassessment corrected errors. The FDA questions whether the analysis complied with the protocol - and whether the approval still rests on substantial evidence. A regulator's job depends on the integrity of the whole evidentiary chain, not just whether a spreadsheet shows edits.
A chain of oversight did not catch the problem early
The trial was reviewed by the FDA, by an advisory committee, and by academic authors whose paper appeared in the New England Journal of Medicine. The journal later retracted that paper at the request of its two academic authors. That sequence has raised questions about how disputed analyses can slip through oversight before a drug reaches the market.
There's also a structural reality. Small biotech companies often develop a drug and then sell it or partner with a larger pharmaceutical company. ChemoCentryx was acquired by Amgen for nearly $4 billion about a year after approval. Smaller firms may have fewer other products to balance against the risks of a single program, which can add pressure to move quickly. That doesn't prove wrongdoing in this case - but it does make clear why transparent trial processes and independent scrutiny matter.
Why Amgen and clinicians disagree about the evidence
Amgen argues that the contested trial procedures don't justify pulling Tavneos from the US market. In a 104-page submission to the FDA, Amgen called ChemoCentryx's reassessment unfortunate but said it didn't amount to manipulation. It also said ChemoCentryx wasn't legally required to report that it revisited the data.
Amgen points to an independent reassessment of the pivotal trial by researchers at the Duke Clinical Research Institute. In that analysis, participants who received Tavneos didn't do worse on one-year remission - but the drug still didn't meet the FDA's standard for showing higher remission rates. Amgen also cites real-world evidence suggesting the drug can reduce steroid use, arguing that keeping treatment options open matters for patients who might otherwise face more steroid exposure.
These arguments aren't the same as a brand-new, definitive clinical trial. "No worse" isn't the same as clear evidence of "better," and real-world evidence can be harder to interpret because it's not controlled like a trial. Patients in routine care may differ in important ways, and their treatments and monitoring may include additional therapies. Real-world studies can add useful context - but design and analysis still matter.
The safety question is separate from the efficacy question
In its April letter, the FDA proposed withdrawal on two grounds: reports of serious, sometimes fatal, liver injury, and the FDA's view that Tavneos no longer has substantial evidence of effectiveness. The FDA's acting CDER director referenced eight deaths and dozens of hospitalizations linked to the drug. That description shouldn't be taken as automatic proof that Tavneos caused every reported event - safety reports still require clinical interpretation - but the reports are serious, and the FDA says they are part of its case.
Some clinicians argue that, with careful monitoring, the treatment may still be worthwhile. Epidemiologist Eric Roberts, who co-authored an independent registry study, described the low rates of drug-induced liver injury found in that study as reassuring. He also pointed to the potential cost of relying on higher steroid doses. The registry evidence offers a different perspective from the FDA's safety reports, but neither side automatically cancels out the other.
So the practical question for clinicians isn't only, "Does Tavneos work?" It's also: which patients, how much benefit, compared with which alternatives, and under what monitoring. Those answers can vary from patient to patient. At the population level, though, regulators still have to decide whether the evidence meets the legal standard for approval.
The evidence gap is not fixed by a plausible mechanism
A drug can make biological sense and still fail to show, in strong evidence, that it improves real outcomes. Likewise, a trial with limitations may contain signals worth studying further without being enough to resolve an approval question. That's why people who question Tavneos's approval have asked for additional peer-reviewed evidence rather than relying on arguments about potential benefit alone.
The FDA asked for post-marketing studies after approval. Researchers critical of the current evidence say Amgen has fallen behind on deadlines. Post-marketing research can answer questions left open at approval - but only if studies are completed, designed well, and reported clearly enough for others to assess what the studies actually found. Delays prolong the uncertainty for patients, clinicians, and regulators.
What withdrawal could mean for patients and regulators
The FDA can't simply erase an approval without following the legal process. Legal scholar Patricia Zettler says the agency must first give notice and offer the manufacturer an opportunity for a public hearing. If the manufacturer can show a real dispute about the facts, a hearing can take place. After that, the FDA can decide whether to withdraw approval, and the company can challenge the decision in court.
That step-by-step process protects the manufacturer's right to respond and gives the agency a record for a major decision. It also takes time. In the meantime, patients and physicians may not know how to interpret the drug's status, how strong the evidence is, or what risks and benefits matter for them. This account doesn't predict what will happen at the scheduled hearing on October 29, 2026, or in any later regulatory action.
Withdrawal could also spill beyond Tavneos. The FDA argues that acting is important to protect public health and uphold the integrity of drug approvals. If a sponsor could get approval based on an analysis a regulator later considers unreliable - without meaningful consequences - other companies might decide the risk is worth taking. On the other hand, withdrawing a treatment some patients and clinicians believe helps could limit options, even if alternative treatments bring their own harms.
A practical framework for patients and clinicians
If you're currently taking Tavneos, a regulatory dispute isn't a reason to stop the drug or switch treatment without medical advice. A clinician can review your disease status, liver monitoring needs, other medications, and the alternatives available to you. The decision is individual, but it should be grounded in a straightforward account of what's known and what remains uncertain.
A useful conversation can focus on four questions:
- What is the treatment goal? Decide whether the priority is controlling active vasculitis, reducing steroid exposure, maintaining remission, or some mix of all three.
- What evidence applies to this patient? Keep trial results, independent analyses, registry findings, and personal experience in separate categories rather than treating them as interchangeable.
- What are the relevant risks and alternatives? Compare Tavneos's potential benefits and safety concerns with the risks of other treatment approaches.
- What would change the decision? Agree on what to monitor - symptoms, test results, and similar triggers - and discuss how new regulatory or clinical evidence will be weighed.
This framework won't settle the regulatory question. But it can make the clinical conversation more useful while the approval status remains contested.
What the case says about drug oversight
Tavneos highlights a weakness that can show up when a drug reaches the market with unresolved questions and later studies don't close them quickly. Approval isn't a lifetime guarantee that the evidence will stay strong forever. Post-approval oversight still depends on the FDA's ability to monitor safety, enforce study requirements, and act when new information shifts the balance.
The case also underscores why data integrity isn't a technicality. Approval depends on a chain of decisions: how a trial is designed, how outcomes are measured, what gets disclosed to regulators, and how independent reviewers assess the results. If trust in that chain breaks, patients lose more than confidence in a single product. They lose confidence that the system can reliably distinguish promising treatments from ones whose benefits haven't been demonstrated.
Conclusion: approval is a decision, not a guarantee
Tavneos raises a difficult but unavoidable regulatory question: what should the FDA do when there are patients who value the drug, but the pivotal evidence is challenged and safety concerns have come to the fore? The answer can't rest on one patient's experience, one company's reading of the data, or a regulator's allegations by themselves. It requires clear evidence, a fair process, and a consistent standard for weighing benefit against risk.
For patients and clinicians, the immediate job is to make treatment decisions with a clear-eyed understanding of uncertainty and alternatives. For regulators and drugmakers, the broader lesson is that trust in clinical evidence is part of the treatment itself. Approval can open access - but it doesn't remove the responsibility to keep checking whether the drug actually works, to monitor harms, and to act when the evidence no longer supports confidence.