The IRIS trial closed enrollment with 1,106 patients across 16 countries. The data safety monitoring board recommended early termination for efficacy: Gleevec was performing clearly enough that keeping patients on the interferon arm was no longer defensible. Patients in that arm were offered the option to cross over. Most did.

The Novartis New Drug Application arrived at FDA on February 27, 2001. This issue covers what happened in the 72 days that followed, and the regulatory infrastructure that made that timeline possible without moving the evidence threshold.

Gleevec's application was a New Drug Application (NDA), governed by section 505 of the Federal Food, Drug, and Cosmetic Act. This pathway applies to small-molecule drugs: compounds made by chemical synthesis with defined molecular structures.

The alternative pathway is the Biologics License Application (BLA), which applies to biologics: large-molecule drugs typically derived from living cells, including proteins, monoclonal antibodies, and gene therapies. Review processes and post-approval requirements differ between pathways, though FDA has worked to harmonize them since the Biologics Price Competition and Innovation Act of 2009.

Imatinib mesylate is a small molecule: a tyrosine kinase inhibitor that competitively occupies the ATP-binding pocket of the BCR-ABL oncoprotein, preventing the kinase from phosphorylating its downstream substrates. NDA was the correct pathway.

The Prescription Drug User Fee Act was enacted in 1992 following a period in which FDA drug review backlogs had grown to 3 or more years. The act created a system in which pharmaceutical companies pay user fees when submitting drug applications, and FDA commits to meeting specified review timeline goals in exchange.

PDUFA fees are substantial. For fiscal year 2026, the fee for a standard NDA that requires clinical data is $4,682,003, about 4.7 million dollars. FDA publishes the schedule annually.

What those fees fund: review staffing, IT infrastructure, and the operational capacity to meet PDUFA review timeline commitments. They do not create financial incentives for individual reviewers tied to approval outcomes.

What the fees do not determine: the outcome of the review.

FDA reviewers are federal civil servants. Their employment and compensation are not contingent on approval rates. The review process includes multiple independent layers: the primary reviewer, a supervisory sign-off, the Office of New Drugs division director, advisory committee consultation for complex or novel applications, and FDA's own statistical analysis of submitted data.

The Complete Response Letter (CRL) is FDA's mechanism for communicating that an application is not approvable as submitted. CRLs specify what is missing: additional safety data, manufacturing process deficiencies, labeling language concerns, clinical data gaps. Receipt of a CRL does not trigger a refund of PDUFA fees.

A documented recent example: in 2025 the FDA issued a Complete Response Letter for the RP1 (vusolimogene oderparepvec) plus nivolumab melanoma application, concluding the single-arm IGNYTE phase 1/2 trial could not be adequately interpreted and that a controlled phase 3 trial was needed. The company had paid the user fee. It did not receive an approval.

These 3 regulatory designations are consistently conflated in public discussion. They operate on different dimensions and can be held simultaneously.

Priority Review: applies to the review timeline. FDA's standard review target for a New Drug Application is 12 months from the date of filing. Priority review shortens that target to 6 months for drugs that offer a significant improvement over available therapy for serious or life-threatening conditions, or that address an unmet medical need. Priority review affects only the clock. It does not change the amount of evidence required, the type of evidence required, or post-approval obligations.

Accelerated Approval: applies to the evidentiary basis at the time of approval. Established under 21 CFR 314.500 for NDAs (and 21 CFR 601.40 for BLAs), it permits FDA to approve drugs for serious conditions based on a surrogate endpoint, which is a biomarker or other measure reasonably likely to predict clinical benefit, rather than requiring direct evidence of benefit at the time of approval. Mandatory post-marketing confirmatory trials are a condition of approval, not elective follow-up. If confirmatory trials fail to verify predicted benefit, FDA can initiate proceedings to withdraw the approval.

Breakthrough Therapy Designation: established under the Food and Drug Administration Safety and Innovation Act of 2012. Applies when preliminary clinical evidence shows that a drug for a serious condition offers substantial improvement over existing therapies on a clinically significant endpoint. Breakthrough designation triggers enhanced FDA engagement during the development process, including more frequent meetings and guidance, and rolling review of completed sections of the application before the full package is submitted. It does not modify what evidence is required at the time of approval or change the post-approval obligations. It changes when and how intensively FDA is involved during development.

Gleevec held priority review and received accelerated approval. Breakthrough therapy designation did not exist in 2001, it was established 11 years later, but the Gleevec review is frequently used as an example of what breakthrough-level evidence looks like in retrospect.

A drug can hold all 3 designations simultaneously. Priority review sets the clock. Accelerated approval sets the evidentiary basis. Breakthrough designation shapes the development interaction. They are not synonyms, and they are not redundant.

Phase III trials have structural limitations that are not signs of failure. They enroll selected patient populations defined by eligibility criteria that exclude patients with certain comorbidities, concomitant medications, age ranges, and organ function thresholds. The population that eventually receives an approved drug is broader. Post-marketing surveillance is the mechanism for identifying signals that emerge in that broader population.

Rare adverse events, typically defined as those occurring in fewer than 1 in 1,000 patients, are often statistically undetectable in Phase III trials, even large ones. A trial enrolling 1,106 patients, as IRIS did, cannot reliably identify an event occurring at that frequency. Post-approval pharmacovigilance programs, voluntary adverse event reporting systems (MedWatch in the US), and mandatory Phase IV commitment studies are the mechanisms for detecting those signals.

Ben Goldacre's work on post-marketing data and publication bias documents the failure modes in this system: reporting is incomplete, publication of negative post-marketing data is less consistent than publication of positive trial results, and the timeframe between signal identification and regulatory action has varied widely across cases. These are legitimate criticisms of how the post-marketing system performs in practice. They are not arguments that the system's design is without purpose.

The Vioxx (rofecoxib) withdrawal in 2004 is the most cited example of a cardiovascular safety signal that post-marketing data eventually identified and that Phase III had not captured at the time of approval. The timeline between signal emergence and regulatory action is a legitimate subject of ongoing debate. The relevant lesson from that case for this issue: FDA approval is a point in the evidentiary record, not the end of it.

The question of whether FDA operates independently of pharmaceutical industry influence functions at 2 levels that are routinely collapsed into a single argument. Separating them produces more accurate analysis.

Level 1: Does industry fee payment determine approval outcomes? The structural design of PDUFA includes explicit separation between fee collection and review decisions. PDUFA agreements are negotiated between FDA and industry trade organizations and specify performance commitments from FDA in terms of timeline targets. They do not specify outcome commitments. Individual reviewers have no financial relationship with the companies whose applications they assess. Advisory committee members undergo conflict-of-interest review and are subject to recusal requirements for financial relationships with the applicant.

The academic literature documents what PDUFA did and did not change. Darrow, Avorn, and Kesselheim (New England Journal of Medicine, 2017) show that user fees sped FDA review and now fund a large share of the agency's drug-review budget, with industry paying roughly three-quarters of the cost of scientific review. The same authors name the real risk in that arrangement: FDA's budgetary dependence on industry could advantage industry when policy is negotiated. That is a structural concern about the system, not evidence that any single approval is purchased.

Level 2: Does the broader regulatory relationship between FDA and the pharmaceutical industry influence policy, rulemaking, and enforcement priorities in ways that matter for public health? This is where the substantive debate lives. The revolving door between FDA and industry creates structural relationships that affect guidance document interpretation, enforcement prioritization, and regulatory framework design over time. These are real concerns with documented examples. Regulatory scholars and public health researchers who raise these concerns are doing legitimate work.

Applying the same sourcing framework that Caulfield's work on health claims suggests: trace the 'FDA is captured' claim to its specific sources before accepting or dismissing it. Some sources raising that claim are academic regulatory critics with peer-reviewed work and disclosed affiliations. Some are commercial entities whose revenue depends on public distrust of FDA-regulated products. Those are different kinds of sources making different kinds of arguments, and the distinction matters for how to weight what they say.

If you work in clinical research or regulatory affairs: The accelerated approval withdrawal actions of 2021 and 2022, when FDA withdrew a number of oncology accelerated approvals whose confirmatory trials had not demonstrated benefit, restarted a debate about how consistently the confirmatory trial requirement is enforced. The pathway's integrity depends on that enforcement being consistent. The Gleevec story is the reference case for what accelerated approval looks like when the surrogate endpoint predicts real benefit and the confirmatory data follows. The debate about other applications of the pathway is a separate question from the pathway's design.

If you are a patient or patient advocate: Accelerated approval gives patients earlier access to drugs when the evidence so far is strong but not yet complete. The tradeoff is real: the full picture of benefit is still being generated after you start the drug. In conditions where existing options are inadequate, earlier access often outweighs the uncertainty. In conditions where effective treatments already exist, the calculus is different. The pathway was designed to move the access timeline earlier when the weight of available evidence justifies it, not to remove the evidentiary standard.

If you are a general reader: The 72-day FDA review of Gleevec was fast because the data in the file was clean and the drug's benefit was clear. The review did not skip steps: it moved through them quickly. Speed and standards are different questions. Priority review is a faster clock. Accelerated approval is a different measurement point with required follow-up. These are not the same thing, and they are not the same as 'the FDA lowered the bar.'

Q4: Who funded this, and do they have a stake in the answer?

Apply this question in both directions. Novartis funded the IRIS trial. That is in the public record, and the structural safeguards of the FDA review process exist partly because funder interest is real. Apply the same question to claims that 'FDA approval is bought.' Who is making that claim, from what platform, and what do they gain if readers believe it?

Q9: What do independent experts with no stake in the outcome say?

The FDA advisory committee that reviewed Gleevec voted unanimously for approval. Advisory committee members undergo conflict-of-interest review. FDA's statistical reviewers conducted an independent analysis of Novartis's submitted data. The DSMB that recommended stopping the IRIS trial early had no financial relationship with Novartis. The independent verification layers exist precisely because financial stakes in outcomes are real, and the design of the review process accounts for that.

Citations verified June 2026. Peer-reviewed sources confirmed via PubMed; regulatory and statutory items cite the primary record.

[1] IRIS trial: O'Brien SG, Guilhot F, Larson RA, et al. N Engl J Med. 2003;348(11):994-1004. https://doi.org/10.1056/NEJMoa022457

[2] Gleevec approval, NDA 21-335 (accelerated approval, May 10, 2001): U.S. FDA Drug Approval Package. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2001/21335_Gleevec.cfm

[3] PDUFA statute and FY2026 fee ($4,682,003 for an NDA requiring clinical data): Prescription Drug User Fee Act of 1992, Pub. L. 102-571; Federal Register, Prescription Drug User Fee Rates for Fiscal Year 2026.

[4] NDA vs BLA pathways and harmonization: Federal Food, Drug, and Cosmetic Act sec. 505; Public Health Service Act sec. 351; Biologics Price Competition and Innovation Act of 2009.

[5] Accelerated approval regulation: 21 CFR 314.500 (drugs) and 21 CFR 601.40 (biologics).

[6] Breakthrough therapy designation: Food and Drug Administration Safety and Innovation Act of 2012 (FDASIA), Pub. L. 112-144.

[7] Vioxx (rofecoxib) cardiovascular signal (APPROVe trial, risk emerging after about 18 months): Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med. 2005;352(11):1092-1102. https://doi.org/10.1056/NEJMoa050493

[8] PDUFA, industry funding, and the structural concern: Darrow JJ, Avorn J, Kesselheim AS. Speed, Safety, and Industry Funding, From PDUFA I to PDUFA VI. N Engl J Med. 2017;377(23):2278-2286. https://doi.org/10.1056/NEJMhle1710706

[9] Accelerated approval enforcement (confirmatory-trial withdrawals): FDA Oncology Center of Excellence, Project Confirm (October 2021); Food and Drug Omnibus Reform Act of 2022 (FDORA).

[10] Example Complete Response Letter (RP1 plus nivolumab, 2025): U.S. FDA CRL database, fda.gov.

[11] Post-marketing data and publication bias: Goldacre B. Bad Pharma. 2012.

The views expressed on Root to Rx are my own and do not represent the views or positions of my employer, or any affiliated organization.

Root to Rx | 228 Park Ave S, #29976, New York, New York 10003, United States

You are receiving this because you subscribed at roottorx.com.