In oncology drug development, FDA rejection is often perceived as a late-stage failure, an FDA Complete Response Letter (CRL) that arrives after many years of pre-clinical development, years of intense clinical work, heavy capital investment, and deep, persistent patient needs. Timelines slip, accelerated approval pathways stall or are halted, and sponsors are left scrambling to understand what went wrong.
But growing FDA transparency tells a very different story.
With more than 200 Complete Response Letters publicly released between 2020 and 2025, a consistent pattern has emerged, particularly in oncology: most FDA rejections are not caused by surprises at submission, but by predictable decisions made much earlier in clinical development.
Having led and advised numerous global oncology development programs at Merck, and supported teams across pharma, biotech, and academic cancer centers, I have seen these patterns repeatedly. In my experience, FDA drug rejections in oncology are rarely driven by surprises at submission, but by predictable decisions made much earlier in clinical development, particularly in programs relying on small datasets, surrogate endpoints, evolving standards of care, and accelerated approval pathways.
FDA Rejection in Oncology Is Rarely a Single Failure
One of the clearest insights from publicly released oncology CRLs is that FDA non-approval decisions almost never cite a single deficiency. Instead, multiple issues compound: insufficient efficacy, unresolved safety signals, trial design limitations, manufacturing readiness gaps, and execution concerns often appear together.
FDA oncology reviewers are not asking:
“Did the sponsor observe tumor shrinkage?”
They are asking:
“Does this development program convincingly demonstrate a favorable and durable benefit–risk profile that supports approval, labeling, and post-marketing responsibility?”
That distinction, between observing antitumor activity and demonstrating clinical benefit, is where many oncology strategies quietly lose alignment.
1. When Antitumor Activity Exists, but Is Not Enough
In oncology drug development, early efficacy signals such as overall response rate (ORR), disease control rate (DCR), or progression-free survival (PFS) are often treated as sufficient proof of approvability. Publicly released FDA Complete Response Letters (CRLs) suggest otherwise.
Analysis of publicly released oncology CRLs, interpreted through regulatory experience, highlights recurring themes including:
- Endpoints that do not convincingly demonstrate meaningful clinical benefit
- Study designs that limit interpretability, including challenges in contextualizing results against existing therapies
- Analytical approaches that introduce uncertainty, such as reliance on exploratory or non-pre-specified analyses
- Safety considerations that require further characterization to support a favorable benefit–risk profile
- Uncertainty around the durability and generalizability of observed antitumor activity, particularly in later-line settings
Example
Publicly released oncology CRLs often reflect FDA concerns about whether observed antitumor activity, such as ORR, can be interpreted as evidence of meaningful clinical benefit. While the CRL language is typically high-level, FDA reviewers frequently express uncertainty when response-based signals cannot be adequately contextualized, generalized, or translated into a favorable and durable benefit–risk profile for the proposed indication.
In the released CRLs, this uncertainty most commonly arises in development programs where the evidentiary package limits FDA’s ability to benchmark observed activity against existing therapies, assess durability of benefit, or extrapolate findings beyond a narrowly defined study population. From the FDA’s perspective, unresolved uncertainty around clinical benefit–risk, particularly outside of high-unmet-need settings, is a sufficient basis to delay approval rather than negotiate it.
How this could have been prevented?
Successful oncology programs begin with early clarity around one fundamental question:
What exact clinical conclusion must the FDA reach for this indication, line of therapy, and patient population, and what data will compel that conclusion?
Trials not designed around this question often generate data that are promising, publishable, and ultimately not label-supportive.
2. Oncology Trial Design Choices That Limit Interpretability
Many FDA rejections in oncology stem from trials that were operationally sound, but strategically constrained.
Analysis of publicly released oncology CRLs, interpreted through regulatory experience, highlights recurring trial design challenges, including:
- Underpowered randomized studies unable to detect clinically meaningful differences
- Control arms misaligned with rapidly evolving standards of care
- Broad or unstable biomarker definitions that shift during development
- Patient populations that do not match the proposed labeling
- Protocol amendments that compromise interpretability of efficacy and/or safety
Example:
Several oncology CRLs reference trials where comparator arms became outdated during enrollment, making it difficult for FDA reviewers to contextualize observed benefit. Others cite biomarker-driven programs where assay changes or cutoffs were modified mid-study, undermining confidence in patient selection.
Once an oncology trial is complete, these limitations are usually irreversible, regardless of signal strength.
How this could have been prevented?
High-performing oncology teams treat protocol development as a regulatory strategy, not merely a clinical exercise. Clinical, regulatory, statistical, and translational leaders align before first patient in, rather than attempting to retrofit regulatory logic after data readout.
3. Manufacturing and CMC: A Persistent FDA Approval Risk in Oncology Drug Development
While clinical efficacy and safety dominate oncology drug development discussions, manufacturing, quality, and Chemistry, Manufacturing, and Controls (CMC) deficiencies remain among the most common causes of FDA Complete Response Letters (CRLs) across therapeutic areas.
Analyses of publicly released FDA CRLs issued between 2020 and 2025 show that:
- A large majority reference quality, manufacturing, or CMC-related deficiencies
- A substantial proportion reference facility inspection findings, GMP compliance issues, or pre-approval inspection outcomes
- Product quality, comparability, and process validation deficiencies are frequently cited contributors to FDA non-approval
Although these trends are not limited to oncology, the risk is amplified in oncology programs involving complex therapeutic modalities, including cell and gene therapies, antibody–drug conjugates (ADCs), biologics, radiopharmaceuticals, and combination products, where manufacturing processes, control strategies, and scale-up activities are often less mature at the time of regulatory submission.
Example:
Multiple oncology drug applications with acceptable clinical benefit–risk profiles have experienced FDA approval delays or CRLs due to incomplete process validation, insufficient comparability between clinical and commercial manufacturing material, or deficiencies identified during FDA pre-approval inspections, underscoring that robust clinical data alone are insufficient to support regulatory approval.
How this could have been prevented?
In oncology drug development, CMC strategy and manufacturing readiness must advance in parallel with clinical development. Programs that defer manufacturing maturity frequently discover, late in development, that otherwise approvable clinical data cannot be supported by an FDA-approvable, inspection-ready, and compliant manufacturing supply chain.
4. Oncology Submissions That Reveal Execution Gaps
A significant number of oncology FDA rejections are not driven by negative data, but by how the data are analyzed, integrated, and presented.
FDA oncology reviewers frequently cite:
- Missing or inconsistent analyses
- Claims that exceed what the data can reasonably support
- Inconsistencies across CTD modules
- Weak justification of accelerated approval or conditional approval pathways
Example:
CRLs often note misalignment between clinical summaries, statistical analyses, and labeling proposals, raising concerns about sponsor execution capability post-approval.
At this stage, FDA review extends beyond the product itself to the sponsor’s ability to execute post-approval obligations.
How this could have been prevented?
Strong oncology submissions are built as coherent regulatory narratives, not document compilations. They anticipate FDA skepticism and proactively address likely questions around benefit–risk, durability, and confirmatory strategy.
5. FDA Feedback That Oncology Teams Did Not Fully Incorporate
Another common and avoidable contributor to oncology FDA rejection is the partial implementation of prior FDA advice.
Recent oncology CRLs show sponsors being asked to:
- Conduct additional randomized or confirmatory trials
- Reanalyze existing datasets using pre-specified methods
- Address issues raised during End-of-Phase 2 or pre-NDA meetings
Example:
Several oncology programs proceeded to submission despite unresolved FDA concerns around trial design or endpoint selection, resulting in predictable requests for additional studies.
How this could have been prevented?
In oncology, FDA interactions are not procedural milestones, they are alignment checkpoints. Successful teams treat FDA feedback as binding design input, not optional commentary.
6. Post-Approval Planning in Oncology Starts Earlier Than You Think
The FDA increasingly evaluates oncology sponsors on post-approval readiness, particularly for first-in-class therapies and accelerated approvals.
Recent oncology CRLs cite concerns including:
- Weak pharmacovigilance and risk management planning
- Vague or unrealistic confirmatory trial proposals
- Absence of a credible real-world evidence strategy
Example:
For accelerated approvals, FDA reviewers frequently question whether sponsors have the infrastructure, timelines, and commitment to deliver confirmatory evidence promptly.
How this could have been prevented?
In modern oncology drug development, approval is not just about launch, it is about lifecycle responsibility. Post-marketing strategy must be visible early and credible.
What Successful Oncology Development Programs Do Differently
Across oncology programs that ultimately secure FDA approval, one principle is consistent:
Approval outcomes are designed into the program early, or lost incrementally along the way.
Successful oncology teams:
- Embed regulatory thinking into clinical trial design
- Identify approval-limiting risks early
- Align clinical, CMC, and regulatory strategies
- Make fewer decisions, but make them deliberately
FDA Approval in Oncology Is Built-in Early
For oncology companies currently running or planning clinical trials, the most dangerous assumption is that FDA approval can be “fixed” at submission.
It cannot.
In oncology drug development, FDA approval is the cumulative result of hundreds of early decisions, many made before the first patient is enrolled.
At Arc Nouvel, we help oncology-focused pharma and biotech teams, particularly small and mid-size companies, identify regulatory risk early, strengthen development strategy, and design programs built for approval, not repair.
