Thanks to Nicholas Sarlis MD PhD at Linked In for highlighting two new oncology policy articles by FDA authors, in J Clin Oncol.
Mehta & Pazdur discuss the balance of accelerated approval, accelerated withdrawal, new 2023 FDA, and some nuances of the interpretation of successive data.
Fallah et al. discuss whether you can run (and enroll) an equipoise clinical trial when one of the arms is already officially labeled "a breakthrough drug."
Agency IQ provides a discuss of some of the 2023 legal changes.
I've clipped an AI summary below.
##
AI CORNER
Here is a summary of the three articles, focusing on their relevance to current policy issues in oncology trials, including accelerated approval, equipoise in trials, and the updated law on expedited withdrawals:
Fallah et al. (2024): "Equipoise Lost? Trial Conduct Challenges in an Era of Breakthrough Therapies"
This article discusses the challenge of maintaining clinical equipoise in oncology trials involving breakthrough therapies. Regulatory mechanisms like breakthrough therapy designation and accelerated approval can heighten perceptions of investigational drug superiority, affecting both patient and investigator behaviors in trials. Key challenges include:
- Early dropout rates: Patients often leave control arms early, seeking the investigational therapy, which can lead to asymmetric censoring and biased trial results.
- Investigator bias: In open-label trials, biases in progression-free survival (PFS) assessments can skew results, especially if patients cross over to investigational arms.
- Mitigation strategies: Recommendations include double-blind designs, better education of trial participants, and sensitivity analyses to account for asymmetric dropout. The authors highlight the risks of using subjective endpoints like PFS in trials where equipoise is difficult to maintain and advocate for trial designs that minimize bias.
Mehta et al. (2024): "Oncology Accelerated Approval Confirmatory Trials: When a Failed Trial Is Not a Failed Drug"
This article reviews the evolution of the FDA's Accelerated Approval (AA) pathway, which allows earlier drug access based on surrogate endpoints. The focus is on challenges in confirmatory trials:
- Withdrawals due to trial failures: About 15% of AA indications in oncology have been withdrawn because confirmatory trials failed to meet primary endpoints, demonstrated negative benefit-risk profiles, or encountered methodological flaws.
- New policies in 2023: The Consolidated Appropriations Act now requires confirmatory trials to be underway at the time of AA and provides mechanisms for streamlined withdrawals.
- Regulatory nuances: FDA decisions consider factors like unmet need and the availability of alternative therapies. Some drugs failing confirmatory trials, such as gemtuzumab ozogamicin, later achieved approval after adjusted dosages or patient selection. The authors argue that failed trials should be scrutinized for methodological flaws rather than automatically signaling drug inefficacy. [For more on these nuances, see Appendix below.]
AgencyIQ (2024): "Accelerated Removal: FDA Making Use of New Authority to Withdraw Ineffective Drugs More Quickly"
This report details recent updates to the accelerated withdrawal process under the Consolidated Appropriations Act of 2023. Key changes include:
- Streamlined procedures: The FDA can now withdraw AA drugs faster under specific conditions, such as failed confirmatory trials or safety concerns. This contrasts with earlier lengthy processes (e.g., the withdrawal of Makena took over two years).
- Public involvement and transparency: The updated law ensures opportunities for sponsor appeals, advisory committee meetings, and public comments.
- Case study – Pepaxto: The expedited withdrawal of Pepaxto in 2024 demonstrated the FDA’s improved ability to act decisively when confirmatory trials indicate harm (e.g., reduced overall survival).
Policy Implications
The articles highlight the tension between early patient access to promising therapies and the risks of premature approval. The 2023 law attempts to address this by:
- Requiring robust confirmatory trials at the time of AA.
- Enhancing the FDA’s authority to expedite withdrawals when drugs fail to meet safety or efficacy benchmarks.
Mehta's emphasis on confirmatory trial quality and AgencyIQ's focus on streamlined withdrawal underscore the importance of designing trials that balance early access with rigorous post-approval scrutiny. Fallah's exploration of equipoise loss adds another layer, emphasizing the need for trial designs that ensure unbiased and interpretable results.
###
Fallah et al.
Fallah J, Mulkey F, Fiero MH, et al. Equipoise Lost? Trial Conduct Challenges in an Era of Breakthrough Therapies. J Clin Oncol. 2024;42(32):3783-3787. doi:10.1200/JCO-24-01200.
Mehta et al.
Mehta GU, Pazdur R. Oncology Accelerated Approval Confirmatory Trials: When a Failed Trial Is Not a Failed Drug. J Clin Oncol. 2024;42(32):3778-3782. doi:10.1200/JCO-24-01654.
AgencyIQ
Gaffney A. Accelerated Removal: FDA Making Use of New Authority to Withdraw Ineffective Drugs More Quickly. AgencyIQ by POLITICO. April 5, 2024. [Link to article].
###
APPENDIX
More on Mehta & Pazdur's and the subtleties of withdrawal decisions.
Yes, Mehta and Pazdur delve into the complexities surrounding withdrawals of accelerated approvals (AAs), emphasizing that hasty action can sometimes undermine patient and public health benefits. Here are the nuanced points they raise about when and why withdrawals may not be immediate:
1. Evaluating Unmet Medical Need
- If a high unmet medical need persists for the patient population, the FDA may delay withdrawal even after confirmatory trials fail. For instance, drugs like lurbinectedin maintained AA despite failing their primary endpoints because they continued to offer meaningful benefit in a landscape with few alternatives.
2. Learning from Serial Data
- Drugs showing promise in initial confirmatory trials but failing due to design limitations or patient selection may merit additional study. Mehta and Pazdur point to examples like gefitinib and gemtuzumab ozogamicin, where confirmatory trials initially failed but subsequent adjustments in dosage, study design, or biomarker targeting led to eventual approvals. These examples highlight the need for a deeper investigation before concluding a drug's lack of efficacy.
3. Balancing Current and Future Benefit-Risk
- Withdrawal decisions weigh both the current benefit-risk profile and the drug's evolving role as part of treatment regimens:
- Drugs demonstrating consistent response rates and durations from initial trials often still hold value for specific subpopulations.
- Sponsors are encouraged to pursue alternative confirmatory trials that address prior design shortcomings.
- New data on biomarkers or mechanisms of action can change the understanding of a drug’s utility.
4. Impact on Ongoing Responders
- For drugs still providing benefits to certain patients, withdrawal could leave these individuals without viable options. This is particularly significant for patients with few alternatives, such as those in refractory stages of disease.
5. Changing Therapeutic Landscapes
- Advances in other treatments during the period between AA and confirmatory trials can influence the interpretation of benefit. For example:
- If new therapies enter the market that outperform the AA drug, maintaining approval may no longer align with public health interests.
- Conversely, if the therapeutic landscape remains static, preserving the AA while revisiting trial designs might be more appropriate.
6. Methodological Issues and Contextual Decisions
- Methodological flaws, such as poor trial powering, patient selection issues, or inappropriate comparators, may lead to premature trial failures. The authors caution against conflating these with true lack of benefit. Thoughtful re-evaluation can ensure these factors do not preemptively eliminate promising drugs.
Conclusion
Mehta and Pazdur argue for judicious decision-making that balances the urgency of withdrawing ineffective therapies with the need for context-sensitive evaluation. They advocate for flexibility to allow further study or alternative trial designs in scenarios where unmet need remains high or evidence suggests potential for improvement with better methods. Their analysis underscores the importance of recognizing AA as an evolving process rather than a binary decision point.
