By Alex Harding, Entrepreneur-in-Residence at Atlas Venture, as part of the From The Trenches feature of LifeSciVC
What will and will not work to regain US strength in phase 1 clinical trials
“I don’t know anyone who runs their phase 1 trials in the US anymore.”
These words, uttered to me recently by a drug development expert at Atlas Venture with decades of experience in industry, are only slightly exaggerated.
Since 2020, the number of US INDs has declined more than 10%, from 2,182 to 1,960. Because that figure captures molecules that already completed first-in-human trials outside the US and are entering the US at later stages of development, I suspect that the underlying decline in US first-in-human trials is even steeper.
Meanwhile, in China, the number of phase 1 trials has increased from 601 in 2020 to 1,168 in 2025. Australian phase 1 clinical trials listed on clinicaltrials.gov have increased 21% over the same period.
It hurts everyone in our industry when filing with the FDA for first-in-human trials—previously the most common and rigorous option, and a speedy approach—is becoming non-viable because the trial process has become so slow and uncertain. At a time when US research institutions are under stress with reduced NIH funding, lower revenues from clinical trial activities apply further stress to US institutions that conduct clinical trials. Even more fundamentally, policy makers should be concerned about the US losing its competitive edge in biotechnology, an industry this country has historically dominated.
This accelerating trend has finally started to get serious attention in Washington and among policy wonks. The Duke Clinical Research Institute (DCRI) convened a think tank in January of this year to examine ways to reverse the shift of clinical trials away from the US, and later published a summary of the conclusions. In June, HHS launched Operation TrialBlazer to create a “Roadmap to Maintaining U.S. Leadership in Early Clinical Research and Development” with a 20-page white paper on its plans to accelerate early clinical trials and improve US competitiveness.
Both of these initiatives yielded some good ideas and some bad ones. I will highlight a few of each:
Bad Ideas
Sponsor Education
The Operation TrialBlazer report takes a patronizing stance toward trial sponsors (eg, companies), asserting that sponsors could be getting their trials started faster if only they were less cautious or better informed about IND requirements:
“Sponsor misunderstanding of upfront Chemistry, Manufacturing, and Controls (CMC) requirements and Pharmacology/Toxicology testing often delay the IND submission and trial initiation due to sponsors generating unnecessary data at a stage where there is still considerable uncertainty about a therapy’s effectiveness.”
For industry-sponsored trials, I doubt “sponsor misunderstanding” of clear FDA requirements is a pervasive issue. Sponsors would not waste time and money on additional preclinical studies without good reason. My conversations with industry regulatory and development experts suggests this problem should be flipped on its head. The prevailing issue is that FDA requirements are inconsistently applied and often overly conservative. So, rather than risking having an IND being put on clinical hold (ie, not allowed to proceed), some sponsors are opting to wait for a more extensive data package—often even including human data generated in other countries—before submitting their IND.
Rather than educating sponsors on regulatory requirements, the FDA needs to issue and apply more consistent guidance so that sponsors don’t feel they must take a risk-averse approach.
Streamlined Protocols, Improved Trial Designs
Operation TrialBlazer also blames sponsors for poor protocol development instead of acknowledging FDA’s regulatory shortcomings.
“An HHS study estimated that simplifying clinical trial protocols and reducing amendments could result in efficiencies that translate into an overall reduction in the estimated cost of bringing a drug to market by up to 22%”
It is true that trial designs have become more complex in recent years, with addition of non-core endpoints, which may indicate in some cases a need to streamline protocols. However, this added complexity often reflects a deliberate shift toward multi-part, modular, adaptive trials that are intended to address specific scientific questions and can enable faster clinical derisking and progression to later-stage trials.
In fact, the FDA was instrumental in overseeing the rise in these more elaborate early-phase trial designs. For example, the adaptive, multi-part, continuously-amended KEYNOTE-001 master study protocol for Keytruda (pembrolizumab) facilitated an accelerated path to approval and patient access for Keytruda across multiple cancer types by enabling fluid incorporation of emerging clinical data into multiple sBLA submissions for the drug. While the KEYNOTE trial framework became excessively complex over the years, it was nevertheless a great example of successful FDA efforts to speed patients’ access to life-saving medications by embracing more innovative trial designs.
More simply, the problem Operation TrialBlazer aims to solve is sponsors shifting more of their early-stage trials outside the US. Sponsors aren’t making this shift because their own trials are too complex.
Operation TrialBlazer also asserts that too many clinical trials are inadequately powered or poorly designed to answer the desired research question.
“Because the clinical trial enterprise has been criticized for conducting small, underpowered studies and given NIH’s role as steward of taxpayer investments in clinical research, NIH has a responsibility to ensure that it supports scientifically sound and rigorously designed clinical trials.”
While it is true that some clinical trials are not rigorously designed to test a scientific hypothesis, I question the merit of adding another layer of regulatory review at the NIH level if the overall goal of Operation TrialBlazer is to accelerate drug development and make the US a more attractive location to conduct early-stage clinical trials.
In short, additional regulatory oversight is not the way to accelerate drug development and lure sponsors back to the US.
Qualified Research Institutions (QRIs)
Operation TrialBlazer has proposed a pilot of a group of academic medical centers and CROs to function as a decentralized network of expert entities to review and advise sponsors on IND submissions, allowing for rolling IND submissions and faster IND reviews.
Superficially, there are some appealing elements in this idea: (1) The FDA staff is stretched thin, and pushing some of the IND review work to QRIs could ease the FDA review burden, (2) rolling submissions could accelerate IND reviews somewhat, and (3) gaining access to expert advisors on an IND package could be useful, particularly as FDA pre-IND feedback has become less reliable due to staffing cuts at the agency.
However, this plan has several serious flaws:
- The QRI concept risks introducing a new layer of bureaucracy to the IND submission process, effectively creating a two-step review process instead of one
- Academic medical centers are notoriously slow and generally lack practical drug development expertise. They would not usually offer credible advice beyond what most industry sponsors already have access to
- The FDA IND review timeline is already quite efficient, at 30 calendar days. This review timeline is substantially faster than China’s and roughly equivalent to Australia’s, and is not a significant bottleneck to drug development. Accelerating the IND review period with a rolling submission framework would not move the needle on a drug development timeline, even if it works as envisioned
Good Ideas
There are several good ideas in Operation TrialBlazer and DCRI that should get more focus from policy makers.
IRB Reform and Site Startup Acceleration
Operation TrialBlazer correctly identifies two bottlenecks in getting clinical trials underway: IRB review and trial site startup time.
IRB review is a significant and avoidable source of delay, particularly when non-central IRBs are used. Academic IRBs meet infrequently to review trial protocols and often issue nit-picky feedback to sponsors that then require an additional round of protocol revision and resubmission, leading to delays over insignificant issues. In contrast, Bellberry Ltd, a private non-profit ethics committee in Australia, conducts approximately 40% of the country’s ethics committee reviews. Its faster, more pragmatic approach helps provide greater consistency and shorter overall timelines to ethics approval. A similar system in the US, where more IRB reviews could occur outside of the academic setting, would make a significant difference to timelines.
Another measure that could further accelerate trial startup time would be if IRB reviews and site startup activities could be conducted in parallel with IND reviews. In fact, this was common practice until a decade or two ago, and allowed the first subjects to be dosed within days of IND clearance in some cases. Now, it takes a minimum of 1-2 months after IND clearance to dose the first subject in the US, and in many cases, it can extend beyond 6 months.
Slow contract negotiation with individual trial sites is another major reason why site startup times have dragged out in recent years. Operation TrialBlazer proposes the “use of master trial agreements” to reduce review periods.
Here we can again use Australia as a model, where all trial sites must use a standardized clinical trial research agreement. The agreement is essentially non-negotiable, removing what can be a protracted legal negotiation process from the trial startup timeline. While the inability to negotiate contract terms may initially appear disadvantageous to sponsors, the standardized agreement is generally viewed favorably because it is balanced and can dramatically accelerate site startup timelines.
Emulation of the Australian Clinical Trial Notification (CTN) Framework
I have touched on a couple of attractive policies in Australia. DCRI takes this a step further in their recommendations:
“Identify which elements of systems like Australia’s Clinical Trial Notification framework (e.g., ethics committee review, fit-for-purpose regulatory submission requirements) can be adapted for the U.S. and establish clear guidance on how these components should be applied to accelerate first-in-human trials while ensuring patient safety and adherence to GCP standards.”
Indeed, I believe there are a number of elements of the Australian framework that could be adopted in the US.
The data package required for a FDA IND is more extensive than what is required in most other countries, including Australia. For high-risk, early-stage trials, the time and cost required to generate the data required for an IND can be difficult to justify. Instead, a leaner data package focused on demonstrating acceptable toxicology and safety data, similar to the information in an Australian ethics committee submission, could be used to determine that the risks to trial participants are acceptable without placing undue burden on sponsors for additional data generation and document preparation.
The Australian ethics committee review process decentralizes trial review so that it is not done by TGA (Australia’s equivalent to the FDA), but rather occurs through private and academic sites. This approach extends beyond Operation TrialBlazer’s concept of QRIs, in that the TGA is not responsible for the review and approval of first in human trials at all, while the QRI network proposed by Operation TrialBlazer would add the QRI review but keep FDA’s final review responsibility in place (effectively adding a new layer of review, not streamlining, as I noted above).
The Australian approach is faster than the FDA process, although it comes with a couple of tradeoffs: (1) greater heterogeneity with decentralized reviews and potentially less experienced reviewers, and (2) no formal mechanism for obtaining HREC feedback prior to submission. The FDA pre-IND feedback process has historically been valued as a source of useful information on FDA’s stance toward future regulatory hurdles.
I propose a hybrid structure that creates more autonomous QRIs that must adhere to common standards maintained by the FDA but have the power to perform the IND review directly (perhaps with a rapid confirmation by the FDA). Preferably, sponsors would also have the option to seek FDA pre-IND feedback to guide decision making for the first-in-human trial and beyond.
Conclusion
Trial sponsors have spoken through their actions in recent years, increasingly favoring other countries for their first-in-human trials. It’s good news that regulators and policy makers in the US are noticing this phenomenon and beginning to take action to address it. Their actions will be most effective if they recognize the shortcomings of FDA that have led to this shift, particularly, overly cautious stances on data requirements, inconsistent application of regulatory principles, and onerous documentation requirements that make the US less attractive to trial sponsors.
The solution to these issues will come from reducing regulatory burden, not increasing it. Historically, FDA has been a leader in regulatory innovation to accelerate bringing life-changing medicines to market, developing important initiatives such as Accelerated Approval, Breakthrough Designation, and Operation Warp Speed. FDA must continue that history of regulatory modernization to bring first-in-human trials back to the US and to maintain the strong position the US holds in the global biotech industry.
Many thanks to Tim Reilly and Jennifer Burg for their very helpful feedback and comments.
