Defining success for early-stage, small biotech trials

Small biotech companies encounter significant challenges in designing and implementing clinical development programmes. These companies have an even lower rate of success than that of large, established companies, due to limited internal experience in clinical development and limited infrastructure.¹

An even greater challenge is their need for external investment. During the early stages of development, in order to progress from one stage to the next, small biotechs must overcome much higher hurdles imposed by investors than those faced internally by large, established companies.

Objectives of early-stage development

The main objective of Phase 1 clinical trials is the initial evaluation of safety in a small number of participants. These trials are usually conducted on healthy volunteers. There are exceptions to this, particularly in oncology, where Phase 1 studies involve patients with the disease. Another objective of Phase 1 is the evaluation of dose range, by monitoring safety and pharmacokinetics (PK) on gradually increasing doses. More recently, Phase 1 trials also include a small extension phase aimed at evaluating early signals of efficacy.

The main objective of Phase 2 clinical trials is to select the optimal dose to advance to Phase 3 studies based on the observed safety/efficacy profile. Phase 2 trials enrol more patients than Phase 1 trials and provide a much better characterisation of safety and efficacy. They need to establish that at least one dose has a sufficiently favourable benefit/risk profile to justify proceeding to Phase 3 of development, which involves larger trials. It is the objective of Phase 3 trials to demonstrate efficacy via statistical significance.

Definition of success for early-stage trials: Current approach

Larger companies that do not depend on external investments follow previously described development objectives. However, smaller companies face much more rigid demands from external investors. Small biotechs are often required to achieve statistical significance in Phase 2, and sometimes even as early as Phase 1 extensions. These demands do not benefit anyone. For patients, potentially life-saving treatments are delayed. For trial sponsors who successfully develop a treatment, revenues are diminished, due to substantial delays in approvals.

The following section outlines approaches that can improve development efficiency without increasing the overall risk of investment.

A more sensible approach

Bayesian statistics

The concept of p-value is very complex. The vast majority of people without formal training in statistics do not understand it, yet, they often believe they do. Misinterpretations of p-values are well documented, with many mistakenly assuming that p-values represent Bayesian posterior probabilities.²

In contrast, posterior probabilities provide an intuitive probabilistic output. They are calculated by combining prior knowledge with newly collected data to derive updated, posterior probabilities. Each time the posterior probability is calculated, it serves as the prior for the next stage of development. Once the new set of data is collected, an updated posterior probability can be determined. This process, known as Bayesian updating, mirrors the natural learning process about a product, expressed in a mathematical framework.

Posterior probabilities also directly answer questions of interest. For example: “What is the probability that the new treatment is at least 20% better than current standard of care?” Therefore, it is time to adopt posterior probabilities as a measurement for sound and clear decision making.

Interim analysis as investment decision points

The current practice is to invest on a trial-by-trial basis. A more sensible approach is to incorporate interim analyses in trials and make investment decisions incrementally. Early-stage trials are particularly suited for this type of investment because regulatory requirements are less rigid. Bayesian updating is an ideal fit for this approach. After each trial or interim analysis, posterior probabilities can be updated. By doing so, decisions can be made based on a lot more comprehensive information.

This approach does not introduce any regulatory barriers. In fact, there are even Phase 3 examples where investment decisions were based on interim analysis using pre-specified criteria that were accepted by regulatory agencies. One such example is the VALOR trial,³ in which Royalty Pharma agreed to make an additional investment if interim data fell within the promising zone, triggering an increase in sample size.

Rethinking success in early-stage trials

Unreasonably high demands for treatments to advance from one stage of development to another can stall progress, delaying potentially life-saving therapies. By integrating well-established statistical methods, such as Bayesian updating and interim analyses, small biotech companies can make more informed, data-driven decisions without increasing investment risks.

To foster innovation while maintaining financial viability, investors should reconsider rigid statistical significance requirements in early trials. Embracing a more flexible, evidence-based approach can accelerate development timelines, benefitting both patients in need and the companies striving to bring novel treatments to market.

About the author

Image

Zoran Antonijevic is vice president of statistical consulting at Bioforum. He has held executive positions in pharmaceutical companies and CROs and designed well over 100 clinical trials in numerous therapeutic areas, many of which included innovative designs. Antonijevic was a long-time chair and leader of the DIA Innovative Design Scientific Working Group. He has authored numerous papers, scientific presentations, courses and trainings, and was editor of the books “Optimisation of Pharmaceutical R&D Programs and Portfolios” and, together with Bob Beckman, “Platform Trials in Drug Development”.