Expert Insights

Expert Insights

Considerations for Designing an Early Phase Oncology Trial

Considerations for Designing an Early Phase Oncology Trial
19 June 2019

Imagine that the compound your company has been developing to treat cancer is about to move out of pre-clinical development and into a first in human Phase I study. There are many questions to consider, but among the most important is which design should be chosen for this crucial study: the conventional Phase I 3 + 3 design followed by expansion cohorts in particular disease areas/mutations or a model-based design such as the Continual Reassessment Method (CRM), Escalation With Overdose Control (EWOC), modified Toxicity Probability Interval-2 (mTPI-2), or another design. The answer to the design question depends on a variety of factors, including operational considerations and assumptions on the dose-toxicity and dose-efficacy curves. The fate of the compound could be dependent on this decision.

Operational Considerations
For many, a key factor in choosing trial design is the operational ease of applying the design. A common misconception surrounding model-based designs is that the use of such a design substantially increases the operational complexity. There is no denying that the conventional 3 + 3 design is easy to implement in practice; it is the key reason why the design is still widely used despite poor statistical properties with respect to identifying the maximum tolerated dose (MTD). The simple design of the 3 + 3 brings with it a trade-off in terms of how dose-escalation decisions are made: each decision is made without regard to what has occurred at previous dose levels. But in our world where ever-expanding data reigns king, wouldn’t you prefer a design that uses the information available at all dose levels to provide a better estimate of the MTD versus only the current dose?

Many people are surprised to learn that a model-based design can be operationalized just as easily as a 3 + 3 designed study if properly planned. The model used for dose-escalation decisions is built prior to patients being dosed so that the only information needed to determine the dose for the next patient (depending on the design chosen) is the dose level(s) and outcome(s) of the previous patient(s). Therefore, the statistician can run the model and have it available for the escalation meeting quickly after the subject reaches the end of the evaluation period. Another advantage of these model-based designs is that some designs can use partial information about the outcome, meaning Enrollment doesn’t have to halt and wait for the observation period to finish before enrolling the next patient.

Assumptions of the Dose-Toxicity and Dose-Efficacy Profiles
The primary aim of Phase I studies in cancer therapies is to establish the MTD based on the protocol-defined dose-limiting toxicities (DLTs).  The 3 + 3 design inherently assumes that as the dose increases so does toxicity and efficacy. The assumption is that, once the MTD is reached, this is the highest dose level safe to give patients, as well as the dose having the highest efficacy amongst safe doses. Assuming this for cytotoxic agents was typically justified. However, in immunotherapies and targeted agents, these assumptions often do not hold (1).

If these assumptions don’t hold for either toxicity or efficacy, then the drug may fail simply due to the choice of a design. Due to different assumptions, the goal for immunotherapy and targeted therapy trials shifts from identifying the MTD to identifying the Optimal Biologic Dose (OBD), defined as the dose with the highest level of efficacy and a tolerable level of toxicity. In these agents, we begin to see further benefits and flexibility of the newer model-based designs. For example, there are model-based methods designed to handle the different assumptions where the dose-toxicity curve may not increase with dose (e.g., drug combination trials where doses of at least 2 drugs are under investigation), as well as methods to handle scenarios where both the dose-efficacy and dose-toxicity curves may not increase with dose.

Ultimately, there is not a one-size-fits-all design. Conversations should occur during the protocol development stage among the scientists, clinical leads, and statistician to determine which design would be utilized best for a certain protocol. Model-based designs require up-front planning. However, they can be more appropriate depending on the operational considerations and the assumptions of the drug. They also can be straightforward to implement and can lead to studies with more reliable results.

We’re here to offer our expertise in these situations. Contact us to find out how we can help you.

Written by: Christopher Tait, PhD, Senior Biostatistician

(1): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6376915/#djy196-B18

x

Submit RFI / RFP

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

Required

x

Contact Synteract

Tell us how to stay in touch with you:

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

Required

x

 

x

Download Resource

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

*Required

x

To complete your download, please choose an option in the Internet Explorer pop-up window at the bottom of this browser.

x

Opt In

By submitting your information to us through this webform, your business details will be added to our database. In accordance with our Privacy Notice, we may then contact you with marketing information about Synteract that might be of interest to you. We will never sell your details to third parties. You can unsubscribe from marketing communications at any time.

*Required