Maximizing Benefit and Improving Tolerability for Patients Through Dose Optimization – Virtual Event Questions and Answers

Friends of Cancer Research (Friends) continues to be committed to improving clinical trial design and drug development to support patients.

In September 2021, Friends hosted an event called Beyond Breakthrough: Optimizing the Breakthrough Therapy Designation that highlighted ongoing discussions between industry leaders and FDA to identify and support best practices for optimizing interactions and facilitating development of drugs with Breakthrough Therapy Designation (BTD). One discussion that emerged was approaches to dose optimization for drugs with Breakthrough Therapy Designation (BTD). The group created a shortened document, or “snapshot,” to facilitate discussions about dosing between industry and FDA. In March 2022, Friends convened invested stakeholders to continue discussions about opportunities for using the document to inform regulatory conversations.  

To complement these discussions, Friends convened a multi-stakeholder group to discuss opportunities for maximizing benefit and improving tolerability for patients through dose optimization, culminating in a panel discussion during the Annual Meeting in November 2021. During the event, over 50 questions were submitted to the panel.  The working group responsible for authoring the Annual Meeting White Paper on dose optimization has provided answers to selected questions below, which are also included as an appendix in the document 

The initial questions below were asked of and answered by FDA representatives, while subsequent questions were answered by remaining members of the working group. As always, sponsors should engage FDA with questions specific to their drug development program.

Questions Answered by FDA 

Does FDA anticipate releasing a draft guidance on the topic of dose optimization for oncology drugs?  

The focus of Project Optimus is to emphasize the importance of dose optimization early and as a component of pre-market drug development. The FDA Oncology Center of Excellence is developing a draft guidance on dose optimization that would be broadly applicable across development programs. Within this framework, FDA is interested in working with stakeholders to ascertain how to apply the general recommendations to their specific development strategies.  

When is it best for sponsors to engage with the FDA regarding dose-finding study design and what factors might influence timing of these discussions?  

Sponsors should recognize the importance of dose optimization and consider initiating discussions about dose-finding and dose selection early in clinical development. FDA is very open to providing feedback during pre-IND meetings and would be open to continued updates as more data from the development program become available. Discussions about dose-finding paradigms do not necessarily need to be tied to milestone meetings. As was mentioned in the White Paper, tools may be explored to facilitate timely exchange of key considerations between sponsors and FDA to support evidence generation for dose optimization. 

Are there opportunities to combine data from different doses if there are no differences in response rates? 

Whether or not response rate data from different doses could be combined depends on the question(s) being asked and answered. There may be opportunities to use data from patients receiving multiple doses to support regulatory decision-making at different stages of development. However, such an approach would have to be discussed with FDA and must be backed by scientific rationale and be based on the totality of data, which includes overall response rate (ORR), duration of response (DOR), dose-toxicity relationships, and supportive pharmacologic data. 

Would patients participating in dose optimization studies be permitted to cross over to an alternate dose?  

Yes, there may be instances where it would be desirable for patients to receive an alternate dose from the dose assigned at trial initiation. For example, emerging safety data could suggest that one of the doses is associated with worse toxicity. In a case like this, patients could be permitted to cross over and receive the preferred dose. Alternately, a prespecified analysis may show that two doses have similar efficacy, but one dose has better tolerability. In this case, it would be reasonable to allow patients to cross over to the better tolerated dose. Sponsors are encouraged to discuss their plans to allow patients to cross over with FDA during the trial design phase. 

Are there specific considerations for drugs being developed under an Expedited Pathway?  

Timely availability of evidence to support dose selection is essential for all drugs, particularly those with expedited development timelines (e.g., a drug with Breakthrough Therapy Designation). It is important that sponsors have a plan for dose optimization early and communicate with FDA at relevant milestones. Randomized evaluation of multiple doses can support dose selection, including for drugs with a rapid development timeline.

Questions Answered by Remaining Panel Members 

What endpoints could help determine the initial dose range for dose-finding studies? How will findings from those studies inform the dose(s) used in subsequent studies?  

The endpoints should support the overall goal of determining a dose that is safe and effective and does not result in unnecessary toxicities. In the initial dose-finding trial, a variety of biomarker endpoints can be used to assess potential efficacy, including both imaging and blood-based biomarkers. Invasive tumor biopsies are unlikely to yield reliable data in this context. Endpoints that are selected should provide sufficient data to achieve the objectives of both initial dose-finding studies and later dose optimization studies. Initial dose-finding studies should focus on identifying a range of active doses for subsequent randomized dose optimization studies, prior to initiation of a registration trial. 

An integrated pharmacokinetic (PK)/ pharmacodynamic (PD) analysis approach may help to interpret early clinical data. It may also be beneficial to review and leverage data from other compounds in the drug class. Models based on preclinical, competitor, or other data may also support a hypothesized target for plasma exposure or effect on a blood-based biomarker.  

The early phase dose-finding studies should attempt to include a comprehensive analysis of the relationship of dose and exposure to both efficacy (using radiographic and blood-based biomarkers) and safety (using toxicity biomarkers, including both clinical events and other conventional toxicity biomarkers). The dose selected for the registration trial should be based on the dose optimization trial, and the lowest dose should be used if there is neither a dose-response nor exposure-response relationship. 

Why is randomization for the dose-finding trial so important? How is this conducted without powering the study? 

Sufficiently sized RCTs are important to understand the relationships between dose and toxicity as well as dose and efficacy. Randomization is critical because there are multiple potential sources of bias, especially regarding patient selection at higher doses. Nonrandomization can result in misleading conclusions because of these biases or intertrial variability.  

Initial dose-finding trials do not necessarily need to be powered to determine statistical superiority but should allow assessment of the general shape of the dose-response curve. The goal of these studies is not to prove whether one dose is superior or inferior to another, but rather to ascertain whether higher doses, usually associated with greater toxicity, are likely to have a superior therapeutic index to lower doses. This assessment can be made with a relatively small sample size, particularly in the context of an extensive preclinical data package. Ultimately, when the assessment of two or more doses leads to comparable efficacy, the lowest effective dose should be used in the registration trial. 

Are there specific considerations for drugs given in combination? 

A dose established for one indication may or may not be the optimal dose for a different population, including for use in combination with other drugs. As such, additional dose optimization should be considered as part of development plans for drugs used in combination, whether both drugs are unapproved, or an unapproved drug is being combined with an approved drug or regimen. Part of the rationale for additional dose-finding studies when drugs are used in combination is that the drugs may have overlapping toxicities that when combined, can become intolerable.  

It may be necessary to have or generate data regarding the contribution of each component in the selected combination. In some cases, it may be feasible to use a lower dose of an originally approved drug without extensive dose-finding studies. Sponsors should consider and discuss with FDA whether the standard of care arm should also utilize a dose lower than the labeled dose.