Early-Phase Dose-Finding Trials in Virology

ABSTRACT

Viruses have taken central place in public health due to the ongoing severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic. Antiviral treatments, and combination of antivirals, can be effective at reducing viral load shortly after infection, improving long-term outcomes. While it is obvious that choosing the right dose of treatment is a fundamental consideration, little has been published in terms of methodology for dose-finding trials in virology. The considerable progress in dose-finding methodology of the last few decades has focused almost entirely on oncology. However, the framework developed in oncology does not apply 11 to virology. While adverse reactions to cytotoxic drugs may be life threatening, for anti-viral agents, we anticipate something different side effects that provoke the cessation of treatment. This would correspond to treatment failure. On the contrary, success would not be yes/no but would correspond to a range of responses, from small, no more than say 20% reduction in viral load to the complete elimination of the virus. Less than total success matters since this may allow the patient to achieve immunemediated clearance. In this presentation, we will introduce a novel methodology whose goal is twofold first, to identify the dose that provides the most favorable distribution of treatment outcomes, and, second, to do this in a way that maximizes the treatment benefit for the patients included in the study. We will compare two modeling approaches in the talk. The first approach relies on the Bayesian Dirichlet-Categorical model to describe the toxicity/efficacy profile of each of the dose levels. The second approach relies on the principles of the continual reassessment model (CRM). We separately model the dose-toxicity curve and the dose-efficacy curve. By representing efficacy with three categories (low, medium, high viral load reduction), we can use the following assumptions dose-efficacy curve is decreasing for the low-response category, and dose-efficacy curve in increasing for the high-response category. By combining the modeled toxicity and efficacy curves, we obtain the center of mass curve over the dose levels of interest. We will compare both approaches via simulations. The first approach described above has been recently published in Statistics in Medicine (doi 10.1002/ sim.8771). The second approach is being currently developed and tested and will be the topic of a future publication.