Modeling and Simulation: Optimizing Phase 3 Study Design to obtain FDA Approval

Modeling and Simulation: Optimizing Phase 3 Study Design to obtain FDA Approval

Modeling and simulation (M&S) is transforming how drugs are developed, and how they undergo regulatory review. Software, such as Phoenix NLME and Trial Simulator, provides the ability to use existing data from animal studies and clinical trials to improve efficiency and productivity, reduce uncertainty, and make more cost-effective choices for drug development at sponsor companies.

Regulatory agencies, more specifically the Food and Drug Administration (FDA) in the United States, are accepting and promoting the use of M&S to help speed up approvals of drug and biological products. In 2018, the FDA started a Model-Informed Drug Development (MIDD) Pilot Program and offered sponsors the opportunity to discuss MIDD approaches for pharmaceutical development programs. Additionally, the FDA created M&S modeling groups to support expanding this field and provide recommendations to sponsors through guidance documents. M&S has become indispensable across regulatory agencies, and these strategies are helping to inform regulatory decisions, predict clinical and population-level outcomes, and protect public health.

How was M&S integrated into drug development and did it help sponsors meet FDA requirements for new drug approval?

One example is Clarus Therapeutics, a small pharmaceutical company that specializes in men’s health, that was developing an oral testosterone (T) replacement product for use in men with T deficiency. When Clarus approached Certara for help, Phase 2 studies were completed, and a Phase 3 study was ongoing. However, Clarus had several concerns with the results coming from the Phase 3 study. First, the mean average T concentration (Cavg) was higher than those observed in the Phase 2 studies. Second, the percentage of men with maximum plasma T concentration values (Cmax) associated with safety concerns were higher than those observed in the Phase 2 studies. Finally, the T Cmax values were likely too high to be acceptable to the FDA per the guidelines for hypogonadal men and would not lead to regulatory approval of the product. Clarus risked not having their product approved if the Phase 3 data did not support adequate efficacy and safety, which would also mean a larger business challenge for their company.

To handle this uncertainty and address these concerns an additional Phase 3 study with a new study design was proposed by Clarus that would bring Cavg (associated with efficacy) and Cmax (associated with safety) down. But would this new study design work? Would it meet the sponsor’s own requirements or those mandated by the FDA? One of the FDA requirements was that a single blood collection point be identified that would function as a relatively accurate predictor of Cavg being in the normal range. Therefore, Certara proposed evaluating various permutations of the Phase 3 study design with M&S to optimize the factors in the design that were most important in achieving the efficacy and safety goals.


To better help Clarus improve their new Phase 3 study design, two regulatory and scientific objectives were important for meeting FDA requirements and establishing efficacy and safety: First, to identify the optimal time window for collecting the “single sample” surrogate (measure of T concentration after dosing). Second, to identify dosing adjustment guidelines throughout the Phase 3 study that could then be translated into the instructions for use and label. Furthermore, the T concentration in the monitored sample should permit assurances that Cavg and Cmax levels were in agreement with the FDA guidelines. The guidelines indicate that pharmacokinetic efficacy endpoints for Cavg are between 300 and 1000 ng/dL and pharmacokinetic safety endpoints for T Cmax are <1500 ng/dL for >85% of men as well as <5% of men with Cmax between 1800 and 2500 ng/dL, and no men with Cmax more than 2500 ng/dL (Figure 2).

Figure 2 PK Safety and efficacy endpoints


What parts of the trial design changed to help achieve the objectives?

Next, it was necessary to consider what parts of the trial design could be modified and evaluated with M&S to meet the two objectives. There were several factors to evaluate, such as the time window during which the “single sample” surrogate could be collected. Would changing the single sample collection at 4-6 hours post-dose to another value improve the percentage of subjects with problematic Cmax? Another factor to change was the T concentration thresholds that prompt the physician to make a dose adjustment. Would reducing or enlarging the threshold for dose adjustment improve the percentage of men with problematic Cmax but still achieve the Cmax within the normal range for more than 75% of men? A final factor to consider was which titration approach would improve the results? The titration approach consists of starting at the lowest dose and titrating upward as needed; or starting in the middle of the dose range and titrating upwards and downwards as needed.


What steps in M&S were needed to answer these questions?

To answer these questions, a population PK (popPK) model was developed in Phoenix NLME to demonstrate that the model is reasonably accurate and predictive to perform trial simulations and modify the study design (Figure 3). The first step was to develop a base model of the Phase 2 and Phase 3 data and evaluate how the covariates of weight, baseline T, and race impact the model. The next step was to demonstrate that the model is reasonably accurate and predictive through goodness of fit checks and visual predictive checks. Finally, simulations were performed in Trial Simulator, so that the parameters in the model can be modified (ie, the study design and various permutations).


Figure 3.Modeling and simulation overview


Check back for my next blog post to see how the Phase 3 study design was changed after reviewing simulations that varied 4 aspects. Also, watch the webinar for more details about how Clarus Therapeutics utilized M&S to improve their Phase 3 study design and ultimately gain FDA approval for their product.

Nastya Kassir

About the Author

Dr. Kassir joined Certara in 2008. She earned a PharmD at the Lebanese University and a doctorate in pharmaceutical sciences at the University of Montreal. Her doctoral research focused on population PK/PD modeling in pediatric transplant patients. At Certara, she is responsible for the analysis of PK/PD data from preclinical and clinical studies and for writing clinical study reports. In addition, Dr. Kassir performs PK/PD and statistical analyses according to state-of-the-art methods and regulatory requirements. She also interprets results and writes clinical study reports. Her expertise is in non-compartmental and compartmental analysis to investigate dose response and population PK/PD relationships to support drug development. She has contributed to numerous drug development decisions and regulatory filings where questions from regulatory authorities (EMEA, FDA, PDMA) were addressed using population PK /PD modeling and simulation. Dr. Kassir has experience in various therapeutic areas and special populations: pediatrics, rare diseases (atypical hemolytic-uremic syndrome, paroxysmal nocturnal hemoglobinuria), Inflammatory diseases (rheumatoid arthritis, psoriasis, Familial Mediterranean Fever), and CNS. To date, Dr. Kassir has authored or co-authored more than 20 manuscripts and scientific communications related to drug development.