Modeling and simulation utilize existing and prospectively collected data from Phase 1 and 2 clinical studies, nonclinical studies, and from the literature to provide important insights on a drug’s effectiveness and safety. This can be accomplished by examining the variability in drug concentrations and the relationship between drug concentrations and clinical outcomes (e.g., measurements of disease activity) or safety and tolerability (e.g., rates of adverse events, changes in vital signs). The insights gained from modeling and simulation can be used to select dose regimens to optimize benefit-risk ratios, inform clinical trial designs, predict trial outcomes, and much more. Importantly, modeling and simulation can also help design Phase 3 studies and interpret Phase 3 results which can sometimes help avoid an unnecessary Phase 3 failure when things do not go as planned.
Phase 3 Clinical Trials
Phase 3 trials are designed to establish efficacy of an investigational drug in a disease state of interest. In Phase 3, the “gold standard” is to conduct two randomized, double-blinded, placebo-controlled (R-DB-PC) trials in a large number of patients (e.g., 300-600 patients) with the disease of interest. If both of these studies show statistically significant and clinically relevant improvements in the endpoint of interest, then false positives due to bias or chance events are considered to be unlikely and effectiveness is considered to have been substantially demonstrated. Drugs in Phase 3 only have about a 60% chance of successful approval. Many Phase 3 trials fail to meet their primary efficacy endpoint because of:
- high placebo response
- lack of adequate power
- having the wrong dose
- random variability
- lack of a relationship between biomarkers used in Phase 2 and the clinical efficacy endpoints used in Phase 3
Sometimes, Phase 3 trials fail for study design or implementation reasons and not due to a true lack of efficacy. When this happens, patients do not receive medications that could potentially meet an unmet medical need. In addition, Phase 3 trials are expensive and resource intensive. Therefore, it is important to make sure that Phase 3 trials are designed appropriately so that valuable resources are not wasted. Modeling and simulation is an important tool that can help avoid unnecessary Phase 3 failures.
Benefits of Modeling & Simulation for Phase 3 Trials
Modeling and simulation can help optimize the design of Phase 3 trials, including choosing the optimal dose and understanding placebo response. Modeling and simulation can also help interpret clinical trial results and can add to the weight of evidence for effectiveness using exposure-response or dose-response analysis. The need to be conservative in making conclusions of efficacy and the need to increase access to patients with unmet medical need as quickly as possible is a challenging balancing act that drug developers often face. Modeling and simulation can help make these critical and ethical decisions with greater confidence. Choosing the optimal dose is crucial for the success of Phase 3 trials. Insights gained from exposure-response analyses for efficacy and safety from Phase 2 studies help with choosing the most appropriate trial design and dose. However, sometimes the optimal dose is not studied in Phase 3. In that event, modeling and simulation techniques such as exposure-response analyses of Phase 2 and 3 data may be used to support the approval of a dose that is not actually studied in a Phase 3 trial. For example, exposure-response analysis of a drug used to treat Cushing’s disease suggested a lower starting dose than originally proposed based on pre-specified criteria. In another example, exposure-response analysis of a drug used to treat ulcerative colitis suggested that higher doses were needed in nonresponders (with a proposed study as a post-marketing commitment). Modeling and simulation also helps drug developers determine the best way to evaluate the endpoint of interest. Understanding the placebo response to the endpoint of interest is also critical because an increased placebo response between Phase 2 and Phase 3 can lead to Phase 3 failures. For example, placebo response modeling helped design a Phase 3 trial for irritable bowel syndrome (IBS) in which the primary endpoint was whether patients had adequate relief of symptoms on a weekly basis. Using combined placebo data from Phase 2 and Phase 3 IBS trials, a longitudinal placebo response model was developed. Simulations were then conducted that showed that the probability of success was lower using this endpoint than using an endpoint that utilized an average response. Therefore, modeling and simulation was able to optimize the endpoint chosen for a subsequent Phase 3 trial.
How Modeling & Simulation Can be Used with a Single Phase 3 Trial to Support Approval
In some cases, the need for a duplicate Phase 3 study can be eliminated with the help of modeling and simulation. If the mechanism of action is well understood, then an exposure-response analysis that convincingly supports causation of benefit can be used as confirmatory evidence in combination with a single well-controlled Phase 3 trial to establish benefit. This strategy can be especially useful in hard-to-study patient populations such as in oncology or rare diseases. One example of this approach being successfully implemented was in the approval of a drug to treat Muckle-Wells syndrome. In this case, a single trial of 31 patients (15 on active treatment, 16 on placebo) along with exposure-response analysis of biomarker data from studies in other indications was used to support drug approval for this indication. Modeling and simulation can also be used to understand when one of two Phase 3 trials fails. Bhattaram et.al described when one of two well-controlled studies for a debilitating neurological condition failed. In this example, the FDA developed a dose-change in symptom score relationship, which provided adequate evidence of efficacy and alleviated the need for an additional clinical trial.
Modeling and simulation can be used to optimize Phase 3 study designs, interpret trial results, and add to the weight of evidence for efficacy. In some cases, modeling and simulation can be used in combination with a single well-controlled Phase 3 study, where the mechanism of action is well understood, to circumvent the need for a second confirmatory Phase 3 study. The use of modeling and simulation to optimize Phase 3 trials requires adequate planning and the collection of appropriate data throughout drug development. Allucent has a diverse and experienced team comprised of experts in designing Phase 3 trials and building custom models tailored to fit your drug development program’s needs. Contact us to learn more about Allucent’s modeling and simulation and study design services.