5 Reasons to Integrate Model-Based Meta-Analyses (MBMA) Into Your Clinical Development Strategy

Posted by Cytel

May 31, 2018 3:48:00 PM

By Esha Senchaudhuri

An important trend in clinical development involves integrating strategic pharmacometric analysis with program level decision-making, to make the most use of available data. This can occur in various forms, from leveraging preclinical data for go-no-go decision making [1], to the need for improved comparative effectiveness frameworks [2].

Here we have five reasons why you should consider utilizing model-based meta-analyses ( MBMAs)  for your program or portfolio development.

1: Exploring Possible Endpoints
We know that devising clinically meaningful endpoints is one of the most important steps in demonstrating that a new drug or device deserves a place in the market. It is also one of the main reasons that Phase 3 submissions are called to resubmit [3]. When planning a trial, particularly one where investigators may be unfamiliar with the new therapeutic, it might be helpful to consider a wide array of endpoints to ensure that the correct endpoint is chosen for the trial. This can be a delicate task.
MBMAs can allow investigators to benefit from knowledge gained across several published clinical trials to showcase the benefits and challenges of confirming safety and efficacy for various endpoints. Trial sponsors can move forward knowing that they have made the best available choice from perspectives as wide ranging as patient safety and expected financial value. When coupled with the correct forecasting and simulation tools, they can also highlight some of the possible pitfalls, allowing trial planners to plan ahead.

2: Constructing Surrogate Endpoints
Surrogate endpoints, which are often biomarkers, are extremely useful for detecting early signals of efficacy and preventing long trials with ineffective compounds. They also help to predict the Phase 3 dose. In some cases, where it is considered unethical to use placebo or standard of care in a control arm, literature-derived placebo or standard of care effect levels can help avoid this roadblock. In opioid and other pain related studies, for example, it is not uncommon to use the published placebo response, as opposed to an actual placebo arm in order to avoid having to give any patient an actual placebo. In general, MBMAs are commonly in use in developing new pain treatments [4][5][6].
The construction of surrogate endpoints requires at least two steps: the actual construction of a surrogate and a justification for why this endpoint should act as a surrogate. MBMAs can help to clarify both steps in this process. By harnessing industry-wide data, MBMAs are better positioned to detect trends and provide trial sponsors with a set of alternative endpoints that may be used as surrogates. Using published results of both the surrogate and the registration endpoint, experts can develop a quantitative account of how the surrogate can predict the registration endpoint.
The statistical experts who assess justification can also benefit from access to the findings of such analyses. After all, the analyses would reveal how the surrogate works and inform judgment as to their suitability.

3: Go-No-Go Rules
The construction of Go-No-Go Rules has been a critical part of recent clinical development strategy. As we highlighted in an earlier post, AstraZeneca improved its pipeline substantially by constructing rules informed by quantitative analyses. These rules are frequently driven by determining the target effect sizes, and published literature can be very useful in this regard. For AstraZeneca, using proprietary data from their preclinical studies was an important step in this process.
MBMAs can also inform the construction of such Go-No-Go Rules , allowing statisticians to benefit from expertise gleaned across the therapeutic sector. This means that studies that move to Phase 3 not only have early phase knowledge validating the enterprise, but can also count on numerous other studies that signal the probability of Phase 3 success.

4: Confirming Safety
When we think about confirming safety, many trial sponsors immediately begin to talk about dose-selection. While clearly an important part of any safety trial, there is also the question of how best to communicate safety risks to regulators. Sometimes, what appears to signal a risk within the context of a trial might not appear so when considered across several trials, (e.g. there might be some bias within a particular study design), or in the broader context of the disease population under Standard of Care. Certain adverse events are also best understood at a wider lens that is captured by the industry at large. MBMAs can help to frame the conversation around a risk-benefit ratio, ensuring that anxious regulators get the best safety data available to them about your product, by class, and in the context of the target population.

5: Confirming Efficacy
According to a 2014 JAMA article, over a quarter of approved studies that were initially rejected, experienced delays due to either choice of study endpoints or failure when compared to other endpoints [3]. The delays experienced by all resubmissions amounted to approximately 435 days [3]. When choice of endpoint is the cause of a rejected submission, investigators are left with the question of how to respond. Sector-wide data about endpoint selection, organized with quantitative insights in mind, can be a useful way to navigate this pivotal decision.

Cytel supports customers with a variety of quantitative pharmacology and pharmacometrics (QPP) projects.  To learn more about our services in this area, click the button below.



[1] Cytel Blog: Maximizing Preclinical Knowledge for Optimal R&d: The AstraZeneca Case Study
[2] Boucher, M. and Bennetts, M. (2016) ‘The many flavors of model-based meta-analysis: Part I – Introduction and landmark data,’ CPT: Pharmacometrics & Systems Pharmacology, 5(2) pp. 54 – 64.
[3] Sacks LV, Shamsuddin HH, Yasinskaya YI, Bouri K, Lanthier ML, Sherman RE. Scientific and Regulatory Reasons for Delay and Denial of FDA Approval of Initial Applications for New Drugs, 2000-2012. JAMA.2014;311(4):378–384. doi:10.1001/jama.2013.282542
[4] Demin, I., et al. "Longitudinal model‐based meta‐analysis in rheumatoid arthritis: an application toward model‐based drug development." Clinical Pharmacology & Therapeutics 92.3 (2012): 352-359.
[5] Mercier F, Claret L, Prins K, Bruno R. A Model-Based Meta-analysis to Compare Efficacy and Tolerability of Tramadol and Tapentadol for the Treatment of Chronic Non-Malignant Pain. Pain and Therapy. 2014;3(1):31-44. doi:10.1007/s40122-014-0023-5.
[6] Colloca L, Enck P, DeGrazia D. Relieving Pain using Dose-Extending Placebos: A Scoping Review. Pain. 2016;157(8):1590-1598. doi:10.1097/j.pain.0000000000000566.


Topics: pharmacometrics, biostatistics, pharmacology, meta-analysis, quantitative decision-making

The Cytel blog keeps you up to speed with the latest developments in biostatistics and clinical biometrics.  Sign up for updates direct to your inbox. You can unsubscribe at any time.


Posts by Topic

see all

Recent Posts