Industry Research: CNS Trials are in Dire Need of Study Optimization

Scientists and experts have conducted extensive research on protocol optimization and the need to enhance study efficiency, and sponsors are starting to look at their study design strategies.  However, which diseases should sponsors focus their resources on?  Industry benchmarks can assist with prioritizing such decisions.

Research suggests that clinical studies are overly complex, and are poorly designed.  For instance, the total median procedures per protocol increased from 105.9 from 2000-2003 to 166.6 from 2008-2011 [1].  Moreover, non-core data (i.e., data supportive of exploratory/secondary endpoints), represented 50% of total clinical trial procedures [1].

In support of this research, we evaluated clinical trial efficiency trends on an industry level.  The ability to access ‘Big Data’ from FDA databases enables business analytics firms to extrapolate metadata and offer strategic insights.  The objective of this analysis is to evaluate clinical trial design inefficiency by gauging industry-wide patient utilization rates for oncology, cardiovascular and neurology trials.

Time Series Analysis: CNS Trials Exhibited the Lowest Efficiency in Patient Utilization

Figure 1 illustrates total clinical trials started in oncology, cardiovascular and neurology indications from 2004-2012.  Trial tapering after 2009 suggests saturation for some of the indications (particularly oncology), and trial growth is unlikely to sustain given current conditions.

Figure 2 illustrates the total number of expected patients enrolled for all clinical trials in a specific disease indication.

Cardiovascular trials utilized the most patients, with oncology and neurology trials following, respectively.  It is important to emphasize that the rate of change for patients enrolled increased continually after 2009, which does not follow tapering clinical trial patterns after 2009 (Figure 1).  Henceforth, we evaluated the rates of change in clinical trials compared to the rate of change in clinical trial subjects, and implemented time series algorithms to determine patient utilization rates from 2004-2012 (Figure 3). 

This algorithm essentially eliminates bias perspectives from the quantity of patients in clinical trials, and disease specific enrollment patterns, and focuses on patient utilization efficiency per trial; in other words, how many patients are needed per clinical trial, with all other factors constant. As illustrated in Figure 3, negative patient utilization ratios indicate that the rate of change in patients is rising more than the rate of change in clinical trials, suggesting inefficiency (and vice versa for positive figures, meaning more efficiency in utilizing patients). 

Figure 3 demonstrates that cancer trials are most efficient at utilizing patients over time (2004-2012), whereas neurology trials are the least efficient.  This may suggest that neurology trials are poorly designed compared to oncology trials, as more patients are needed in order to achieve desired regulatory and study outcomes.  Additionally, research suggests that CNS studies exhibited a 13.4% non-core procedure incidence rate [3].

What Should Sponsors & CROs Do to Improve Trial Design Efficiency?

There are several methods that sponsors and CROs can readily implement to improve trial design. (1) Use Industry Benchmarks:  By accessing advanced analytical assessments, sponsors and CROs can leverage customized industry-based analytical benchmarks with solid expertise in clinical operations to gauge their study design. (2) Conduct Analytical Enrollment Viability: Sponsors and CROs can now ‘Validate’ protocols on real-time patient data and optimize inclusion/exclusion criteria, so that protocols ‘Match’ patient populations; this method reduces timeline slippage, and costs associated with amendments. (3) Optimize Post-Marketing Endpoints in Study Design: According to the aforementioned literature, it is important to include secondary endpoints in study design, however, it is wise to plan post-marketing data analyses in advance in order to minimize unnecessary data collection during clinical trials.

* assumes that all trials in all phases are increasing at similar rates.

References:

[1] http://www.iom.edu/~/media/Files/Activity%20Files/Quality/VSRT/LST%20Workshop/Presentations/Getz.pdf

[2] http://www.appliedclinicaltrialsonline.com/appliedclinicaltrials/Blogs/How-to-Fix-the-Protocol-Complexity-Problem/ArticleStandard/Article/detail/836600?contextCategoryId=49914

[3] http://www.appliedclinicaltrialsonline.com/appliedclinicaltrials/article/articleDetail.jsp?id=808505&pageID=1&sk=8ace64eaca8a49eb0be76bfbde57d39b&date=