How Poor Medication Adherence is Damaging Your Clinical Trial Data and How to Fix it, Fast Key Challenges Solved Get Stronger Data Forecast Trial Costs Improve Statistical Power Power Move Forward Faster Faster

Introduction As a clinical trial professional, you already know that 50 percent of new drug compounds fail to demonstrate efficacy in therapeutic trials¹. The root cause of most failures is a lack of necessary efficacy data to support clinical trial claims. Typically, low quality or underpowered data can be directly tied back to an inability to control and report on patient adherence to drug protocol. With average trial adherence rates of only 43 to 78 percent² (and worse for complex regimens) why are we surprised that trials are kicked back for lack of supporting efficacy data? So what can we do? How can we improve our understanding of patient adherence, and its impact on dosing, efficacy and safety? How can we improve adherence, increase trial data power and improve odds of compound approval? Here are four ways that clinical trial managers, investigators, researchers and pharmaceutical sponsors can take control of compliance, improve data analysis and reporting, and achieve higher quality clinical outcomes.

“Data from non-compliant patients can affect trial results to such an extent that they can make or break a candidate drug.” Martin Lamb, head of biz dev at Clinical Trial Services

Solve key challenges in order to: Get stronger data; Forecast trial costs; Improve statistical power; and Move forward faster. 

1. Tartaglia LA., Complementary new approaches enable repositioning of failed drug candidates. Expert Opin Investig Drugs 2006;15:1295-8 2. Lars Osterberg, M.D., and Terrence Blaschke, M.D., Adherence to Medication, N Engl J Med 2005; 353:487-497, August 4, 2005

2

1

Use the Right Tools to Monitor, Analyze and Improve Data

For efficacy results to be justifiable, clinical researchers need strong, repeatable data. Medication adherence is the main determinant of the quality of that data. Research shows that medication adherence in clinical trials is both poor and highly variable among groups and individual patients, however the tools we generally use to monitor adherence (pill counts, journals, MPR) are highly flawed and often manipulated or misleading. For clinical trials, it’s is critical that we move toward new adherence platforms that support robust data collection and encourage better adherence. Real-time access to data that confirms whether participants are taking the correct dose, on schedule, helps researchers work with participants to stay on track – or remove them from a study. The ability to document behaviors, timelines and other data adds credibility to these decisions, which is not possible through self-reporting, pill counts or other traditional tools.

Considerations Adherence platforms don’t have to be complicated or expensive. And they can add huge value to a trial without requiring participants to do anything differently. Because trials rely on data quality, look for a platform that offers a real-time monitoring solution. The platform should record the time and amount of each dose regardless of patient location, transmit that data to a secure location and potentially offer patient reminders (visual and sound), and via text, email or phone.

In 2011, Adler & Lynch electronically monitored a one pill per day regimen.

Only

34%

of patients managed to take their medications as prescribed.

However, 99% of self-reported patient diaries and 97% of pill counts reported adherence sufficient enough to avoid taking action.

Trial teams should monitor behaviors at the individual participant level to build stronger data sets around adherence and its implications for dosing, efficacy and safety. With this level of data, teams can easily identify participants that are not adhering, and remove them quickly, decreasing wasted time and costs associated with unusable participant data.

3

2

Understand Power and Sample Size

Considerations

Power and sample size are critical components of trial data and justification. But even the best-designed trials can be sunk by low rates of adherence that decrease the required power. This is a huge issue because incomplete data means decisions around dosing efficacy and safety might be made from faulty information and may prevent or delay a trial from moving forward. We all know the basic rules for recruiting: avoid enrolling participants likely to have low adherence rates and focus on trial design and settings (location, etc.) that limit barriers to adherence. Assuming a trial is designed to these basics, what else can we do to manage adherence so we achieve the expected power and sample size? After all, the average FDA clinical trial costs $16,000 per subject and 22 percent of drugs launched in the U.S. still require expensive post-launch dose adjustments.

Calculate Up Front: Based on the sample size required for statistical validity, start by calculating the potential impact of non-adherence. Think about key elements, such as potential patient fall-out or removal due to poor adherence, assumptions for average adherence, variance among individuals/groups and or manipulation/false reporting. Real time information on the time and dose on a patient-by-patient basis offers critical visibility into data and enables correction of non-adherence through counseling and/or removing “professional” or “trouble” patients. By improving adherence by 10 or even 50 percentage points, it’s likely possible to reduce sample size and save costs. This visibility could also help make dosing titration decisions faster or prove a link between adherence and efficacy/safety allowing researchers to make a supportable decision to submit findings to regulators.

Variable implementation creates drug-specific issues of efficacy and safety As adherence rates drop, required sample size increases exponetially

Occasional toxicity

250% Impact to Sample Size required to maintain statistical power

200% 150% 100%

Periodic loss of effectiveness

50% 250% 20%

30% Reduction in adherence

Patient Non-adherence in Clinical Trials: Could There Be a Link to Post Marketing Patient Safety? Dorothy L. Smith, PharmD, Drug Information Journal, 46 (I) 27-34, 2012

40%

Chart Blaschke, Osterberg, Vrignes, Urquhart, 2012, Ann Rev Pharmacol Toxicol 52:275-301 *Lars Osterberg, M.D., and Terrence Blaschke, M.D., Adherence to Medication; N Engl J Med 2005; 353:487-497 August 4, 2005

4

3

Manage Dosing for Best Results

Phase 2 trials are designed to determine correct dosing, define drug-drug or drug-food interactions and collect other data required for the design of Phase 3. But when participants don’t adhere to trial requirements, skip dosages or take medications off schedule, the efficacy data required to support recommendations around accurate dosing and interactions can be highly compromised. Weak efficacy data is a key reason why less than 40 percent of trials move to Phase 3. Ideally, clinical trials want to limit the flexibility around factors like dosing and intervals. According to a 2005 study by Frost & Sullivan, simple forgetfulness is the number one barrier to compliance-accounting for up to 70 percent.

Considerations Up until recently, clinical trial tools for collecting data have been limited and flawed. Data collected using participant feedback or basic monitoring can be inaccurate (knowingly or unknowingly) or simply not provided. With each uncertainty, data is questioned, power reduced and trials fail. Systems like the CleverCap adherence platform can accurately record the multiple data points required to identify strong or poor adherence and critical dosing and timing data needed to understand interactions. When evaluating a platform look for advanced data functions such as: • Ability to record the exact time and amount of each dose regardless of patient location, and transmit real-time activity to a central location where trial managers can access it (via any connected device). • Ability to provide visual and sound reminders and via text, email and phone to potentially improve medication adherence. – Access restrictions to improve patient safety and potential Pill count automation, on an intra-visit basis, which provides a full-time series of dosing patterns. • Robust tracking and summary reports on-demand for analyzing changes and trends. • Ability to measure individual patient medication utilization vs. the prescribed regimen.

In a Phase 3 FDA trial, the CleverCap platform reported

95% adherence

with 76% of patients taking their medications within 30 minutes of the time prescribed.

Patients who were adherent, but missed the dosing window, did so by only one hour on average.

5

4

Move toward DataDriven Adaptive Trials

‘Adaptive’ trials are designed to change course as they progress, relying on data to make changes to protocol instead of waiting for the next trial. When components of this data are accessible in real time, such as whether participants are adherent to protocol, researchers can more easily eliminate uncertainties and focus on true indicators and results, adapting with more confidence. For example, if a particular medication combination or dosage appears to be more successful, researchers might increase the proportion of participants receiving the more successful dosage. Currently researchers have limited tools for making dosing decisions as they do not know if efficacy/safety issues are drug or adherence related. This helps hone in on specific positive efficacy results while reducing ineffective combinations. By having stronger (more powered) data collection on the more effective combinations, a trial can move into its next phase with a more targeted desired outcome and design.

Considerations By using adaptive techniques, researchers can reduce the time spent addressing outcome variations from inconsistent adherence after the trial has ended. With access to real adherence data, researchers know which participant data needs to be set aside for further review or segmentation. Each new segment of data has the potential to: • Improve the accuracy of trial results—and reviewer confidence; • Increase patient safety and drug efficacy; • Reduce enrollment requirements and expenses; • Speed up trial completion and submission; and • Ultimately, improve the chance of compound approval.

Reasons for clinical attrition Failures in phase 2

1%

19%

Efficacy Strategic

51% 29%

S tratPharmacokinetics/ bioavailability Safety

6

Reliability/Timely Feedback/ Granularity of Dosing Data Capture

INTUITIVE AND PASSIVE

CleverCapTM

Prescription Refill Rates (MPRs)

GOLD STANDARD

It helps teams visualize and monitor adherence as well as potentially improve patient behavior via visual, sound, text, email and/or phone reminders.

Blister Packs Printed Calendar Post Visit Self-Reports

Traditional MEMs

Microchips on Pills + Patch/App and/or Wearable

With patient alert features activated, CleverCap has demonstrated up to 60 percentage points of improved adherence rates, resulting in improved adherence rates which result in better trials, stronger data and faster progression through R&D.

eDiaries Directly Observed Therapy Solutions (DoT)

LOW

CleverCap records the time and date each pill is dispensed and reports it to CleverCap’s cloudbased Monitoring, Reporting and Analytics platform. Clinical teams can easily access this timely and accurate data using easy-to-understand reports and detailed metrics via any web browser. Teams may also elect to receive real time alerts regarding missed doses or overdoses according to customizable parameters.

Blister Packs Electronics App Based Self-Reports

INVASIVE TO PATIENT LIFE STYLE

Ease of implementation

Site Visit PK Samples Site Pill Counts

CleverCap™ is the GOLD STANDARD of medication adherence systems for clinical trial teams. It provides dose-by-dose, pill-by-pill monitoring, reporting and analysis of patient medication adherence.

HIGH

Size of circle = Technology Costs, Implementation and Scalability Risks

Data Accuracy / Reliability

For more information contact: Moses Zonana 617.372.0111 or [email protected] clevercap.org

7