How AbbVie Is Introducing Digital with Real Impact into Clinical Trials to Optimize Research
AbbVie’s Dr Michelle Crouthamel is focused on identifying clinical trial tech with real impact for reducing patient burden and impacting trial efficiency. She outlines how she identifies and implements digital health technology into clinical trials to make measurements more precise, for the advancement of medicine.
How do you view digital health technologies as you solve for problems in 21st Century drug development?
Despite the advancement of technologies and data analytics today, many clinical outcome measurements are not quite “21st century”. In clinical research, we are still based a lot on the numerical scale, whether it is patient-reported, clinician-reported, or caregiver-reported outcomes. Lots of clinical trials show placebo effects ranging from 20-50% coming from these scales. In the 21st century, doctors and patients expect “precision medicine” from pharma, i.e. who likely will benefit from the drug and by how much.
So when we think about the “21st-century” outcome measurements, it’s about using technology and data science to improve and make measurements more precise and more objective. We owe that to our patients.
"If we can get broad agreement across patient advocacies, industry, tech vendors, the medical community, and regulators, it will get us to a point of transforming clinical research faster."
What are some areas of improvement from current measurements?
I have two examples. The first is from our work in vitiligo. The current clinical assessment is called “Facial-VASI score,” in which physicians calculate body surface area using the patient’s palm and thumb, and via visual inspection by the investigator to see what degree of depigmentation the patient has. That method lacks precision and introduces a lot of subjectivity into the assessment. When we see an assessment like that, we realize there is an opportunity to improve. In this case, there are imaging technologies that can provide much greater precision on the Body Surface Area (BSA) affected and the actual intensity of the pigmentation area.
Another example is atopic dermatitis. This is a very prevalent autoimmune disease for young children, and the primary patient burden is itchiness. We can’t measure “itchiness” because it’s a sensation, but we can measure the actual scratching event. Up to this point, you could only understand that by asking how much a child scratched during their sleep. How accurate can that be?
Once identified, what are the steps at AbbVie to create an appropriate digital measurement?
The first step is identifying the “right problem” and how the potential digital health technology can help you quantify it better than a numerical scale. There are a lot of things you can measure but not everything you can measure is clinically meaningful. It’s important to establish internal alignment on the intended use of that data, justify what you believe is important to measure, and how to improve the measurement. You also need to see if there are “mature” technologies available that can measure what you want to measure. The next step is to discuss them openly with the regulator, providing scientific justification on why we think this is important to measure in this population and the technology that we propose using, to avoid pursuing something that won’t be eventually accepted by the agency.
While searching and evaluating specific technology, look at how validated the technology is. Digital Medicine Society published a V3 framework: verification, analytical validation, and clinical validation. We typically look for technology that is already V1- or V2-ready, so we can move fast into clinical validation in the population we’re interested in. Then we’ll apply that to the clinical trial: collecting data to understand how the digital measure compares to the gold standard to prove clinical validity. Then, after collecting the data, we circle back to regulators and seek further feedback. If the clinical validation is successful, you may negotiate with regulators using it as a ranked endpoint in pivotal trials.
"The sponsor also needs to set parameters around missing data, because you don’t want the site to feel as though they need to babysit the technology constantly or provide endless tech support."
What are the external challenges in your work that you’re addressing?
I would say it’s a combination of regulatory alignment, maturity of technology, and implementation into clinical trials. One of the areas I keep seeing people struggle to get regulatory alignment is sleep measurements.
Why is it so crucial to get sleep measurements right?
If you speak to patients in the autoimmune disease space, 80-90% of patients will say that pain and inflammation actually affect their sleep. However, if you engage with regulators, they challenge that sleep is not directly connected to the core of the disease, therefore, it may not justify as a meaningful endpoint to be measured.
In atopic dermatitis, we see evidence of people having 20-30 scratch events per hour over the course of their entire sleep, and that their sleep is significantly disrupted. Some people have their total sleep time down to 2-4 hours per night. If the medicine can actually reduce inflammation and therefore reduce scratch events and then people can sleep better, wouldn't that be a meaningful outcome to the patient?
Many people would say yes, but we still need that alignment with regulators because sleep is considered a distal concept, and an indirect outcome measure for autoimmune therapy. If we truly believe we should be measuring things that are meaningful to the patient and the medicine can alleviate that burden for the patient, we need to gain alignment and support from regulators.
"It’s about combatting the myth or preconceived notion that adding DHT in trials will increase burdens for the patients and sites. There are patient research suggesting that patients don’t mind DHT."
How do you deal with potential technology immaturity?
If you believe that there is the right technology that can improve outcome measures for your program, you can develop or co-develop it with the vendor. Although we have many technologies available today compared to 10 years ago, fully validated DHTs are still limited. Be prepared to develop capabilities internally and know-how to design and conduct validation studies. Just as we invested in PRO development 10-20 years ago, we can invest in new technology that could go a long way in the future.
A big value of DHT is that it provides access to and democratizes health technologies so that patients are empowered to manage their disease more effectively. Eventually, many of these sensors could be on the shelf in your local pharmacy: a parent could buy it, or a physician would prescribe it for disease management.
How do you communicate the value of digital alternatives to tried-and-true traditional measurements?
If you’re trying to convince senior leaders to adopt a digital endpoint, show them how using it could have practical implications for the study. We have many examples showing that a digital endpoint has much tighter data, reduced variability, and very minimal placebo effects. That could lead to a significant reduction in sample size, which translates into smaller and faster clinical trials and lower cost overall. The DHT-enable outcome data can inform decision-making on things like pipeline progression, dosing selection, etc.
And from a cross-functional perspective, it’s about combatting the myth or preconceived notion that adding DHT in trials will increase burdens for the patients and sites. There are patient research suggesting that patients don’t mind DHT. In fact, many prefer wearing, for example, a device for 30 days rather than answering the same questionnaire every day for 30 days. We all should be mindful of the burden to patients and sites but it is important to collaborate with the operational team and minimize potential burden, so they’re not fearful that implementing this will cause issues for the trial.
How do you address potential burdens or hesitations from sites about using digital biomarkers?
Invest in better training for the site and provide regular support whenever they have questions. Also, quantify how much time it will actually take to implement the new device, versus the tasks that a site typically has to perform, to demonstrate how it reduces the burden to the site. We have a Digital Lab where we test the usability of technologies to proactively pre-identify potential issues and address them in the site training. It is important to partner with technology vendors to improve usability. If you do that, you will likely gain support from the site and the buy-in.
The sponsor also needs to set parameters around missing data. When using traditional measurements, you might only need 5-6 data points over time. A DHT might have 20,000 data points every day. The sponsor needs to lay out how much missing data is meaningful and tolerable in the protocol, and when the site should reach out for intervention. That’s a crucial piece because you don’t want the site to feel as though they need to babysit the technology constantly or provide endless tech support.
You’re passionate about the industry working together to accelerate the adoption of digital biomarkers in clinical trials. Why is this such an important conversation to have?
Digital health technologies offer promises of improving outcome measures and enabling decentralized trials. To realize these values, technology validation is a foundational element that requires investment. We need to work across companies to achieve industry standardization. For example, if we all believe that scratch is an important measurement for atopic dermatitis in how we measure drug efficacy, we need to standardize that.
And to do so, all the different players have to agree, including patients, HCPs, regulators, and payers.
We do not want the development of these tools to become a barrier to how we develop medicine. If we can get broad agreement across patient advocacies, industry, tech vendors, the medical community, and regulators, it will get us to a point of transforming clinical research faster.
