Clinical innovation: using digital solutions to deliver the next wave of medicines

In a typical year, AstraZeneca conducts over 240 global clinical trials, involving more than 123,000 patients in around 60 countries. For clinical innovation to deliver life-changing and potentially curative new medicines to patients as quickly and safely as possible, diverse digital and other technologies are increasingly being used to optimise clinical trials to get more medicines to more patients, faster than ever before. The move towards precision medicine is necessitating changes to the design and execution of clinical trials. The upsurge in digital healthcare during the COVID-19 pandemic has catalysed changes in clinical trial recruitment and participation. The ambition is to use technological and digital solutions to reduce the burden for patients and trialists, so that clinical trial participation ultimately becomes part of daily practice.

Designing clinical trials with patients in mind

Incorporating the patient voice into the initial planning process of our trials ensures that our designs are manageable and understandable, as well as feasible and practical.

Much can be learned from previous clinical research. Through Merlin, an internal AstraZeneca artificial intelligence (AI) and predictive analysis tool, some of our study teams are aiming to optimise the design and cost of new trials based on experience from previous study designs, including reducing patient and investigator burden and carbon emissions. The data leveraged by Merlin can help to increase patient recruitment and create diverse patient cohorts that are more representative of the patients typically seen in clinical practice.

Optimising clinical trial design

To get medicines to patients faster, clinical trials need to be more efficient, with fewer delays and lower costs. Adaptive trials, basket studies, platform trials, synthetic control arms and dose optimisation studies are some of the options for achieving these goals.

Introducing real world control arms to clinical trials has the potential to reduce the need for placebos and the burden of study participation for both patients and investigators. Comparing novel agents with routine care could require fewer patients for the overall study. Using control data from matched patients in earlier studies could reduce the need for control arms altogether. Such options would require significant changes to current clinical trial regulations, but there is undoubted interest in their potential advantages for reducing trial burden.

Developing novel endpoints

Through advances in AI and digital technologies, novel endpoints are being identified to inform decision making and better capture the whole disease burden of patients in clinical trials, while reflecting the science and the needs of payers.

In oncology, novel biomarkers based on circulating tumour DNA (ctDNA) or circulating free DNA (cfDNA), are increasingly used to guide patient selection for clinical trials. These technologies are creating opportunities for earlier detection and treatment, and for ongoing monitoring for cancer recurrence before relapse becomes apparent in traditional imaging.

In some asthma and chronic obstructive pulmonary disease (COPD) trials, we are now using CompEx: a novel composite endpoint developed at AstraZeneca that combines exacerbations with other indicators of worsening asthma or COPD.^1-3 This reduces the size and duration of studies needed when only exacerbations are recorded.

In chronic kidney disease (CKD), using a novel endpoint developed by academic researchers has reduced the time it takes to answer important questions about the treatment efficacy. Instead of using an event-based primary endpoint to study the impact of potassium-removing therapy in patients with CKD in the Phase III STABILIZE‑CKD trial, we are using reduction in estimated glomerular filtration rate (eGFR) over time as an indicator of slowing of disease progression. In this way, it is possible to include patients with earlier stage disease when few major events, such as need for dialysis or transplantation or death, would be expected to happen.

In late-stage cardiovascular trials, Automating Identification Detection Adjudication (AIDA) has been developed to accelerate the classification and confirmation of events compared with standard procedures carried out by physicians.⁴ Following a study showing high consistency between automated and expert adjudication of cardiovascular (CV) events (ischaemic stroke, transient ischaemic attack), we are now using the system in several studies, including DAPA‑MI, a registry-based trial in patients following a heart attack.

Reducing the environmental impact

Reducing the environmental impact is another key goal of designing clinical trials with patients in mind and this can happen at multiple levels.⁵ These include decreasing face-to-face meetings, reducing the number of wasted lab kits, shortening shipping times and cutting back on single-use plastic. We conducted a clinical trial lifecycle assessment to identify scope for reducing our trials’ carbon footprint and are now applying this information to reduce the environmental footprint of our studies. Indeed, the design of the DAPA-MI trial resulted in 45 percent fewer emissions compared to similar studies with more standard designs.

Improving the clinical trials experience

We are going beyond site-level recruitment, focusing our efforts on outreach and engaging with patients to ensure awareness of clinical trials is an option for treatment.

Having become familiar with online healthcare and volunteering for vaccine trials during the pandemic, patients are increasingly learning about clinical trials and accessing local participating centres through websites such as Breast Cancer Study Locator. It is hoped that this approach will increase the proportion of eligible patients who choose to participate in trials, from the current three percent and broaden the trial participant diversity.

Digital technologies may also facilitate patient recruitment through collaborations with healthcare services and academia, to identify patients in disease registries and longitudinal cohorts who could be eligible for clinical trials. These resources could also be used for patient follow up. In the DAPA-MI trial, treating physicians in registries can join the study and integrate it within their routine clinical practice, with automated data collection and reduced administrative workload.

Data analytics can identify patients who meet inclusion criteria for a clinical trial from large real‑world datasets, collected from multiple healthcare institutions. Approximately one third of US patients recruited to the Serena-6 trial in metastatic breast cancer have been recruited in this way.

It is also essential to work closely with research coordinators and investigators to understand how new studies can be integrated into clinical research workflows with minimal disruption, as is gaining patient insights during protocol design on the practicalities of participation. What is an acceptable number of clinic visits and duration for appointments? How many investigations and treatments can reasonably be carried out during a visit? What are the logistics of moving between departments, especially for someone who may feel unwell? Trial design must be flexible to accommodate patient needs, variations in infrastructure of participating centres, and patient preferences for in-clinic visits versus online consultations at home.

Harnessing data and digital solutions to augment clinical trial outcomes

Since well before the pandemic, AstraZeneca researchers have been testing clinical biomedical devices, including spirometers to test lung function of COPD patients in clinical trials, ‘home lab tests’ to monitor parameters such as creatinine as an indicator of CKD and ‘wearable devices’ to potentially monitor heart rate and blood pressure. We now estimate that up to 70 percent of data currently collected during hospital visits could be collected from patients at home via online questionnaires and monitoring devices. Today this is approximately only 10 percent.

Patients will not want to juggle multiple devices and apps to report data. Optimising the quality of the patient experience is essential if we are to achieve the benefits digital solutions can offer.

We plan to use Unify, a single app designed by collaboration of patients, healthcare professionals and AstraZeneca to simplify the trial experience for all participants, in 70-80 percent of our studies. Already available in nearly 30 countries and 65 languages, the app links information the patient needs about a clinical trial, including clinic visits and virtual consultations, medication reminders and patient reported outcomes. Trial investigators and clinicians use the same app to connect with patients, eg, for virtual consultations and to coach patients on using devices such as spirometers at home and to access data and support treatment adherence.

As we become confident in the viability and integrity of devices and apps that can be used at home, we move closer to the point of incorporating them into trials at scale. This could significantly reduce clinic visits for patients, administration for trialists and enable us to include patients who would have previously missed out on opportunities due to living too far from participating centres.

Where next for digital solutions in clinical trials?

Clinical innovation is not about a single app or sensor, it is about a different way of working with trial sites that recognises the value of patients and all those who provide their care.

Changing the way clinical trials have been performed for many years does not come without challenges and risks. Consulting widely with regulators as well as with patients and clinicians is vital to ensure the smooth integration of digital solutions into clinical trials.

In the longer term, digital healthcare has enormous potential not only for clinical innovation in clinical trials but in routine patient care. It offers opportunities for earlier diagnosis, faster treatment based on precision medicine and patient-friendly monitoring and ultimately, improved outcomes. 

References

1. Fuhlbrigge A, Bengtsson T, Peterson S, et al. A novel endpoint for exacerbations in asthma to accelerate clinical development: a post-hoc analysis of randomised controlled trials. Lancet Respir Med. 2017 Jul;5(7):577-590.
2. Jauhiainen A, Scheepers L, Fuhlbrigge A, et al. Impact of season and geography on CompEx Asthma: a composite end-point for exacerbations. ERJ Open Res. 2020 Oct 19;6(4):00246-2020.
3. Vogelmeier C, Fuhlbrigge A, Jauhiainen A, et al. COPDCompEx: A novel composite endpoint for COPD exacerbations to enable faster clinical development. Respir Med. 2020 Nov;173:106175.
4. Lea H, Meeson A, Nampally S, et al. Can machine learning augment clinician adjudication of events in cardiovascular trials? A case study of major adverse cardiovascular events (MACE) across CVRM trials. European Heart Journal. 2021 Oct;42(1).
5. Sustainable Healthcare Coalition. A strategy to reduce the carbon footprint of clinical trials. Lancet 2021; 398: 281-282.

Click the link below to download the full Issue 6 2022 for free!

European Pharmaceutical Review Issue 6 2022