Modernizing Clinical Trials: Digital Technologies And The Cloud

Published by FiercePharmaModernizing Clinical Trials:Digital Technologiesand the Cloud1

2Modernizing Clinical Trials: Digital Technologies and the CloudDeveloping and bringing a new drug tomarket is costly, with estimates rangingfrom 650 million to 2.9 billion, a 145percent increase over the last ten years.1Execution of clinical trials is historically the mostexpensive part of the process, accounting for aroundtwo-thirds of total R&D costs 2 , and taking upwardsof 10 to 15 years to complete. Factors contributingto these high costs include enrollment of patientcohorts that are not optimal for the trial, unproductivetrial sites, low patient engagement and difficulty insharing and analyzing resulting data.companies continue to struggle with the challengesThe problems begin with the limitations of sponsors’of enrollment, engagement and data sharing. Againstability to identify patients in the active disease statethis backdrop, pioneering companies are turning to theand determine the best sites to enroll them. By oneAWS Cloud to modernize their clinical trials with thecalculation, around 30% of clinical trial sites neveranalytic capability to optimize studies by predictingenroll a subject. These unproductive trial sites delaythe right patients to enroll, securely coordinatingdata generation and cause a large amount of wastedand sharing data and incorporating regulated digitaleffort, which contributes to the high cost of drugtechnologies.3development.Similarly, once a trial is underway, sponsors need agileHOW SCALABLE CAPACITY CANSTREAMLINE TRIAL DEVELOPMENTtools to incorporate the use of mobile technologies,Companies such as Bristol-Myers Squibb are performingsuch as wearables and mobile devices, which couldin silico clinical trial simulations using cloud-based high-generate a wealth of data while promoting betterperformance computing (HPC) to help them optimizepatient engagement. To compound matters, thethe design of early-phase trials. Bristol-Myers wasdifficulty associated with sharing, aggregating andable to reduce total analysis time by 98% by runninganalyzing trial data can limit the speed and efficiencysimulations in support of a pediatric phase 1 study onof studies at each step in the process.the AWS Cloud instead of their on-premises system.4 Itcan be challenging to scale the computational powerThe efficient generation of high-quality clinical trialnecessary to run trial simulations on-site, as it ties updata is critical to success for new therapeutics. Yet,resources and causes bottlenecks for other researchers.However, virtually unlimited HPC resources can G THE GLOBAL BIOMEDICAL PULSE TheBiopharmaceutical Investment & Competitiveness (BCI)Sur vey – 2015. Available at: https://www.pugatch-consilium.com/repor ts/BCI%202015%20 -%20Measuring%20the%20Biomedical%20Pulse.pdf. (Accessed: 5th September 2018)Non-Enrolling Sites Come at a Price Geeks Talk Clinical.Available at: a-price. (Accessed: 5th September 2018)Published by FiercePharmaaccessed on-demand in the AWS Cloud.“Because of the compute capacity that we’re able toderive from AWS, we can now hand out dedicated4AWS Case Study: Bristol-Myers Squibb. Amazon Web Services,Inc. Available at: ol-myers-squibb/. (Accessed: 11th December 2018)

3Modernizing Clinical Trials: Digital Technologies and the Cloudcompute environments to our PK scientists. Instead ofcomplexity and patient recruitment and retention. 6being able to run hundreds, they can now run thousandsThe current process of protocol design involves aof clinical trials to prepare these optimized designlabor-intensive process by which researchers useruns,” Russell Towell, a Senior Solutions Specialist atbenchmark studies and historical data to predict theBristol-Myers Squibb, said. 5 By using these cloud-protocol feasibility based on its projected rate ofbased trial simulations, they were able to reducerecruitment, which are often immediately outdatedthe number of subjects by one-third, and reducegiven an ever-evolving competitive landscape.the length of the study by almost a year.To aid in these effor ts, companies are looking toUtilization of highly scalable, on-demand cloud-artificial intelligence and machine learning (AI/ML)based computation power for in silico clinical trialto help expedite and optimize patient recruitment.simulations can help in the upfront optimization ofFor example, Knowledgent has built an Intelligentthese trials, decreasing patient and trial site burdenTrial Planning (ITP) application on the AWS Cloud thatand potentially speeding time to completion.uses AI/ML to predict the feasibility of clinical trialsOPTIMIZING PATIENT AND TRIAL SITESELECTION WITH ADVANCED ANALYTICSand forecast recruitment timelines. The ITP platformenables study design teams at pharma organizationsto run prediction analysis in minutes, not weeks,The average clinical development program takes10 to 15 years to complete, in part due to protocol5Bristol-Myers Squibb on AWS - Customer Success Story. (2013).https://www.youtube.com/watch?v Vi96Wr x ASgoPublished by FiercePharma6How long a new drug takes to go through clinical trials.Cancer Research UK (2014). Available at: -through-clinical-trials. (Accessed: 11thDecember 2018)

Modernizing Clinical Trials: Digital Technologies and the Cloudplanning helps to facilitate smarter trial planning byUTILIZING HISTORICAL CLINICALTRIAL PATIENT DATA TO INFORMCLINICAL DEVELOPMENTenabling researchers to determine the most optimalComputational power and AI/ML is only one part ofsites, countries and/or protocol combinations.the puzzle, though. These resources can only yieldallowing them to iterate faster and more frequently.Powered by machine learning, real-time scenariomeaningful insights when turned on extensive, timely“Just by eliminating poor per forming sites, trialand accurate data, and when informed by a thoroughteams have the potential to reduce their trial costunderstanding of the natural history of a disease. Pharmaby 20%. And by making data-driven decisions that arecompanies have looked to their prior experience,significantly more accurate, we can plan and executemedical literature and emerging real-world data (RWD)clinical trials faster, leading to hundreds of thousandsto meet this need.in cost savings for every month saved in a trial,” AriYacobi, Chief Data Scientist at Knowledgent, said.However, none of these resources are per fec t.Published literature is static and covers just a fewTo accomplish this type of predictive power largedata elements about one trial at a time. RWD isamounts of data needs to be securely stored andfar more voluminous but can be undermined byquickly shared, a benefit of utilizing a cloud-baseddifferences between patient populations, unsystematicplatform. The ITP application leverages the securedata collection and a limited geographic coverage.storage capacity of AWS, as well as other AWS services,Finally, the sponsor’s own historical clinical trial datato clean, aggregate and integrate the data and ensureis inherently limited to the scope of its earlier studiesthat data scientists have the ability to query it. Theand is laborious to standardize for meta-analysis. Alluse of the highly scalable cloud-based computationalof these shortcomings hinder researchers’ ability topower and advanced machine learning algorithms onmake data-driven decisions.AWS Cloud can help pharma companies to efficientlyand effectively predict and plan clinical trial design andRecognizing the need for fit-for-purpose data addressrecruitment, decreasing overall clinical trial timelines.these challenges, Medidata has made available a poolPublished by FiercePharma4

5Modernizing Clinical Trials: Digital Technologies and the Cloudof standardized clinical data covering thousands ofproducts. Subsequently, these mobile technologiestrials from multiple sponsors that enrolled millionsrely on the cloud for secure and reliable transfer ofof patients in all parts of the world. Medidata hasdata to trial sponsors.stored the data on the AWS Cloud and through theirSynthetic Control Database is providing a uniqueAs the Clinical Trials Transformation Initiative (CTTI)patient data source that is available, actionable andnoted in recent recommendations, 7 systems thatconsumable to improve decision making in trialsensure the authenticity, integrity and confidentialityfrom design through submission.of data over its entire lifecycle are essential to theeffective adoption of mobile technologies. TheseBecause Medidata’s database is built on the AWSsystems must pull in data from multiple sourcesCloud, researchers can interact with the Syntheticand make it available to authorized users, withoutControl Database via a highly customizable and flexiblecompromising security.visualization tool built entirely with AWS services.Among the many benefits of using the SyntheticAWS IoT services like AWS IoT Analytics and AWSControl Database, drug sponsors can more preciselyIoT Device Management excel at connecting suchcalculate sample sizes, improve statistical accuracy,physical devices to the cloud for secure data collection,and understand the natural history of disease andmanagement and analysis. Teams across pharma andbackground rates of serious adverse events. Companieshealthcare companies want to incorporate wearablesthat use the resource to improve decision-makingor other devices into their trials and in some cases arestand to increase accuracy, reduce costs and shortenalready pulling IoT and mobile device data into securetimelines in clinical trials.cloud environments for analysis. For example, withUNLOCKING THE POTENTIAL OFMOBILE TECHNOLOGIESthe support of uMotif software on the AWS Cloud,Manchester University in the United Kingdom ran aresearch program looking at the effect weather had onClinical trial sponsors have traditionally generatedpeople suffering from rheumatoid arthritis and chronicmost of their data when patients visited trial sites,pain. According to uMotif CEO Bruce Hellman, 13,500but have lacked the ability to continually assess thepatient participants used the app on their tablets oreffect of their therapies on people as they went aboutsmartphones, provided 38 million data points andtheir everyday lives. The emergence of smartphones,wearables, ingestibles and other devices has givensponsors new capabilities, enabling them to seea fuller picture of the safety and efficacy of theirPublished by FiercePharma7CTTI Recommendations: Advancing the Use of MobileTechnologies for Data Capture & Improved Clinical Trials.Available at: ations.pdf#Pg2Ln13. (Accessed: 7th September 2018)