Modernizing Clinical Trials with Amazon Web Services (AWS)

While digitally transforming clinical research holds great promise, it is a complex undertaking. Managing the volume and variability of healthcare data is challenging for life sciences organizations looking to modernize clinical trials. From IoT to data management to high-performance computing and machine learning, leading life sciences organizations are using Amazon Web Services (AWS) to develop scalable, global, predictable, and secure solutions to modernize clinical trials while mitigating risks.

Modernizing Clinical Trials with Amazon Web Services (AWS)
Modernizing Clinical Trials with Amazon Web Services (AWS)

Making sense of such a variety of data not only requires deep domain expertise but also understanding of how to leverage the scalability of the cloud. Life science companies are working with AWS Partner Network (APN) Partners to develop tools that enable companies to meet the vast data storage and computing needs of precision medicine trials, pressure-test the feasibility of study protocols using AI and machine learning, and integrate mobile technologies into studies to ease patient burdens and reduce costs.

Content Summary

Abstract
A risk-averse industry faces a failing R&D model
Life science organizations and partners are building solutions on AWS
How AWS partners are modernizing clinical trials
Feasibility testing featuring Knowledgent, now part of Accenture
Mobile technologies featuring Philips and Ypsomed
Precision medicine featuring WuXi NextCODE and GMI
Get ready. Start. Modernize

Abstract

While digitally transforming clinical research holds great promise, it is a complex undertaking. Managing the volume and variability of healthcare data is challenging for life sciences organizations looking to modernize clinical trials. From IoT to data management to high-performance computing and machine learning, leading life sciences organizations are using Amazon Web Services (AWS) to develop scalable, global, predictable, and secure solutions to modernize clinical trials while mitigating risks.

Making sense of such a variety of data not only requires deep domain expertise but also understanding of how to leverage the scalability of the cloud. Life science companies are working with AWS Partner Network (APN) Partners to develop tools that enable companies to meet the vast data storage and computing needs of precision medicine trials, pressure-test the feasibility of study protocols using AI and machine learning, and integrate mobile technologies into studies to ease patient burdens and reduce costs.

A risk-averse industry faces a failing R&D model

The pharmaceutical industry’s legacy R&D model is under pressure to find new solutions to fill unmet needs and deliver cost-effective benefits. Game-changing treatments, including a cure for Hepatitis C, cancer immunotherapy, re-programmed T-cells, and gene therapy, have helped patients worldwide—but they’ve also strained healthcare budgets and made clinical trials more complex and costly. Also, trial design and execution can be slowed down by rigid design rules, unrealistic burdens on patients and investigative sites, paper-based data collection, and inefficient monitoring.

As a result, the return on investment (ROI) of clinical research has precipitously decreased. For the top 12 biopharma companies, ROI on R&D sunk to 1.9% in 2018, from 10.1% in 2010. Just one in ten product candidates makes it from Phase I trials to approval. Moreover, even though risk should be decreasing as a product advances through development, a majority (58%) of late-stage pivotal efficacy trials (often called Phase III trials) fail. Recruiting and retaining patients is increasingly challenging: one in five studies terminate or produce compromised data because enrollment lags.

The clinical research enterprise now faces an imperative: adopt new approaches and technologies to more effectively harness the massive amounts of data generated before and during clinical trials to streamline the drug development process. Innovations including precision medicine, advanced analytics, and mobile technologies have the potential to streamline trials, reduce costs, and increase success rates.

Life science organizations and partners are building solutions on AWS

Leading life sciences organizations are using the AWS Cloud at a global scale to support the complex regulatory and compliance requirements of the life sciences industry. AWS can deliver the high availability critical to physicians and patients, the scalability to handle sudden surges, and the storage space to enable data mining for research purposes.

Pharmaceutical companies and researchers who design and conduct clinical trials can leverage AWS’ breadth of functionality to scale as-needed and modernize their analytics platform. This increased flexibility enables them to quickly explore their data sets and generate new insights they were not able to previously. Using fine-grained security permissions, they can give access to relevant data to collaborators, to mine the data all while aligning to the security and compliance requirements of their organization. Life science organizations and partners are building solutions on AWS

Leading life science organizations choose to run their clinical trials on AWS for several reasons, including its:

  • Flexible consumption and contract models that enable virtually infinite scaling
  • Global data availability from datacenters in 22+ Regions and 69+ Availability Zones
  • Ability to meet GxP regulatory requirements
  • Advanced analytics and AI/ML services for deep learning capabilities
  • A broad set of security services that enable data protection, identity and access management, logging, auditing, and traceability.

Additionally, AWS Partner Network (APN) Partners specialize in creating custom solutions for life science organizations. The solutions offered by APN Partners can accelerate precision medicine, make trial designs more feasible, and use smart devices to ensure that patients receive the right treatment at the right time.

How AWS partners are modernizing clinical trials

Feasibility testing featuring Knowledgent, now part of Accenture

It is difficult to accurately predict patient enrollment rates at investigative sites because planners must find recent trials similar to their prospective study to benchmark performance expectations. They then must identify changes in the therapeutic landscape such as competing trials, new drug approvals, reimbursement decisions, regulatory changes, and investigator enthusiasm. Finally, they must pinpoint differences between the benchmark studies and the new study, such as the size of the eligible patient population, study burdens, and site locations. The process is manual, time- and resource-intensive, and often relies on subjective judgments.

Challenge
Clinical researchers must understand precisely how inclusion/exclusion criteria in clinical protocols will impact the timelines of their trials. To optimize site selection, they need an analytic model that uses a fixed set of consistently relevant factors plus the capability to incorporate variable factors, and they need answers in a timely fashion.

AWS Partner solution
To help bring more precision to clinical trails, Knowledgent (now part of Accenture) created the Intelligent Trial Planning (ITP) application on AWS. ITP uses AI and machine learning to predict the feasibility of a clinical trial and the recruiting timeline, and it relies on Amazon S3 for external and internal clinical trial raw data storage. AWS Glue then cleans, aggregates, integrates, and extracts features for downstream analysis. All that data can then be queried using Amazon Athena in S3. Amazon EMR with Jupyter notebooks then enables clinical researchers to run machine learning models and view predictions through Amazon API Gateway for a serverless web front-end.

Results
The ITP application helped customers to augment the trial planning process rather than eliminate expert knowledge—after the rate of recruitment (RoR) prediction is complete, users can adjust, override, or enter additional information. The software:

  • Mines data from disparate sources such as a client’s Clinical Trial Management System (CTMS), clinicaltrials.gov, and other external databases
  • Provides a rate of recruitment prediction in minutes, versus weeks
  • Enables frequent iterations using larger datasets powered by machine learning
  • Improves accuracy with scenario planning (analyzing site/country/protocol combinations)
  • Eliminates poor-performing sites and can cut trial costs by 20% on average
  • Informed country selections with data about the competitive landscape (i.e., competing trials)
  • Refined results with further SME advice and empirical testing, but subjective inputs lessened

Looking ahead
Advanced analytics can make the process of site identification faster and cheaper. Pinpointing target patient populations and identifying high-performing sites within reach of those patients can help keep trials on track and avoid costly delays in development—for example, delays in clinical trials cost developers an average of $325,000 per month.

Mobile technologies featuring Philips and Ypsomed

Using mobile technologies (MTs) for data capture and transmission can streamline clinical trials and make them more patient-and site-friendly, boosting recruitment and retention. For example, smart drug delivery devices can improve protocol adherence, ensuring that more patients take the proper dose on schedule, enhancing the quality of trial data. Currently, investigators and study personnel track dosage counts and other patient-recorded data about treatments by manually transferring it to the study database after each patient visit. In between site visits, study coordinators make personal phone calls to remind patients when to take investigational drugs or answer questions. Simply put, it’s inefficient.

Pharmaceutical sponsors and research organizations that want to remotely supervise patients face a slew of device-related technical complexities. For example, how do you continuously monitor smart devices across regions once deployed? How do you control access to different smart devices? For smart devices to work in the setting of a formal clinical trial, many need integration with a dedicated therapy app. Investigative sites must be able to configure and remotely push updates while relying on secure device-to-cloud communications. Finally, there needs to be a built-in mechanism to ensure the integrity of medication adherence and smart device performance data.

Challenge
Ypsomed, a developer and manufacturer of injection and infusion systems, developed a solution for pharmaceutical sponsors and clinical research organizations (CRO) conducting international clinical trials with protocols that require patients to self-inject their medications at home, on a strict schedule. In this setting, the CRO needs smart devices that both monitor whether patients are administering the study medication correctly and allow study coordinators to communicate with them in real-time. The CRO also needs a remote device management system that functions globally, provides secure device-to-cloud communication, stores data at scale, and is compliant with applicable privacy and security regulations—such as HIPAA and GDPR—across multiple geographies.

AWS Partner solution
Ypsomed built its connected drug delivery solution YDS SmartServices for clinical trial medication adherence monitoring and smart device management on Philips HealthSuite digital platform (HSDP) on AWS. The solution integrates connected medication injection devices, such as SmartPilot for YpsoMate with a full device management solution. AWS provides a global, scalable platform for HSDP to provide 24×7 support for Ypsomed through continuous monitoring and regional deployments. Utilizing Amazon EC2 for computing capacity, Amazon S3 for storage, and AWS IoT Core for device connectivity, HSDP helped Ypsomed secure device-to-cloud connectivity and data storage to meet their regulatory requirements.

Results
The Ypsomed connected drug delivery solution:

  • Was developed from scratch in five months using HSDP’s PaaS and 24×7 operations support, reducing Ypsomed’s overhead, infrastructure, and staffing costs
  • Enables clinical trial participants to self-administer medication using the SmartPilot device and CROs to leverage the YDS SmartServices to more effectively monitor and device usage remotely
  • Segregates the encrypted injection, patient, and device performance-oriented data for CROs
  • Empowers CROs to build their clinical solutions on top of the system backbone or integrate with an existing clinical dashboard through a standard interface
  • Facilitates ingesting data from multiple healthcare and consumer sources (HL7, DICOM, IoT) and provides an audit trail on data elements
  • Enables life cycle management processes for the devices, independent of therapy areas
  • Manages both device and cloud to implement real end-to-end security and to meet relevant requirements around data privacy and security

Looking ahead
The consequences of poor protocol adherence include failure to confirm efficacy, risk of harms due to misleading labels, impaired development of breakthrough drugs, and treatment failures. However, to improve adherence, smart devices need to be part of a digital system that enables real-time tracking of the injection date, time, and success or failure (for example use errors). The system must also guide patients throughout the injection process and transmit relevant data across protocols and interfaces seamlessly. Solving the multi-dimensional challenges of building an integrated system requires both technical and clinical experts. Providers who lead in developing solutions—and companies who lead in adopting them—will gain a competitive advantage over those who wait and watch as smart devices and other mobile technologies such as sensors and wearables transform clinical trials.

Precision medicine featuring WuXi NextCODE and GMI

Precision medicine has been hailed for its potential to transform drug development—and for oncology and rare diseases, it has already done so. Genomic discovery is a crucial enabler for precision medicine in the future. However, handling and interpreting massive amounts of genetic data within the traditional drug development infrastructure has proven difficult. Both deep domain expertise and unprecedented computing power are required. Standard workflows and tools can’t process or interpret the sheer volume of next-generation sequencing (NGS) and other -omics data.

Challenge
Global biopharmaceutical companies need a better-informed discovery process to improve drug development and follow-on clinical trials. The average drug costs $2.87 billion (in 2013 dollars) to bring to market, according to a recent study by the Tufts Center for the Study of Drug Development, and only about 63% of product candidates make it to Phase 2 while just 30% advance to Phase 3, according to a 2016 analysis by the Biotechnology Innovation Organization, Biomedtracker, and Amplion. For patients to see greater benefit from the discovery, the industry needed a better means to drive insight.

AWS Partner solution
Life-sciences startup Genomics Medicine Ireland Limited (GMI), and AWS APN Partner WuXi NextCODE, the global contract genomics organization, announced the launch of a long-term strategic alliance with one of the top global biopharma companies to conduct population genomics research in Ireland aimed at advancing the discovery and development of novel therapeutic approaches to a range of serious diseases.

GMI has the largest whole-genome sequencing (WGS) lab in Ireland, and one of the largest in Europe, and works with a network of Irish hospitals. WuXi NextCODE’s purpose-built informatics platform—the genomically ordered relational database (GORdb)—stores the data, while the Sequence Miner and Clinical Sequence Analyzer applications are used to analyze it. GMI’s WGS results are delivered and stored in Amazon S3 Buckets. Amazon EC2 Spot Instances—among other AWS services such as AWS Lambda and RDS—are employed to run a secondary analysis of sequencing. AWS exclusively hosts GMI’s sample collection and processing systems. GMI also hosts an instance of GORdb, Sequence Miner and Clinical Sequence Analyzer on AWS, and provides it as a managed service. In addition to its power and flexibility, AWS’s security model ensures GMI can provide access to sensitive data in a way that is highly secure and fully compliant with European data protection laws.

Results
In this one alliance alone, the GMI-WuXi NextCODE genomics and biopharma program will:

  • Focus on major chronic diseases within oncology, neuroscience, and immunology that affect hundreds of thousands of people in Ireland and hundreds of millions worldwide
  • Sequence 45,000 genomes from volunteer participants across Ireland to seek novel insights into the biological processes that underlie complex disease.
  • Use the research database developed by GMI to identify new molecular approaches for therapeutic drug discovery and development as well as to develop companion diagnostics

Looking ahead
Precision medicine has the potential to help people stay healthier longer by ensuring they receive a drug that is right for their genetic makeup, at the right time, and in the correct dose. Genomics is a major driver for precision medicine because it can characterize populations genetically to understand disease risk, predict individual response to medications, and point scientists toward new treatments and cures. Pharmaceutical companies are embracing genomics to boost drug development success rates and design better clinical trials. For example, when genetic biomarker tests are used to develop drugs, the probability a compound will make it from Phase I to market doubles. Tightly defined patient populations, selected via biomarkers, can identify the patients most likely to benefit from a drug, meaning smaller trials with fewer patients to demonstrate effectiveness. Patients who are unlikely to benefit from a treatment won’t receive it, reducing unnecessary costs and the burdens associated with lack of efficacy and adverse effects. The cloud is the technological foundation of genomically-driven precision medicine: it provides an environment that supports the continual creation of data and keeps it secure.

Get ready. Start. Modernize

In March 2019 then-FDA Commissioner Scott Gottlieb, M.D., said modernizing clinical trials is one of the agency’s top priorities. He also acknowledged that the “clinical research enterprise” is fantastically complex: it is heavily regulated, resource-intensive, and risky. “Efforts to streamline medical product development based on advancing science can be frustrated by legacy business models that discourage collaboration and data sharing, and the adoption of disruptive technologies that make clinical research more effective,” he observed.

Adopting new trial designs, analytic techniques, and technologies is how companies will modernize clinical trials—but they need to be smart about it. One way to mitigate risks is to rely on solutions developed by companies that combine deep domain expertise and technical computing prowess. The innovative solutions offered by APN Partners, backed by the scalable and secure power of AWS, are already transforming clinical research.

Source: AWS