Pharma: Ripe for Digital Disruption

Abstract

Digital technology is transforming the life sciences sector: from AI technology mining compound libraries for drug discovery, and software and electroceuticals as therapeutic solutions, to sensors in chips monitoring adherence and absorption. This disruption will only accelerate. Our article considers some of the legal issues with these transformative technologies.

Introduction

While the life sciences sector has incrementally adopted digital technologies over the years, the rate of digital transformation over the next five years looks to be unprecedented. Collaborations and acquisitions in the pursuit of digital transformation have accelerated significantly. In 2018, $9.5billion was invested in the digital health sector in over 698 deals¹. From the FDA approval of Otsuka’s sensor embedded drug Abilify Mycite² and the approval of Pear Therapeutics’ app for the treatment of opioid abuse, through to Takeda’s partnership with Emulate Inc into the use of organs on chips for drug discovery and development⁴, every aspect of the pharmaceutical sector is ripe for digital disruption.

Areas of particular interest for transformation include: (i) the drug discovery and development phase; (ii) interactions with healthcare professionals and patients; and (iii) complementary or standalone therapeutic software treatments.

Drug discovery and development phase

Traditional methods of drug discovery and development are expected to last ten years and cost more than $2billion⁵. However, the use of artificial intelligence, organs on chips, apps and wearables could significantly reduce these timescales and costs.

Artificial intelligence (“AI”) is already being deployed by a number of pharmaceutical companies for a whole range of purposes, including drug discovery. The ability to mine huge volumes of public data, as well as significant private data sets (e.g. compound libraries and clinical trial data), represents an unprecedented opportunity to identify potential targets, drug candidates or even new indications for existing drugs, faster and more cost-effectively.

These newly identified targets or candidates can then be developed and tested, with the data this generates being processed by the AI algorithm to further refine its learning. Pharmaceutical companies including the participants in the MELLODDY consortium⁶ (comprising 10 major pharmaceutical companies), Astellas⁷and Takeda⁸ have all announced collaborations with artificial intelligence businesses for drug discovery and development. The success stories for using AI are compelling: the Barrow Neurological Institute successfully used IBM Watson to analyse all RNA-binding proteins in the human genome, genomic data and published materials and identified an additional five amyotrophic lateral sclerosis RNA-binding proteins which were previously unlinked to amyotrophic lateral sclerosis.⁹

While the development of organs-on-chips is nascent, if successful they offer an entirely new approach to clinical trials. California based business Emulate Inc, has developed a liver chip, lung chip and intestine chip and is currently performing studies with the FDA to assess the use of organs-on-chips for toxicology studies^x.

The possibility of replacing animal trials, and even human trials in due course, with organs on chips could transform the drug development process. Likewise, the use of AI to mine patient data to identify optimal trial subjects, as well as trial subject “matching” apps (where interested trial subjects can be matched to potential trials and apply via an app), look set to materially alter the trial recruitment process. The use of apps, wearables and remote devices in trials offers the potential for continuous monitoring of, and increased interaction with, trial subjects with limited interference in the trial subjects day-to-day life, which may improve trial recruitment and retention, as well as the quality of data generated.

Interactions with healthcare professionals and patients

Digitalisation also allows pharmaceutical companies to interact with health care professionals (“HCPs”) and patients in new ways. This might include electronic “pop up” reminders for HCPs¹¹, new ways of imparting information to patients¹², or complementary apps to support patients receiving certain therapeutic treatments¹³. Digital transformation could disrupt drug detailing which has not seen significant change in approach for many years. Indeed this digital approach reflects a new generation of HCPs with limited time for detailing meetings and, in some cases, decreasing interest in face to face interactions with the sales force. At the same time, this type of continual remote patient-support offers a new way to develop patient-centric care, while also potentially reducing the need for acute care intervention (e.g. by offering remote monitoring or support, or improving medication adherence) which will improve the patient’s quality of life, as well as reducing costs for healthcare systems.

Therapeutic software solutions

In addition to transforming how drug products are developed and administered, digital health has the potential to see traditional drug products replaced with therapeutic software solutions. While historically these types of digital solutions have been focused on wellness with diet apps, fitness wearables and relaxation apps, or helping patients to manage chronic conditions with symptom journals and trackers, this is now shifting towards true therapeutic solutions.

For example, Pear Therapeutics’ app for the treatment of opioid abuse and virtual reality with bio-data-driven applications to treat acute and chronic pain¹⁴. A particular area of focus is the treatment of mental health diseases, with software platforms offering AI-driven triage facilities, coupled with treatment techniques and access to human physicians interacting via messaging or telephone, at any time.

Challenges

A recent study¹⁵ performed by our law firm, Simmons & Simmons LLP, asked more than 400 international c-suite executives across the life sciences and technology sector about the opportunities and challenges presented by digital health. Notably, 71% of all respondents stated that digital health will transform patient care, with 63% putting digital transformation at the top of its agenda.

However, only 11% of all digital health opportunities that come to an organisation’s attention enter detailed due diligence and just a third of those are executed. Life sciences respondents indicated that half of their digital health collaborations in the last 3 years did not meet their stated objectives. Perhaps unsurprisingly, respondents indicated that two of the main challenges for digital transformation projects result from data protection and life sciences regulation.

Data protection

Data protection concerns arising from purported transfers of medical data have recently been in the news - see, for example, Google’s reported: “Project Nightingale” partnership with a large US healthcare provider¹⁶ and the FT’s investigation into the flow of data from health websites in the UK¹⁷.

Most digital transformation relies on the mining of vast amounts of sensitive personal data (or “special category data”), i.e. “personal data revealing racial or ethnic origin…the processing of genetic data…data concerning health…”.

Consequently, the control and processing of identifiable sensitive personal data will be subject to strict data protection laws such as the EU General Data Protection Regulation (“GDPR”), which will need to be complied with. The GDPR (which came into force on 25 May 2018 and applies both to arrangements entered into before and after that date) strengthens existing EU data protection obligations and introduces new requirements. Importantly, it has significantly strengthened the sanctions which may be imposed by the data protection authorities, namely, fines of up to €20million or 4% of global annual turnover (whichever is higher) for the most serious breaches.¹⁸

The GDPR implements a number of changes, including requiring data controllers to factor in and plan for data protection from the outset (“privacy by design and default”), as well as perform privacy impact assessments prior to processing to ensure the necessity and proportionality of the processing of the data, and that potential risks for data subjects are mitigated. Other points to consider are:

• consent: often consent is the only statutory basis on which identifiable sensitive personal data can be processed. In such cases, the consent must be explicit, unambiguous and distinct. However, there are often difficulties with consent:
• consent must be given for each data processing procedure performed (rather than a general consent). However, it may be difficult to predict from the outset what these different procedures and purposes might be e.g. in the future the business may wish to mine the data for a specific (or unknown) purpose.
• consent may be withdrawn by the data subject at any time¹⁹ after which no further processing of this data is permitted. It will, therefore, be important to ensure that such personal data can be segregated from any other data for prospective processing.
  Consequently, it may be better for a business to rely on another statutory basis for the processing of special category data. Businesses should also consider the extent to which they can rely on the principle of compatibility under Article 5 of GDPR to facilitate secondary use of data and avoid having to rely on a new legal basis for that use²⁰.
• pseudonymisation: Importantly the pseudonymisation of data (i.e. “the processing of personal data in such a manner that the personal data can no longer be attributed to a specific data subject without the use of additional information”) does not constitute anonymised data and, therefore, the requirements of the GDPR continue to apply to pseudonymised data. It is very difficult to truly anonymise data to the level required by the GDPR and, therefore, businesses should be cautious before determining that GDPR would not apply.
• right to be forgotten: A data subject may request to be “forgotten” (i.e. to have all data held about that data subject deleted) if certain circumstances apply. In such cases, the business must “without undue delay” erase all data related to that individual. The business should factor its compliance with such an obligation into any data collection model.

Life sciences regulation

In addition, businesses will need to ensure that their new ways of interacting with HCPs and patients, as well as new products, comply with applicable life sciences regulations.

For example, in the EU (and the UK) promotion of prescription-only medicines to the public is prohibited and, therefore, businesses will need to ensure that any complementary apps supporting patients do not promote the related (or any other) prescription-only medicine. Equally, consideration should be given to whether any automated interaction with HCPs constitutes promotion such that it will need to comply with applicable advertising laws or interferes with any post-market pharmacovigilance obligations.

Under the EU Medical Devices Regulation 2017/745²¹, software which has a medical purpose may qualify, and be regulated as, medical devices in their own right (sometimes referred to as “software as a medical device” or “SaMD”). Businesses will, therefore, need to carefully consider the purpose and functionality of software it develops, as well as the claims made about it, in order to determine whether or not it would qualify as a medical device. If so, then consideration should be given to the classification of the device.

Previously, in the EU, most SaMD was classified as a Class I device which is subject to a relatively straight-forward market access pathway involving self-certification of the device’s compliance. However, increasingly SaMD is being accorded a higher classification which requires a more involved market access pathway including assessment by a regulated third party (“notified body”). For example, apps focused on the treatment of acute mental health conditions to prevent suicide, could have very serious effects if defective and therefore may be subject to a higher classification.

Businesses should also be mindful as to how classification might change with revised iterations of the software. If a software solution does qualify as a medical device, then, not only would the business need to comply with market access requirements, but also extensive post-market obligations regarding traceability, supply chain management/visibility, adverse event monitoring and reporting and advertising.

With a 32% increase in investment in digital health in 2018 compared to 2017²², digital health looks likely to go from strength to strength. However, in many cases, these developments can only be achieved through collaboration between pharma businesses and new and incumbent tech businesses. In doing so, pharma will need to navigate a myriad of legal and cultural issues in order to realise its digital transformation including IP ownership, regulatory compliance, product liability, data protection and cybersecurity issues.

This article is based on an article which was first published in Pharma Focus Asia.

¹ Pharmaphorum: 2018 breaks records for digital health investment report
² Abilify Mycite
³ Reuters: FDA clears Pear Therapeutics' mobile app to help treat opioid abuse
⁴ Emulate Inc and Takeda partner on organs-on-chips for drug discovery and development
⁵ Policy & Medicine: A tough road: cost to develop one new drug is $2.6billion; approval rate for drugs entering clinical development is less than 12%
⁶ Machine Learning Ledger Orchestration for Drug Discovery announced in June 2019
⁷ Biovista Inc. announces drug repositioning collaboration Astellas Pharma NuMedii announces new indications discovery collaboration with Astellas
⁸ Numerate and Takeda enter agreement to generate novel clinical candidates using AI-driven drug discovery Takeda and institute ConvergeHEALTH apply AI to test how treatment-resistant depression responds to medication
Recursion announces extension of AI-enabled drug discovery collaboration with Takeda
⁹ Artificial intelligence in neurodegenerative disease research: use of IBM Watson to identify additional RNA-binding proteins altered in amyotrophic lateral sclerosis by N Bakkar and others
¹⁰ FDA collaborates with Emulate, Inc, on use of organs-on-chips as a toxicology testing platform
¹¹ The effects of on-screen, point of care computer reminders on processes and outcomes of care by K Shojania and others
¹² For example: Patient education using virtual reality increases knowledge and positive experience for breast cancer patients undergoing radiation therapy by YA Jimenez and others
¹³ For example: BTG's IO Loop app
¹⁴ For example: Samsung Electronics, Travelers Companies Inc, Cedar-Sinai, Bayer and AppliedVR explore a digital pain-reduction kit that uses therapeutic virtual reality (VR) and wearable technology as a non-pharmacological supplement to managing pain
¹⁵ Simmons & Simmons 2019 Digital Fusion Report
¹⁶ For example: The Guardian - Google medical data Project Nightingale
BBC - Google probed over US patient data deal
¹⁷ FT investigation into use of data from health websites
¹⁸ Notably, GDPR and the UK Data Protection Act 2018 will continue to apply in the UK after Brexit.
¹⁹ Article 7(3) GDPR
²⁰ European Data Protection Board Opinion 3/2019 of 23 January 2019
²¹ The UK Government has stated that “key elements” of the EU Medical Devices Regulation 2017/745 will be implemented into English law after Brexit.
²² Pharmaphorum: 2018 breaks records for digital health investment report