The future of precision medicine part 5: keeping pace with regulation

The first four instalments in this series of articles on precision medicine focused on its challenges and opportunities in general, the importance of large quantities of high quality, diverse data, and some of the particular issues that will be encountered in the translation of research and theory to clinical practice.

This article will consider an important issue that will need to be addressed, namely the creation of appropriate regulatory frameworks.

Creative interpretation by tribunals and regulatory bodies tasked with applying laws and regulations often plays an important role in providing the flexibility to cater for the issues of the future, particularly when society, technology and the issues they give rise to have moved far beyond those the original drafters could possibly have conceived of.

However, interpretation can go only so far—it is essential to the development of precision medicine that the regulatory frameworks for assessment and approval of treatments, and the closely related science of clinical trials, keep pace with the advances being made on the diagnostics and treatment front.

As things stand, neither is well adapted nor fit for purpose as far as precision medicine approaches are concerned.

“Equally as important as the framework for the approval of treatments will be the development of standards which companion diagnostics will be required to meet.”

As we begin to see more treatments indicated not for a specific condition or type of cancer (for example), but rather a particular genetic feature or biomarker (eg, a particular fusion protein or cells displaying a particular receptor which is targeted by the treatment), developing a good understanding of what quality and quantity of evidence are necessary to demonstrate the safety and efficacy of a new treatment will be crucial.

The regulators and industry seem to be in agreement on this.

Good models

The former US Food and Drug Administration (FDA) commissioner Scott Gottlieb, speaking at the World Economic Forum in 2017, emphasised that there is a need to develop good natural history models of disease and also the potential for industry to better target development programmes to highly responding patient subgroups in order to present more convincing evidence of clinical efficacy in those populations to regulators.

Gottlieb noted that new legislation such as the 21st Century Cures Act has given the FDA the tools to allow it to approve products with strong, early proof of concept and addressing an unmet need, subject to post‑registration monitoring of safety to discharge concerns about long-term risk that will not be able to be dealt with at the pre-market stage.

Novartis chief executive officer Vas Narasimhan, who trained as a doctor and came into the role from a background as head of Novartis’ global drug development team, stressed that although he sees a big role for artificial intelligence (AI) in speeding up drug discovery and clinical development, it’s crucial for trials to get the basics right in terms of dose, patient population and an understanding of the target product profile and endpoints. Without those fundamentals, no amount of AI will save a development programme.

Recognising that precision medicine treatments represent a challenge to the conventional frameworks for assessment and approval of products, Gottlieb encouraged companies not to be afraid to engage in dialogue with the FDA from an early stage (even preclinical).

Citing CAR-T, CRISPR-Cas9 and gene therapies by way of example, he explained that those early conversations are an opportunity to educate the regulators and can help define expectations and avoid problems later on when the company is seeking an approval.

Panellists at the 2017 Westminster Health Forum, including Krishna Prasad, group manager at the UK’s Medicines and Healthcare Products Regulatory Agency (MHRA), also highlighted the serious questions to be considered as treatments become more specialised and patient subgroups potentially more and more granularly stratified.

These include: