Adapting to the Future of Licensing

Currently, authorization is the magic moment in the drug approval process. A “yes” answer catapults a therapy from the experimental stage into routine use by thousands of patients – including potential off-label indications. At this point, the medicine is considered to be safe and effective; however, there can be unpredictable effects in real-world patients because assumptions on safety are based on data from narrowly selected patient populations enrolled in clinical trials. For example, patients over 60 and those with pre-existing conditions are often excluded. To address these challenges and others, a new approach is under discussion by regulators, drug developers, payers and patient advocates worldwide: adaptive licensing.

Adaptive licensing (also known as staggered approval or progressive licensing) is an iterative process that may begin with the early authorization of a medicine in a restricted patient population, followed by proactive monitoring of newly treated patients. If the safety and efficacy profile in these patients is positive, the marketing authorization can be adapted to expand the patient population little by little, based on accumulating data. I sometimes describe it as building out from the center of an onion, where layers of data from broader patient populations are progressively added. With each layer, the regulators make new decisions on whether to expand, or indeed contract, the patient population, as supported by the data.

It also offers advantages in terms of monitoring safety and efficacy after the initial marketing approval. In the current system, the patient populations and conditions in clinical trials are very controlled, but this all changes after approval. In fact, very few data are collected to evaluate the safety or efficacy of the drug in a real-world setting, beyond the passive reporting of adverse events. With adaptive licensing, the initial patients are followed much more closely and there is the potential to supplement clinical trial data with real-world data, which is useful for regulators, payers, sponsors, and patients alike. In today’s system, these stakeholders do not have a great deal of confidence that they will get the postmarketing data they want and need. If everyone agrees upfront what data are going to be collected, there are much clearer assurances that this
will occur.

Can it work?

So far, I’ve focused on the potential benefits of adaptive licensing, but there are a number of hurdles to overcome before such a system becomes the norm. First of all, the interactions among drug companies, regulators, payers and others will need to be much more transparent and collaborative. Frequent and extensive dialogue within and among stakeholder groups will be required to build trust and arrive at an acceptable solution for all. This will be particularly important for payers and health technology assessment (HTA) bodies, who already take a variety of approaches to assess comparative effectiveness of traditionally developed medicine. These differences will almost certainly be magnified when considering adaptively developed medicines.

Adaptive licensing involves an element of risk sharing – all stakeholder groups are going to have to compromise, to give up something in order to get medicines to the patients who need it. For instance, regulators and payers will need to be comfortable with making initial decisions based on a smaller dataset and feel confident that additional data will be forthcoming; payers and sponsors will need to come up with more flexible pricing strategies that can be modified as additional data are accumulated; patients will need to accept more uncertainty about a medicine early in its lifespan in order to obtain its potential benefits; and sponsors must be willing to accept a small initial market in order to accelerate the availability of medicines to the patients who need them most. There aren’t many particularly good examples of this type of multi-stakeholder interaction and collaboration in drug development. To make this work, we must get each stakeholder group to focus on achieving a greater good.

At the New Drug Development Paradigms (NEWDIGS) program, led by the MIT Center for Biomedical Innovation, we wanted to move beyond theory and explore how adaptive licensing could work in practice; we wanted to see whether this level of stakeholder collaboration was realistic. With that in mind, we have been holding quarterly two-day workshops over the past four years. The goal of these workshops is to get stakeholders together early to talk about the common goal they all share and work together to improve the efficiency of that process.

In these workshops, trial sponsors discuss real-life case studies of drugs in development. They describe the current regulatory pathway, and propose how an adaptive licensing approach could be applied. To date, 14 different candidate medicines have been discussed across a wide range of therapeutic areas and drug types – small molecules, biologics, combination therapies and vaccines.
The discussions at the workshops are carried out under confidentiality agreements. Proprietary information about the candidate medicine stays in the room, although the ideas developed during the workshop can be shared more broadly. In the room, there are multiple pharma companies, regulators from a number of agencies, payers and HTA representatives from several jurisdictions, patient advocates and physicians. A company typically presents a proposal, and then everyone weighs in with suggestions on how the plan could be refined to make it more robust from a scientific standpoint, or how it might be refined to more effectively get regulators and payers the information they need, as they need it. The conversations are surprisingly interactive, with a great deal of idea cross-fertilization, which is impressive to see in an industry that is so often focused on the proprietary nature of what they are doing. The sponsor participants, regardless of whether they have chosen to pursue adaptive licensing for their product, also get a lot out of the discussions, benefitting from the diverse perspectives that are voiced. Indeed, several companies have brought their development candidates back for two, three, even four follow-up sessions, as new information is accumulated and/or to continue to refine their approach.

Supporting implementation

Having seen such a positive reaction to the workshops, we hope to broaden stakeholder familiarity with this type of collaborative environment with a recently launched program called the Janus Initiative. From development to clinical outcomes, the Janus Initiative is being designed to bring together a diverse group of experts (not only from multiple organizations but also from multiple functional areas within those organizations) to capture the key information about a therapeutic program. Specific factors that will be considered include: patient population structure, clinical trial programs, registration, reimbursement, adoption, public health impact, and even caregiver impact.

Janus collects multiple stakeholder views on the strength of evidence for each input and the range of estimates, using both retrospective and prospective case studies. If consensus cannot be reached, enhanced visualizations of the various perspectives facilitate stakeholder dialogue to define the range of scenarios to explore. Janus employs a linked series of simulation models to quantify the effects of adaptive licensing across time and according to each stakeholder’s perspective. Wherever possible, Janus will leverage and link existing tools rather than create entirely new ones.

The successful adoption of adaptive licensing will also require proactive collection and analysis of safety and efficacy data across the lifespan of a medicine, not just up until the time of authorization. Therefore, a more diverse dataset, beyond that collected in randomized, controlled trials, will be required, including real-world data from observational studies and/or patient registries, as well as emerging data sources such as social media, electronic medical records, and wearable sensors. The NEWDIGS Data Program has been established to evaluate data sources, quality, and requirements to support adaptive licensing, as well as the technologies, policies, and processes that will be needed for their analysis.

One step at a time

Until we understand more about what adaptive licensing is, how it works and the potential benefits and drawbacks, the initial drug candidates (such as those in the pilot program now underway in Europe) will likely be medicines to address an unmet medical need. In this context, an unmet medical need is not necessarily a first-in-class molecule, but could be a ‘rescue’ therapy for a group of patients where all other available therapies have failed. For these therapies, the benefit/risk ratio is clearly weighted in favor of intervention. An example would be a therapy for Duchene’s muscular dystrophy, a progressive degeneration of the muscles affecting 1 in 3600 boys with an average life expectancy of 25; if the treatment could be made available in three or four years rather than 10, many lives could be saved.

As we learn more, I believe that adaptive licensing will be applicable to a broader range of medical products, not just those for unmet medical needs. By definition, adaptive licensing is very flexible – the development program is designed specifically for each product – so it gives us the opportunity to put together more rational development plans rather than following established precedents or being purely reactive. It is easy to simply stick with what we know, rather than do what makes sense from a patient’s perspective. With adaptive licensing, industry could put the focus back on the needs of the patient.

Regulators on board

Regulators are already moving in this direction. Several national regulatory agencies have been discussing the potential of adaptive licensing for a number of years, but there was a big step earlier this year when the EMA launched a pilot scheme. Such bold moves are likely to move the discussion of adaptive licensing forward, with an emphasis on its practical aspects. In particular, the EMA wants to understand how future adaptive licensing pathways might work for different types of products and indications. Two medicines have been chosen to take part in the pilot, and although the EMA can’t reveal the names, it says it has selected the products based on several criteria including: unmet clinical need, an early stage of development that will allow for actionable input, positive prospect of expanding from a restricted indication to broader populations, and the potential for real-world data in fulfilling expansion requirements. As the pilots progress, the European Commission will be examining the legal and policy aspects of adaptive licensing.

A new and more sustainable paradigm in drug development has been talked about for a number of years and now seem to be poised for launch. I hope that the experiences gained in EMA’s adaptive licensing pilots, NEWDIGS Janus Initiative, and the work of other groups, will provide stakeholders with the foundation on which to build and refine drug development policies and procedures of the future, and the confidence to adopt them widely.