Protecting public health: Is real-world evidence the key?

pharmafile | April 8, 2019 | Feature | Business Services, Manufacturing and Production, Medical Communications, Research and Development, Sales and Marketing |  feature, pharma, real world data, real world evidence, real-world data, real-world evidence 

In often tough regulatory climates and with threats to public health are growing in sceptical corners of society, the need to deliver treatments and preventative therapies is a key concern for industry and national healthcare bodies. Matt Fellows explores how incorporating wider use of real-world evidence into these systems could be the solution we need.

There are few matters, if any, more important than public health. This places an immeasurable responsibility on researchers and clinicians all the way up to the chief executives of Big Pharma to maintain the highest standards of safety and efficacy in the drugs they make available to the public. In regions such as the US and Europe, a robust system of peer-reviewed research and tight regulation goes a long way to ensuring these standards are achieved, but given the sheer breadth and heterogeneity of patients across the world, is it really feasible to expect the humble randomised control trial (RCT), the gold standard in clinical research, to provide all the data and knowledge required to provide for all their needs in all situations?

Operations within the industry continue to move unerringly towards an ever more global perspective, and this, coupled with a drive to always push for better practice and patient outcomes, has revealed the cracks in the system through which some patients will fall, and where medical knowledge is left wanting. And, with an ever-growing industry consensus, real-world evidence (RWE) is one of the prime solutions to meet these challenges and maintain the highest standards when it comes to public health.

While not a new methodology, RWE is gaining increasing attention within the industry for the value it adds to the robust data gathered through a well-designed RCT, while covering bases that simply cannot be covered in a controlled setting.

Deepening value

Novo Nordisk is just one of the names in Big Pharma which has devoted itself to realising these benefits. The company recently revealed new real-world data (RWD) from its REFLECT study, which sought to better evaluate the efficacy of its insulin degludec therapy Tresiba in the world beyond the closed doors of a clinical study. Novo Nordisk’s Director of Clinical, Medical and Regulatory, Dr Avideh Nazeri, spoke to Pharmafocus to explain how real-world studies were key to achieving proof-of-concept:

“This was a safety and effectiveness study in essence, meaning we really wanted to test how a new medicine is working in a real life setting. Of course, in RCTs, you’ve got two elements which are critical to the results: one is randomisation, and it’s critical because it takes away all the biases, but it also has its own limitations, just like many other things in life. One of them is that in real life situations, doctors don’t just randomly put patients on therapies; they look at their patients, they individualise, they find what their needs are, and then they put them on to therapy. So, one of the benefits of real world studies such as REFLECT is that you remove that randomisation part; in essence, you are creating a bias, but that bias is to the benefit of the patient.

“The other part when you talk about RCTs is the element of control,” she continued. “RCTs are really protocol-driven and you almost micromanage the patients; they have regular visits; there are countless lab tests going on, whereas in a real world setting, you kind of loosen up on that a little bit and you go with their routine care. An element of REFLECT was that we were also not introducing any unnecessary tests, or anything that a doctor and a nurse would have done above and beyond their routine patient care. For instance, one of the questions is that the routine care in Germany may be slightly different from routine care in the UK, so real world studies really have that holistic element situated in real life.

“In REFLECT, the primary endpoint was to first look at the rates of hypoglycaemia, because when you start on a new insulin, one of the common concerns is that switching therapy may cause either more hypoglycaemia or require more monitoring and surveillance of the patients. So they switch patients over to Tresiba – nothing extraordinary, nothing additional – and see how it goes, carrying on for 12 months.

“We knew from the older RCTs that insulin degludec led to the reduction of hypoglycaemia, and again that was reconfirmed: you had a significant reduction of hypoglycaemia overall. Not only that, but one of the potentials of an insulin which gives you less hypoglycaemia is that potentially you can safely adjust your insulin dose to get to a better glycaemic result.

“That latter result was one of the things that we never identified in RCTs, because the design of RCTs are what we call non-inferiority, meaning everybody has to aim for the same glycaemic targets in two arms: one with one insulin and one with the other insulin, so you never actually do see any difference in glycaemic control in any insulin trials – you go with the same glycaemic goal, and then you see any differences in hypoglycaemia. The beauty of REFLECT is that it is a single arm, observational study, and you compare the patients before and after, so it was very reassuring to see not only that rates of hypoglycaemia in both type 1 and type 2 patients were reduced, but there was also a significantly improved situation in terms of the glycaemic control, both in HbA1c and fasting plasma glucose in both cohorts. That is quite unique, and that’s why we are excited about REFLECT: we have proof-of-concept that the medicines that are working in the very controlled setting of RCTs can work equally in normal, real life, when people are amongst their family, they visit the hospital less, they are less micromanaged, and health care professionals are not following protocol as they go.”

By recognising the value of RWE and designing and enacting a study to effectively harvest it, Novo was able to not only vindicate its earlier research and hypotheses, but also reveal previously unseen insights. In the past, with a focus set squarely on the value of RCTs alone, achieving this breadth of insight into the wide spectrum of patient need would just not be possible on the same level because of the method’s inherent drawbacks, as Dr Nazeri explains:

“The historic model of drug development, if you want to be really honest, didn’t include anything on real-world data for years and years; you go with RCTs to fulfil your regulatory requirements, and even in that situation, you will further develop the data in terms of effectiveness and safety through pharmacovigilance efforts. In the past model, you only would have relied on adverse event reporting, because everybody in drug discovery and development knows that your ticket of entry is market; you will not be able to identify the full potential throughout the initial development.

“Let’s not forget the RCTs always have a long list of inclusion and exclusion criteria, which is critical for that phase of the development of a drug,” she adds, “but in real life, doctors have to choose medicines for all their patients – they don’t exclude anybody. So I do think, in essence, that it does expand the scope of evidence generation across the board, and it isn’t limited to selected groups which fit the inclusion and exclusion criteria; it deals with real patients that have so many comorbidities and may belong to so many different parts of a country and communities.”

Fighting the sceptics

From diabetes to another key issue for public health, and one which, today, is under threat perhaps more than any other, with potentially disastrous consequences: vaccination. By now we have all heard the fractured whispers at the corners of social media: vaccines cause autism – in particular, the measles, mumps and rubella (MMR) jab. They don’t, as has been proven numerous times before, but with one more for good measure, researchers at the Statens Serum Institut in Copenhagen decided to double down and prove once and for all that there is no connection between the two, and RWE formed the bedrock of their strategy.

“The link between MMR vaccination and autism can be traced back to the 1998 Wakefield paper,” lead author Dr Anders Peter Hviid told Pharmafocus. “Although the paper was later retracted, the link has gained worldwide attention and has fueled a strong anti-vaccine movement which persists to this day. We did a study back in 2002 which was published in the New England Journal of Medicine, where we found no support for an association between the MMR vaccine and autism among more than half a million Danish children. However, despite our earlier study and other observational studies reporting no association, the idea that vaccines can cause autism is still being promoted by some politicians and celebrities, and on social media, anti-vaccine misinformation features prominently. This causes a lot of parental concern, and while true anti-vaccination beliefs are likely a small, but very vocal, minority, vaccine hesitancy is growing. We are now seeing the impact that vaccine hesitancy has, with more and more frequent measles outbreaks in Europe and North America. We felt that it was time to revisit the link, and to address some of the criticisms of our original 2002 study.”

So, to fight this crucial battle effectively, how would the team gather a broad enough, meaningful set of data that could put the myths to rest? Luckily, being based in Denmark, they were not short of sources and channels to procure these data.

“In Denmark, each inhabitant is assigned a unique personal ID. This ID is used in all nationwide registers which comprise a wealth of demographic, socio-economic and health related information,” Dr Hviid explained. “These unique register data can be used for research. This enables studies encompassing the whole Danish population with individual level information on exposures and outcomes. This type of large-scale real-world data is only available in Denmark and some of the other Nordic countries. For the current study, we were able to include 657,461 children – almost all children born in Denmark 1999-2010 – with information on MMR vaccination date, date of possible autism diagnosis and other covariates of interest for each child.

“We found no difference in autism hazard rates between MMR-vaccinated and unvaccinated children (Hazard Ratio, 0.93, 95% CI, 0.85 to 1.02). This corroborates with our 2002 study and other observational studies of the link. We also addressed the idea that MMR might increase the risk of autism in children at higher risk of autism than the average child. We looked at children with autistic siblings – a proxy measure for genetic susceptibility – and children with autism risk factors such as older parents, prematurity, low Apgar score at birth, etcetera. We found no support for an association in children with autism risk factors. We also looked at the risk of autism in time-periods after vaccination; we found no support for any clustering of cases. Finally, we also looked at other vaccines given before the MMR, and found no support for an association in children with no vaccinations in infancy; this analysis was in response to the claims that no study exist utilising a completely unvaccinated control group.

“RWD is essential in situations such as this, where we need to evaluate the safety of a vaccine or a drug with respect to rare outcomes and outcomes occurring years after exposure. It is not feasible to conduct a clinical trial of MMR vaccination with the number of children and years of follow-up needed to evaluate this link. Furthermore, it would not be ethical to withhold the MMR vaccine for a control group.”

As Dr Hviid stresses, for a public health issue of this magnitude, the use of RWE is an imperative; vaccine hesitancy was listed by the WHO as one of the top ten threats to global health in 2019, and without proper and precision-executed science and hard data, the problem will only grow.

So, 17 years after their last study, does Dr Hviid think the added value of RWE will help to settle the ‘debate’? Unfortunately, but not surprisingly, he is somewhat reticent.

“There is a general trend towards anti-authority and anti-intellectualism in today’s society. This is fueled by the internet where fake news and false information feature prominently and can be difficult to identify for lay people,” he remarked. “I think the true anti-vaxxers will never be satisfied. There are already several “critical analyses” of our study in anti-vaccine blogs and YouTube videos. However, our study has received massive global attention, and I am hopeful that our message has reached many concerned parents around the world.”

Building bridges

As well as fleshing out robust clinical data and grounding it through a pragmatic lens, RWE may hold the key to bringing industry stakeholders together to more effectively tackle public health concerns and also to secure additional regulatory approval for existing therapies, as Dr Nazeri notes:

“One of the advantages right now that people are looking at in certain parts of the world is the payers negotiations and health technology assessments (HTAs). HTAs often look at the whole value of medicines beyond RCTs. I think maybe in the next 10 years the point will come when regulators also look at RWE and then try to do some product labels expansions. Right now, we can’t typically use the data from real world to get an additional indication or label expansion, but this could be a future promise for RWE. Look at populations where it’s quite difficult to do RCTs: paediatrics, for example; everybody knows that recruiting babies or kids into clinical trials is quite challenging, and that is the number one reason why medicines are sometimes introduced in children quite late, because first you have to launch in adults and then you have years of trying to recruit patients into paediatric trials. So in future, if we can really make real-world evidence robust and work with regulators in parallel to allow us, if a drug is safe, to breach those license gaps – that would have great potential. However, right now that doesn’t exist.”

So how does the industry work to make these potentials a reality? It certainly has its work cut out, but the public need is clear and the bar has been set; with the growing threat from vaccine sceptics which is wreaking very real, tangible and direct consequences across Europe and the US as incidence of conditions such as measles surge, the pressure is on to bridge the fissures in the research and regulatory landscape. Dr Nazeri gives her thoughts on how this may be achieved as we move forward, and the key is collaboration:

“I do think that the value of RWE needs to be communicated to healthcare professionals, because, for years and years, they have all been very much into the idea that the highest level of evidence is only RCTs; unless people really get involved in RWE, they may be a bit sceptical. You just need to really reiterate how simple this whole design is. There is a little bit of that further training that we need to do, but otherwise I do think that from a patient perspective they don’t really mind – you do the same thing, almost, that you would have done in a clinical trial: you get consent, it’s quite transparent, they see their own regular doctors, they don’t need to do anything additional.

“We have equal responsibility as industry, but we really need regulators to identify the framework they’re looking at when they’re assessing RWD, because right now it seems like they’re also trying to develop new ways of looking at it. The utter magnitude of RWE would become big data when you have aggregated registries – some of them are not necessarily very clean data. With big data there is a little bit of reservation: are they even meaningful? What do they mean? Who needs to assess them? Are your local regulators the right people, or do you need to have additional tools? So we’re not quite there yet; I think regulators are still trying to understand the role of industry, pharmaceutical companies, and also academia to come up with a partnership and define what kind of RWE would really add value from a safety and effectiveness perspective. I think that can only happen if there is really collaboration and a dialogue.

“It would be really great to push for a partnership and collaboration whereby they would come to us and propose the questions: ‘these are the type of data that we want you to collect in the real world’, rather than the model right now, where we come up with the design – which is usually very simple like REFLECT – and we think it’s the right way to do things. In the future, it would be really great if, during the development phase for new drugs, you could open it to academia or payers or patient organisations. They could then submit their questions and we could then design a real-world study to answer those questions.”

It’s clear that RWE forms a crucial cornerstone as part of a wider, robust strategy to protect public health. There is much more to be done to forge a collaborative path forward that puts the best interests of the public, both individually and collectively, at its heart. With the world crying out for better, evidence-based solutions to today’s health crises, industry must step up to meet these needs. Dr Hviid puts it succinctly: “RWD/RWE is absolutely key in protecting public health. Vaccines and drugs must be studied under the noisy and heterogeneous conditions of real-world use.”