To approve or not approve in desperate times?
We are just four months into the worst virus pandemic in modern times. The influenza pandemic of 2009 was nowhere near the severity of the COVID- 19 pandemic, from both a public health and economic point of view. No vaccines or therapeutics are approved and available to the world population to fight the SARS-COV-2 virus.
It is now spreading rapidly to all corners of the world and the number of infections and deaths continue to increase. The containment measures put in place with lock down of societies and economies have had a dramatic effect on the world economy. World leaders are now easing restrictions to bring life back to “normal”, get people back to work and the economy back on track. This is not an easy decision, since such a move could bring a second wave of the virus into communities with a potentially heavy load on health care systems and many more deaths. A successful “opening-up-strategy” will therefore be dependent on communities continuing social distancing and hygiene measures in combination with the availability of therapeutics, vaccines, increased testing for the virus, and efficient contact tracing.
There is now a surge of research into new vaccines and therapeutics by companies and academic institutions all over the world. This has already resulted in several interesting potential new vaccine candidates and drugs that are in clinical testing. However, by normal regulatory standards the clinical development, with large randomised double-blinded clinical trials, will take more than a year for therapeutics, and several years for vaccines, to generate evidence regarding safety and efficacy.
Regulators are taking many measures to speed up the development of vaccines and therapeutics, such as rapid scientific advice to developers, using fast track, emergency use, rolling reviews, monitoring systems, etc that will facilitate the time to market. Intensive world-wide collaboration between regulators, academics and health authorities is ongoing, working on harmonising evidence requirements for vaccines and therapeutics. One such forum is the International Coalition of Medicines Regulatory Authorities (ICMRA).
The main question for regulators is how much evidence on safety and efficacy they will need to have to approve early, and how much greater uncertainty in this pandemic situation are we prepared to accept, compared with a non-pandemic situation? Getting this right will be crucial. Such assessment must be based not only on strict criteria on safety and efficacy but also epidemiologic considerations on how to best manage the pandemic. (A crucial question is to what extent economic or political consideration will be a part of these decisions. This is of course not in the mandate of independent regulatory bodies and I refrain to further elaborate on this.)
In this respect the situation is somewhat different between therapeutics and vaccines. Most of the therapeutics in development are repurposing of old drugs or drugs in clinical development for other indications. For these drugs there is already pre-clinical and, in some cases, clinical data available that could be extrapolated and used for treatment of COVID-19. From a regulatory point of view, conditional approvals based on strong phase 2 data could be an option followed by confirmatory phase 3 studies and post authorisation safety and effective studies. Looking at the large number of drugs in development and the speed of development, the first such drugs could hopefully be ready for approval in a couple of months if efficacy can be confirmed.
The situation is different for a vaccine and compared with the situation during the 2009 pandemic we are in a more difficult situation. In 2009 we had long experience of influenza vaccines and it was relatively easy to add a new influenza strain – H1N1 – to the existing platforms of vaccines. This is not the situation today, where we have a virus with no approved vaccine to date and previous research for corona vaccines is very limited.
There are currently eight vaccines in early clinical development and more that 100 vaccines in preclinical development. Normal regulatory standards would require evidence of safety and efficacy through phase 1-3 clinical trials involving thousands of test subjects and stretched over several years. In the present desperate situation, there is a need to consider early approval with less robust data on safety and efficacy followed with long term safety assessment post marketing
One such approach under discussion is the use of conditional approval of vaccines based on strong phase 2 data demonstrating that the vaccine produces a sufficient level of naturalising antibodies in test subjects. This would then be followed at a later date by completion of large phase 3 studies in order to confirm efficacy and safety. This is a risky approach since experience in vaccine development has shown that only producing antibodies does not necessarily correlate with preventative efficacy in a wider population. Long-term safety is also difficult to judge with only a small set of patients in the phase 2 studies. Should something not work out as intended, this could have dangerous consequences, considering that large groups of people would get these vaccines.
Another approach is to have an adaptive trial design seamlessly moving through stage 1-3 where an early approval could be considered once a certain level of evidence has been generated.
There could be an issue with early approvals in that it could be difficult to finalise the clinical trials since individuals may terminate their participation not to risk ending up in a placebo arm. It could also be a question of whether real world follow-up studies will give a clear answer on the efficacy of the vaccine compared with robust phase 3 studies.
One way to generate data early on efficacy is to perform active challenge trials i.e. exposing healthy subjects to the virus in order to generate clinical data on protection. This is controversial and it is questionable if it would be ethical in the present situation. But positive data from such a study in combination with reduced number on study participants in phase 3 could reduce the time for development and approval. So far, we have not seen these kinds of studies but WHO has recently in a guidance document .lined out the ethical criteria for such a trial to go ahead.
There is a lot of speculation and expectations from media, politicians and the scientific community of when therapeutics and vaccines will be available. Regulators will be under pressure to consider whether an early approval or not is the way forward. They also need to explain the basis for such go-no-go decisions in a transparent and understandable way. In 2009, during my time as executive director of EMA, we started a small research project to test the usability of multicriteria decision analysis (MCDA) methodology together with colleagues from ECDC and London school of economics, to support a decision on whether to approve 2009 H1N1 vaccines early versus wait for more data. Modelling the efficacy, safety and the severity of the pandemic and the underlying uncertainties we came to the conclusion that it was appropriate to approve the 2009 pandemic vaccine early rather than wait until a complete set of evidence was available. It should be pointed out that this research project was independent from the EMA and CHMP official opinion.
Our model was simplistic and if it was to be used for SARS-COV-2 vaccines it must be further refined and cover all aspects of the present epidemic. The work was published in European Journal Public Health in 2013. Whatever decisions are made regarding vaccines against the SARS-COV-2 virus we need to ensure that they are based on the most appropriate scientific evidence and, in that respect, randomised clinical trials must be the golden standard to prove efficacy and safety. No one will benefit from an early approval of a vaccine that in the end will prove to lack efficacy and or have serious side effects. The MCDA model of decision making that I described could be a supporting tool for regulators and public health authorities to find out what is the right development and approval strategy for a new vaccine.
By Thomas Lönngren