As next generation Covid-19 vaccine developers consider whether ethical and practical clinical efficacy trials can be conducted, they might consider whether vaccine efficacy can instead be inferred. We consider the alternative routes for vaccine developers of using an immune correlate of protection or immunobridging studies and some potential risks associated with such approaches.
Setting the scene
At the time of writing, the European Medicines Agency (EMA) and the Medicines & Healthcare products Regulatory Agency (MHRA) have granted temporary / conditional marketing authorisations for three Covid-19 vaccines, two of which employ mRNA technologies. According to the World Health Organisation (WHO), more than 200 Covid-19 vaccines are currently under development (see their Covid-19 candidate vaccines tracker here) and the coming months will likely see a considerable number of developers apply for regulatory approval of their Covid-19 vaccines across the world.
However, in the present Covid-19 pandemic, second-generation Covid-19 vaccine developers may be faced with a number of ethical and practical concerns when designing clinical trials to combat the novel virus. For example, is it ethical to carry out placebo-controlled efficacy trials, which might involve withholding approved Covid-19 treatments from trial participants?[i] In view of the widespread vaccine rollouts taking place in a number of countries to target the pandemic, it also might not be possible in practice to find a sufficient number of unvaccinated volunteers to take part in clinical trials.[ii] Further, vaccine developers may be concerned that regulators will require their candidate vaccines to be measured against a “gold standard” of immunisation demonstrated by first generation vaccines and supply of those first generation vaccines may be limited.
Faced with such issues, it should be questioned whether alternative solutions to full efficacy trials exist so that second generation vaccine developers might take advantage of pre-existing bodies of research on Covid-19.
Regulatory approval requirements for vaccines
EU Directive 2001/83 (the Directive) sets out the general obligations for vaccine developers to provide clinical trial results in their marketing authorisation application dossier (their Dossier).[iii] Generally, a developer of a medicinal product must include in their Dossier the results of clinical trials.[iv] However, Article 10 of the Directive (and the equivalent provisions in UK legislation) also includes a mechanism by which developers of generic, hybrid and biosimilar medicines are permitted to, rather than carry out full clinical trials, cross-refer to data relied upon in support of a previous authorisation. Such abridged applications may only be accepted for review after the expiry of the 8 year data exclusivity period that typically protects the previous authorisation.
These specific derogations may be of limited use to vaccine developers whose product does not neatly fall within the above categories. For them, a closer look at EMA and WHO guidance suggests that further alternatives might be available which would allow them to submit applications based on data from head-to-head trials or on data submitted by vaccine developers whose product was subsequently approved.
Alternatives to efficacy trials
The EMA Draft Guideline on Clinical Evaluation of New Vaccines (the Guideline) provides that a new vaccine can be authorised without an efficacy trial in two circumstances: (i) where there is a well-established immune correlate of protection (ICP); or (ii) where immunobridging takes place.
Immune correlate of protection
An ICP is “a type and amount of immunological response that correlates with vaccine-induced protection against an infectious disease and that is considered predictive of clinical efficacy”.[v] It is a measurable biomarker that can be used to interpret the immune responses to a specific antigenic component.
Being able to rely on an ICP would simplify and accelerate vaccine development by allowing developers, instead of proving efficacy through lengthy Phase III efficacy trials, to demonstrate that their product achieves an accepted immune response in patients from which efficacy can be inferred.[vi]
However, an ICP must first be established and accepted which may take considerable time and may not be feasible. Often ICPs have only been determined from vaccine efficacy trials with long-term follow-up of subjects post-marketing.[vii] Furthermore, only a limited number of infectious diseases have a widely accepted ICP[viii] and to date there is no established ICP for Covid-19 (although several bodies are seeking to identify one).[ix] Even where an ICP is identified, it may be of limited assistance. For example, where an ICP is found for a particular disease, it may not be applicable beyond a specific population or type of vaccination (such as mRNA).
Immunobridging
Immunobridging refers to where vaccine efficacy can be inferred by demonstrating a non-inferior immune response between the candidate vaccine and a licensed vaccine for which efficacy and/or effectiveness against a specific disease has been estimated.[x]
As immunobridging requires cross-referring to the efficacy data of a licensed vaccine, next generation vaccine developers will have to give careful consideration to whether such a strategy would be prohibited by data exclusivity rules where vaccine development and initial authorisations have been taking place within the last 8 years, such as Covid-19 vaccines. An interesting issue to be considered is to what extent reliance upon an ICP would avoid any such data exclusivity issues if, for example, the ICP has been specifically established on the basis of data submitted in respect of a third party licensed vaccine. Aside from data exclusivity issues, there may be difficulties in obtaining supplies of a licensed vaccine for an immunobridging study given the current demand for Covid-19 vaccines and supply constraints.
Immunobridging strategies may therefore be principally seen in the context of estimating efficacy in specific groups which have been excluded from Covid-19 vaccine clinical trial participation and for supporting major changes to manufacturing processes or other modifications of existing vaccines.[xi] These modifications may be to address variant strains of SARS-CoV-2. The EMA[xii] and the ACCESS Consortium of regulatory authorities (Australia, Canada, Singapore, Switzerland and UK)[xiii] have recently issued guidance on the requirements for modifications. The EMA’s view is that the primary approach, in the absence of an ICP, is to conduct head-to-head immunobridging studies comparing the original and variant vaccines in vaccine naïve patients. If such a study is would not be in the best interest of patients because subjects should not receive the original vaccine as primary vaccination, the EMA has proposed that immune responses with the variant vaccine could be compared against prior generated data on the immune response to the original vaccine. In contrast the ACCESS Consortium’s guidance commences on the basis that a head to head comparative study may not be in the best interest of patients and that a stand-alone immunogenicity and reactogenicity study of the variant vaccine would be appropriate and a comparison conducted of immune measures in sera from individuals vaccinated with the parent vaccine.
Takeaways
In November 2020, the EMA stated that it expects that “at least one well-designed large-scale phase 3 efficacy trial would be required to support the marketing authorisation of a Covid-19 vaccine;[xiv] however, ethically and practically the landscape is rapidly changing as more Covid-19 vaccines become available. Vaccine developers facing challenges in designing full efficacy trials will need to consider how they can still satisfy the data requirements of their marketing authorisation application.
While regulators’ guidance refer to approaches based on relying on an ICP or immunobridging for inferring efficacy, such approaches may not be feasible for various reasons including the lack of an accepted ICP or potential data exclusivity risks. While an ICP is yet to be established for Covid-19, it may have a considerable impact in future and it can be hoped that scientific progress in establishing one will be speedy as has been the case generally for the development of Covid-19 vaccines.
Where neither ICPs nor immunobridging can be used to develop a vaccine, the EMA Guideline states that a vaccine efficacy trial should be conducted “whenever this is feasible”.[xv] Vaccine developers should give careful consideration to whether an efficacy trial can be conducted without large-scale placebo controlled trials by, for example, using human challenge studies if appropriate. Where such trials would not be feasible, the extent of the data that might be acceptable to support a marketing authorisation will have to be considered on a case-by-case basis and vaccine developers would be well advised to proactively engage in discussions with regulators.
This article was co-authored by Eline D’Joos and Sian Avery.
[i] See EMA Draft Guideline on Clinical Evaluation of New Vaccines (April 2018), section 5.2 (Efficacy Trial Designs), which suggests that a placebo might only be acceptable in situations where there is no licensed vaccine; see also https://www.nature.com/articles/s41591-021-01230-y.
[ii] See EMA Draft Guideline on Clinical Evaluation of New Vaccines, section 4.2.1 (Randomised Control Trials).
[iii] While the Directive and certain scientific guidelines issued by the EMA at the time of writing continues to be relevant for the UK, the MHRA may decide in the future to diverge from EU approval requirements (and it already has taken minor steps to deviate from the EU’s approach by granting temporary authorisations to supply rather than conditional marketing authorisations). The Medicines and Medical Devices Act received Royal Assent on 11 February 2021 which confers powers on the Secretary of State for Health and Social Care to amend existing secondary legislation on the regulation of medicines.
[iv] Articles 8(1), 8(3)(i) and 12 of the Directive.
[v] EMA Draft Guideline on Clinical Evaluation of Vaccines (26 April 2018), section 4.2.
[vi] ICPs can also be helpful for making modifications to a licenced vaccine. For example, there is a well-established ICP which assists in modifications of the annual flu vaccine. The flu vaccine has to be altered every year due to genetic changes in the influenza virus and it would not be feasible to do a large-scale clinical efficacy trial each year. See also https://www.cell.com/trends/microbiology/pdf/S0966-842X(18)30180-X.pdf; https://www.npr.org/sections/health-shots/2020/12/04/942939137/immune-response-in-animals-good-news-for-covid-19-vaccine-development?t=1612794317670; and https://www.nature.com/articles/s41591-021-01230-y; https://onlinelibrary.wiley.com/doi/full/10.1111/irv.12706#irv12706-sec-0004-title.
[vii] An ICP can be validated by a significant consensus within the scientific community or may be formally established and accepted by regulatory boards; however, to date, there is limited guidance available as to what the regulatory expectations are with respect to the adequacy of evidence for an ICP to be established.
[viii] For example, influenza, polio and hepatitis B. See section 5.4.1 World Health Organisation Draft Guidelines on Clinical Evaluation of Vaccines: Regulatory Expectations (27 January 2016).
[ix] For example, CEPI has established a network of laboratories to assess and compare immunological responses generated by COVID-19 vaccines (see here) and COVAX has been involved in seeking to identify an ICP (see here).
[x] See World Health Organisation Draft Guidelines on Clinical Evaluation of Vaccines: Regulatory Expectations (27 January 2016). WHO recommends that bridging studies should take the form of comparative immunogenicity trials designed to demonstrate non-inferiority, but warns that where a candidate vaccine contains additional subtypes of an organism, this will not be straightforward: see section 5.6.2.1, World Health Organisation Draft Guidelines on Clinical Evaluation of Vaccines: Regulatory Expectations (27 January 2016).
[xi] Such as infants, those with co-morbidities, pregnant and lactating women, etc. See ‘Considerations for Evaluation of Covid-19 Vaccines: Points to consider for manufacturers of COVID-19 vaccines’ (WHO, 25 November 2020), section 3.3.13.
[xii] EMA Reflection paper on the regulatory requirements for vaccines intended to provide protection against variant strain(s) of SARS-CoV-2 (February 2021)
[xiii] https://www.gov.uk/government/publications/access-consortium-guidance-on-strain-changes-in-authorised-covid-19-vaccines/guidance-on-strain-changes-in-authorised-covid-19-vaccines
[xiv] EMA considerations on COVID-19 vaccine approval (16 November 2020).
[xv] EMA Draft Guideline on Clinical Evaluation of Vaccines (26 April 2018), section 5.1.
