DRUG DEVELOPMENT – Human Challenge Studies in Vaccine Development


The human challenge model (HCM) for the study of pathogenic organisms comprises the deliberate exposure of humans to known or putative disease-causing material. The following review describes how the concept is being applied to help develop vaccines against a number of common diseases, including influenza, rhinoviruses, respiratory syncytial virus (RSV), cholera, malaria, dengue, and Salmonella typhi.


Since the 1990s, the human challenge model (HCM) has been used to provide early performance data through proof-of-concept (PoC) and mode of action (MoA) studies to accelerate the clinical development of vaccines. Governmental and commercial organizations are developing challenge models to accelerate vaccine development programs, eg, for siRNA and mAB trials in respiratory tract infections (influenza and RSV), and to gain a better understanding of the underlying pathological processes that drive immune responses. Such exploratory models started gaining favor in regulated clinical trials in the early 1990s and were first mentioned in regulatory guidance in 2010.

Users of HCM frequently emulate previous protocol designs, as these have proven value and offer a route toward a standardized model. To ensure statistically significant data, it is essential to employ previously characterized human challenge stocks with known attack rates, consistent symptomology, and predictable adverse event (AE) rates, ie, Virus Emergent Adverse Effects (VEAEs). However, despite the many years of use, there are still no agreed guidelines referencing the quality requirements for the challenge agent or performance of a human challenge study (HCS). It is generally accepted that challenge agents should be manufactured under cGMP conditions when used in the framework of a clinical trial; however, for experiments that are not considered a clinical trial, there is no formal advice. The FDA has published an expectation regarding challenge agents, and the EMA and several EU national agencies have indicated that a virus should comply with the quality requirements of an investigational medicinal product (IMP), but there are, as yet, no references to human challenge in the ICH-GCP regulations. It would assist the HCS community if an agreement could be reached as to quality requirements for challenge agents and its classification [IMP /Non-Investigational Medicinal Product (NIMP)]. The International Association for Biological Standardization has debated this subject, and some consensus was reached but no “white paper” or other draft guideline has, to date, been issued.

Regulatory-wise, HCS are accepted for the following:

• as PoC studies for influenza and other upper respiratory tractinfections. In such Early Phase studies, protective (vaccine) or curative (vaccine/drug) efficacy is being assessed. There are currently no specific guidelines regarding efficacy markers (correlates of protection) for PoC in HCS studies, although the FDA guideline on influenza studies mentions haemagglutination and tissue culture infective dose 50% (TCID50)

• as a method for determining optimal dosage (to identify the correct individual dose, dose range, or schedule for field studies)


For a challenge study to prove effective, it is essential that subjects are confirmed to be susceptible to the challenge agent prior to entering a trial to ensure a high attack rate (rate of infection). Unfortunately, despite a general agreement that measurement of the haemagglutination antibody (HA) is an effective method for assessing immunity to influenza, variabilities in the sensitivity of the HA assay, and changes in agglutination potential of viruses currently circulating worldwide, may serve to lower the predictability of attack rates from 100% to 80% or less. Prior to enrollment in a challenge trial, subjects should be pre-screened in order to assess the relative naivety of the cohort to the challenge agent. Subjects should also be quarantined for at least 48 hours prior to admission to avoid cross-contamination by concomitant diseases, or the entrance of a secondary infectious agent into the unit. Subjects are usually kept separated and are tested for adventitious agents prior to enrollment to reduce the chance of drop-outs or failures due to infection control failures.

Interventions in an HCS may be drug-related (dosing post-inoculation with the challenge agent), preventative-vaccine-related (vaccination up to 3 months prior to challenge), or therapeutic-vaccine-related (vaccinated just prior or at the time of inoculation). Other considerations affecting dose-response may be the dose or titre of the challenge virus, or the type/suitability of the virus to the model being tested, eg, strain-specific interventions.

Identifying the most appropriate outputs (objectives) may be crucial to the success or failure of an intervention, eg, are the correlates of protection (CoP) known? Can the assays be standardized against a World Health Organization (WHO) or other recognized control? What bioavailability requirements are there, and can they be related directly to the therapy/challenge agent, eg, cytokine stimulation? Are the symptoms or any AEs because of VEAEs or treatment emergent (TEAE)?

Wider ethical considerations concerning the HCM (extant to the subject informed consent) come into play when potential threat(s) to the subject, staff and the public collide. For example, can subjects be released into the community following a challenge study if they are still qRT-PCR or antigen positive? What is the relative risk associated with such subjects? Further work is required to assess the relative transmission rates of differing challenge agents and the mechanisms underpinning infectivity. To date, safety parameters are largely based on theory or observational studies and the principles of barrier nursing.

From a regulatory point of view, appropriate quarantine measures (isolation and supervision) are required to be put in place to assure the authorities that not only are trial staff and the community safe from immediate disease, but also that the likelihood of long-term sequelae within the subject cohort are low. Quarantine conditions are also part of the quality envelope, protecting subjects and staff from cross-infection, ensuring observation and interventional analyses may be performed in a timely manner, thus optimizing the accuracy or purity of measurements.


Late phase field trial data demonstrating the relative efficacies of the two predominant influenza vaccines – trivalent, cold-adapted influenza vaccines (CAIV-T – live) and trivalent influenza vaccines (TIV – lyophilized) – has been proven to be reproducible in the HCM. Measurements of efficacy (attack rates, incidence and severity of disease and antibody response) are similar for vaccine classes and viral serotypes.

Influenza, rhinoviruses, respiratory syncytial virus (RSV), cholera, malaria, dengue, and Salmonella typhi amongst others have all been trialed in Human Challenge Trials (HCTs) against potential vaccine candidates. The only limiting factors have proven to be the availability of a suitable agent (cGMP) and the potential for adverse events or long-term sequelae associated with certain agents.

The general regulatory principles that apply for influenza challenge trials also apply for other pathogens and vaccines. The agent of disease must possess the necessary characteristics, ie, high attack rate, short (acute) disease state, consistent symptomology, and available therapeutics or support measures to be considered as a challenge agent. An additional regulatory complication, depending on the virus and its GMO classification (type II/III), may be the difficulty in finding a suitable clinical pharmacology unit (CPU) with the appropriate quarantine facilities.


In the field of influenza, susceptible populations or cohorts may be challenged with differing strains or serotypes of influenza to enable the performance of PoC studies for antiviral drugs and vaccines; the establishment of relevant etiology/ies; assessments concerning the immunogenicity of candidate vaccines; and the kinetics of immune responses as well as the development of correlations between clinical trial data and resistance to infection. The development of such correlates has proven pivotal in the rapid evaluation of vaccine efficacy against seasonally variant influenza strains.

Thus, the HCM model may be used to accelerate both validation and discovery programs through concepts related to protective antibody levels and insights into novel mechanisms of immunity.

Regulators have voiced criticisms concerning the limitations of the HCM, including the fidelity of manufactured challenge agents to wild-type stains; reduced pathogenicity and symptomology; high viral counts in inocula and the volumes applied and route of inoculation; excessive attenuation of agents during the manufacturing process (passage associated); and the potential for mutations in surface proteins and genomic sequences. However, to ensure subject safety and to comply with infection control measures, subject/pathogen balance must be maintained in favor of the subject. Should challenge agents cause significant morbidity or long-term sequelae, the value of the model as a reliable and safe model would be diminished and the willingness of subjects to participate would suffer.

To try to address some of the issues raised by regulatory authorities, circulating and naturally attenuated strains of organisms are used as far as is possible (for example in malaria studies and Plasmodium falciparum) to best represent the naturally occurring disease state. However, both regulatory and ethical considerations may ultimately serve to limit the use of highly pathogenic or chronic disease-causing agents. It may also be the case that challenging subjects with replication defective agents, eg, HIV or HCV, may not be an acceptable solution for investigating such viral diseases until other in vivo models have been exhausted.


HCS may play a significant role in the Marketing Authorization Application (MAA) for both vaccines and drugs. PoC and limited safety data from HCS may also serve to accelerate and enhance later clinical trial applications (CTAs). However, regulators are rightly wary of placing too much emphasis on HCS data due to the limited cohort sizes (approx. 80 subjects); restricted safety database (inherent to studies with limited subject numbers and short timelines); a controlled study environment not strictly representative of the “natural habitat;” and the manner of transmission [large volumes of virus may have the potential to overwhelm natural (innate) immunity processes].

It has been suggested that performing larger challenge studies (200+ volunteers) or supplementing the HCS with an early phase safety study, would increase the statistical significance of the data. However, emulating natural routes of infection is currently limited by our knowledge of viral epidemiology and the relative significance of droplet versus contact transmission rates.

Although it is inherent to the model that HCS studies should take place in a controlled environment, it may be possible to make subject inclusion/exclusion criteria less stringent in order to better mimic natural populations. Also, where correlates of protection demonstrate strain or intra-assay variability [the predominant strains of influenza A circulating since 2010 demonstrate poor haemagglutination in haemagglutination inhibition assays (HAI) – a primary marker for immunity], it may be an appropriate strategy to use HCS as a direct in vivo efficacy model. Finally, for some indications in which the population to be challenged is similar to the population that will receive the vaccine, eg, vaccines for travelers, HCS could be considered for use as a pivotal study.


More work is required to identify and define more effective/predictive correlates of protection. The lack of understanding behind some vaccine models must serve to lessen the impact of many studies, both early and late phase, if such poorly understood or inconsistent correlates alone are used as the primary outcome. Although it has been stated that “efficacy trumps all,” that efficacy cannot be later translated into advances in knowledge or understanding if the CoP can only be defined as “it worked” and not directly related to a measurable marker. Both the academic and commercial sectors require improved correlates to measure the effect of therapeutics of all types on disease states.


It is becoming widely accepted that HCMs can tangibly accelerate certain pipelines and add directly to the body of knowledge relating to a product and its interactions, both with the challenge agent and the host. As the HCM becomes popularized as part of the route to licensing, accessibility to the model and relevant challenge agents may become more of an issue than acceptability. Efforts to make cGMP challenge agents more widely available are being supported by the NIH; however, historically, private investors have funded and thus driven the model, bridging the transition of HCM from academia to industry. Given the considerable investment required, it is unlikely that many new players will enter the field in the short-term and that the challenge will remain a premium service with limited access.

With a global need for new antibiotics and antivirals and a low take-up rate by industry for less-rewarding avenues and indications, it may fall to state institutions, including regulators alike to promote novel means of bringing investigational products rapidly to market, where safety is not compromised, and early efficacy data can enable the rapid prioritization of promising candidates.

In conclusion, although further evolution of the model may be required before it can address all the regulatory strictures, the HCM has already provided efficacy data for a large number clinical trials, principally for upper-respiratory tract infections, and has been instrumental in providing timely non-inferiority data for a range of new anti-influenza drugs. Requests for HCS continue to increase year over year, and it is to be expected that as early phase trials are potentiated, the number and size of late-phase studies may be reduced in accordance with the value of the EP data. Human challenge studies are a relatively new phenomenon in EP trials, and their contribution has still tobe fully realized.

Bruno Speder holds a degree in Bioengineering and a degree in Business Economics from the University of Ghent, and has an additional degree in Health Economics from the EHSAL Management School. He joined SGS Life Science Services in 2008, and has held several positions in the regulatory group before assuming his current position as Head of Clinical Regulatory Affairs. In this role, Mr. Speder is involved in all the regulatory aspects of drug development, with a focus on regulatory support to sponsors in early development phases.

Adrian Wildfire is Project Director, Infectious Diseases and Viral Infection Unit, at SGS Life Science Services and has nearly 30 years of experience in communicable diseases. He has trained and worked within the fields of bacteriology, virology, parasitology, and mycology, obtaining his Fellowship in Medical Microbiology in 1990 and a Masters in Parasitology in 1998. He is the author of numerous papers and has been employed by a range of infectious disease key opinion leaders in tuberculosis, HIV, and influenza/upper respiratory tract infections at various hospitals, academic institutions, and CRO organizations. These include the Royal Postgraduate Medical School (Imperial College), the St. Stephen’s Centre (Chelsea and Westminster Hospital) and Retroscreen.