Health authorities will soon ask millions in the United States to take booster vaccines for COVID-19. Vaccines are developed by drug makers but require approvals from regulators validating their safety and efficacy for human use. In stunning news leaving many health experts scratching their heads, regulators at the U.S. Food and Drug Administration (FDA) signaled their willingness to accept efficacy data generated in animals on the latest COVID-19 subvariants, Omicron BA.4 and BA.5, in lieu of data produced in humans. The results will be examined in conjunction with safety data reported for an earlier Omicron subvariant, BA.1.
The desire to speed up regulatory acceptance in time for a planned Fall 2022 booster rollout seems to be the key driver behind these unprecedented steps by the agency. On August 17, White House COVID-19 Response Coordinator Ashish Jha asserted that boosters will be available to teens and adults “in a few short weeks.” On August 31, the FDA approved the COVID-19 boosters in a rush, clearing the process for final authorization by the Centers for Disease Control and Prevention (CDC). But substituting results in animals for critical health data in humans is a very misguided practice.
Vaccines are among the most reliable public health interventions against infectious diseases. Throughout the years, immunization by vaccination has been crucial in protecting human health. A good example showing the benefit of vaccines is the drastic differences documented in the number of COVID-19 deaths between unvaccinated and vaccinated individuals. That said, the protection offered by vaccines wanes over time and for several reasons. For instance, new variants can go undetected by the immune system and “outsmart” the immunity offered by earlier immunization. The immune system itself weakens with age or in the presence of other medical conditions simultaneously, as shown in COVID-19 patients.
By design, COVID-19 boosters seek to enhance one’s protection from coronavirus and confer immunity against new emerging variants. Viruses, including coronaviruses, evolve constantly, sometimes faster than a vaccine can stop them. More than 12 variants of COVID-19 have been identified to date. This includes Omicron, which is of great concern to public health officials and the focus of intense vaccine development efforts worldwide. Adding to the complexity, Omicron has many subvariants like the commonly discussed Omicron BA.1, BA.4 and BA.5.
During the COVID-19 pandemic, health officials often used the term “flattening the curve” as a key public health strategy. The curve is a visual representation of the number of patients needing care over time. With emerging COVID-19 variants and subvariants, keeping the curve flat through effective immunization boosters is essential to reduce the pressure on the health care system, enable the development of the most effective vaccines and prevent another pandemic.
On August 15, the U.K. became the first country worldwide to approve a dual COVID-19 vaccine targeting both Omicron BA.1 and the original virus, SARS-CoV-2. This approval was based on robust safety and efficacy data obtained from a clinical trial investigating Omicron BA.1. Today, BA.5 is the most prevalent Omicron subvariant in the U.S., accounting for almost 90 percent of new COVID-19 infections, according to the CDC. It is also believed to be the most contagious.
To this end, U.S. regulators have opted not to adopt the strategy of their U.K. counterparts. Instead, in June 2022, the FDA asked drug makers to develop multivariant booster vaccines, i.e., ones that protect against multiple variants at once, including the most recent BA.4 and BA.5 subvariants. Drug makers often find themselves in a race against time to tailor vaccines to the latest variants and obtain regulatory acceptance. Normally, this entails generating preclinical data, usually in artificial animal models, followed by decisive safety and efficacy data in humans in the form of clinical trials, also known as first in humans (FIH) trials. This type of trial remains the gold standard needed for approvals of pharmaceutical drugs and vaccines.
Numerous issues are at play here. First, compared to the original virus or amongst themselves, subvariants like BA.1, BA.4 and BA.5 can vary considerably in terms of their acute effects in humans, like the havoc they wreak on the immune system and their capacity to evolve and spread. Almost all new COVID-19 variants, including the Omicron subvariants, have not been fully characterized yet. Simply put, we do not know enough about these subvariants to bypass testing them in humans, prior to making their booster vaccines available to millions. We are also just beginning to understand their long-haul effects.
Second, substituting the efficacy data in mice, or any artificial animal model, for human data is extraordinary under any circumstances. It is not a secret that scientific findings in animals often fail to translate to medical discoveries in humans. In fact, 90 percent of experimental drugs tested on animals do not work in humans due to either safety or efficacy issues. In most contexts, animal data used to inform clinical applications need to be met with the utmost skepticism.
Separately, it is hard to reconcile this FDA approach with the landmark decision taken at the March 2020 meeting of the International Coalition of Medicines Regulatory Authorities (ICMRA) — more specifically during the Global regulatory workshop on COVID-19 vaccine development. There, global regulators stipulated that, “It is not required to demonstrate the efficacy of the SARS-CoV-2 vaccine candidate in animal challenge models prior to proceeding to FIH clinical trials.” That important policy decision was a vital step to accelerate the development of successful vaccines in the first place. Ostensibly, now the reverse is being promoted by the FDA.
In retrospect, much of the COVID-19 animal data generated during the pandemic produced more questions than answers. Had it not been for the world’s full attention on developing vaccines, relying only on artificial animal models for directions and “next steps” could have been catastrophic from the standpoint of saving time or generating actionable information. For instance, studying COVID-19 in mice, the most widespread model in the world, is misleading. Mice — unlike humans — are not naturally susceptible to COVID-19 due to a key structural difference between the two species in their host-pathogen interaction. For example, a cell surface receptor (ACE2) that serves as a doorway for coronaviruses to infect cells is unresponsive in mice. Murine models had to be artificially modified or “humanized” before they could be infected.
Furthermore, non-human primates (NHPs), considered closer to humans than other species, showed very mild clinical symptoms in response to SARS-CoV-2, practically rendering them unreliable as research models for COVID-19. No fatalities from disease burden were registered in SARS-CoV-2 infected NHPs, making them also inappropriate for modeling COVID-19-induced multisystem organ failure, a hallmark of COVID-19-related deaths. In fact, an objective examination of the many artificial animal models used to study COVID-19 reveals disparate characteristics — with hardly any capturing the sequela of COVID-19 infections in humans. Radical structural, physiological, anatomic, digestive, genomic, metabolic and behavioral differences underlie those discrepancies among species.
In fairness, the FDA has taken concrete steps to explore innovative technologies that could reduce the reliance on artificial animal models. These include advancing the science and applications of Microphysiological Systems (MPS) — a category of sophisticated research tools, like Organ-Chips, also known as alternative methods to animal experimentation. These advanced systems are made possible by the convergence of many disciplines such as biophysics, regenerative medicine and electrical engineering. For instance, the FDA was a co-host of an inaugural World Summit focusing on MPS earlier in 2022. The agency also sustains programs like the Predictive Toxicology Roadmap and ISTAND with somewhat similar objectives. More recently, a new initiative, although poorly funded, has been conceived by the FDA “to implement a cross-agency New Alternative Methods Program.” In its budget proposal for fiscal year 2023, the agency expressly stated that, “New alternative methods have the potential to provide both more timely and more predictive information to accelerate product development and enhance emergency preparedness.” In addition to Organ-Chips, alternative methods include artificial intelligence (AI) or machine learning approaches as well as three-dimensional (3D) bioprinting techniques, which are sophisticated engineering methods for the fabrication of biomedical parts that replicate the features of natural human tissues.
All this is taking place in a climate of legislative momentum seeking to advance research methodologies that are based on human biology (e.g., the FDA Modernization Act). Efforts with similar goals are underway at leading biotech companies, academic institutions and nonprofit organizations. In this context, the decision to rely solely on animal efficacy data in the regulatory acceptance of booster vaccines is out of sync with the growing demands for human relevancy as a priority. It may also reflect competing dogmas within the agency regarding the reliance on animals for regulatory approvals.
In conclusion, the seemingly mix-and-match approach proposed by the FDA, substituting efficacy data in other species for critical data in humans and safety data on one Omicron subvariant for another, sets a negative precedent and might backfire. COVID-19 vaccination is critical for public health but relying only on animal data is a needless gamble. This reliance is nothing short of cutting corners and never a clever health policy. Besides, any reports of health complications or deaths that could have been avoided with human data will be used by vaccination detractors to generate more skepticism among the public, and undermine an effective public health measure. For the planned Fall 2022 boosters, the FDA must explore alternative scenarios like obtaining, at minimum, robust data from representative human cohorts or allowing time for clinical trial results on Omicron subvariants to be completed. In the long term, the agency needs to bolster MPS technologies or similar human-based approaches as viable alternatives to animal experimentation. Short of that, the FDA will not be looking forward.