The quest for a COVID-19 vaccine: Promising first results

Nearly 140 vaccine developers have been in the race to come up with a COVID-19 vaccine. Among them, four of the five who were first off the blocks, soon after Chinese scientists made public (January 11) the genetic sequence of the causative virus SARS-CoV-2 (Wuhan strain), have turned in the results of their performances in the first important round, called the phase 1/phase 2 of human or clinical trials. And, all of the candidate vaccines look promising enough to enter the next, critical round, the third phase of clinical trials. (The preclinical stage involves animal trials that are meant to delineate the broad action profile of a drug or a vaccine, its general safety and to identify its toxicity patterns.)

The phase1/2 results show that all the candidate vaccines have the potential to protect vaccinated individuals against the disease. But do they actually protect vaccinated individuals? And if they do, how much protection do they confer, and how long will the conferred protection last—months, a year or years? Is a booster dose required even to mount a significant initial immune response? The answers to these questions are what the results of phase 3 trials are expected to provide. While three of these front runners have published their results in peer-reviewed journals, the fourth is yet to be publish its results in such a journal but the authors have made them public through medRxiv , a “not yet peer-reviewed” open e-print web repository.

The earliest of the four to take off were Moderna, a biotechnology company in the United States that is collaborating with the National Institute of Allergy and Infectious Diseases (NIAID) of the U.S. National Institutes of Health (NIH), and CanSino Biologics, a Chinese biotech company. These two started their phase 1 trials in March itself. The other two—the Oxford Vaccine Group (OVG) in the United Kingdom that has partnered with the British-Swedish company AstraZeneca, and a collaboration between the U.S. multinational Pfizer and the German BioNTech, a relatively new company specialising in mRNA (messenger-RNA) based human drugs and therapeutics—began their trials only in April. But all four started the vaccine development work soon after the genome sequence of the SARS-CoV-2 was published.

These four candidate vaccines deploy advanced vaccine technologies and none of them has the whole virus (live attenuated or inactivated), as is common in older vaccines. To be sure, no vaccine based on these technologies has so far been licensed for public use for any other disease; all earlier attempts were only up to preclinical or early clinical stages.

Moderna-NIAID and Pfizer-BioNTech have used the mRNA platform to code for antigens—the relevant proteins of any pathogen, in this case SARS-CoV-2 (an RNA virus)—so that the vaccinated individual mounts an appropriate immune response against it. Upon the delivery of the vaccine into the body, the host cells translate the RNA sequence to produce the encoded antigens, which then stimulate the body’s adaptive immune system to produce antibodies against the pathogen. According to the promoters of this technology, RNA/mRNA vaccines provide flexibility in the design and expression of vaccine antigens that can mimic antigen structure and expression during natural infection. While RNA is required for protein synthesis, it does not integrate into the genome, it is transiently expressed and is metabolised and eliminated by the body’s natural mechanisms and is, therefore, considered safe. Importantly, the RNA-vaccine manufacturing platform has the ability to rapidly produce large quantities of vaccine doses irrespective of the encoded pathogen antigen. The CanSino and OVG candidate vaccines contain a viral vector to deliver the genetic sequence of the identified antigenic component of the pathogen that is integrated into the vector by recombinant technology. So, when a person is vaccinated, the necessary immune response is generated. In recent times, adenoviruses have emerged as the vector of choice for vaccination because they do not integrate with the genome of the host and are also non-replicating.

Both the OVG and the CanSino vaccine candidates have a genetically altered adenovirus as the viral vector, but with a crucial difference. While the Oxford group uses a chimpanzee adenovirus (ChAd), which does not infect humans, CanSino uses the human-infecting adenovirus called adenovirus5 (Ad5). Most humans thus normally carry antibodies to human adenoviruses, and this, as we shall see later, can have a confounding effect in the interpretation of trial results of vaccines with human adenoviruses as the carrier. The true antibody response of such a vaccine would be hard to delineate.

The initial (phase 1/phase 2) results of the trials of all four vaccine candidates were published in July, with the Pfizer-BioNTech trial results still to be peer-reviewed. The pace at which these four vaccine developers have successfully poised themselves to begin phase 3 of the trials in July itself, just four to five months after the start of their phase 1 trials, is indeed remarkable considering that these stages of vaccine development are known to take several years (Figure 1). The urgency dictated by the unprecedented scale of the COVID-19 pandemic has prompted scientists, industry and regulators to fast-track the development of vaccines by combining, or having concurrent running of, phase 1 and 2 trials, and with fewer trial recruits than normal.

Moderna’s vaccine basically consists of mRNA instructions to build the coronavirus’s spike (S) protein. The S protein (which has two subunits, S1 and S2) is the crucial antigenic component of the COVID-19 virus (SARS-CoV-2) that enables it to bind to human cells, gain entry into them and use the host’s replication machinery to multiply itself (Figure 2). The S protein is, in fact, the primary target of most candidate COVID-19 vaccines. Emerging vaccine technologies have enabled delivery through suitable platforms of just the relevant antigens of the pathogen, in this case the S protein alone—not the virus’ entire DNA or RNA—which enables human cells to produce the foreign protein in sufficient quantities so that the immune system is suitably primed to respond when infected by the actual SARS-CoV-2. Called mRNA-1273, Moderna’s vaccine consists of a suitably stabilised mRNA that encodes the SARS-CoV-2 S protein, which is encapsulated in a lipid nanoparticle. The first results of the phase 1 trials of this vaccine were announced through a press release in May, but it lacked details whereby experts could judge its real efficacy. The details were only published on July 14 in The New England Journal of Medicine , but as a “preliminary report”. But even now many experts would reserve their judgment until complete details are published.

“Experience with the mRNA platform for other candidate vaccines [Moderna has used it to develop bird flu vaccines],” the authors write in their paper, “and rapid manufacturing allowed the deployment of a first-in-human clinical vaccine candidate in record time.” In fact, the first trial participants were vaccinated on March 16, just 66 days after the genomic sequence of the virus was posted. The NIH conducted the trials. The results showed that the recruits (between 18 and 55 years of age) for the trial who received the vaccine made more neutralising antibodies than were seen in convalescent COVID-19 patients. But a second booster dose, four weeks after the first, was required before the vaccine produced a substantive immune response. This preliminary report is stated to be the first of three reports of data from the phase 1 trials. The second will include data from adults older than 55 years, and a final report will include results on the safety aspects and on how long the immunity is expected to last. A durability study will be based on following up the participants’ immunogenicity profile over a period of one year of the phase 1 trial.

The first part of the study enrolled 45 healthy individuals of 18-55 years of age, who were divided equally into three groups and received two injections of the trial vaccine 28 days apart at 25 micrograms, 100 micrograms and 250 micrograms respectively. The vaccine was administered as a 0.5 ml injection in the upper arm muscle. According to the authors, the two-dose regimen was generally without serious toxicity; systemic adverse events after the first vaccination were all mild or moderate. However, greater reactogenicity was seen after the second vaccination, particularly in the 250 microgram group, and in all the three dose groups, local injection-site reactions were primarily mild.

To assess the immunogenicity of the engineered vaccine, through suitable assays the study measured the binding antibody response to the virus’ entire S protein and specifically to the receptor-binding domain (RBD), which is a part of the subunit S1. The vaccine-induced neutralising antibody response was also assessed through an assay that used a pseudotyped lentivirus (a lentivirus combined with the coronavirus’ envelope and S-proteins) and also a wild-type SARS-COV-2-based assay. (B-cells produce a diversity of antibodies and not all antibodies produced can neutralise the pathogen to prevent it from infecting cells; some just bind to a pathogen’s antigen and signal the other arms of the immune system to act.) These assays tell us whether the vaccine can also produce antibodies that actually neutralise the virus.

The study compared the vaccine-elicited immune response to that induced by actual infection in recovered COVID-19 patients for which the serum specimens of 41 convalescing patients were used. The study notes that the binding antibody titres increased rapidly after the first vaccination, with seroconversion (when antibodies can be detected in body fluids) being attained by all participants by day 15. It was seen that binding antibody titres increased with the dose for both the first and the second vaccination. RBD-specific antibody measures, according to the study, were similar in pattern and magnitude.

While the median value of the antibody titres after the first vaccination (for both the S protein and the RBD) were similar in magnitude to that in convalescent serum specimens, the titre on day 57 (after trial participants got a booster shot on day 29) for the S protein antibody, even for the lowest dose of 25 micrograms, was more than double the median value in the convalescent serum samples. These responses, according to the authors, are quite robust. As regards the neutralising antibodies, while responses (which were dose-dependent) were detected only in fewer than half of the volunteers before the booster dose, neutralising antibodies could be detected (on day 43) in all the participants with the lowest response seen in the 25 microgram group and similar responses in the other two higher dose groups. These responses, according to the paper, were similar to the upper half of the distribution of values in the serum samples. More importantly, on day 43, the neutralising activity against the wild-type SARS-CoV-2 assay reduced virus infectivity by 80 per cent or more in all the participants. And these too were either similar or higher than the values seen in the convalescent serum samples. This, the paper notes, supports the need for a vaccination schedule that includes a booster dose. According to experts, while the vaccine does seem to elicit antibodies that neutralise the virus, what we do not know is how much we actually need for protection.

T-cell responses

As important as the above antibody responses is the observation of T-cell responses in the trial. T-cells are certain types of white blood cells and part of the immune system. While the so-called CD8 T-cells directly target infected cells and eliminate them, the CD4 T-cells, also called helper T-cells, trigger other parts of the immune system to act in achieving the same goal. The 25 microgram and 100 microgram doses elicited the CD4 T-cell response. The CD8 T-cell response could, however, be elicited only at a low level in the 100 microgram group and that too only after the second vaccination.

“The hallmark of a vaccine,” Dr Anthony Fauci, the director of the NIAID and the top government medical expert in the U.S. efforts to combat the disease, has been quoted as saying, “is one that can actually mimic natural infection and induce the kind of response that you would get with natural infection. And it looks like, at least in this limited, small number of individuals, that is exactly what’s happening. The data really look quite good [and] there were no serious adverse events.”

Of the three doses evaluated,” says the study report, “the 100-microgram dose elicited high neutralisation responses and… CD4 T cell responses, coupled with a reactogenicity profile that is more favorable than that of the higher dose. These safety and immunogenicity findings support advancement of the mRNA-1273 vaccine to later-stage clinical trials.” Following the observation of some serious adverse events in the 250 microgram group, the researchers have dropped the idea of using this high dose in the forthcoming trials.

A phase 2 trial of the vaccine in 600 healthy adults, evaluating doses of 50 micrograms and 100 micrograms is currently in progress. A large phase 3 efficacy trial, expected to evaluate a 100 microgram dose in 30,000 participants, is scheduled begin on July 27, according to the U.S. government registry of clinical trials.

Pfizer-BioNTech’s vaccine

Because Pfizer-BioNTech has used similar technology in its vaccine, the phase 1/2 trial results of this U.S.-German vaccine almost mirror those of the Moderna-NIAID vaccine described above and have invariably invited comparisons.

On July 1, Pfizer-BioNTech announced the results of the phase 1/2 trials of its vaccine called BNT 162b1 in the U.S., which were conducted between May 4 and June 19. The trial assessed safety, reactogenicity and immunogenicity, with the last parameter being assessed only in terms of antibody (IgG) titre measurement. Concurrent trials with similar cohorts were carried out in Germany between April 23 and May 22 but with broader immunogenicity measures that included T-cell responses as well.

While Moderna did not use placebo controls in its study, both the U.S. and German trials were placebo-controlled, single (observer)-blind studies to evaluate safety, tolerability and immunogenicity with increasing dose levels of the vaccine. The results of both have been posted on medRxiv , the U.S. trial results on July 1 and the German trial results on July 20. BNT162b1 is the most advanced of the four mRNA vaccines against SARS-CoV-2 that the Pfizer-BioNTech collaboration is studying under a programme called “Project Lightspeed”. Like the Moderna-NIAID vaccine, BNT162b1 too is a modified lipid-nanoparticle-encapsulated mRNA candidate that encodes an optimised RBD antigen only. The Moderna vaccine coded for the entire S protein. According to the U.S. trial results, the doses administered were well tolerated and generated dose-dependent immunogenicity as measured by RBD-binding and virus-neutralising antibody titres.

The U.S. study had 45 healthy subjects (18-55 years of age), 24 of whom received two injections 21 days apart of 10 micrograms or 30 micrograms (12+12), 12 received a single dose of 100 micrograms and the remaining 9 received two doses of a placebo also 21 days apart. In all the 24 subjects, because of the strong booster effect, dose-dependent increase of RBD-binding IgG antibody concentration was observed seven days after the second dose. The (geometric) mean concentrations of the two doses were found to be respectively 8 and 46.3 times the value of the (geometric) mean concentration of 38 convalescent serum samples. Similarly, the highest neutralising antibody titres for the two doses were seen seven days after the second dose, which were respectively 1.8 and 2.8 times the values in the serum samples. As regards the antibody responses in the single-dose 100 microgram group, the IgG concentrations and neutralising antibody titres were respectively 3 and 0.35 times the corresponding mean values of serum samples.

According to the study, at the 10 microgram or 30 microgram dose levels, adverse reactions, including low grade fever, were more common after the second dose than after the first. Local reactions and systemic events after injection were dose-dependent, generally mild to moderate and transient, and no serious adverse events were reported. The trial did not give the 100 microgram group a second dose because of the greater number of local reactions and systemic events after a single dose itself and also because 100 micrograms did not show any significant increase in immunogenicity compared with the 30 microgram dose. The German trial was evaluated with 60 healthy volunteers (18-55 years) with a two-dose vaccination at 1 microgram, 10 micrograms, 30 micrograms and 50 micrograms dose levels given on day 1 and day 22. The remaining 12 received a single dose of 60 micrograms. According to the trial’s preliminary results, the vaccine elicited high, dose-dependent virus-neutralising titers and RBD-binding IgG concentrations after the second dose. Measured on day 43, the neutralising antibody titres were 0.7 (1 microgram) to 3.5 (50 micrograms) times compared with the mean value of a panel of convalescent sera. On the basis of the reactogenicity at the 50 microgram level, the 60 microgram group was not administered a second dose. Significantly, the sera of vaccinated subjects were also found to broadly neutralise in pseudovirus assays across 16 SARS-CoV-2 RBD variants publicly documented and against the dominant D614G strain (see “Vaccine scenarios”, Frontline , July 3).

The German trial also assessed the T-cell responses in the participants. The observed responses, according to the published results, demonstrated a high level of CD4 and CD8 T-cell responses against the virus. Although the strength of T-cell responses varied between subjects, there was no clear dose-dependence from 1 microgram to 50 micrograms, which indicates that even low mRNA dose vaccination could elicit significant RBD-specific CD4 and CD8 T-cell responses. This contrasts with the Moderna trials, which saw only a low level CD8 T-cell response at lower doses. The other significant difference between Moderna-NIAID’s mRNA-1273 vaccine and Pfizer-BioNTech’s BNT162b1 vaccine is that the latter induced significant immunogenicity at comparatively lower dose levels. However, as Dr Fauci remarked to the online publication “STAT” when asked to compare the two: “I don’t think you could say anything about one being better than the other. They both induce good responses. Let’s see what happens in the real world.”

The two companies will use the preliminary data from both the German and U.S. phase 1/2 studies to determine a dose level to progress to a large, global phase 2b/3 safety and efficacy trial. “That trial may involve up to 30,000 healthy participants and is anticipated to begin in late July 2020,” BioNTech stated in a July 20 press release.

Oxford-AstraZeneca’s vaccine

Of the four vaccines being discussed here, the results of the Oxford vaccine trials seem to have been the most awaited in the media, though the reasons for that are not clear. Sarah Gilbert, a professor of vaccinology at Oxford University who leads the OVG, had recently developed a simian adenovirus-based recombinant viral vector vaccine for the Middle East respiratory syndrome (MERS), which is caused by another coronavirus. In fact, trials with the MERS vaccine had shown that a single dose of an adenovirus-vectored vaccine, which encodes the S protein of MERS-CoV, protected non-human primates against MERS. Given that experience, Sarah Gilbert began working on a vaccine against COVID-19 in January (when the virus was called nCoV-19) on a non-replicating simian adenovirus vector vaccine that would express SARS-CoV-2’s S protein and named it ChAdOx1 nCoV-19 (Figure 3). The code for the complete S protein is integrated into the vector genome.

With the vaccine’s phase 1/2 trials showing promising results, which were published in The Lancet on July 20, the OVG is already conducting one part of its phase 3 trials in Brazil, South Africa and the U.K. Although the Oxford vaccine trials began a little late, it is the first vaccine to have already begun its phase 3 trials. The larger phase 3 trial, with 30,000 volunteers, will begin soon in the U.S. The phase 1/2 results showed that there were no safety concerns and that the vaccine induced significant cell-mediated and humoral (antibody) immune responses. The vaccine elicited T-cell activity within 14 days and an antibody response within 28 days of vaccination. These responses were strongest after a booster dose, with all the participants having virus-neutralising activity. Before the phase 1/2 trials, the OVG had found in preclinical studies in rhesus macaques that a single vaccination with the vaccine elicited antibodies and a cellular immune response. Protection against lower respiratory tract infection was also observed in vaccinated non-human primates after a high-dose SARS-CoV-2 challenge.

The U.K. phase1/2 trials were conducted between April 23 and May 21 at five sites in the U.K. This randomised single-blind placebo-controlled trial with the ChAdOx vaccine nCoV-19 (also christened AZD1222) had 1,077 recruits. Of them, 10 were assigned to a non-randomised, unblended vaccine prime-boost group. The rest of the participants were randomly assigned in a 1:1 (vaccine/placebo) ratio and either received ChAdOx1 nCoV-19 at a dose of 50 billion (5 × 10 ¹⁰ ) viral particles or the placebo, which was the meningococcal conjugate vaccine (MenACWY). (The dose was decided on the basis of OVG's experience with its ChAdOx MERS vaccine.) Both the vaccine and the placebo were given as a single intramuscular injection. The prime-boost group received a two-dose schedule, with the booster-dose given 28 days after the first vaccination. The protocol also included giving a paracetomol to the participants as a prophylactic to prevent post-vaccination fever.

According to the study, post-vaccination, while local and systemic reactions (pain, feeling feverish, and so on) were more in the ChAdOx1 nCoV-19 group than in the placebo vaccine group, no serious adverse reaction event was reported. The results with ChAdOx1 nCoV-19 showed that a single dose of the vaccine elicited an increase in spike-specific (binding) antibodies by day 28, which was boosted with the second dose. Neutralising antibodies were seen in all the participants after a booster dose on day 42 as measured by different assays, including a pseudovirus-based assay. The IgG titres too increased with a two-dose regimen, and the paper says that further work on a two-dose regimen is underway. The trial also showed that ChAdOx1 nCoV-19 resulted in a marked increase in spike-specific T-cell responses (including CD4 and CD8 cells) as early as day 7 with peaking on day 14 and remained elevated up to day 56. “However,” the paper says, “a boost in cellular response was not observed following the second dose.”

According to the paper, older age groups with co-morbidities, health care workers, and those with higher risk for SARS-CoV-2 exposure are being recruited for further ChAdOx nCoV-19 trials, given as a single-dose or two-dose regimen, in the U.K. and elsewhere. The study concludes that ChAdOx1 nCoV-19 is safe, tolerated and immunogenic, while reactogenicity is reduced with paracetamol. A single dose elicited both antibodies and cellular responses against SARS-CoV-2, with a booster dose augmenting antibody titres. The preliminary phase 1/2 results supported the ongoing phase 2 and 3 trials.

CanSino’s vaccine

In March, a team of scientists from several Chinese institutes carried out a limited phase 1 trial of the recombinant viral-vector vaccine with the genetic sequence of the S protein integrated with the human adenovirus5 (Ad5); this is in contrast to ChAdOx1, which has a chimpanzee adenovirus. The trial was limited in the sense that it was restricted to a single centre and was open-labelled (as against blinded) and non-randomised. It was aimed at studying the safety, tolerability, reactogenicity and immunogenicity profiles at three different doses given to healthy Chinese adults: 50 billion, 100 billion and 150 billion (5×10 ¹⁰ , 1×10 ¹¹ and 1.5×10 ¹¹ ) viral particles.

The results of this trial, which were reported in May, showed that the vaccine was safe and well tolerated with promising immunogenicity. However, given the risk of severe adverse reactions at the high dose, the highest dose of 1.5×10 ¹¹ was dropped in the phase 2 trial. This study was aimed at further evaluating the immunogenicity and safety in a larger population and to determine the appropriate dose for the phase 3 efficacy study. The phase 2 trial also removed the age limit of 55 years to include older and more susceptible people in the study. The results of this study were published in The Lancet in July according to which CanSino’s candidate vaccine has a good safety profile, with only mild, transient adverse events related to vaccination and no serious adverse events, and good immunogenicity.

Between April 11 and 16, 508 volunteers were recruited for the study with a mean age of about 40 years. On the premise that a higher antigenic dose elicits greater immunogenicity, the participants were randomly assigned in the ratio 2:1:1 to receive 1×10 ¹¹ (D1), 5×10 ¹⁰ (D2) viral particles and the placebo respectively. The study found that a single injection of the vaccine at doses D1 and D2 induced comparable specific immune responses to the spike glycoprotein at day 28, with no significant differences noted between the two groups. Seroconversion of neutralising antibodies was observed in 59 per cent and 47 per cent of the participants, and that of binding antibodies in 96 per cent and 97 per cent of the participants in the D1 and D2 groups respectively. Positive specific T-cell responses measured by a suitable assay were found in 90 per cent and 88 per cent of the participants in the D1 and D2 groups respectively. At day 28 post-vaccination, 95 per cent and 91 per cent of the participants in the D1 and D2 groups respectively showed either a cellular or antibody response.

A limitation of this study is that all the recruits for the trial were from Wuhan. More importantly, as pointed out earlier in the article and as the authors themselves note, pre-existing immunity to the Ad5 vector and increasing age could significantly reduce the immune responses, particularly humoral, to the vaccine. So, they have argued that, for participants with high pre-existing anti-Ad5 immunity one injection of the vaccine might be inadequate to elicit a high level of antibodies, particularly for people 55 years or older because they are likely to have higher baseline levels of ant-Ad5 neutralizing antibodies. This, the authors say, indicates that this population might be more tolerant of higher dose or a booster dose regimen of the vaccine than those who are young and naive to Ad5. On the basis of their previous experience with the Ad5-viral-vector Ebola vaccine, the authors say that a flexible additional dose (between three and six months) might be a potential solution to provide enhancement of immune responses. On the basis of the conclusion drawn from the above results of the phase 2 trials that a single-dose immunisation schedule of CanSino’s vaccine at 5×10 ¹⁰ viral particles is an appropriate regimen for healthy adults, the promoters plan to start an international multi-centre, randomised, double-blind, controlled phase 3 efficacy trial with that regimen soon.

As none of the above vaccines has been tested for more than few weeks, their durability is yet to be determined. More importantly, controlled laboratory studies on immune responses do not guarantee protection against COVID-19. It is also impossible to compare the performances of the four vaccines on the basis of the above results alone because of the different trial protocols and the different formats of data presentation. Phase 3 trials with thousands of participants take a long time. Realistically, therefore, none of the above front runners is likely to bring a vaccine for public use before early to mid 2021.