It’s been a while since I went through the whole vaccine landscape (I’ve been putting it off!), but there’s a lot to catch up on. I’m going to incorporate some slightly reworked material from my July post in the introduction to each vaccine class, for reference, but everything on the candidates themselves is updated information. That earlier post includes more background on each specific candidate that I won’t be repeating here. Search for “update:” to see information added after this post was first published.
Update: (September 17) see here for a good roundup of lesser-known vaccine candidates.
This class uses some other infectious virus, but with its original genetic material removed. In its place goes genetic instructions to make coronavirus proteins, and when your infected cells do that, these proteins will set off an immune response. Note that this is different than being infected with a “real” virus, whose instructions are (naturally enough) to produce more virus, which go off and infect more cells. No, in this case each viral particle that you’re injected with will be able to infect one cell, and that’s it. An advantage of this approach is that it should appear to your immune system like a pretty realistic viral attack, and set off a full range of responses. A disadvantage is that this technique (as far as I can tell) has only once been used in human therapy (the Ebola vaccine), although it’s had a lot of work from a lot of groups over the years. Things have now accelerated, a phrase that you can just keep using these days. Another disadvantage is that if you pick a viral vector that infects people anyway, some of your patients may already have antibodies to that one. That can mean that your attempt to repurpose it might crash and burn as your carefully engineered vector gets attacked by antibodies and eaten by immune cells before it can even do its work. It also means that booster shots would have an uphill battle, since the antibodies from that first dose will be waiting for the second one. Antibodies to the viral payload: good. Antibodies to the viral vector itself: not so much.
Oxford/AstraZeneca: ChAdOx1-nCov19/AZD122: this one has been in thousands of patients in the UK, Brazil, and South Africa, and the team recently announced a trial in the US as well. And even though I just finished up talking about how booster shots with a viral vector can be tricky, that’s exactly what they’re studying: two doses, four weeks apart. Clinical results released in late July looked promising. There have been a number of supply deals announced, with this being the latest one, and both the EU and the US have paid to reserve hundreds of millions of doses if the clinical results are sufficient for a rollout. Update: (September 9) this trial has been paused as of September 8 on reports of a serious adverse event, likely transverse myelitis. More here, but this is an evolving story. Update: (September 17) the trial has resumed in the UK, but remains paused as of today in the US. See here for more. And a trial is starting to look at this one as an inhaled formulation as well.
CanSino/AMMS: we had clinical data from this one at about the same time as the Oxford publication. And you could indeed see the preexisting-immunity problem with this adenovirus (the more common human-infecting Ad5, as opposed to the Oxford chimpanzee adenovirus); the paper itself stated that this was a big concern. Update: (September 17) the company is defending their choice of vectors. CanSino has just cancelled plans for a Phase III trial in Canada, for reasons that frankly are still obscure – probably politics, but to be honest, Canada is doing pretty well with the pandemic and is probably not a good place to collect Phase III data to start with. But they’re starting a 5,000 patient trial in Saudi Arabia, and have announced a Phase III in Russia, with plans to go into Pakistan and Mexico as well. Now those populations could have plenty of coronavirus cases, unfortunately. But I remain skeptical that the Ad5 platform is going to come through – existing immunity varies in different populations around the world, but in general it looks like too many people already have antibodies to the vector itself. Update: (September 15) more on the role of China’s People’s Liberation Army in the development of this candidate.
Johnson & Johnson (Janssen): these folks have another obscure adenovirus (Ad26) in an attempt to avoid those problems. They’ve announced a Phase III trial starting this month that looks like the largest in the field (up to 60,000 patients) in locations all around the world. They’ve also been signing deals for reserving hundreds of millions of doses, should things work out, and continue to expand their manufacturing capacity. From what I can see, they’re the only ones at present who are not running a booster-shot trial, which could be an interested logistical advantage for them. Update: (September 15) looks like a trial has started in Spain as well. Considering how the coronavirus has come roaring back there, that’s good timing. But because of the AstraZeneca/Oxford problems, there are reports of people backing out.
Gamaleya Research Institute: this one is a two-shot dosing program, but one of them is an Ad5 vector and the other is an Ad26. This is the Russian vaccine that made so many headlines not long ago, and for the moment, I have no more to say than I said at the time: its “approval” in Russia was nothing more than a publicity stunt. Update: (September 10) the institute has published some data, but there are questions about their numbers in turn.
Reithera: this one is a gorilla adenovirus, for the same immune-evading reasons as the other unusual vectors above. The company has announced that the first human volunteer was dosed late in August in an Italian Phase I trial, and that they hope to go into Phase II/III by the end of the year.
Altimmune: this is another adenovirus vector, but this time administered intranasally (which can be a real logistic advantage over injections). The company also says that it expects the candidate to be storable at room temperature, another real advantage given what we’ve been hearing recently, but we’ll see how that holds up. They and their partners at Alabama-Birmingham have announced what look like solid results in mouse models, but they’re obviously behind the more highly publicized vaccine candidates when it comes to human dosing.
Merck/Themis: this is the first non-adenovirus vector candidate on this list – it uses the measles virus, actually. The company has said that it plans to start human trials in this quarter, which means this month, but we have very little data otherwise. Merck’s Roger Perlmutter has said, though, that with this one and the IAVI candidate (next paragraph!) that they’re trying for a one-dose vaccine rather than a booster shot regimen. Update: (September 9) note that in contrast to the other viral-vector candidates, this one and the Merck IAVI one below are both competent to replicate in the body. Update: (September 17) human dosing of this candidate has started in Belgium.
Merck/IAVI: the vector in this case is VSV, the virus used in the Ebola vaccine (which, as mentioned above, is the only one so far in humans that uses this viral-vector delivery idea). This one is said to be going into humans by the end of the year. The hope is that this one could be orally administered in a “swish-and-swallow” manner, which would certainly be an advantage. Update: (September 9) see note immediately above. Update: (September 30) Merck has begun their clinical trial on this one as well, in a single-dose protocol.
Vaxart: this company has also announced work on an oral coronavirus vaccine candidate, but information has been very hard to come by since then. I cannot find any data, nor any idea of when a trial might start. What I do find is lots of fans of the stock – not necessarily the company’s fault, because coronavirus therapy stocks have of course been an insane circus all year long, but it does make the signal/noise rather poor. We’ll see if some solid information appears eventually. Update: (September 17) the company has received clearance to start Phase I trials.
Washington University: a recent paper in Cell describes this chimpanzee adenovirus candidate. Interestingly, while an immune response was certainly induced in mice when the vaccine was given by injection, but an even more robust protective effect was seen when it was given intranasally. You have to wonder if the eventual coronavirus vaccine that people will take routinely will be a nasal one, but it’s for sure that whatever rolls out in the next few months will not be.
MediciNova: this Japanese company has an influenza-virus vector platform that they’re also using to try to develop an intranasal coronavirus vaccine. They’ve announced some mouse-model data; we’ll see if this one can make a place for itself in human trials.
Update: (September 17) Xiamen/HKU/Beijing Wantei: this is another influenza-vector intranasal vaccine candidate, which has just been approved for Phase I trials in China.
These take DNA or RNA coding for coronavirus proteins and inject that directly into the bloodstream (all of the ones below are mRNA, except for Inovio, Genexine, Zydus Cadila, and AnGes). “Directly” isn’t quite the right word, though – for these things to work, they have to be formulated and modified to survive destruction in the blood, to be taken up through cell membranes, and to be used for protein production once they’re inside. There have been extensive experiments in animal models over the years, but this is another category where no existing human vaccine uses the technology (yet!) Advantages include fast development and (possibly) ease of manufacture, depending on how exotic the final form turns out to be, and lack of an existing immune response to the vaccine itself (as seen with some of the viral vectors above). The big disadvantage is, well, once again no one has taken these things into humans yet. Another one is that it looks like the ones under development will need to be stored under rather stringent cold-chain conditions.
Moderna: well, you can’t open up a news site without reading something about Moderna. So I’ll highlight that the company has released some data on their vaccine candidate’s performance in older patients (looks good). The company says that it is on target to complete enrollment in its Phase III trial by the end of this month, and (in response to some worries) is now publishing data on the diversity of its patient population. Like all the front-runners, we are waiting on efficacy data, and everything we have to say now will be outweighed by what those numbers tell us. Update: (September 17) the company has released its entire clinical trial protocol, the first of the companies involved in this to do so).
Pfizer/BioNTech: this one is also on track to complete enrollment late this month, and the company continues to stick with its optimistic timelines. Their CEO said today that they might have results in October, but that means the very end of October, surely. Update: (September 17) BioNTech has announced that they’ve substantially increased their production capacity, buying a plant in Germany from Novartis.
Inovio: I have been searching, but I have no new science results to talk about with this DNA vaccine candidate. It’s a stock-market fight, and another short-selling firm just came out with a very pessimistic report. And sure, that’s what short-sellers do, but they’re not always wrong, either. I have never been optimistic about these folks, and I have seen no reason to change my opinion. Hammer me for it, fanboys. Update: (September 17) more details from the judge on the recent court case that went against Inovio. Update: (September 28) the FDA has put the company’s Phase II/III trial on hold, with issues around the delivery device and perhaps more.
Curevac: they have a Phase I trial underway in Germany and Belgium, and it’ll be interesting to compare their data to Moderna and Pfizer/BioNTech when it completes. The company is already negotiating with the EU about supplying its candidate if it works, and says that its timeline would put that in the middle of next year. Update: (September 17) more funding from the German government.
Imperial College: I believe that this is the leading self-amplifying mRNA candidate left, after Pfizer and BioNTech dropped theirs. They’ve shown good results in mice, and a Phase I trial is underway in England, with reports of trials starting in Uganda by the end of the year. An advantage of the self-amplifying platform is (as you’d figure) that it takes lower dosages, but it’s not something that (to my knowledge) has been tried in humans until now. Update: (September 17) a trial is beginning studying this candidate and the Oxford/AZ one as inhaled formulations.
Arcturus/Duke-NUS: here’s the other self-amplifying candidate that I’m aware of, which went into humans just a couple of weeks ago in Singapore. Arcturus has both an mRNA platform and a lipid-based delivery platform, which have come together under the current conditions to get into the clinic very quickly indeed. Update: (September 9) here’s a preprint with early data on this candidate.
Sanofi/Translate: some slides have emerged in a regulatory filing that show immune responses in animal models. No idea when this one is going into humans, although it’s supposed to be soon.
Zydus Cadila: this DNA candidate has been through phase I, and early in August received approval to go into Phase II trials. Update: (September 17) these are apparently underway. To the best of my knowledge, though, we have no data other than the company’s statement that it was well tolerated in Phase I. The CEO has said that they hope to launch the vaccine itself next March.
Genexine: this DNA vaccine candidate has now been into primate models, and is approved to start clinical trials in Korea – but I can’t find a firm date when that’s supposed to happen.
AMMS/Abogen/Walvax: this one (China’s lone mRNA candidate, from what I can see) has been approved for human trials, but other than that, details are very hard to come by. I don’t even know if it’s been dosed yet, or what the timeline is.
AnGes/Osaka/Takara: this is a Japanese collaboration on another DNA vaccine candidate, which has begun recruiting for human trials. This one combines a DNA plasmid with a skin-penetrating jet injection device, so it’ll be a more complicated development to get through.
Recombinant protein vaccines
Here we get to a technique that really is used for human vaccines. The previous two categories force your own cells to make viral antigen proteins, but here you’re making those proteins industrially and just injecting them directly. The advantage can be that such protein production can be accomplished in many different ways and is already done on a large scale. That said, every new protein is a new project, with its own idiosyncrasies. A disadvantage is that this technique sometimes does not produce enough of a robust immune response by itself (at reasonable doses of protein, anyway), and needs added “adjuvants” as part of the vaccine formulation. These are substances that increase immunologic reaction – through mechanisms that honestly have not always been so well understood over the years (more here) and you’ll see these in the entries below. Update: (September 17): for an overview of this area, see here.
Novavax: this company seems to be the leader in this area at the moment. They have just published results from their Phase I/II trials, and the antibody responses look good after the second dose of their adjuvant-containing vaccine. The results were, in fact, quite similar in the 5-microgram and 25-microgram dosage groups, which highlights the work that the adjuvant is doing. CD4+ T cells were also induced (Th1 phenotype, as in the other vaccines that have reported so far), but there were no data on CD8+. It looks like this one will be handled under standard refrigerator temperature shipping and storage. The safety profile looks good so far as well. So like the others in human trials, we’re going to have to wait for the key efficacy numbers to sort things out – in an interesting development, though, Novavax is the only vaccine developer (so far) to explicitly say that they would be willing to take the FDA up on Commissioner Hahn’s recent proposal to have companies file for approval before their Phase III trials are complete. Update: (September 17) the company has announced a production agreement with the Serum Institute in India to increase supplies.
Sanofi/GSK: coming up fast now is this Big Pharma competitor. The companies have announced today that they’ve started a Phase I/II trial at 11 sites across the US, with a Phase III to start in December. This one is using GSK’s adjuvant technology, which is already in human vaccines, and will also be shipped and stored at refrigerator temperatures (and not freezers or worse, like the mRNA candidates).
Update: (September 9) Finlay Institute (Cuba): the Cuban government has begun human trials with a protein subunit vaccine (Sobreana-01). Over six hundred people are being enrolled in a Phase I trial as of late August.
Clover Biopharmaceuticals: this company has already been in human trials for a while now, but there’s absolutely nothing new to report, from what I have seen. Not sure what’s going on here, and I hope to see some results soon.
Queensland/CFL/GSK: this one has been in human trials since back in July (no word yet). A preclinical animal study has been done in hamsters, though. It’s worth noting that this one uses a so-called “molecular clamp” technique to try to keep the viral protein in a more antigenic conformation, which is different from the other candidates in this area.Zydus Cadila: this Indian company is also in human trials.
Vaxine: this Australian company has partnered with Oxford Expression Technologies to try out recombinant Spike protein with Vaxine’s proprietary adjuvant. Phase II trials are supposed to begin this month.
Medigen: this is a Taiwan-based company with yet another mixture of Spike proteins and adjuvant from Dynavax. A human trial is set to begin this month.
Adimmune: another Taiwan effort with a Spike protein, and this one has already begun human dosing there.
UBI: still another Taiwan protein vaccine, but this one (from what I can see) is a mixture of several antigens. It has just received permission from Taiwan’s regulatory authorities to start human dosing.
Zhifei Biological Products: as far as I can see, there’s been nothing substantial about this one since the announcement of Phase II trials back in early July.
Stabilitech: no updates on this one, either, that I have been able to find.
Attenuated Virus Vaccines:
This is another well-precedented vaccination technique. It involves producing a weakened form of the actual infectious virus, one that is not capable of causing damage but can still set off the immune system. There are several ways to do this, and it’s a bit of an art form involving taking the virus through a huge number of replications in living cells as you select for variants that are less and less harmful. An advantage is that such vaccines can be quite effective at raising a response – ideally, the immune system reacts exactly as it would to the real pathogen, except you avoid all the getting-sick part. A disadvantage is that part about it being an art form: balancing the lack of harm with immunogenicity is not something that can always be achieved. Some viruses have a wider window for this sort of thing than others, and it’s not easy (or possible, really) to know if this is a feasible pathway up front. That may well be one reason why (at the moment) I know of only one company pursuing this route.
Meissa Vaccines: this small company announced back earlier in the summer that it had had a pre-IND meeting with the FDA about its attenuated coronavirus program, and was doing preclinical studies. If there has been more news since then, I have not been able to track it down.
Inactivated Virus Vaccines:
This is also a technique that’s been used in medical practice for many years, and it’s another inactivation step beyond the attenuated viruses. Heat or chemical agents are used to damage the virus to the point that it can no longer infect cells at all, but the plan is for there to be enough of the viral material left unaltered to still raise an immune response. Not the most high-tech approach, but it can definitely work. Many times, though, vaccines of this don’t provide enough of a response in a single shot, so you will very likely looking at a booster vaccine schedule. Interestingly, the Chinese groups seem to have this field to themselves; I’m not aware of any inactivated-virus vaccine for the pandemic that’s in serious development anywhere else.
SinoPharm/Wuhan Institute of Biological Products: this candidate was the subject of a recent publication on a Phase I human trial, and that (while a bit light on details), was notable for a rather mild side effect profile. We’ll wait on efficacy data. This one is current being tested in the UAE. Update: (September17) the UAE has now approved use of this vaccine before clinical trials are complete.
SinoPharm/Beijing Institute: this one has sometimes been hard to distinguish from the other SinoPharm candidate, and I have been unable to locate any new information on it.
Sinovac: no new scientific data on this one, from what I can find. They’re in Phase III trials in Brazil and Indonesia, but that hasn’t stopped the Chinese government from authorizing its use in high-risk groups of government employees. Both governments have signed agreements to purchase the vaccine if it continues to show good clinical data. Update: (September 17) the company says that thousands of its employees have been given the vaccine as well. And there are reports that Indonesia is trying to develop its own vaccine rather than be completely reliant on this one.
Institute of Medical Biology (China): nothing new on this one since it was said to be entering clinical trials in June.
Bharat Biotech: this candidate is currently in Phase I/II trials at several locations in India, and the company has also received permission to conduct a separate trial with intradermal dosing (which should, in theory, use less vaccine per dose). What we’re not seeing is any more of the craziness back in July that would have had the vaccine theoretically launched two weeks ago, which is a relief. Update: (September 17) the company has provided data on this candidate’s prophylactic activity in a primate model.
Update: (September 17) Valneva: this French company is going to start Phase I trials of an inactivated-coronavirus vaccine, along with adjuvant from Dynavax, in December.
Here’s yet another category, which can be thought of as a “stripped virus”. A VLP has most or all of the surface proteins of the real virus, but doesn’t have the genetic payload inside, and therefore cannot replicate. But the immune response that develops to the surface antigens is still available. This technique is already used for vaccines against HPV and Hepatitis B, so it’s proven that it can work well. You have several options for preparing such VLPs, mixing and matching material from the natural virus (or more than one natural virus) and recombinant proteins.
Mitsubishi-Tanabe/Medicago: this is a plant-based candidate, coming from tobacco leaves (see the July post linked in the first paragraph today for more). Some plant-derived proteins can have very different immunogenic profiles than the “same” protein produced by other methods (different glycosylation patterns, generally). They’re currently in Phase I, with plans (I typed “plants” the first time) to go on to Phase II/III in October. They’re using adjuvants from both GSK and from Dynavax to go along with their proteins.
Kentucky BioProcessing: this is another tobacco-leaf protein production company, from R.J. Reynolds. I’m not sure whether to put them here in the VLPs or up in the recombinant protein section, but I’ll leave them here next to Medicago for convenience. They have completely preclinical testing on their candidate and are awaiting FDA approval to go into human trials.
Oxford/SpyBioTech: here’s a new one, based on a protein engineering system known as “Spytag/Spycatcher“. Molecular biology aficionados will know this one already – it’s an efficient protein conjugation method, with these two domains engineered to form a new covalent bond to each other when you express them in your desired targets and let them mix. That’s been used to put Spike antigen proteins (the RBD part) onto a known VLP called mi3, a technique that was already being investigated before the coronavirus epidemic even showed up. Mice given the RBD itself raised a detectable but not very strong immune response, but when the RBD was conjugated to the mi3 particles, that construct set off a much more potent and wide-ranging antibody response. I don’t know if this can get into human trials in time to help, but it looks quite interesting.
Dang, that’s a lot of vaccine candidates. And as you can see, it’s a long-tail distribution – there are some big ones that everyone knows about, but a lot of people are bringing a lot of technologies to bear on the problem. This makes me think that we’re going to have a multichapter story, in the end. There will be the first vaccines approved, then the second wave, then the improvements on those, until we have (with luck, hard work, skill, and lots of money) tossed this virus out of the human population and back to the bats, pangolins, or whoever had it in the first place.
An excellent side effect is that vaccine technology will never be the same after this – it’s going to be like aircraft design before and after World War II, and for many of the same reasons. This whole pandemic has been awful, in many different ways, but we’re going to come out of it stronger and more capable than when we went in.