Making sense of the deluge of early COVID-19 vaccine data
A roundup of early COVID-19 vaccine data shows it’s not going to be smooth sailing
It’s been a big couple of weeks for COVID-19 vaccine data, and the findings are simultaneously encouraging and disheartening. While they hint that some level of protection is achievable -- by multiple candidates -- more than one red flag has arisen.
Low levels of neutralizing antibodies, pre-existing immunity to viral vectors, and a lack of the sterilizing immunity required for herd immunity will likely temper at least some of the enthusiasm.
The past three weeks have seen early human data from two of the leading COVID-19 vaccines, non-human primate studies on an additional three, including one prototype compound, and mouse data from a sixth.
Both of the clinical reports came this week. Moderna Inc. (NASDAQ:MRNA) and CanSino Biologics Inc. (HKEX:6185) have been neck and neck in the race to develop a vaccine. Both started Phase I testing of their respective vaccines in mid-March. CanSino edged ahead on April 12 to become the first COVID-19 vaccine developer to start Phase II testing.
Both companies reported that their vaccines induced neutralizing antibodies in healthy volunteers and that the results were dose-dependent.
However, Moderna did not disclose the neutralization titers for its vaccine mRNA-1273, or how they were measured, making it impossible to compare with results from CanSino’s candidate, Ad5-nCoV. Instead, Moderna said its neutralizing titers were similar to mean titers from convalescent plasma. But convalescent plasma is highly variable, and it is not yet clear how different titers relate to patients’ disease severity or levels of protection. FDA recommends a titer of 160 for convalescent plasma.
Moderna rose $13.31 (20%) to $80 on the data report on Monday, raised $1.3 billion in a follow-on financing that evening, and finished the week at $69, with a market cap of $25.6 billion.
In a Lancet study Friday, CanSino reported mean neutralizing titers of 34 in its high-dose group -- well below FDA’s recommended titer for convalescent plasma.
Another problem was that half of the subjects in the study had pre-existing antibodies against the vaccine’s Ad5 vector, which muted T cell responses to treatment.
CanSino rose HK$8 to HK$222 on Friday, a day most stocks on the Hong Kong exchange tumbled. Its market cap is HK$49.4 billion ($6.4 billion).
Moderna’s mRNA-1273 is an mRNA vaccine and doesn’t use a viral vector.
University of Oxford’s clinical vaccine ChAdOx1 nCoV-19, which is now in the hands of AstraZeneca plc (LSE:AZN; NYSE:AZN) as AZD1222, is also delivered in an adenoviral vector. Phase I data from AZD1222 have not yet been reported.
At least 16 preclinical COVID-19 vaccines are also delivered in viral vectors.
Moderna has amended the doses in its upcoming Phase II trial based on the Phase I findings; it expects to begin a Phase III trial in July.
CanSino has not disclosed a timeline for Phase II data or start of Phase III.
Three preclinical papers published in the past two weeks also provide insight into the performance of vaccines now in the clinic.
Though protocol differences make the results hard to compare directly, the studies provide clues about the extent of the protection the vaccines may provide.
The first results came in a May 6 Science study from Sinovac Biotech Ltd. (NASDAQ:SVA) demonstrating its inactivated SARS-Cov-2 vaccine PiCoVacc induced neutralizing antibodies in non-human primates and protected the animals against infection with SARS-CoV-2. Animals were dosed intramuscularly.
The high dose of the vaccine led to undetectable viral loads in the lungs, throat or anus at seven days post challenge. None of the vaccinated monkeys developed pneumonia, whereas all the control monkeys did.
However, Sinovac delivered three doses of PiCoVacc to the animals, while its Phase I/II trial in humans is administering two doses, raising the question of whether it will be as effective in the trial. Also unclear is how the high dose in the Science study (6 µg) relates to the high dose in the clinic (1,200 SU/0.5 mL), which makes interpretation difficult.
A May 13 bioRxiv preprint with data from a different clinical COVID-19 vaccine raised yet another question about the Sinovac data.
The May 13 study, which was conducted by the University of Oxford and NIH’s National Institute of Allergy and Infectious Diseases (NIAID) on Oxford’s AZD1222, showed that non-human primates can be protected from pneumonia but still show other clinical symptoms -- symptoms that Sinovac didn’t measure.
On the face of it, Sinovac’s neutralizing antibody titers looked better than Oxford’s.
Both vaccines induced neutralizing antibodies in all vaccinated animals, but Oxford’s titers ranged from 5 to 40, while Sinovac’s high dose yielded a mean titer level of 50. Oxford’s was dosed once, intramuscularly.
And if Sinovac had dosed its monkeys twice instead of three times, its titers may have been lower. Conversely, one of the arms in Oxford’s Phase I/II trial is testing two doses of its vaccine, which could see titers rise.
The Oxford finding that drew the most attention was the fact that the neutralizing antibodies induced by its vaccine didn’t protect the animals from infection.
After viral challenge, the vaccinated monkeys showed similar viral genomic loads in nasal swabs as control animals. Viral subgenomic RNA, a sign of virus replication, was also detected in nasal swabs from vaccinated monkeys (see “Sinovac Vs. Oxford”).
Table: Sinovac vs. Oxford
The findings suggested the vaccinated monkeys would still be capable of infecting others, casting doubt on whether mass vaccination with the Oxford vaccine would lead to the herd immunity some see as essential for defeating COVID-19.
The good news was that the Oxford vaccine reduced viral load in lung fluid and decreased symptoms of clinical disease, suggesting it may make the disease less deadly to humans (see “Oxford Vaccine Could Reduce COVID-19 Symptoms, Won’t Provide Herd Immunity”).
Like Sinovac’s data, none of the vaccinated Oxford monkeys developed pneumonia; however, Oxford measured additional clinical symptoms and found three of its six vaccinated monkeys developed an elevated respiratory rate. Overall clinical scores rose in the vaccinated animals, although not as high or for as long as the controls. Sinovac did not report respiratory rate or clinical scores.
Inovio Pharmaceuticals Inc. (NASDAQ:INO) added mouse data to the mix in a Nature Communications paper on May 20. Sinovac and Oxford also reported neutralizing antibody titers from mice enabling a rough comparison, which suggests Inovio’s INO-4800 vaccine is at least on par with other two, if not more potent than Oxford’s (see “A Tale of Three Mice”).
INO-4800 is in Phase I testing. Inovio expects to have data from challenge studies in multiple animal models in the coming weeks, and preliminary Phase I data next month. It plans to start a Phase II/III trial in July or August.
Table: A tale of three mice
Homing in on best version of the spike
A Science paper published May 20 addressed one of the biggest questions in COVID-19 vaccine development: which form of the spike protein makes the most potent antigen.
Led by Harvard’s Beth Israel Deaconess Medical Center, the research team compared the ability of five DNA vaccines encoding variants of the spike to induce neutralizing antibodies in non-human primates and protect the animals from viral challenge.
Figure: Antigen match-up
The full-length spike performed best at both, beating out smaller spike peptides, including the receptor binding domain, which is the subject of several vaccine programs.
The full-length spike also produced higher neutralizing antibody titers than convalescent sera from monkeys and humans (see “Antigen Match-Up”).
None of the vaccine variants produced fully sterilizing immunity, suggesting the lack of sterilizing immunity in Oxford’s study may be a problem broader than for just that candidate.
Beth Israel Deaconess is collaborating with Johnson & Johnson (NYSE:JNJ) on a preclinical AAV-based COVID-19 vaccine. The pharma told BioCentury the vaccines in the paper do not read through to its lead candidate, for which it plans to report preclinical data “in the very near future.”
The pharma expects to start clinical trials by September.