After all the attention paid to Moderna and Pfizer’s new Covid-19 vaccines, some health experts have expressed caution about the companies’ confident assertions that the vaccines are 95 percent effective for all populations.
This week, researchers at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) showed that the vaccines’ effectiveness may vary depending on a person’s race, and said that the vaccines should be tested robustly across populations with diverse genetic backgrounds.
Using advanced machine learning AI methods that examined a form of vaccine similar to Moderna and Pfizer’s, the team found that the number of people whose cellular immune system is not predicted to robustly respond to the vaccine ranged from less than half of one percent of white participants to nearly 10 percent of Asian participants. (Participants self-reported their race.)
“There are obviously many other factors to consider, but our preliminary results suggest that, on average, people of Black or Asian ancestry could have a slightly increased risk of vaccine ineffectiveness,” says MIT professor David Gifford, senior author of a new paper outlining the findings. “Our work shows that clinical trials need to carefully consider ancestry in their study designs to ensure that efficacy is measured across an appropriate population.”
Gifford and PhD students Ge Liu and Brandon Carter also presented a promising new machine learning-based approach for improving the vaccines’ effectiveness with certain populations: adding a small number of additional Covid-19 peptides to a given dose of the vaccine.
The team’s “augmentation” vaccines use vaccine components that have been observed to cause cellular immune system responses in COVID-19 patients. As shown below, by adding somewhere between 5 and 20 additional peptides to a particular dose, the team says that preliminary tests show that they improved the vaccine’s effectiveness to nearly 100 percent in all populations.
The researchers deemed a vaccine to be most effective if at least six COVID-19 vaccine peptides were shown to be displayed by the proteins that regulate the person’s immune system (known as HLA alleles).
The project builds off the team’s Opti-Vax system, which aims to design new vaccines and augment existing vaccine designs so that they better cover a more diverse range of populations.
As next steps, the researchers are collaborating with other researchers to test their vaccine designs in animal models. They seek to understand if the pieces of the virus they propose to cause immunity will offer better protection than the vaccines currently in clinical trials.
This work was supported in part by Schmidt Futures, Google Cloud, and the National Institutes of Health.