Facial Paralysis After COVID-19 Vaccines: What We Know

Pfizer-BioNTech and Moderna COVID-19 vaccines did not have a higher reported rate of facial paralysis than other viral vaccines, an analysis of pharmacovigilance data suggested.

The mRNA vaccines showed no higher safety signal for either narrowly or broadly defined facial paralysis in VigiBase, the World Health Organization pharmacovigilance database, reported Charles Khouri, PharmD, of University Grenoble Alpes in France, and co-authors, in JAMA Internal Medicine.

"While spontaneously reversible in the vast majority of cases, facial paralyses are serious and debilitating adverse drug reactions," Khouri told MedPage Today. "Should a link be established, this adverse reaction has the potential to hamper the vaccination campaign."

During the phase III Pfizer-BioNTech and Moderna trials, seven cases of facial paralysis or Bell's palsy were reported in the vaccine groups (7 of 35,654), and one case was seen in the placebo groups (1 of 35,611). A causal relationship was not established, but the FDA recommended that vaccine recipients be monitored.

The VigiBase findings are "welcome news as a slight imbalance was apparent in the phase III clinical trials of both mRNA vaccines, suggesting a potential safety signal," noted Gregory Poland, MD, of the Mayo Clinic in Rochester, Minnesota, who wasn't involved with the analysis.

"However, no clear signals have been detected in any of the U.S. vaccine safety surveillance systems as suggesting any elevated rate over any other vaccine or background," Poland told MedPage Today. "In fact, the only time facial paralysis has been correlated with vaccines has been in association with those that used E. coli heat-labile toxin as an adjuvant, e.g., the intranasal inactivated influenza vaccine."

Facial paralysis can occur in several settings, including viral infections, traumatic injury, cancer, or during pregnancy. Bell's palsy, the most common cause of facial paralysis, is a form of temporary paralysis or weakness on one side of the face that stems from dysfunction of cranial nerve VII (facial nerve). Symptoms appear suddenly over a few days and generally start to improve after a few weeks.

In their study, Khouri and colleagues compared the mRNA vaccines against two control groups -- all other viral vaccines and influenza vaccines only -- using two facial paralysis definitions, one narrow and one broad. The narrow definition was simply the term "facial paralysis" in the VigiBase records; the broad definition included facial nerve disorder, facial paralysis, facial paresis, facial spasm, oculofacial paralysis, and VIIth nerve injury.

Of 133,883 cases of adverse reactions with mRNA COVID-19 vaccines reported by March 9, the researchers identified 844 (0.6%) events related to facial paralysis. These included 683 cases of facial paralysis, 168 cases of facial paresis, 25 cases of facial spasms, and 13 cases of facial nerve disorders (some adverse events were co-reported). Overall, 749 cases were reported with Pfizer-BioNTech vaccines and 95 cases with Moderna. Most cases involved women (67.8%); median age was 49.

Median time to onset was 2 days. Most reported cases came from the U.S. (37%) or the U.K. (23.2%). More than half of cases (56%) were deemed serious. About a quarter of reports (23.9%) indicated the patient had not yet recovered, 19.8% said patients had recovered, 13.3% said they were recovering, and the status of most other patients was unknown.

The researchers also identified 5,734 (0.5%) cases of facial paralysis among 1,265,182 cases of adverse drug reactions reported with other viral vaccines, and 2,087 (0.7%) cases among 314,980 reports for influenza vaccines. They used disproportionality analyses to compare groups, noting that a signal would be significant if the lower boundary of the information component 95% credible interval (IC₀₂₅) was greater than 0.

For broad and narrow definitions of facial paralysis, Khouri and colleagues did not detect any signal of disproportionality for mRNA vaccines versus other viral vaccines (IC₀₂₅ = -0.01 and IC₀₂₅ = -0.06) or influenza vaccines alone (IC₀₂₅ = -1.36 and IC₀₂₅ = -0.32).

"It's important to note that VigiBase is a pharmacovigilance database," observed Al Ozonoff, PhD, of Harvard Medical School in Boston, who also wasn't involved with the study. "It's a very valuable resource but it does have some limitations."

VigiBase aggregates data from many different systems, Ozonoff pointed out: "For example, one-third of the reports in this study come from the U.S., and those U.S. reports come from VAERS [Vaccine Adverse Event Reporting System]. These typically are voluntary reports and there's no systematic case ascertainment, although typically some ascertainment is done. For these reasons, it's not possible to draw any generalizable or population-based estimates from these data," he told MedPage Today.

Pharmacovigilance databases represent a subset of cases and "they're almost always dramatically underreported in terms of the incidence of events," Ozonoff added. "I think we simply don't have enough data right now to come to a firm conclusion."

Why incidence of facial paralysis reports varied in the phase III trials is unclear.

"Given the large amount of screened and collected adverse events in clinical trials, a serendipitous difference found for a given adverse drug reaction does not prove a causative link," Khouri noted. "Purpose-designated, post-approval studies are needed to confirm or infirm these signals."

Comment