HK data validate cerebrovascular safety of BNT162b2, CoronaVac vaccines

A Hong Kong study finds no association between the BNT162b2 and CoronaVac vaccines and subclinical progression of MRI cerebrovascular disease biomarkers in community-dwelling adults.

“In the primary analysis, compared with the unvaccinated subgroup, BNT162b2 and CoronaVac were not significantly associated with the composite primary outcome,” said Assistant Professor Yiu Ming Bonaventure Ip from The Chinese University of Hong Kong (CUHK) at ISC 2024.

On multivariable logistic regression for the composite primary outcome* (adjusted for age, sex, COVID-19 infection history, and baseline radiologic cerebrovascular findings), the adjusted odds ratio (aOR) for BNT162b2 vs unvaccinated was 0.77 (p=0.472). For the comparison between CoronaVac and unvaccinated, the aOR was 0.51 (p=0.071). [Ip YMB, et al, ISC 2024, abstract LB24]

However, dyslipidaemia (aOR, 1.81; p=0.041) and increased white cell count (aOR, 1.34; p=0.005) had a significant association with the composite primary endpoint. “[These imply that] metabolic risk factors were stronger risk factors for cerebrovascular disease progression than COVID-19 vaccines,” noted Ip.

Evidence has shown an association between acute cerebrovascular events and severe COVID-19 infection, noted Ip. “Native-like mimicry of SARS-CoV-2 spike protein by vaccine may trigger fibrotic remodelling, vasoconstriction, inflammation, and hypercoagulability through interaction with ACE2 receptor, ANG II T2 receptor, platelet factor 4, etc.”

“Whether the production of spike protein through vaccination brings about similar risks of cerebrovascular events need to be further studied,” Ip continued. As such, they sought to determine the risk of stroke and progression of radiologic cerebrovascular disease markers in BNT162b2 and CoronaVac recipients in Hong Kong.

A total of 415 participants were recruited from the CUHK Brain Health Longitudinal Study. Of these, 342 were vaccinated (n=190 [BNT162b2] and 152 [CoronaVac]), while 73 were unvaccinated. They should be free from neurologic disease and must have baseline MRI assessment and blood test prior to the pandemic. About two-thirds of participants were about 63 years.

Overall, 26 percent developed the primary outcome, 14 percent had COVID-19 infection on serological exam, and none had severe COVID-19 infection.

Secondary endpoints

Compared with the unvaccinated group, vaccination was not associated with white matter hyperintensity (WMH) progression after adjusting for age, sex, infection history, and baseline WMH ratio (aOR, 0.58;p=0.258 [BNT162b2] and aOR, 0.44; p=0.11 [CoronaVac]) and cerebral microbleed (CMB) after adjusting for age, sex, infection history, and baseline CMB (aOR, 0.69;p=0.532 and aOR, 1.19; p=0.761, respectively).

Other components of the primary outcome were excluded owing to the low event rates.

sVNT

Of note was the inverse association between sVNT** and risk of primary composite outcome in the BNT162b2 subgroup (aOR, 0.96; p=0.001). Although the reason was unclear, they postulated that the poor neutralizing antibody (NAb) response may delay antibody-mediated clearance of spike protein, leading to cerebrovascular insult, Ip said. However, this should be interpreted with caution as there was no spike protein level measured nor was there any clinical event that occurred.

“The risk of cerebrovascular disease progression in BNT162b2 recipients with low sVNT may inform vaccination strategies in individuals at risk of poor response to mRNA vaccines,” Ip said. “This potential signal requires further study, especially among immunocompromised individuals who might mount a poor NAb response to mRNA-based vaccines.”

Implications

Taken together, the findings provide important long-term safety data of COVID-19 vaccines on cerebrovascular health. “Cardiovascular risk factor control is still more important than short-term predisposition by COVID-19 or COVID-19 vaccines,” said Ip.

“Whether COVID-19 vaccine protected participants against cerebrovascular damage mediated by natural COVID-19 infection requires further study,” he continued.

Apart from the observational study design, other limitations are the small population of unvaccinated individuals, as well as healthy vaccinee bias. Also, as the BNT162b2 group included recipients of any BNT162b2 dose, the cumulative effects of homologous BNT162b2 vaccines on the primary outcome may have been diluted by heterologous BNT162b2 vaccines, added Ip.

“Longer-term monitoring for clinical events is needed to substantiate our study findings on radiologic cerebrovascular events,” he said.