Why would nasally-transferred coronavirus only affect the left side of the brain?

WE GET QUESTIONS!

Q – “I survived a mild case of COVID. Should I be worried about the volume of gray matter in olfactory-related structures in the left hemisphere of my brain?”

A – Most of what you've read on social media may be overstated.

One of the scariest things about SARS-CoV-2 (other than possible death) is that it affects multiple organs, including the brain. The vast majority of studies have compared measures in COVID survivors to those obtained from participants without COVID. These cross-sectional studies cannot determine whether pre-existing differences can account for disease-related 'changes'.

An important new preprint by Douaud and colleagues reported results from a longitudinal study that obtained MRIs from participants before and during the pandemic. Carefully matched cohorts of COVID+ (n=394) and COVID- (n=388) people were given a second scan 3 years after their initial entry into the study. The majority of patients were not hospitalized. The authors hypothesized that brain areas related to smell and taste, senses which show notable decrements in infected patients, would be altered in the COVID-19 population.

 

from Vaughan & Jackson (2014). The “piriform axis” is an unusual oblique angle that shows primary olfactory cortex (Pir, piriform cortex) in the orbitofrontal region of the frontal lobe and nearby medial temporal lobe structures related to emotion and memory (Am, amygdala and Hip, hippocampus).

The frightening entrée of the virus into the brain may be through the nose. In mice, the S1 subunit of the coronavirus spike protein crosses the protective blood brain barrier and ends up in the olfactory bulb within 10-30 min (Rhea et al., 2021). After that, we don't really know what happens. So it's a stretch to suggest that neurotropism (viral infection of brain cells) causes alterations at a macro level in humans that can be detected by structural MRI. Anyway, that's the hypothesis.

What was unclear to me, however, was the number of other analyses conducted as part of the study. Other than the smell/taste regions of interest (ROIs), there were 2,360 distinct measures of brain structure or function (including resting-state fMRI and task-related fMRI, which we assume were not significantly different in the patients).

Sticking to the ROIs, the authors ran permutation tests that corrected for multiple comparisons and found that only the left hemisphere was affected. WHY?? Despite the Proustian speculation...

“...where the left hemisphere seem [sic] to be more involved in the emotional aspect of olfactory memory (the famous madeleine de Proust association that seemed particularly targeted in COVID-19).”

...it's mysterious why the virus would have deleterious effects on the brain by only invading the left nostril.

 

Fig. 1. (Douaud et al., 2021). The three main regions showing significant loss of grey matter (thickness, volume) between the two time-points specifically for the COVID patients are the parahippocampal gyrus, the lateral orbitofrontal cortex, and the superior insula. All results were localised to the left hemisphere.

The other notable aspect of the results was the massive overlap in distribution between the COVID+ and COVID- participants (see snarky commentary in the first figure). The authors helpfully showed the full spread of values for the time 1 vs. time 2 difference (clustered around zero). But did gray matter volume and thickness actually increase in some people? Although the group differences were statistically significant, were there any functional consequences? Severity of smell/taste loss? Effects on memory or emotion? We don't know.

COVID-19 can have persistent, disabling effects in some people, including young and previously healthy individuals (the “long haulers”, see Davis et al., 2021). There is no doubt about the reality of Long Covid.

On the other hand, alarmist coverage of preliminary neuroimaging findings is not helpful. The differences in Fig. 1 of Douaud et al. do not depict a shrinkage of 4 standard deviations, despite what some widely circulated tweets may claim. As one of the authors explains:

Yes it's explicit in the Methods that this is an r2z (this Z is needed for the permutation testing) that takes into account N - so think of it about being more like SE than SD. Some people have wrongly thought of it as a 4SD effect in the data, which it isn't.

— Stephen Smith (@fmrib_steve) June 20, 2021

 

In fact, the exploratory analysis showed the largest loss of brain volume was non-specific (and not discussed). And perhaps not due to neurotropic invasion?

 

Longitudinal studies are extremely valuable, and the authors are to be commended for this. We'll wait for future papers to verify these findings, but for now I'm not totally convinced.

References

Davis HE, Assaf GS, McCorkell L, Wei H, Low RJ, Re'em Y, Redfield S, Austin JP, Akrami A. (2021). Characterizing long COVID in an international cohort: 7 months of symptoms and their impact. EClinicalMedicine. 2021 Jul 15. Epub ahead of print. PMID: 34308300.

 

Douaud G, Lee S, Alfaro-Almagro F, Arthofer C, Wang C, Lange F, Andersson JL, Griffanti L, Duff E, Jbabdi S, Taschler B. (preprint). Brain imaging before and after COVID-19 in UK Biobank. medRxiv. June 20, 2021.

Rhea EM, Logsdon AF, Hansen KM, Williams LM, Reed MJ, Baumann KK, Holden SJ, Raber J, Banks WA, Erickson MA. (2021). The S1 protein of SARS-CoV-2 crosses the blood–brain barrier in mice. Nature Neuroscience 24(3):368-78.