Azelastine is generic now, and this was an academic study. Little interest by companies to test off-patent drugs. It was done in Germany (zero interest by US to do COVID-19 research now) Trial was n=450 split into azelastine and placebo groups. That's large by academic standards

Subjects took azelastine 3x/day for 8 weeks. SARS2 antigen testing was done 2x weekly by study personnel, so not self-reported which is good. Those with symptoms but SARS2 negativity were tested for other viruses. Solid study. SARS2 rate 2.2% treatment v 6.7% placebo, p = 0.02

So the study was large but the number of cases was small. P value was 0.02. Actually the study isn't too small, but sized just right for a predicted effect size of 40%. What was measured here was 67%. The power analysis used to select n=450 is very elegant; this is how it's done

There's concern that cases in dropouts (11 Az, 13 placebo) could have changed the result. There were 24 dropouts and case rate in untreated non-dropouts was 1 in 15, so we'd expect 1 case in the dropouts (2 if there was no benefit). Possible but unlikely to get 2 Az, 0 placebo:

And *even if* we saw unexpectedly 2 COVID19 cases in the 11 Az dropouts and 0 COVID19 cases in the 13 placebo dropouts, then the rate of infection would become 7/227 (3.1%) in Az and 15/223 (6.7%) placebo, or a 54% reduction. Per Fisher's exact test, the p value would be 0.125.

That's higher than the arbitrarily set criteria of 0.05 for reporting something is "significant", but that's not the same as saying there's no evidence of effect, let alone proof of no effect. Rather it would mean there's 87.5% certainty that a 54% reduction is not from chance.

Incidentally as mentioned there was a lower incidence of rhinoirus infections as well (1.8% vs 6.3%), suggesting Az works through some common immunomodulatory effect. Since it's better than placebo then it's not a simple effect of irrigation.

So overall I'd say the study appears to be persuasive of a protective effect of azelastine. Either it reached our arbitrary conventionally agreed-upon formal threshold of 95% certainty already, or it just missed that in a more far-fetched "they just got lucky" scenario.

But as with everything, nothing is 100% sure. So another even larger study would be nice to get certainty to 95% even with worst-case dropout scenarios factored in. And small correction; study was in JAMA Internal Medicine, not JAMA itself https://jamanetwork.com/journa...

This is actually the 3rd RCT of azelastine for COVID19. This one is for prevention, but earlier ones involved patients first testing PCR+ for SARS2 and then randomized to Az or placebo. Az reduced viral load 1 week later (p values < 0.01) suggesting it can work post-infection.

Here is the first of those trials from 2023 https://www.nature.com/article...

And here's the second one from 2024. In these post-infection studies, the drop in titers isn't huge and there's not faster symptom resolution. This 2024 study is 5x(!) daily; the other studies are 3x daily. https://www.mdpi.com/1999-4915...

The authors suggest the mechanism of action is direct antiviral based on in vitro testing with micromolar concentrations (inhibiting entry or the protease, the target we work on) but such testing is prone to false effects. In reality drugs rapidly drop below micromolar in vivo.

Most importantly, direct effects on SARSCoV2 wouldn't explain the reduction in rhinovirus infections. As azelastine is a potent inhibitor of H1 receptor on mast cells, my guess is inhibition of mast cells leads to better antiviral clearance by neutrophils and macrophages.

As the data suggest azelastine has some effect within a few days, it may work well for post-exposure prophylaxis, e.g. after being near someone with COVID-19 or the common cold. Then people wouldn't need to take it 3x/day for weeks. That would be a good next trial to do.

In contrast, the benefit or harm or nasal steroids is unclear. Early studies on immune-naive patients showed less severe disease in people taking nasal steroids, consistent with severe pulmonary disease also being immunological and ameliorated with dexamethasone. However...

just in July there was a study of allergy patients taking or not taking nasal steroids. This was a prospective case-control study (not blinded, not placebo-controlled) with study personnel testing for SARS2 prospectively and biweekly (that's nice) https://www.jacionline.org/art...

Results were that nasal steroid use was associated with 88% higher infection risk (p = 0.01) in adults but not children. Hard to rule out confounding factors, e.g. those with worse allergies might be more susceptible to begin with, but regardless nasal steroids don't seem to help

BTW azelastine OTC is 206 mcg/puff compared to 140 mcg in the study. The package says you can do 2 puffs (412mcg) each time and go 24h in between doses for allergies, but to match the study's daily dose, you'd want 1 puff each 12h

@michaelzlin What is the rationale? is there any mechanism of action of azelastine in preventing the SARS-Cov2 infection?

@mino_cappmont It was observed to be associated with reduced viral titers in databases but there's no known molecular rationale

@michaelzlin So is this used daily? Only after expected exposure?

@suetheob Those are the big unknowns

@michaelzlin Is this a good substitute for common antihistamines like Zyrtec for seasonal allergies, in addition to the benefits of preventing Covid and common cold? Thanks. (You know, killing two birds with one stone😁)

@michele_hiatus It's similar to Zyrtec in mechanism of action so should substitute fine, but it's shorter acting and apparently has a taste.

@michaelzlin https://x.com/skominers/status...

@michaelzlin 🙏 for sharing your analysis of this study, Dr. Lin.

@michaelzlin Plenty of nurses hawking various nose sprays on TikTok but never mention masking as prevention.

@michaelzlin This is the 4th azelastine COVID-19 study. Many low-cost and widely available treatments applied directly to the respiratory tract reduce COVID-19 risk. Effective treatments were known very early and could have curtailed the pandemic.

@michaelzlin So I guess the question is - do you need to use it that frequently? Or just before a high risk event.

@michaelzlin Important to note that 99.1% of the trial participants were vaccinated, so this study can’t say anything about replacing the need for vaccines.

@michaelzlin Suppressing your mucosal immunity prevents infection?