RE: Latest research IFN23 Apr 2021 16:16
It is clear that there is a confusion here.
Autoantibodies against ifns exist in some people whether they had this virus or not. They exists in some conditions and never in otherwise healthy individuals. It is a type of autoimmune disorder. See examples below.
https://pubmed.ncbi.nlm.nih.gov/16784312/
https://pubmed.ncbi.nlm.nih.gov/27059145/
What they found with covid, is that about 10% of hospitalised patients with severe disease had autoantibodies against ifn alpha and omega while no one asymptomatic had them. Later, they found some against lamda as well but it seems not so much in ifnb (only very few). We see something similar even in people that never had covid, some have autoantibodies against ifns but again ifnb is one of the least common.
Going back to the covid patients, what this suggests is that if ifn a function hadn't been impaired, then they would not be as likely to have severe covid (since none of the asymptomatic had them). Whether these existed from before is still under investigation (it would have to be a study similar to vaccines where they follow them from before they get infected). Assuming that it is definitely the virus that causes this autoimmune condition, one could in fact argue that the virus does not induce against ifn b because it cares less about it. But given that immunity is probably the least straightforward part of human biology and that in other conditions there is also less preference for auto antibodies against ifnb, it is perhaps because it is simply not as easy to induce autoantibodies against ifnb.
But, IN ANY CASE, when in almost all cases when ifn a is inhibited the patient do worse, and they haven't found asymptomatic cases where autoantibodies against ifn a exist, then we can conclude that ifn a is beneficial when it is there. If can't make this argument for ifn a, then you can't make it for ifn b either based on this imformation.
If that is clear, then it should be clear why ifnb should have an advantage specifically in patients with autoantibodies against other ifns as well as why there is a higher chance it would best work there (because they are already missing at least one other and by the way, alpha and beta share the same receptor).