A proportion of people who are diagnosed with RA have no autoantibodies. In general they do not respond to rituximab, which makes sense. We have always known that it is not always possible to distinguish RA from other sorts of inflammatory arthritis like psoriatic arthritis, especially when there is no psoriasis to see so it seems likely that a proportion of 'RA non-responders' do not have a B cell related disease. Then there are people whose joints are so badly damaged that they do not get much benefit from any drugs. Of the remainder with autoantibodies most respond well but there also seems to be a small group with a lot of inflammation whose disease responds very slowly and is hard to get fully under control. This is not actually that surprising. Remember that rituximab does not kill the cells that actually make the autoantibodies (i.e. plasma cells), it kills younger B cells that might one day become plasma cells, so there is a lag in the effect. If the plasma cells making autoantibodies live long enough, and we know that some can live for up to ten years, then repeated courses of rituximab for several years may not reduce antibody levels enough to control inflammation. We tend to see rheumatoid factor levels dropping to about a half the original level in responders, which often seems to be enough, but clearly it might not always be. Even for B cells we have reason to think that some can hide away in lymph nodes and survive rituximab. So it is not that difficult to explain non-responses. What is needed is a something that actually kills all autoantibody producing plasma cells. The problem with that is that it needs to be specific for autoantibody producing plasma cells rather than useful plasma cells and we do not know how to do that.