Thanks @MelbME @DMissa and @jnmaciuch for your discussion here online.
Science at work in realtime, I love it.
Trial Report Plasma cell targeting with the anti-CD38 antibody daratumumab in ME/CFS -a clinical pilot study, 2025, Fluge et al
Science at work in realtime, I love it.
hotblack
Senior Member (Voting Rights)
They used SF-36 Physical Function (SF-36 PF) and DePaul Questionnaire-Short Form (DSQ-SF). Is that what you need?
Edit: was just about to post a better link but @Utsikt got there with the relevant text.
Last edited:
OrganicChilli
Senior Member (Voting Rights)
Patients also recorded their perceived Function level in per cent according to a table with examples (Supplementary protocol).
This one—don’t see it in the supp.
This one—don’t see it in the supp.
hotblack
Senior Member (Voting Rights)
Ah, yes, I couldn’t find that either. It says there’s an example in the supplementary materials but I can’t see it, only the results (in Table 1.xslx)
jnmaciuch
Senior Member (Voting Rights)
Continuing a discussion from another thread, since this one seems like the most appropriate place to discuss actual figures.
This was Sasha's original comment:
The figure in question:
And this was rync97's reply:
Actually I think what Sasha is saying about baseline severity still holds. The distribution of severe patients in the responder groups is either random happenstance (highly possible due to only n=2), or it tells us something about how underlying disease severity, reflected in both severity label and questionnaire scores, predicts responsiveness.
This was Sasha's original comment:
The figure in question:
And this was rync97's reply:
Actually I think what Sasha is saying about baseline severity still holds. The distribution of severe patients in the responder groups is either random happenstance (highly possible due to only n=2), or it tells us something about how underlying disease severity, reflected in both severity label and questionnaire scores, predicts responsiveness.
jnmaciuch
Senior Member (Voting Rights)
Also, the fact that NK levels are not significantly different throughout the entire course of treatment is an interesting point. If we choose to assume that these preliminary results show real improvement in the responder group, then this pretty much means we can discount any theories about efficacy of dara being mediated by simply the number of NK cells available to kill plasma cells. The only way it could make sense is if all the relevant cell killing happens before the 2w timepoint, which seems less likely given that improvements really only started at the 8w or 20w mark depending on if you're looking at SF-36 scores or step count. Not impossible if you have legitimate ways to explain a substantially delayed effect, but still much less likely.
The difference in baseline NK cells could still end up being a real predictor of responsiveness, but a theory relying on sheer number of NK cells available to kill plasma cells ends up being quite weak given this data.
OrganicChilli
Senior Member (Voting Rights)
What could be a possible explanation?
Could NK cell cytotoxicity be important?
Could you try to explain your point in a different way—not fully understanding. I do think that most of the “killing” occurs in the first few weeks.
jnmaciuch
Senior Member (Voting Rights)
I guess one possibility is if autoantibodies were involved in a feedback loop with something else, so it would take a few weeks for the other part of that loop to ramp down on its own. I don't really have solid ideas on what that other part of the loop could be, though, and I've already said in other threads why I'm not too sold on autoantibodies as an explanation for ME/CFS to begin with.
Possibly. In a world where we knew for certain that dara was effective and that it was the killing of plasma cells by NKs that mediated its effectiveness, the most logical explanation would be something about NK cell functionality that happens to also correlate with NK cell count at baseline. Judging from the papers linked earlier in this thread, even in the case of multiple myeloma this seems like a more robust explanation.
It's the fact that by week 2, there's no significant difference in NK cell level, and it stays that way throughout treatment course.
If better response is mediated solely by a difference in the number of NK cells [edit: available for cell killing], then the only period of time where that would have been significantly different is within the first week or so. Which would mean a very fast turnaround time to clear out all the bad plasma cells. Maybe that's possible, maybe that's what is happening. My point is just that a "pure numbers" theory overall seems like a weaker explanation given that it has to explain some counterintuitive trends in this data.
Jonathan Edwards
Senior Member (Voting Rights)
I think it is reasonably plausible that all the relevant killing occurs within hours*. The plasma cells that are susceptible may die off leaving resistant cells. If NK numbers fall then right at the beginning would be the best time.
If relevant plasma cells are killed then you would expect their antibodies to decline over a period of 1-3 months. Antibody half-life is probably very variable but 30 days may be a ballpark**. If the antibody is feeding in to a signalling pathway a bit like our Qeios suggestion then dying down of T cell populations might take a couple of weeks maybe. Subjects then need to have confidence that they really can do more so they may not appreciate their freedom from PEM for a bit.
I am not necessarily arguing for this but I don't see the delay as a big problem.
* B cell killing with rituximab is probably over in hours - there are no B cells the next day if I remember rightly.
** Functional life of antibody populations is a bit of a mystery.Most monoclonals are supposed to last about ten days but the suppression of B cell genesis by rituximab usually lasts 6 months and in some cases 2 years.
If relevant plasma cells are killed then you would expect their antibodies to decline over a period of 1-3 months. Antibody half-life is probably very variable but 30 days may be a ballpark**. If the antibody is feeding in to a signalling pathway a bit like our Qeios suggestion then dying down of T cell populations might take a couple of weeks maybe. Subjects then need to have confidence that they really can do more so they may not appreciate their freedom from PEM for a bit.
I am not necessarily arguing for this but I don't see the delay as a big problem.
* B cell killing with rituximab is probably over in hours - there are no B cells the next day if I remember rightly.
** Functional life of antibody populations is a bit of a mystery.Most monoclonals are supposed to last about ten days but the suppression of B cell genesis by rituximab usually lasts 6 months and in some cases 2 years.
Kitty
Senior Member (Voting Rights)
But with a cohort this small, it introduces a seed of doubt if two conditions need to apply.
Unless there's a known link between the functionality of this type of cell and the numbers of them, e.g. more get killed off if they're tagged as poor quality?
Jonathan Edwards
Senior Member (Voting Rights)
Not sure about that. I share the scepticism about the NK numbers simply being an index of plasma cell killing efficacy in this context.
Utsikt
Senior Member (Voting Rights)
Have anyone tried to look at the relationship between NK cells and severity?
If the NK cell count doesn’t appear to be correlated with severity, and NK cell count for some reason is a proxy for whatever determines the response to Dara, then the severity of any given patient shouldn’t matter.
If the NK cell count doesn’t appear to be correlated with severity, and NK cell count for some reason is a proxy for whatever determines the response to Dara, then the severity of any given patient shouldn’t matter.
ryanc97
Senior Member (Voting Rights)
Any correlation between severity and response is confounded by the correlation of NK cells and severity.
Last edited:
ryanc97
Senior Member (Voting Rights)
My take is the impact of the doses exponentially decays. The delay would be because of the existing supply of AABs being cleared out. So after the first dose the plasma cells are wiped out and the production line stops, and it takes some time for the body to consume the remaining AAB supply. Once the body 'consumes' the supply, the improvements come.
And it just so happens in ME, assuming the rate of production/consumption of AABs are similar, once production is cut off, consumption of the existing supply takes 6 weeks.
I just realized if you could find similar response curves in Lupus with Dara, that would validate it. But problem in Lupus is the Dara study they did, they didnt measure the response in terms of symptoms, but rather the actual Lupus AAB.
Last edited:
Kitty
Senior Member (Voting Rights)
If you mean in this study (rather than generally) there might not be much.
Although there are marked differences in average SF36 scores between responders and non-responders, there's no significant difference in step count and DSQ-SF. It's possible you're right, and severity hasn't much to do with it. The only thing that stands out between responders/non-responders is the difference in NK counts.
[Edited to add a couple of words for clarity]
jnmaciuch
Senior Member (Voting Rights)
That may be the explanation, though I'd expect dara to be considerably slower than rituximab considering the expression of CD38 was usually at least an order of magnitude lower than CD20 in my old data sets. I think it's really only in cancerous conditions where it may reach the same levels as CD20.
Maybe a few days of marginally higher NK cells is enough to overcome this obstacle--I haven't seen any relevant data to know one way or the other. Though I've been wondering why F&M didn't also measure plasma cells if they were already doing flow for NK cells.