Trial Report A pilot study of disulfiram for individuals with persistent symptoms despite prior antibiotic treatment for Lyme disease, 2025, Kuvaldina et al

A pilot study of disulfiram for individuals with persistent symptoms despite prior antibiotic treatment for Lyme disease

Mara Kuvaldina, Jessica Preston, Denise McClellan, Martina Pavlicova, Thomas H. Brannagan, Brian A. Fallon

[Line breaks added]

Introduction
In vitro studies report that disulfiram is effective in killing Borrelia burgdorferi. Case series suggest disulfiram may help to reduce the symptoms of patients with persistent symptoms despite prior antibiotic treatment for Lyme disease. This pilot study assessed safety, tolerability, and signs of clinical response.

Materials and methods
Participants with a history of previously treated Lyme disease and persistent fatigue were randomly assigned in a double-blinded fashion to either Group A (disulfiram for 4 weeks and placebo for 4 weeks) or Group B (disulfiram for 8 weeks). Primary outcome endpoint was at 10 weeks with a follow-up at 14 weeks. The primary aim was to assess safety and tolerability. A clinical aim assessed signs of clinical improvement using well-validated measures, focusing on improvement in fatigue and quality of life. Target enrollment was 24 participants.

Results
940 individuals were screened, 11 were enrolled and nine participated in the trial. Dosing started low and increased based on response and tolerance to a maximum of 500 mg daily.

Safety. Two participants discontinued medication due to clinical worsening, one of whom was briefly hospitalized. Three additional participants were withdrawn from treatment due to lab test abnormalities.

Tolerability. Only three of nine participants completed the full course of treatment (two in Group A and one in Group B). Lower doses were better tolerated than the highest dose.

Clinical response. Of nine participants, clinically meaningful improvement was noted in fatigue for six and in quality of life for four. Among the six fatigue responders, improvement was also noted on a multiple domain symptom index (six of six), overall symptom burden (five of six), and functional impairment (four of six).

The study was terminated early due to end of project funding, higher than expected adverse events, and recognition that sufficient information was gathered to inform future studies.

Conclusions and relevance
This study reveals the risks associated with disulfiram, especially at higher doses, while suggesting potential clinical benefits among some participants. Efficacy could not be assessed given the small sample size and the lack of a placebo-control group.

Clinical trial registration: NCT03891667

Link | PDF (Frontiers in Medicine) [Open Access]

 

It’s good that they state this outright.

There also seems to be a contradiction in the paper. They say both that this was a placebo-controlled trial, and that it was not a placebo-controlled trial.

 

I don't think there's necessarily a contradiction. They can't say whether improvement from baseline to end was due to disulfiram or not because everyone got disulfiram but it's possible they can still get some form of control via a placebo after week 4 by having two groups and taking measurements at week 4 and 10. I'm not sure whether any of that was actually done but it would seem theoretically possible, for instance if if Group A showed more benefit than Group B during weeks 5–8, that benefit can be attributed more confidently to continued disulfiram. Since, the authors didn't seem to look at response to blinding and overall the sample size is negligible, it's just as equally possible that the study provides no meaningful data besides the treatment having side effects. Presumably at that amount of treatment adverse effects and side effects it might be rather easy to tell who is receiving placebo and who isn't after week 4.

 

But they say that it was a placebo-controlled trial and that they didn’t have a placebo control. Per definition, you can’t have the former without the latter.

 

I think you may be able to call it a placebo-controlled trial because they had a group where a placebo offered control even though it's not the classic “placebo vs. drug” simply because the placebo offers some control during the second half of the trial, so it's more than a dose-duration study with 2 groups.

In that sense they aren't doing what you're saying they are doing, but it isn't clear to me what they are doing either. Perhaps they also don't know?

 

I tried to look up what their design is called, and this document describes something I believe is similar as a randomised withdrawal trial (section 2.1.5.2.4).
https://www.ema.europa.eu/en/docume...e-control-group-clinical-trials-step-5_en.pdf

If that’s the case, their assertion that they do not have a placebo-control group is wrong.

The issue isn’t that they did not have a placebo-control group (because they technically did), but that the placebo arm was designed so that they got the treatment first. Therefore, you can’t properly isolate the effect of the treatment because it might have a long-lasting or permanent effect.

So it seems to me like their two statements are contradictory, and that the last statement is wrong.

 

I don't know enough much about the language, someone knowledgable about clinical trials would know and the notion of a randomised withdrawal trial according to the link above indeed seems similar (though not quite the same) but in that case the nomenclature of "placebo-controlled trial" still seems fitting to me.

Yes, a long-lasting effect or permanent effect would confuse the picture but as with many other things that would depend entirely on the study goals and observations. One might think that the authors will have had suspicions about whether there is a long-lasting or permanent effect and one might think they are more so looking at the scenario I initially described and the reader would understand that it's those things that are being controlled for rather than anything else. Who knows. The trial design is anyways probably not of particular relevance here given the results.

My reading is that it's fine as it is. We seem to agree that it's a placebo-controlled trial (although not in the "placebo vs drug sense" but rather that there is a placebo that offers some control) and I think saying that there is no placebo-control group is also fine (because there is no placebo-group that would offer control, of course reading things differently one can argue that there is a control group and that the control is afforded by a placebo, but I find the other reading more natural).

In any case we're just discussing nomenclature here so the discussion is probably not very fruitful.