‘As to diseases, make a habit of two things - to help, or at least to do no harm’ - Hippocrates.

It’s pretty much a given that all therapeutic interventions are accompanied by some level of risk. The tricky thing is balancing this assessed risk against therapeutic efficacy.

When we first realised that small cone direct moxa had been used to treat TB in Japan in the 1930s (and that it therefore might have a part to play in today’s struggle to contain the parts of the pandemic that are drug-resistant and often incurable) we knew much less about the complexities of today’s manifestations of the disease (nor, in truth, the immensity of it) than we do now. What we did realise was that the only way to find out what part this simple treatment might play was by proper investigation.

What we also learnt very quickly was how unfavourably TB interacts with HIV/AIDS, and how challenging it is to treat the two diseases together. From a biomedical perspective, this makes treating co-infected cases a far riskier and more challenging proposition than those infected solely with tuberculosis.

We also realised that we would have to take this risk into consideration ourselves in the course of our investigations - after all the moxa specialists in Japan in the 1930s weren’t treating any TB patients co-infected with these two deadliest of infections together. HIV was still 50 years away. In the earliest stages we were even advised to exclude anyone who might be co-infected from any study we might develop: it might make the venture too complex and potentially also too risky.

Given that our focus was from the outset committed to Africa where prevalence, incidence and mortality rates from TB are by far the worst in the world, we quickly realised that excluding co-infected patients wasn’t just going to be very difficult - it was practically-speaking impossible because whenever both diseases co-exist the diagnostics for either disease proves less reliable than normal. So rather than excluding co-infected cases, it looked more practical to assume that any TB patient we encountered might well be HIV positive whether or not they were diagnosed as such. The bottom line was that, if the treatment was to play any sort of role at all in Africa at all, it simply had to help to some degree these many co-infected cases - or , in Hippocrates’s terms, at least it should ‘do no harm’.

So our focus from the outset was alerted to the risks associated with this. Our first pilot study in Uganda (a country recognised as having high incidence of both diseases) didn’t just look at reported patients’ responses, it was also constantly alert to possible adverse effects.

Ensuring that the moxa cones were appropriately small seemed the best start - that they weren’t just rice-sized, but half-grain rice sized - and of dry rice at that. We developed simple ways of doing as a part of the patients’ introduction to the therapy. Then there was dosage, and we paid very careful attention to the reports of documentation in the earlier Japanese literature. The mantra we paid most attention to was: if the patient is very sick, keep the dosage small; if the patient is or becomes stronger, build up the dose.

With these simple principles we completed three successive pilots two of which were in South Africa. No HIV positive patients were excluded, and when we reviewed the findings we could find only one significant complaint coming from the patients - that in some cases their appetite had improved beyond the resource they had available to satisfy it. In what is such a terrible wasting disease (one which used to be called ‘consumption’) this hardly seemed a negative effect, as challenging as it was for the patients who reported it.

So we went on - sponsoring and helping design the Phase II RCT that has been run by Makerere University‘s School of Health Sciences. Unsurprisingly the medical staff who have been involved in this trial have been just as alert to the possible complications as we were, but they too saw no reason to exclude the HIV cases from the study design (for the same sorts of reasons).

This RCT has been far more rigorously monitored than the preliminary pilots so we can be a lot more confident in its findings. To date (and the study is now fully enrolled with the last patients now two months into the eight of the treatment period, and with the majority already completed) we're pretty confident that we are going to see the study winding up with no evidence of problems in the co-infected group (comprising a good 30% of the whole).

So what was it that we were we worried about? Well the treatment involves burning the tiny cones of moxa down to the skin so that the patient feels an instant of heat before snuffing it out. We were certain that this is an essential part of the treatment, not just because the literature stated this to be so, but also because our own hypotheses about how an immune response to the treatment might occur concurred with the idea. But if the patient had to 'feel' the tiny burn, then they were quite possibly also going to develop small blisters - and if these occurred then we had to recognise the risk of infections occurring in immuno-compromised patients.

Small cones did reduce this risk, of course, as did careful dosage, but the third key strategy to minimise it was to careful monitor the patients. This responsibility devolved to the clinic’s ‘TB Focal Person’ Rehema Kigongo and her attention to each and every patient has been meticulous.

So far we haven’t seen any instances of infections at the site of the moxa burns - even in patients with known immune deficiency, and nor have we had to withdraw any patients from the study because of adverse effects from the moxa. This may well be because the moxa may create a localised immune response at the site of burn at the same time as it may promote a systemic one. What we can say for sure is that if we complete the study with the same report this alone will be a huge finding. Essentially it will give a green light for further studies which may (and we hope will) follow on, specifically now looking at drug-resistant cases. Because if co-infected cases of drug-susceptible TB disease can be safely treated with moxa (and also appear to benefit from it) then there seems no logical reason not to investigate responses in drug-resistant cases in a similar fashion (even those with functionally untreatable extensively drug-resistant XDR-TB) - and that this could even be done with some confidence.

Due care and attention certainly can never be waived. Diligent attention to patient safety remains an absolute necessity, but we hope that health practitioners anywhere who are engaged in the struggle to contain this terrible untreatable part of the pandemic (and having to finally give up on patients who are termed ‘therapeutically futile’ because of the extent of their drug resistance) will seriously consider the possibilities of further investigations.

If he were alive today Hippocrates himself would surely recognise their terrible dilemma. Twenty-five centuries ago he called the disease phthisis (or 'wasting'), and he even counselled his students not to treat it because the patient would fail to recover whatever efforts were made to save them.

We’re now weeks away, we hope, from our next major milestone - when the responses of all 180 patients from the critical first two months of their treatment will be analysed by a biostatistician. Depending what this data then tells us we’ll be ready to respond - and we hope that the father of medicine, Hippocrates, might at least approve of where we've got to so far.