Does a nitric oxide generating dressing heal diabetic foot ulcers faster?

A new randomized controlled trial (RCT) has found a nitric oxide generating dressing heals diabetic foot ulcers (DFUs) faster than control dressings. The RCT, led by the world-renowned Professor Mike Edmonds and many of the UK’s most well-known DFU researchers, was recently published in one of world’s leading wound journals, Wound Repair and Regeneration. But, again is this the full story?

For those regularly reading our latest research summaries, you may be thinking we are going crazy as only a few weeks ago we reported the same study by the same authors. But in this case we aren’t crazy, as this is a completely new study on a completely new dressing by the same Prof Edmonds who led the RCT of sucrose octasulfate dressing and who defined “diabetic foot attack” in the last months.

So what do we know about RCTs on wound dressings for DFUs?

As we discussed in our summary of Prof Edmond’s last RCT, there had been no large high-quality RCTs demonstrating the effectiveness of wound dressings to heal DFUs, until Prof Edmond’s recent RCT. Prior to that RCT, this lack of evidence had led the International Diabetic Foot Guidelines to recommend we “select dressings principally on the basis of exudate control, comfort and cost” as “the evidence to support the adoption of any particular (dressing) intervention is poor, because the available studies are small and at high risk of bias.”

Well ‘when it rains it pours’, as in the space of weeks here is another large RCT on a wound dressing to heal DFUs. This time the dressing is a nitric oxide generating dressing. Nitric oxide has been known to stimulate new blood vessel development, help with inflammation and there is a suggestion it is also antimicrobial in laboratory studies. However, to date it has proven difficult to satisfactorily integrate nitric oxide into a dressing because of its short half-life in dressings (i.e. it runs out too fast when applied).  So, what did this RCT do and was it as high-quality as Prof Edmond’s last RCT?

What did this new study do then?

Well this was a large multi-centre RCT investigating a dressing that for the first time, according to the manufacturer, generates a sustained release of nitric oxide (NO) on the wound. For this study they recruited patients from 10 diabetic foot clinics across the UK. To be included, patients needed to have a superficial DFU of >2.5cm2 and a palpable pulse or ankle brachial index (ABI) >0.5.

Patients were then randomized (by computer) to receive the NO dressing every 2 days (intervention group) or receive a dressing chosen by their treating clinician as often as they needed (control group). Otherwise according to the authors, both groups received normal DFU care as decided by their treating clinicians, such as debridement, offloading and antibiotics.

Patients were then followed every 1-2 weeks until 12 weeks and again at 24 weeks. At each follow up visit, photos of the DFUs were recorded using the Silhouette wound imaging system. These images were then assessed by a clinician ‘blinded’ to the patients so they couldn’t be biased. This also allowed the authors to determine the % ulcer area reduction, who healed completely and any adverse events.

So what did they find?

They randomized 147 included patients into two groups (72 in the intervention; 75 control). There were no significant differences between the groups for demographic, comorbidity, DFU and treatment characteristics according to the authors; but, the exact details of these characteristics were a little sketchy. After that, it was also a little challenging to work out how many patients started and completed the study, as the authors included patients with multiple DFUs as separate DFUs and reported multiple different DFU drop outs at multiple different time points before the 12 weeks were completed.

Overall, after 12 weeks they found the NO dressing group had statistically larger % ulcer area reductions, DFUs healed and serious adverse events compared to the control group. Again though this was a challenge to read the way it was reported with multiple different outcomes (% ulcer area reduction, % ulcers healed, % ulcer recurrence), at multiple different time points (4 weeks, 12 weeks), for multiple different subgroups (DFU durations, size, infection status), for multiple different protocols (intention-to-treat and per-protocol) and versions of these protocols.

What was good or not so good about this study?

Whilst this large RCT had many strengths – recruited large numbers of patients in multiple diabetic foot clinics, computer generated randomization and blinded outcome assessor – there were a few limitations. These included: i) there was a lack of detail for baseline characteristics and in the ones reported a lack of criteria that defined them; ii) there was no standardized protocol for other DFU care, so we just don’t know exactly which group got exactly what debridement, offloading, antibiotics, and in the control group what dressings were used; iii) the patient and treating clinicians weren’t blinded, so they knew which treatment they were getting and inherently that can bias results; and iv) the study was funded by a wound dressing company which according to the authors was involved in all aspects of the study including designing, analyzing data and writing the paper which also has the potential to introduce bias.

What does that all mean?

Well it seems that the NO dressings compared to control dressings, healed DFUs quicker with fewer serious side effects. But, in this paper these findings were a little unclear, due to either a lack of detail or the somewhat confusing way the study was reported. This suggests that although this large RCT shows some nice promising findings for the NO dressing’s effect on healing DFUs, we should interpret these findings cautiously until further RCTs are conducted to see if they show similar results.

As an aside, there have now been two recent large RCTs from the same lead author into two different wound dressings that both say they heal DFUs quicker than controls. From our read the previous RCT was of much higher-quality than this new RCT; but, that is just our interpretation. Recently we also highlighted a new tool designed to help you rate the quality of diabetic foot studies.

So this raises an interesting exercise, why don’t you have a crack at using this diabetic foot study quality appraisal tool to rate the 21 important study items in these two similar studies, and see what you find?

We quickly did and scored the previous RCT  20/21 (95%) and this new one 12/21 (57%); which might start to demonstrate the value of using such a tool. But, more importantly, what did you find???