Does Continuous Diffusion of Oxygen Improve Diabetic Foot Ulcer Healing?

Abstract

The objective of this article is to evaluate the study design, rational and results of Continuous Diffusion of Oxygen Therapy (CDOT) to heal diabetic foot ulcers (DFUs). The study was a double-blinded placebo controlled randomized clinical study to evaluate DFUs over a 12-week period. The primary outcome was the proportion of ulcers that healed and the secondary outcome that was reported was the time to ulcer healing. In the per protocol analysis and in the intent to treat analysis, a higher proportion of DFUs healed in the CDOT group (per protocol 46% vs 22%, P = .02, intent to treat 31.5% vs 15.1%, P = .03). CDOT patients healed ulcers faster compared to the in the sham treatment arm (P = .026)

Topical oxygen to improve wound healing and appropriate devices to provide this therapy have been controversial. There is certainly a bias by physicians because this concept does not fit our current belief system that oxygen applied to the surface of a wound will not have an effect on healing. The authors provide the background and rationale for their novel approach and Continuous Diffusion of Oxygen Therapy (CDOT) device. In a study using a rabbit ear wound model, ¹ Said and colleagues delivered continuous oxygen at 3mL/h for 8 days. Semiquantitative analysis showed increased deposition of glycosaminoglycan. At day 5 there was a 91% increase in epithelial coverage and by day 8 there was a 156% increase in epithelial coverage. Real-time PCR analysis for changes in AP-1, an indicator of keratinocyte transcriptional activation, showed differences between the oxygen and control group that approached statistical significance (P = .07). Likewise, Rao and colleagues ² evaluated topical oxygen in a rat model of ischemic wounds over a 2 week period. The rate of wound healing, granulation tissue formation, accumulation of collagen fibers, and new blood vessels were significantly greater in the topical oxygen therapy group. In a small randomized clinical trial, Yu and colleagues ³ evaluated 20 patients with UT Ulcer grades 1-3 for 8 weeks. In a per protocol analysis, the authors reported a significant difference in the rate of wound closure in patients that received topical oxygen therapy.

A larger clinical study was needed to help understand the efficacy of topical oxygen therapy. The study is unusual because it includes a true placebo control in the form of a nonfunctioning CDOT device. The methodology and study design for phase 3 and 4 randomized clinical trials for diabetic foot ulcers (DFUs) are usually very similar. Many randomized clinical trials (RCTs) that evaluate medical devices for DFUs do not have sham controls. Negative pressure wound therapy studies^4,5 and cellular and tissue based product studies, such as the Grafix, Integra, Dermagraft, and Apligraft RCTs,^6-9 do not have a sham treatment arm.

Usually DFU RCTs exclude patients based on end stage renal disease, glycated hemoglobin, ulcer size, ulcer location, ulcer duration, and peripheral vascular disease. The primary outcome is usually the proportion of wounds that heal, and secondary outcomes include the time to heal and adverse events.

There are several aspects of the study that deviate from standard approaches, including lack of details concerning the standard of care, ulcer location, ulcer duration, and ulcer classification and exclusion criteria. Neiderauer and colleagues report that moist wound therapy, debridement, and off-loading were provided to both groups, but no details are provided. This is especially important for off-loading. The type of off-loading is strongly related to wound healing. The details of the kind of off-loading provided to each treatment should have been clearly stated. ¹⁰

Likewise, there is simply no data provided concerning the ulcer location, duration, or UT ulcer classification. I am especially interested to see whether the CDOT technology was successful on wounds on the sole of the foot, and whether the type and duration of diabetes and chronic kidney disease were similar in both treatment arms.

The most serious concern about the article was that there was not an intent-to-treat analysis and no adverse event data were reported. It is unusual for an article to be accepted for publication without an intent to treat analysis. The article states that “subjects who failed to meet eligibility criteria, withdrew for any reason, or who completed but were not among the first 50% to complete each treatment arm were excluded.” However, the data are provided in consort diagram, so an intent-to-treat analysis could be performed. The authors randomized 146 subjects and 23 “dropped” from each treatment arm. According to the per-protocol analysis 46% of patients in the treatment group and 22% of patients in the sham group healed (P = .2). The per-protocol analysis did not include the 46 subjects that “dropped” in the analysis. The intent to treat analysis in which every subject that was randomized was included in the analysis showed that 31.5% of CDOT patients healed, and 15.1% of sham patients healed (P = .03). The two-tailed Fisher’s exact test shows similar significant outcomes as the per-protocol analysis.

Most RCTs report adverse events in detail. There was no difference in adverse events in the two treatment arms (CDOT 16.4% [n = 12] vs Sham 19.2% [n = 14]). It would be helpful to know the adverse events that were related to infection. In theory, topical oxygen might reduce the incidence and severity of infection. In many other DFU studies, infection is the most common foot specific adverse event.

The results of this study are promising. Among patients that received CDOT, there was a higher proportion of DFUs that healed, the rate of healing was faster, and there was no difference in adverse events in the treatment groups.