• LecturehallSeparating Fact from Fiction: Does Topical Oxygen Therapy Work? Results of the TWO2 Trial
  • Lecture Transcript
  • TAPE STARTS – [00:00]

    Male Speaker: There's been a lot of talk for very many years about obviously, oxygen therapies and topical oxygen therapies and a lot of misunderstanding despite all the clinical use of these modalities. So I entitled this talk Separating Fact from Fiction to show you some of the latest data on topical oxygen therapy to bring some hard science to a nebulous field that had a lot of anecdotal case reports in that for the most part.

    So hopefully, we'll have a little bit better understanding of topical oxygen and this is really a talk that presents the results of our topical oxygen trial which we actually completed. Now, it's been a year ago and this paper has been submitted for publication just within the last few weeks. This research was sponsored by AOTI. Some of you might have seen the workshop that was held yesterday. Anyways, very interesting. We're not the only topical oxygen modality on the market, but this is just one that had a very interesting trial that I wanted to present to you.

    So our learning objectives have been published before, but it's basically to give you an overview of oxygen therapies and specifically, of the latest results of the clinical trial of topical oxygen therapy.

    For many years, I've used this graph to show the interrelated underlying metabolic disturbances in a diabetic foot. Not all patients have every one of these underlying pathophysiological aspects, but many patients do and we have to understand all the interrelated problems in the diabetic lower extremity.


    Obviously neuropathy is always our primary concern because it is the most common underlying complication of diabetes. But not just sensory, we have to recognize the important role of both motor neuropathy as well as autonomic neuropathy in this pathway.

    And of course vascular disease is not just macro vascular disease, although that is indeed a very important metabolic dysfunction. Also, micro vascular and micro neurovascular dysfunction leading to that impaired Lewis triple flare response that we know occurs in the diabetic foot.

    And when we apply trauma to this high risk system, then we can see with an associated impaired response to infection how these wounds can just propagate into a vicious cycle, become infected, develop amputation after they become necrotic. And by understanding the multiple pathways or multiple interrelated dysfunctions here, we can better understand the complexity of the diabetic foot.

    And I've always believed that if you can really understand the underlying pathophysiology, you'll be much more successful at managing these problems through various interventions. So I think that's always key.

    So now my original premise is that as we all know oxygen is essential for wound repair and angiogenesis. ATP oxidative phosphorylation is – provides the energy to drive these biologic processes that really undergo a great deal of activity during wound repair and regeneration. NADPH oxidase provides that respiratory burst that produces superoxide and other reactive oxygen species that enable bacterial killing.

    Also, these reactive oxygen species provide for intracellular and extracellular signaling and growth factor activity that is so critical to healing.


    And of course oxygen is essential for collagen production and acting as an enzyme cofactor that enables the tissue regeneration that we so depend upon.

    So many times we just kind of forget the very, very important role of oxygen. But it's good for us to review the integral components within the wound repair process that are really dependent upon oxygen. And why oxygen is an important thing that we need to consider when we're healing is chronic wounds.

    Many years we've known that the center of chronic wounds are hypoxic, whereas the PO2 in the arterial blood is about 100 millimeters of mercury. In the center of a hypoxic wound, there's a low PO2 of only about 10 millimeters of mercury. On the periphery of the wound, it's about 60 millimeters of mercury.

    So we see that this hypoxic center really can have a detrimental effect on healing – it's a stimulus, for sure for angiogenesis. But it's also, needs to be addressed in our agents to provide for wound repair.

    So it's very important to understand how hypoxia can drive these processes and how by addressing the hypoxia, we could help stimulate the wound repair.

    This paper was published in 2017, Wound Repair and Regeneration. And it was really a review of looking at a number of oxygen papers whether they'd be scientific reviews, experimental studies, hyperbaric studies or even topical oxygen therapies.

    And they did reveal that there were a lot of positive outcomes, but most of them were anecdotal and were not rigidly tested.


    And the vast methodological diversity makes it really difficult to make any final consensus statement about different types of therapies because there's such heterogeneity in all of these studies. Different inclusion criteria, different outcomes, different lengths of outcomes, 12 weeks, 12 months, how do you compare one to another?

    And so they recommended that further randomized clinical studies are warranted to achieve more solid and consistent data. And this is the guiding principle now that we really need to see in all wound repair studies. Decrease the heterogeneity, come up with some consensus for what do we need to achieve in a wound repair study so that we can compare one type of modality if you will to another, so we can have some kind of a comparative effectiveness or at least comparative effectiveness studies.

    So about oxygen therapies. We're all familiar with hyperbaric oxygen. We know that. Many of you use hyperbaric oxygen therapies. It's well-established, very well researched for many years, but the data is just inconsistent. And it's efficacy and healing chronic, the diabetic foot ulcers is disputed because of the heterogeneity of the studies and the inconsistencies of the studies.

    This is very evidence – much evidenced by the recently failed Dutch DAMOCLES HBO trial, which was published in Diabetes Care in 2018. Very large study, looking at 12-month outcomes and they were no better than a standard care. It wasn't even a sham controlled trial, which was really surprising for an HBO therapy. And this mirrors many other trials for HBO, that just didn't have the same outcomes that we would expect in a rigidly applied robust study protocol.


    So because of the inconsistencies, although we know theoretically HBO really can have some marvelous effect and when it's indicated, it can produce great results. But the HBO is not endorsed by the ADA Standards of Care in 2019 because of the mixed data. It's also not endorsed by the NICE Guidelines in the UK for the same reasons.

    Now, topical oxygen therapy is still controversial despite 50 years of use. You can go back 50 years and see topical oxygen being applied. I think the first paper was in Lancet, many years go back in the '60s. There are recently over 30 clinical reports and positive reviews – is that me? I thought it was me.

    Okay, so there's over 30 clinical reports and positive reviews in the literature. But many of these weren't controlled, many of them are anecdotal. So we don't really know how do – how do we really rigidly apply our scientific criteria to these results? So many, many questions are still available or still raised.

    So high quality RCTs are really required to establish efficacy. Not that clinical studies are not warranted, but when you're really trying to rigidly test something, you need to have really a rigid protocol and rigid standards. We'll be talking more about that particular subject later this afternoon.

    So this came from Dave Margolis' review. Dave Margolis is very well-known in this area out of the University of Pennsylvania. And he did a very large review, of over 6,000 patients who had undergone HBO.

    And he found, much as I've said that use of HBO neither improved the likelihood that a wound would heal or prevent an amputation in a large cohort of patients – in a large number of patients who were in the CMS database, Medicare patients.


    So he was saying that the usefulness of HBO needs to be reevaluated.

    And subsequent to this 2013 paper, we see several trials, the Federko trial, which failed, as well as the Santema trial in 2018, which also failed to heal diabetic foot ulcers in rigid designs.

    So there's been some growing interest in topical oxygen, this is globally. We produced this paper in the Journal of Wound Care, this was a supplement on wound care therapy. I think it was 2017, this came out. This was a global panel of persons looking at this, including HBO, including topical oxygen and even oxygen producing dressings.

    And there was some good evidence in this regard and a very interesting document. And the study that I'm going to be talking about was still ongoing at that time, but it received a level 1B evidence rating for the design of the study.

    And again, other papers have called for the need for very rigorous ongoing studies to further document the efficacy of both HBO and the oxygen containing dressings, as well as topical oxygen therapies.

    So the hardest thing to understand is what effect does oxygen have at the tissue level itself? And where can hypoxia impede healing and where can elevating the oxygen levels really help?


    So this can be somewhat of a difficult sly to understand. But if we break it up into sections, we can see that the normal tissue PO2 should be about 100 millimeters of mercury, somewhere in that area. And at this level, we can see a very, very high level of cytochrome oxidase or ATP cellular metabolism occurring. We can see a high level of collagen production and a reasonable level of NAPDH oxidase activity.

    Remember, oxygen is a cofactor, a very important component of all these processes. So they run normally at a 100 millimeters of mercury.

    But what happens in that hypoxic wound where your level of oxygen goes down to about 10 millimeters of mercury, look what happens along this curve, how it is so deficient. Look at the effects that it's going got have on ATP cellular metabolism on collagen production where we're going from levels up here to levels way down here. It's driving very, very much down. So the cellular and enzymatic activity, the collagen production and the ability to fight infection, far reduced at such low levels of PO2.

    So here, when we boost that level of oxygen, you know, this could be with HBO2 or even with topical oxygen where we go about in this regard, 50 millimeters above your standard atmospheric pressure of 760 millimeters. We can see that we are really driving this process to its maximum. So that's really the underlying theory as to why topical oxygen might really be beneficial even though it's just topically applied, not systemic applied.

    So we have to have some kind of animal science to say, well, is our theory really functioning at the organism or at the tissue level?


    This paper by Fries that was published in 2005 really looked at chronic wounds and applied topical oxygen and to see the effects of healing. Now, what they did was create wounds on the back of pigs here and they had control wounds on the animals as well as treated wounds. So they exposed the open derma wounds to topical oxygen treatment, and they found that there is an increase in the tissue PO2 2 millimeters below the level of the tissue they implanted, Clark electrodes to measure the oxygen levels in that tissue.

    They also found histological evidence for improved tissue regeneration, repeated treatment, accelerated wound closure and oxygen treated wounds showed sign of improved angiogenesis and tissue oxygenation.

    Now, let me show you this graph down the bottom here. This was with the oxygen electrodes implanted 2 millimeters into the wound. And remember that at baseline, your hypoxic wound is going to be less than 10 millimeters of mercury. And just after four minutes – after four minutes, the oxygen levels in that tissue increased by tenfold up to 40 millimeters of mercury.

    So this was some real evidence that topically applied oxygen can indeed raise the tissue PO2 levels of oxygen. Okay. And here we see differences in healing rates between the control, the sham treated side as well as the active treated side. And we also see differences in tissue improvement and regeneration here. This is the actively treated side. We have a more normal epidermis as we can see in these two panels.


    And this is the control side that has an abnormal appearing type of tissue histology. So this paper I believe really showed that topically applied oxygen can certainly have positive effects at the tissue level. Despite the fact that many people say it must be applied systemically to have any effect at all, this paper in 2005 really showed, yes, it can have some very positive effects.

    So there are many different topical oxygen devices available on the market. You've seen in many of these here even at this meeting or previous meetings, you have continuous diffusion of oxygen, you can have some low level of constant pressure or you can some intermittent or cycle pressurized topical oxygen.

    And this is the agent that we attested in our study, figuring that the cyclical pressurized oxygen in a chamber would demonstrate those positive effects that we saw in that Fries study from 2005, even though that study did not use this particular therapy. I think it actually used this therapy.

    So probably back in 2011, we started designing this protocol recognizing that we were going to come up against a great deal of controversy, a great deal of animosity let's day, especially by the hyperbaric community, by building podiums with hyperbarists who were rigidly opposed to their hair bristles when you even mention topical oxygen.

    So it was in that framework that we set up this study originally. It had to be rigorous. It had to be sham controlled. It had to be blinded. It had to follow a very rigid protocol because we knew that only by doing a really rigid protocol would we have a positive effect in this regard.


    So we called it a state-of-the-art sham controlled RCT, non-healing chronic diabetic foot ulcers using this cyclical pressurized oxygen therapy at home. And it was the best standard of care that we could use, the best rigidly applied offloading. We used 17 centers of excellence in the United States as well as abroad. So it's a multicenter, multinational study done here in several VA and non-VA centers as well as some recognized centers overseas. And you can see some of these centers there.

    So our study design included a group sequential design, a total of 220 patients. But build into the protocol were two a priori interim analysis. This is becoming a much more common type of a design to use because it can shorten your study by decreasing the number of patients, but it has to be built into the protocol for it to be valid.

    All of our patients had to meet a stringent eligibility criteria and as you'll know now, most studies have like a two-week run in period where you've given them the best standard of care and only those patients who have healed less than 30% – less than 30% are allowed to proceed to randomization because those patients who in two weeks healed 30%or greater are expected to heal just by normal standard of care.

    And so way of separating people that are most likely predisposed to healing to keep them out of this study so you're really just including those chronic wounds that need some type of an advanced therapy. And again, once they got into the randomization phase, it was double-blinded, meaning investigators or patients didn't know what they're receiving and was sham controlled with an identical unit, same lights, same sounds adjusted and provide for any topical oxygen therapy.


    Here, you can see the particular unit. So the patients administered this themselves at home 90 minutes a day. They came to the study sites weekly for debridement and evaluation. And of course, it was a 12-week outcome period where our primary outcome was 100% healing at 12 weeks, which really should be the standard for diabetic foot ulcers studies now.

    This is a little bit about the group sequential design. And I think this is important to understand. What is it? In normal studies, if we had the 220 patients, which were the final number of patients, you'd need only about a 10% difference in effect to reach significance. At 146 patients in the a priori analysis, this should be the second interim analysis, you need a 20% difference. But at the first interim analysis of just 73 patients, you would need a 30% difference to achieve significance.

    And the significance level in group sequential designs were always reduced because of the multiple reviews. So instead of a P value of 0.05, they can only achieve significance if they're below P 0.2 or excuse me, 0.02 or 2%. So it's a more rigid criteria.

    What we found at the first interim analysis, we were delighted to find that we achieved our primary outcome. So we were mandated by the protocol to terminate the trial. And what we found was that almost 42% of the patients in the active group healed compared to 13.5% in the control group. And this was very highly significant.


    On looking at the multivariable analysis, we looked for the various covariates or cofactors that might have an effect on this, and the only significant variable was the treatment effect, sham versus active therapy.

    So we met our outcomes at the first 73 patients. What this meant, we had 73 patients complete their 12-week of treatment and then we're mandated to stop enrollment and stop the trial because we achieved our outcomes here.

    These are just some of the patients – typical patients that you would see. We did allow amputation patients in if they fit the criteria otherwise. And so these would be fairly typical patients. But we allowed UT grade 1 and 2 and we allowed some modest degrees of ischemia into our trial too because we want it to be more real world, so we allowed ABIs of down to 0.7. We allowed various parameters of renal insufficiency. We allowed HP hemoglobin A1c of up to 12. So we pushed the limits so that we would get more complicated patients into the trial.

    And so if we look at the Kaplan-Meier curve here, we can see a very nice separation starting right from almost week one and you can see those curves are separate, the log-rank test and Cox's proportional hazards show that it was a highly significant difference between those two groups just on this small subset of patients that were analyzed in that first mandated interim analysis.

    And so the patients who were randomized to the active group, active group compared to control have 4 times likelihood, 4 times higher likelihood of healing than did the sham group. So this was very, very positive.

    Down here, we say that about 23% of our patients failed the screening, which indicates that these were allowing only the more difficult patients to come into the trial, whereas in many other DFU trials where they allow very, very perfused patients to come in.


    They let them right into the trial and they're expected to heal anyway. So we feel that we had more complicated patients only enter our trial.

    Now, if we look at those patients who did not heal, but had larger wounds over 4 centimeters in diameter, we see that the active group had an average reduction of 64% over 12 weeks. These are the ones who did not heal, but they had a percent wound area reduction of 64% active, and an enlargement of 2% in the sham group at 12 weeks in these larger wounds. This was highly significant also.

    So this is a secondary measure of outcome that you'll often see in studies, percent wound area reduction. And there was a significant difference in that when we look at those other patients who hadn't healed within those first 12 weeks. So that was a very important parameter as well.

    So our compliance was good, because that's always a problem. If you look at the problem with HBO studies, people aren't compliant because it's really a very difficult chore to go to a center for 40 days, when they might live far away from the centers, so the compliance isn't good. We found about a 94%, 95% compliance with both the therapy as well as the offloading device. And we chose a very good offloading device that itself had undergone clinical trial to show that it was equally effective as total contact cast, which is why we used this offloading boot, which comes from Italy.

    Twelve-month follow up, that's important too. We followed our patients for a full 12 months before we were able to finally analyze them. And we looked at the recurrent ulcers in both groups. And there was no really difference. Those who healed on the active or the sham, pretty much there was no difference in those who remained healed and also those that recurred after 12 months.


    So now if we look at this, look at the percentage diabetic foot ulcers healed within the 12 weeks, we can see here there's a 41.7 versus 13.5, this was the original primary outcome. We also looked at this at 12 months which I think is important for the wound healing studies as well. And here we see it was 56% healed active versus 27% at 12 months. And this is what most of the HBO trials are looking at as their primary outcome.

    So we saw significance at 12 weeks and as well as 12 months. And if we look at the margin of effect between active and sham, 66 margin or a – percent margin of effect at 12 weeks or 309% relative performance, 41% divided by 13.5.

    Now, if we look at the 12 months, we see a 52% margin of effect at 12 months or a 207% relative performance. So this is pretty rigid data that supports the value of this type of topical oxygen therapy.

    So we utilized one of the most rigorous protocols undertaken in the diabetic foot ulcer that was designed to remove the standard of care failure and blinding biases that we've seen in other studies, topical as well as hyperbaric, because we knew we had to be very rigorous to address all the concerns that others would have.

    And so this was a very robust RCT. We demonstrated significantly superior in healing recalcitrant complex diabetic foot ulcers at 12 weeks compared to sham, placebo alone. Even the larger DFUs that were not healed were on a trajectory to heal more significantly in the active group compared to the sham group.

    We found that there is more robust long-term healing.


    It's an easy-to-use home care therapy with high patient compliance, as we said, about 95% compliance with this home therapy. And so we believe this should be considered a frontline adjunctive therapy.

    Now, you always must be adjunctive to good wound care therapy as an option for those patients that might be hard to heal diabetic foot ulcers, those kind of patients that we deal with on a regular basis, a real life basis every day.

    So we have to thank our co-investigators. I don't know if any of our co-investigators are here in the room. But these are some very, very well-known researchers and diabetic foot specialists in both the Europeans centers as well as here in America. So with that, I want to thank for your attention. Hopefully, that was of interest to you.

    TAPE ENDS - [29:01]