A new electron paramagnetic resonance device to measure transcutaneous oxygen in humans

We read with great interest the recent paper by Kmiec et al proposing a new method based on electron paramagnetic resonance (EPR) to measure transcutaneous oxygen pressure (TcpO2) in humans. 1 On a technical point of view, we are a bit surprised by the fact that, in the first paragraph of the Introduction, the authors mention near infrared spectroscopy (NIRS) among the techniques that measure oxygen pressure, although similar suggestions are found in the literature.2 To the best of our knowledge, NIRS measures oxygen saturation, but not oxygen partial pressure. This is an issue because there is a strictly nonlinear relationship between oxygen pressure in biological tissues and hemoglobin or myoglobin saturation. On a clinical point of view, we congratulate the authors for their idea and agree that EPR could be a promising technique with potential applications in the clinic. However we believe that a few points need to be clarified to support this assumption. First, beyond the use of TcpO2 at rest in severe peripheral artery disease (PAD), the authors could have referred to the use of TcpO2 during exercise in patients with arterial claudication.3-5 This point is of particular interest because exercise requires fast response time and motion‐insensitive devices. Concerning this latter point, in paragraph 2.3, the authors indicate that flexible cables were used to reduce motion artifacts. If EPR is motion sensitive, this is a major issue for exercise use. It would be useful that the authors provide recordings performed during movements of the subjects and explain whether or not the raw signal was denoised to improve the analysis. Second, we disagree with the authors when they claim that the electrochemical approach “requires heating of the skin while making measurements,” and that “heating is an obligatory requirement for TcOM (transcutaneous oxygen monitor) operation.” TcpO2 devices are perfectly able to record oxygen pressure value at the surface of the skin, but, as underlined by the authors, heating improves diffusion and increases local blood flow. The basal vasoconstriction makes the flow to the skin very low in ambient moderate temperature conditions.6 If one aims at estimating perfusion defect at the area of interest, the issue is to get rid of the skin vasoconstrictor tone by local heating. Local heating is of interest to reduce the variability of TcpO2 measurements 7 and improve the sensitivity of tcpO2 in the detection of severe limb ischemia.8 Third, it appears that 60 to 120 minutes were required to reach equilibrium when probes were positioned on the patient. This very long delay might limit EPR use in clinical routine. Could this delay be reduced by the use of some liquid between the skin and EPR probe? We agree that this 1‐ to 2‐hour delay might result from oxygen trapped under the probe, as suggested by the authors. Nevertheless, this might also result from a very slow response of the EPR electrode to oxygen content changes. Indeed, in Figures 3 and 4, it seems that the half‐time responses to changes were extremely long. If it was actually the case, this would not allow accurate exercise recordings. We urge the authors to provide the results of simultaneous response of electrochemical and EPR devices to acute changes induced by tourniquet or oxygen breathing. Fourth and last, the authors indicate that higher TcpO2 values after 10‐minute tourniquet ischemia was attrubited to hyperemia. A limb ischemia induced by cuff occlusion of more than 10 minutes is a quite painful experience. It could be that local oxygen pressure increase actually resulted from the increase in systemic oxygen pressure with hyperventilation attributed to pain or any other changes in experimental conditions. We advocate that the interpretation of local TcpO2 changes should always account for eventual systemic oxygen pressure changes, for example by the monitoring of chest values.5,9