Proceedings of The Physiological Society

Physiology 2016 (Dublin, Ireland) (2016) Proc Physiol Soc 37, PCA069

Poster Communications

The Roughton-Forster extrapolation of DLCO versus PO2 does not give its diffusive component

M. Kang1, D. Grebenkov1, H. Guénard2, I. Katz3, B. Sapoval1

1. Ecole Polytechnique, Palaiseau, France. 2. Université Bordeaux 2, Bordeaux, France. 3. Air Liquide Santé International, Jouy-en-Josas, France.


In 1957, Roughton and Forster proposed a decomposition of carbon monoxide diffusing capacity (DLCO), also called the transfer factor (TLCO), as a sum of two resistances in series namely a membrane plus plasma resistance (1/DmCO) and a blood resistance (1/θVc). Based on oxygen partial pressure (PO2) dependence of DLCO at large pressure, Forster et al. (1957) suggested that the diffusive component (DmCO) could be obtained by linear extrapolation of 1/DLCO to zero PO2, where the blood resistance becomes zero. Here, we mathematically examine the same process, i.e., capture of CO by haemoglobin (Hb) after diffusion from the gas, with a minimal model of the physico-chemical phenomena responsible for capture, namely diffusion and reaction. In the frame of diffusion-reaction, the diffusive component DmCO is supposed to be the value of DLCO when the reaction time of CO and Hb is so fast that the red blood cell plays a role of a sink for CO. The diffusion-reaction equations have been solved for three different morphology of red blood cells: flat-parallel surfaces, spherical RBCs in a cylindrical capillary and biconcave RBCs in a cylindrical capillary. The results, obtained from analytical solutions for the flat case and numerical (Finite Element Method) solutions for the spherical and biconcave cases, demonstrate properties that are common to the three different RBC morphologies. First, 1/DLCO versus the CO-Hb reaction time (τco) is only approximately linear. Consequently, the linear extrapolation method proposed by Roughton and Forster does not give 1/DLCO(τco=0). Therefore, the supposed correspondence between DmCO and the value of DLCO extrapolated from large PO2 is erroneous. The figure below illustrates the dependence of 1/DLCO(τco) as a function of τco in the case of biconcave RBC in a cylindrical capillary. One observes that for small τco corresponding to small PO2 the dependence is not linear. Note that for standard PO2 = 100 mmHg, the value of τco = 0.5 ms (Kang and Sapoval, 2016). As similar results are found for the other morphologies, one has to conclude that the classical interpretation of DLCO is not soundly based. Thus, an alternative interpretation of DLCO is needed. Such an alternative has been proposed recently by Kang and Sapoval (2016). The new interpretation is based on the idea that for a CO molecule to be captured, it must travel by diffusion from the gas to the Hb molecule during a duration called δ, then it has to react which needs on average a time τco. So the time for capture is simply δ + τco.

Where applicable, experiments conform with Society ethical requirements