Letter in Reply to Gueret et al: Carbon Dioxide Removal: Low Bicarbonate or H+ (Cl-) Addition?

Abstract

To the Editor: We thank Gueret et al. for their interest in our article,1 and appreciate their many remarks. Most can be addressed by reiterating that our manuscript reports the results of a modeling experiment for blood plasma and a separate in vitro experiment, using an electrolyte solution designed to resemble the electrolyte composition of plasma (artificial plasma). Therefore, the modeling work used values for the Ka of weak acids (Atot) and strong ion difference (SID) based on blood plasma. For the artificial plasma, we simply added phosphate because it is a component of plasma. We agree with Gueret et al that phosphate was certainly removed during dialysis, as were all small molecules with a diffusion gradient, including bicarbonate and dissolved CO2. This is a well-established phenomenon.2 In response to Gueret et al’s remarks about SID being calculated from only sodium and chloride, it is unclear which SID they are referring to. The SID we report in Table 1 refers to the values used in our modeling, and in Figure 6, we present the post-filter SID values from our in vitro experiment; calculated using sodium, chloride, potassium, calcium, and magnesium. These values were measured by flame spectrometry and presented in Supplementary Table S3. Furthermore, the small differences between post-filter sodium values measured by flame spectrometry and the concentration reported in Table 2 are explained by standard measurement errors (small weighing errors when constructing the solutions and small measurement errors with flame spectrometry). We welcome Gueret et al’s question about pH and bicarbonate at the start of each experiment because it gives us the opportunity to highlight the information contained in Supplementary Table S2, which confirms starting pH is the same. Since pH and pCO2 are stable, we did not think it was necessary to provide calculated bicarbonate as well. The authors of the letter also draw attention to some really important findings in Figure 4, by noting our two bicarbonate solutions (0 mmol/L and 16 mmol/L) have a similar effect on pCO2. We thank the authors highlighting this fascinating phenomenon because the modeling shown in Figure 3 predicts this is exactly what would happen if bicarbonate is removed with dialysis, we discussed this in detail in the manuscript. Gueret et al conclude their letter by asking how we excluded “that the effect of the dialysis solution composition on CO2 removal was not secondary to the transfer of CO2 and bicarbonate through the membrane according to the difference in pCO2 and bicarbonate concentration”. Contrary to excluding this effect, our study was designed to show this is indeed the case, as described in the discussion and conclusion of our manuscript. While we welcome, encourage, and enjoy the discussion, Gueret et al do not provide a reasonable argument to support their claim that our “study doesn’t allow to definitively concluding about the use of Stewart theory on CO2 removal according to the bicarbonate concentration in dialysate solution.” However, we do thank the authors for an additional opportunity to highlight supplementary data published with our manuscript.