Abstract
The transfer factor of the lung for nitric oxide (TL,NO) is a new test for pulmonary gas exchange. The procedure is similar to the already well-established transfer factor of the lung for carbon monoxide (TL,CO). Physiologically, TL,NO predominantly measures the diffusion pathway from the alveoli to capillary plasma. In the Roughton–Forster equation, TL,NO acts as a surrogate for the membrane diffusing capacity (DM). The red blood cell resistance to carbon monoxide uptake accounts for ∼50% of the total resistance from gas to blood, but it is much less for nitric oxide.
TL,NO and TL,CO can be measured simultaneously with the single breath technique, and DM and pulmonary capillary blood volume (Vc) can be estimated. TL,NO, unlike TL,CO, is independent of oxygen tension and haematocrit. The TL,NO/TL,CO ratio is weighted towards the DM/Vc ratio and to α; where α is the ratio of physical diffusivities of NO to CO (α=1.97). The TL,NO/TL,CO ratio is increased in heavy smokers, with and without computed tomography evidence of emphysema, and reduced in the voluntary restriction of lung expansion; it is expected to be reduced in chronic heart failure. The TL,NO/TL,CO ratio is a new index of gas exchange that may, more than derivations from them of DM and Vc with their in-built assumptions, give additional insights into pulmonary pathology.
The classical technique for measuring gas transfer from the alveolus to the pulmonary capillary blood is the single breath transfer factor of the lung for carbon monoxide (TL,CO), but known in North America as the diffusing capacity of the lung for carbon monoxide (DL,CO). In the last two decades, the single breath measurement of diffusing capacity of the lung for nitric oxide (TL,NO or DL,NO) has been introduced [1, 2]. Since the work of Roughton and Forster [3], the model for gas transfer from alveolus to blood consists of two resistances in series:
where 1/TL is the total resistance to gas transfer (mmol−1·min·kPa in SI units or mL−1·min·mmHg in traditional units), 1/DM is the resistance to passive diffusion across the alveolar–capillary membrane and intracapillary plasma (DM is the membrane diffusing capacity), and 1/Θbl·Vc is the resistance to gas transfer of the red blood cell, which includes, for reactive gases such as carbon monoxide (CO) and nitric oxide (NO), chemical combination with the red blood cell haemoglobin (Hb) (1/Θbl is the resistance of red blood cells to gas transfer, e.g. CO or NO, per mL of blood and Vc is the pulmonary capillary blood volume measured in mL). Θbl is the specific transfer conductance of blood (measured in vitro) for a specified gas.
For CO these two resistances (1/DM and 1/Θbl·Vc) are approximately equal. For NO, the total resistance to alveolar–capillary transfer (1/TL) is much less, ∼20–25% of that for CO, thus TL,NO is four to five times greater than TL,CO, and the resistance resides mostly in the 1/DM component. This occurs for two reasons: 1) the physical diffusivity of NO is approximately twice that of CO and its resistance (1/DM) is half; and 2) the rate of combination of NO with blood in vitro is considerably faster than for CO [4]. Because the blood cell resistance for NO is low compared to the membrane resistance, the measurement of TL,NO has been regarded as a surrogate for DM. In essence, TL,NO measures DM and TL,CO measures DM and Θbl·Vc; thus, the TL,NO/TL,CO ratio will be weighted towards the DM/Θbl·Vc ratio times the ratio of diffusivities for NO and CO.
In this issue of the series we review measurements of the TL,NO/TL,CO ratio that have been reported in normal subjects and in various respiratory and pulmonary vascular conditions. It should be noted that the TL,NO/TL,CO ratio is equivalent to the ratio of the transfer coefficients for NO (KNO) versus CO (KCO) because TL = K × alveolar volume (VA), where K is the rate of uptake per min·mmHg−1 for NO tension (PNO) or CO tension (PCO), and VA is common to TL,NO and TL,CO.
PHYSIOLOGICAL DETERMINANTS OF TL,NO AND KNO
There are important differences in the way NO and CO are handled by tissues and blood namely: 1) the diffusivity (solubility/MW2) of NO in plasma is 1.97 times that of CO, and 2) the rate of NO uptake per mmHg of NO tension per mL of blood, i.e. its specific conductance (Θ) [4], is 5.75 times faster than the uptake of CO at a PO2 of 100 mmHg [3]. The chemical reactions of NO and CO with blood are also different. For example, NO reacts directly with the oxygen of oxyhaemoglobin to form a nitrate plus a deoxygenated form of Hb called methaemoglobin (metHb) in which the iron atoms of the haem ring are oxidised from the ferrous (Fe++) to the ferric (Fe+++) form [5]:
CO does not react with O2 but competes with oxygen for the Fe++ site on the haem ring:
The increased affinity of CO for Hb (∼220 times that for O2) is due to the different angles of attachment of CO and O2 to the haem ring [6]. NO and CO are tightly bound to Hb through their extremely slow dissociation constants. Unlike NO, the rate of reaction of CO with oxyhaemoglobin is PO2 dependent; once Hb is saturated with oxygen, the specific resistance reaction rate (1/Θ) is linearly related to PO2. This is the basis of the Roughton–Forster formulation (equation 1). TL,NO, on the other hand, is independent of the level of alveolar PO2 (PA,O2) [7] because NO reacts directly with haemoglobin (equation 2) rather than competing with oxygen for Hb binding sites (equation 3).
IS THERE SIGNIFICANT BLOOD RESISTANCE TO NO UPTAKE?
Investigators have cited the rapid reaction of NO with Hb (250 times faster than CO) as a reason for considering TL,NO to be a surrogate for DM [1]. The assumption that Θbl,NO for red blood cells is infinite cannot, in theory, be correct because of the advancing front phenomenon, i.e. the reaction rate of NO with Hb is so high that, according to Morris and Gibson [8], “effectively every molecule of NO which enters the reaction radius is captured [instantaneously] by a heme group. The observed rate [Θbl] would then be a measure of the rate of diffusion to the site.” This means that a diffusion pathway, either across the red blood cell membrane or within the substance of the cell, or both, is an essential component of Θbl,NO.
Experimentally, red blood cell lysis (by the addition of water to blood in a membrane oxygenator model of NO and CO transfer [9]), or red blood cell substitution, in anaesthetised dogs, with cell-free haem based oxyglobin [10] increased TL,NO substantially, but hardly altered TL,CO. This suggested, for NO uptake, that there was significant resistance in the red blood cell membrane, or its interior, or in a stagnant layer of plasma immediately surrounding the cell, and separate from any resistance stemming from the chemical combination with haemoglobin; conversely, most of the red blood cell resistance to CO uptake appeared to be associated with the haemoglobin molecule itself. Unlike TL,CO, TL,NO is unaffected by changes in PA,O2 [7]; as already mentioned, this is not surprising considering the chemistry involved (equation 2), but it supports the notion that the red blood cell resistance to NO uptake is independent of the haemoglobin molecule. In addition, TL,NO but not TL,CO seems to be relatively independent of the haemoglobin concentration in blood [11]. Borland et al. [10] estimated that 37% of the resistance to NO uptake lies in the 1/Θbl·Vc component (∼50–60% for CO uptake), but this figure must be treated with caution as it involved exchange transfusion in dogs, substituting bovine Hb-glutamer-200 (a cell-free blood substitute) for whole blood. To conclude, significant blood resistance to NO uptake exists, both for theoretical reasons and from experimental data, but in absolute terms 1/Θbl,NO is a small fraction (<16%) of 1/Θbl,CO. Thus, it is not inappropriate to regard the TL,NO, much more than the TL,CO, as weighted towards DM.
PHYSIOLOGICAL DETERMINANTS OF THE TL,NO/TL,CO RATIO
If, as a simplifying assumption, TL,NO “operationally” equals DM, the Roughton–Forster equation can be rewritten for NO and CO as follows:
where α (=1.97) is the ratio of membrane diffusivities of NO to CO in plasma. Assuming that 1/Θbl,NO was negligible, Guenard et al. [1] showed that DM and Vc could be calculated from a single breath manoeuvre with CO and NO as test gases, using a value for ΘCO appropriate for the single breath PA,O2:
where Vc is calculated in mL and DM,CO is calculated separately from equations 4 and 6. This was a more convenient solution than the Roughton and Forster [3] two-step approach at two different PA,O2 values. Reasonable values were found in normal subjects for DM,CO and Vc [1], but the estimates for Vc are dependent on the values chosen for Θbl,CO as explained in the Appendix.
Equation 7 can be rearranged (equation 1, adapted for CO uptake, and divided by equation 6) as follows:
TL,NO/TL,CO = α(1 + DM,CO/ΘCO·Vc)(8)
This illustrates the dependence of the ratio on DM,CO/Vc since α and Θbl,CO (at a given PO2) are fixed quantities.
Alternatively, if there is finite resistance to red blood cell NO uptake [10], 1/DM,NO must decrease, for a fixed value of 1/TL,NO, when 1/Θbl,NO·Vc increases from zero, as equation 4 reverts to equation 1. Thus, DM,NO will now exceed TL,NO. This increase in DM,NO (TL,NO, TL,CO and DM,CO being unchanged) “forces” α (in equations 5 and 8) to increase, even though it is a physical constant. Nevertheless, the dependence on the DM,CO/Vc ratio in equation 8 will remain.
Glénet et al. [12] have presented a diffusion model (in two dimensions) for the TL,NO/TL,CO ratio, which is a rectangular box whose height and width define the thickness of the alveolar–capillary membranes and the thickness of the blood sheet; they show that the TL,NO/TL,CO ratio is related to the tissue diffusivity (for NO) and inversely to the product (approximately the area of the box) of the thickness of the blood and tissue sheets, and to Θbl,CO. The sheet is thicker at functional residual capacity, mainly due to increased blood thickness (Vc/VA), and thinner with continuous positive pressure breathing or haemodilution; in all cases the TL,NO/TL,CO ratio changed appropriately. Thus, one would predict that in extrapulmonary restriction the TL,NO/TL,CO ratio (∼KNO/KCO) would fall and that this might be clinically useful, and this prediction is supported by measurements in normal subjects at different levels of lung expansion (fig. 1b).
DM,CO AND VC FROM SIMULTANEOUS SINGLE BREATH TL,NO AND TL,CO
Using equation 6, DM,CO can be calculated if TL,NO and α are known, on the assumption that the blood resistance to NO uptake (1/Θbl,NO·Vc) is 0, Vc can then be derived from the Roughton–Forster equations if ΘCO at a PO2 of 100 mmHg is known (equation 7). Nevertheless, there are several uncertainties in this calculation of Vc. There are seven separate equations [15], differing in slope and intercept, for the expression 1/ΘCO=α·PO2+β, all measured in vitro under different experimental conditions, with α being a temperature and pH-dependent coefficient linked to the reaction of CO with Hb. β is related to λ, the ratio of the permeability of the red blood cell membrane to the interior of the cell, but may also depend on stagnant layers of plasma adjacent to the cell [16]. Thus, 1/ΘCO a PO2 100 mmHg (13.3 kPa) may vary from 0.82 to 1.71 min−1·mmHg−1. Another variable is the DM,NO/DM,CO ratio (α) which, on physical principles, should be in the range 1.93–1.97. Investigators have “forced” α to 2.42 [17] or 2.08–2.26 [18] to give a “best fit” with the DM,CO and Vc calculated from the oxygen two-step Roughton–Forster TL,CO method. Since α is defined as the physical diffusivity ratio of NO/CO, this approach cannot be correct physiologically. A third uncertainty is the DM,NO/TL,NO ratio, generally assumed on the basis of the zero blood cell resistance to NO uptake to be 1.0 [1], although values of 1.57 have been measured experimentally [10], albeit under rather artificial conditions of red cell substitution with cell-free haem oxyglobin. The dependence of estimates of pulmonary capillary volume (Vc), on ΘCO and the NO red blood cell resistance proportion, for fixed values of TL,NO and the TL,NO/TL,CO ratio, is shown in the Appendix where TL,NO at rest (144 mL·min−1·mmHg−1) is taken from Zavorsky et al. [19] and the TL,NO/TL,CO ratio (4.5) from the average of eight studies.
In the Appendix we show that calculations of Vc, from simultaneous TL,NO and TL,CO measurements, using equation 7, are very dependent on the choice of 1/Θbl,CO and that DM,CO is dependent on the value chosen for the blood resistance fraction of NO uptake ((1/Θbl,NO·Vc)/(1/TL,NO)). We propose, therefore, that calculations of DM,CO and Vc from simultaneous measurement of TL,NO and TL,CO be set aside until there is more consensus concerning the 1/ΘCO–PO2 relationship and the DM,NO/TL,NO ratio. The TL,NO/TL,CO ratio avoids these uncertainties and assumptions; it also has the advantage that it represents the KNO/KCO ratio (VA being common to both measurements), which, as rate constants, have a direct bearing on gas exchange efficiency.
TL,NO AND TL,NO/TL,CO: NORMAL VALUES AND EFFECTS OF AGEING, LUNG VOLUME AND EXERCISE
We present a literature review of simultaneous measurements of TL,NO and TL,CO in normal subjects in table 1. Although there is a wide spectrum in the mean values between studies (for example, the subjects in Zavorsky et al. [19] were probably more athletic), it is more pertinent to relate reference values for TL,NO to TL,CO values measured at the same time, as TL,NO/TL,CO ratios. In two large European series [13, 20] (table 1), the TL,NO/TL,CO ratio averaged 4.45 and 4.8, respectively, and in a North American study [21] averaged 5.16. The average value of eight smaller studies [1, 12, 19, 22–26], weighted for numbers, was 4.5. At the present time, each laboratory should establish its own standard for the TL,NO/TL,CO ratio in healthy subjects, although the current consensus is that the ratio is in the range of 4.3–4.9.
Ageing
In the age range 25–55 yrs, van der Lee et al. [13] found the TL,NO/TL,CO ratio increased by 0.33% per year, but three other studies [12, 20, 21] found no change in the ratio with ageing. Thus, TL,NO and TL,CO seem to decline with ageing at essentially the same rate.
Lung volume
TL,NO is more sensitive to VA deflation than TL,CO. For example, from VA,max to VA,50%max the TL,NO declines by 43% versus 29% for TL,CO (fig. 1a) [13]. The explanation is that the fall in TL,CO is buffered by a greater increase in KCO (+42%) than KNO (+14%) (fig. 1b) [14]. This is due to a greater decrease in DM than Vc as lung volume decreases. In other words, a rise in Vc/VA is the principal reason for the increase of KCO [14]. If KNO (∼TL,NO/VA) reflects DM/VA, the effects of volume change on KNO (fig. 1b) should be similar to DM,CO/VA, as calculated from the Roughton–Forster DL,CO analysis [14]. In fact, at VA,50%max (as a fraction of the value at VA,max), the KNO ratio (1.14) from van der Lee et al. [13] is almost the same as the DM,CO/VA ratio (1.12) from the data of Stam et al. [14], although there was considerable inter-subject variability. Figure 1b shows that volume change affects TL,NO/VA (∼KNO) and DM,CO/VA in a very similar way, quite differently from TL,CO/VA (∼KCO), lending further support to the notion that TL,NO is “effectively” measuring DM.
The rise in KCO as lung volume and expansion diminishes is the reason for the fall in the TL,NO/TL,CO ratio (fig. 1a) when lung volume is lowered, and this fall may be a useful marker of extrapulmonary restriction versus other pathologies (fig. 2).
Exercise
Zavorsky et al. [19] have summarised the data from seven studies on the effect of moderate-to-heavy (maximum oxygen uptake 46.5 mL·min−1·kg−1) exercise. There was a linear increase in TL,NO and TL,CO , which were highly correlated. The TL,NO/TL,CO ratio decreased by an average of 9% (range -2 to -16%). DM and Vc both increased on exercise [17], but TL,NO will not share the increase in Vc caused by capillary recruitment and distension, so the TL,NO/TL,CO ratio will fall.
Breath holding time
Dressel et al. [28] found slightly higher TL,NO values at very low breath holding times of 4 s; this effect has not been reproduced by other researchers. No significant differences were seen between 6- and 8-s breath-holding times [28]. Although there are advantages in sticking to the usual 10 s, for the sake of comparison with previous single breath TL,CO estimations, the sensitivity and response time of some NO analysers (table 2) will force some researchers into accepting a 6- or 8-s breath holding time.
MEASUREMENT OF TL,NO AND KNO: TECHNICAL MATTERS
Most investigators use the single breath technique with breath holding as described for the TL,CO (DL,CO) by Ogilvie et al. [30], with the breath-hold time estimated according to Jones and Meade [31] or Graham et al. [32]. table 2 summarises the technical aspects from the principal reference studies. NO oxidises to NO2 when in contact with air, so it is stored in a nitrogen tank and dispensed just before use. This reaction is rather slow; therefore, mixing the NO with air in the inspiratory bag does not immediately lead to significant NO2 formation. NO reacts with certain plastics and connections to and from the dispensing bag, and these connections should be made of polytetrafluoroethylene (e.g. Teflon™; DuPont, Wilmington, DE, USA) or stainless steel. Borland and Higenbottam [2] showed that there is no interaction between NO and CO. The commercially available combined TL,NO and TL,CO apparatus has similar values for TL,CO as the traditional TL,CO apparatus when the same subjects are tested on both [32].
Because the rate of uptake from alveolar gas (∼KNO) is four to five times faster than for CO (∼KCO), breath holding times have, in general, been shorter than the 10 s that is the usual for TL,CO. Nevertheless, note that the very sensitive chemiluminescence NO analyser used by van der Lee et al. [13] allows them to extend the breath hold time to the usual 10 s, and this increases the accuracy of both the TL,NO and the TL,CO measurements. Endogenous levels of NO and CO are usually ignored. For normal populations a TL,CO and KCO correction for Hb is waived; for clinical studies, a Hb correction to a standard [Hb] is recommended but it is not required for TL,NO and KNO [11]. Smoking is generally forbidden for 24 h before testing because of its effects in raising plasma CO tension (“back-pressure” effect) and increasing HbCO (“anaemia” effect), but smoking and CO do not affect the TL,NO.
THE TL,NO/TL,CO RATIO (∼KNO/KCO) IN DISEASE
The TL,NO/TL,CO ratio can be normal, increased or decreased. A normal TL,NO/TL,CO ratio does not exclude a pathophysiological state, because both the TL,NO and TL,CO can be lowered equally, but it is unlikely that a pathological process will affect both components proportionately. According to equation 8, the TL,NO/TL,CO ratio is mainly influenced by the DM,CO/ΘCO·Vc ratio, or the ratio of the membrane to red blood cell conductance for CO. Figure 2 illustrates the clinical situations in which the TL,NO/TL,CO ratio is increased or decreased, and table 3 lists situations where the ratio is high or low with an explanation in terms of alterations in the pulmonary microcirculation versus changes in alveolar surface area.
Increase in the TL,NO/TL,CO ratio
TL,NO is independent of Hb level, but the TL,CO falls in anaemia; therefore, the TL,NO/TL,CO ratio, uncorrected for the Hb concentration, increases (fig. 2) [11]. Similarly, TL,NO is independent of PA,O2, but the TL,CO falls as PA,O2 increases; therefore, the TL,NO/TL,CO ratio increases. In 26 patients with pulmonary vascular disease [24] (77% had a diagnosis of chronic thromboembolic pulmonary hypertension), the TL,NO/TL,CO ratio was slightly increased (112%), but this was no more sensitive than the reduction in TL,NO, TL,CO, KNO or KCO. In a subgroup (n=36) of heavy smokers (n=236) with computed tomography (CT)-proven emphysema [27], 92% had a low KNO compared to 78% who had a low forced expiratory volume at 1 s/forced vital capacity (FVC) ratio (<0.7 being considered abnormal). The area under the receiver operating characteristic curve (ROC) (most right and least wrong: maximum=1.0) for the detection of CT-based emphysema was 0.894 for KNO and 0.822 for KCO. The negative predictive value of KNO was much greater than its positive predictive value. The TL,NO/TL,CO ratio was raised in this cohort of heavy smokers (4.9 versus 4.36), but the ratio did not differentiate between those with CT-diagnosed emphysema and those without.
Decrease in the TL,NO/TL,CO ratio
The small (∼10%) fall in the TL,NO /TL,CO ratio with exercise is consistent with an increase in pulmonary capillary diameters (increase in Vc versus DM, and fall in the DM/ΘCO·Vc ratio). Pulmonary capillary recruitment, which also occurs, increases surface area (DM) as well as Vc, and this limits the fall in the TL,NO/TL,CO ratio. With deflation of the lung in normal subjects the TL,NO/TL,CO ratio falls [13, 24], so a TL,NO/TL,CO ratio decrease should be a marker for extrapulmonary restriction.
In 25 nonsmoking patients with stage II–III sarcoidosis [29] the TL,NO/TL,CO ratio (∼KNO/KCO ratio) determined by a rebreathing technique was reduced (85% predicted) in keeping with the low DM/Vc ratio (79% normal). TL,NO was more reduced than KNO (34% pred normal versus 60%), a similar pattern to TL,CO and KCO, which suggests that loss of alveolar membrane surface area (loss of alveolar units) exceeded membrane thickening. If all ventilated units were equally involved in the membrane thickening, we would expect KNO and TL,NO, as % pred, to be equally reduced. On exercise [29], recruitment of diffusing capacity (as % of resting values) was similar for normal subjects and patients with sarcoidosis, with a decrease (-15%) in the TL,NO/TL,CO ratio, consistent with capillary dilatation on exercise, which would not be “seen” by NO diffusion. In another study of 41 patients with diffuse interstitial lung disease (66% had sarcoidosis) the TL,NO/TL,CO ratio increased [24]; we speculate that these patients may have had more end-stage disease and fibrosis.
SUGGESTIONS FOR FUTURE RESEARCH
Chronic heart failure
A reduction in DM,CO with normal or elevated Vc is a characteristic finding in chronic heart failure, at least in the early stages [35, 36]. Therefore, a decreased TL,NO/TL,CO ratio would be expected in the New York Heart Association (NYHA) grades I and II. As pulmonary hypertension intervenes in NYHA grades III and IV, the TL,NO/TL,CO ratio might return to normal or increase.
Extrapulmonary restriction
The interpretation of the TL,CO in extrapulmonary restriction is complicated by the rise in KCO (∼TL,CO/VA) to >120% pred when alveolar expansion diminishes (fig. 1b). The TL,NO/VA (∼KNO) is relatively independent of volume expansion, and this would make the interpretation of the TL,NO in extrapulmonary restriction more straightforward. In addition, the expected fall in the TL,NO/TL,CO ratio would add diagnostic usefulness to the finding of a raised KCO per se.
Interstitial lung disease
Conventionally, DM,CO and Vc are reduced equally in interstitial lung disease. table 3 shows that sarcoidosis with end-stage disease and fibrosis [24] had a raised TL,NO/TL,CO ratio, but sarcoidosis without fibrosis [29] had a reduced ratio. An increased ratio suggests that Vc is more compromised than the alveolar–capillary membranes, whereas greater membrane involvement would lead to a reduced TL,NO/TL,CO ratio. Thus, replacement of inflammation by fibrosis might be associated with a TL,NO/TL,CO ratio, which rises from normal or less than normal to a value >100% pred. Similarly, the development of vascular remodelling with pulmonary hypertension in scleroderma (systemic sclerosis), for example, might also see the TL,NO/TL,CO ratio rise above normal.
Chronic obstructive pulmonary disease
Further studies in chronic obstructive pulmonary disease, in relation to high-resolution CT quantitation of emphysema would be welcome. Studies of the ratio in bronchiectasis and obliterative bronchiolitis (e.g. post bone-marrow transplant) would be of interest.
CONCLUSION
The TL,NO is a relatively new pulmonary function test, similar in many ways to the more established TL,CO. It differs from the TL,CO in being independent of PO2 and haematocrit. Physiologically, the TL,NO behaves as if most of its transfer resistance lies in the thickness of the pulmonary membranes and blood, with red blood cell access including the binding of NO to Hb to form metHb being relatively unimportant. The TL,NO/TL,CO ratio is weighted towards the DM/Vc ratio and α, the ratio of diffusivities in plasma of NO to CO (α=1.97). The normal ratio lies between 4.3 and 4.9. The TL,NO/TL,CO ratio is reduced in extrapulmonary restriction and is predicted to be reduced in chronic heart failure. The TL,NO/TL,CO ratio is increased in interstitial and pulmonary vascular disease, and in heavy smokers, but it is not yet known if it will predict the onset of emphysema. The TL,NO/TL,CO ratio provides an alternative way of investigating the blood gas barrier and alveolar–capillary pathology.
Acknowledgments
We would like to thank N. Pride (National Heart and Lung Institute, Imperial College, London, UK) for his review of a draft of the manuscript and D. Simmonds (Medical Artist, London) for figures.
APPENDIX: see table 4
Calculations were made using the Roughton–Forster equation (1/TL=1/DM + 1/Θbl·Vc) with fixed values for 1/TL,NO (1/144) and 1/TL,CO (1/32). 1/DM,NO was calculated from 1/TL,NO on the assumption of: 1) zero red blood cell resistance (1/DM,NO = 1/TL,NO) (table 4 data sets A–C); or 2) with a red blood cell resistance equal to 37% of the total resistance (1/DM,NO = 1/TL,NO × 0.63) (table 4 data sets D–G) [10]. DM,CO was calculated from DM,NO using α, the NO/CO physical diffusivity ratio (1.97). 1/TL,CO (given) - 1/DM,CO (derived)=1/Θbl,CO·Vc, from which Vc was estimated from various equations for the 1/Θbl,CO versus PO2 relationship (at PO2 100 mmHg). The red blood cell resistance proportion for CO uptake ((1/Θbl,CO·Vc)/(1/TL,CO)) was calculated. Finally, Vc was derived from 1/Θbl,NO·Vc (= 1/TL,NO - 1/DM,NO) using the value by Carlsen and Comroe [4] for Θbl,NO (4.5 mL·min−1·mmHg−1·mL−1), (table 4 data set G).
Comment
The effect (table 4, C to B) of an increase in 1/ΘCO of one unit is to increase estimates of Vc from 47 to 99 mL (+106%) or (table 4, F to E) from 37 to 76 mL (+105%). ΘNO becoming finite (table 4, D to F) increases DM,CO, but decreases Vc by 20% (table 4, A versus D and B versus E). Even the highest values of DM,CO and Vc (117 and 99, respectively) fall short of morphometric estimates [37] at rest of Vc (180 mL) and DM,CO (463 mL min−1 mmHg−1, but corrected down to 272 mL min−1 mmHg−1 [38]). These calculations highlight the uncertainties in deriving DM,CO and Vc from simultaneous measurements of TL,NO and TL,CO.
Footnotes
Previous articles in this series: No. 1: Naeije R, Vachiery J-L, Yerly P, et al. The transpulmonary pressure gradient for the diagnosis of pulmonary vascular diseases. Eur Respir J 2013; 41: 217–223.
Statement of Interest
None declared.
- Received May 25, 2012.
- Accepted August 6, 2012.
- ©ERS 2013