Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide?

G Model

ARTICLE IN PRESS

RESPNB-2725; No. of Pages 8

Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Respiratory Physiology & Neurobiology journal homepage: www.elsevier.com/locate/resphysiol

Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Gerald S. Zavorsky a,∗ , Ivo van der Lee b a b

Department of Respiratory Therapy, Georgia State University, Atlanta, GA, United States Spaarne Hospital, Department of Pulmonary Diseases, Hoofddorp, The Netherlands

a r t i c l e

i n f o

Article history: Received 27 September 2016 Received in revised form 17 November 2016 Accepted 18 November 2016 Available online xxx Keywords: Transfer factor of the lung for CO Transfer factor of the lung for NO Lung diffusing capacity Pulmonary function tests Nitrogen monoxide

a b s t r a c t Pulmonary diffusing capacity for carbon monoxide (DLCO) has been an important pulmonary function test used since the 1950’s. It measures the uptake of CO from the alveolar space into pulmonary capillary blood, following the same path as oxygen. It’s used to evaluate/follow the progress of various lung diseases. In the eighties, a new test was developed similar to the DLCO test: pulmonary diffusing capacity for nitric oxide (DLNO). About 81–90% of the variance in DLNO is shared by DLCO in patients with cardiopulmonary disease and in healthy subjects. When DLNO is abnormally low, so is DLCO, and when DLNO is normal, so is DLCO (Kappa Statistic = 0.69, n = 251). The probability that DLNO and DLCO will be abnormally low when a cardiopulmonary disease is present (sensitivity) is 79% and 68%, respectively. The DLNO test avoids many technical issues associated with the measurement of DLCO: (1) DLNO is relatively unaffected by inspired oxygen concentration or ambient pressure, (2) DLNO is unaffected by carboxyhemoglobin, (3) DLNO is minimally affected by hemoglobin (Hb) concentration, thus correcting for Hb is not needed. (4) DLNO is more affected by lung volume compared to DLCO, thus DLNO divided by alveolar volume (KNO) is a better measure than KCO in those with restrictive lung disease, and (5) DLNO is a more stable measure over time compared to DLCO. Therefore, DLNO has several advantages over DLCO in the management of patients and could replace the DLCO test in most cases moving forward. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The determination of pulmonary diffusing capacity for carbon monoxide (DLCO) is a crucial element in the daily practice of pulmonary physicians. It is used as a screening tool for multiple pulmonary pathophysiological disorders. More than that, the DLCO is a tool in the follow up of patients with interstitial lung disease, pulmonary hypertension, obstructive lung disease, and some orphan lung disease. In subjects with “dyspnea e causa ignoti”, the combined spirometry and DLCO helps the clinician in guiding towards the right diagnostic track (Hughes and Pride, 2012). Although the DLCO measurement is recommended in daily practice, the underlying physiological principles are not fully understood by many physicians who use this measurement. The DLCO is the uptake of CO that can pass through the lung per mmHg of partial pressure per minute. In physical terms, it is a conductance;

∗ Corresponding author at: Department of Respiratory Therapy, Georgia State University, Urban Life Building, Room 1229 (12th Floor), Atlanta, GA, United States. E-mail address: [email protected] (G.S. Zavorsky).

the higher the conductance, the larger the uptake of CO. Focusing on the alveolar-capillary membrane, the flow depends on the transport as a free gas in the alveoli, the alveolar-capillary membrane, the plasma, and on its reaction with haemoglobin after its diffusion in the red blood cell. This reaction is the motor diffusion process. The carboxyhemoglobin formation is highly dependent on the presence of oxygen. Thus the interpretation of DLCO demands more reasoning than simple raw data like height and weight. This schematic description of CO transport led Roughton and Forster to split the conductance of CO into two separate conductance’s in series (Roughton and Forster, 1957). The first conductance is due to the transport of free CO through the alveolar-capillary membrane, otherwise known as alveolarcapillary membrane diffusing capacity for CO (DmCO) and the blood conductance (DbCO) characterized by the rate of reaction of CO with haemoglobin (Hb) in lung capillaries, usually ␪CO (the specific conductance in the blood for CO). The concentration of Hb in the pulmonary capillaries is proportionate to pulmonary capillary lung volume (Vc). The Vc is the clinically pertinent parameter. Thus DbCO is ␪CO·Vc provided that the concentration in Hb is normal. In its early definition, the only obstacle to CO transport as a free gas

http://dx.doi.org/10.1016/j.resp.2016.11.008 1569-9048/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model RESPNB-2725; No. of Pages 8 2

ARTICLE IN PRESS G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

was considered to be the alveolar-capillary membrane, explaining the use of “m” in DmCO. In fact the flow of CO can be hampered by a long path in distended alveoli like in emphysema and by an increase in the amount of plasma between the inner part of the capillary and the inner part of the red cell. The driving force in this diffusion process is the pressure gradient over the alveolar-capillary membrane, which is determined (on the capillary side of the membrane) by the reaction of the gas with Hb. A slow rate of reaction with Hb would favour a relative limitation of diffusion with the blood, and a high rate reaction with Hb would favour a relative limitation of the alveolar-capillary membrane. For example, oxygen has a relatively high rate of reaction with Hb in hypoxia and a low rate in hyperoxia, therefore oxygen transfer is more limited by the alveolar-capillary membrane in hypoxia than hyperoxia. If one compares nitric oxide (NO) and CO, NO would be more sensitive to an abnormality of the alveolar-capillary membrane than CO.1 Conversely, CO would be more sensitive to an abnormality of the pulmonary capillaries than NO. A quantitative analysis of the sensitivities of these transfers shows that the transfer of NO is nearly equally sensitive to membrane and blood conductance as CO is clearly mainly sensitive to the blood conductance (Martinot et al., 2015). It has been typical but incorrect to consider DLCO as a sensitive marker of a defect of the alveolar-capillary membrane. In 1983-84, the first abstracts were published describing a novel measure of assessing the transfer of a gas from the alveoli to the blood, called the gas transfer factor of the lung for nitric oxide, or TLNO (Borland et al., 1983, 1984).2 In 1987, the first peer reviewed paper on TLNO was published originating from France (Guénard et al., 1987) and the British group finally published their work soon after (Borland and Higenbottam, 1989). In the first peer-reviewed paper on DLNO, Guénard and colleagues determined that when inspiring a small amount of NO (∼8 ppm) along with the traditional diffusion mixture (0.3% CO, 10% He, 21% O2 , balance N2 ) simultaneously, one could estimate pulmonary capillary blood volume (Vc) and alveolar-capillary membrane diffusing capacity for CO (DmCO) (Guénard et al., 1987). This offered an advantage over the traditional two-step Roughton and Forster technique (Roughton and Forster, 1957), as the one-step NO-CO method could obtain Vc and DmCO in a single manoeuvre, reducing the carboxyhemoglobin build-up, and having similar gas distribution throughout the lung over the same cardiac output. By knowing a few assumptions, like the diffusivity ratio of NO to CO, and knowing the specific conductance of the blood for NO (␪NO) and CO (␪CO), and estimating the haemoglobin concentration of the patient and the alveolar oxygen pressure during a breath-hold manoeuvre (∼100 mmHg), Vc and DmCO could be calculated (See Fig. 1A and B). As the DLNO/DLCO ratio is weighted towards the DmCO/Vc ratio (Hughes and van der Lee, 2013), it has been argued that the calculation of DmCO and Vc may not be necessary and that the DLNO/DLCO ratio is a good substitute for the DmCO/Vc ratio (Hughes and van der Lee, 2013) However, it has also been argued that the DmCO and Vc components: “. . .have always been more of a physiological understanding of the pathophysiological basis of disease than a clinical interest in altering treatment. . ..clinicians have done quite well by following the DLCO as a global index for patient management in both pul-

1 NO has a ∼200 fold higher reaction velocity to hemoglobin compared to CO (Johnson et al., 1996). 2 The “transfer factor of the lung for nitric oxide” (TLNO) was European terminology. It is equivalent to the North American terminology “pulmonary diffusing capacity of the lung for nitric oxide” (DLNO). The TLNO = DLNO when expressed in the same units.

monary vascular and parenchymal diseases. Partitioning DmCO and Vc is not likely to alter treatment.” (Dr. Connie Hsia, University of Texas Southwestern Medical Center, December 2004, personal communication). Thus, if the DLNO/DLCO ratio (DmCO/Vc ratio) has not been proven to be an important component in patient management, then is either test, measured separately, equally effective in patient management? Over the past 40 years, several studies were published that demonstrated that the DLNO (or TLNO in European terminology) had several advantages over the DLCO. The purpose of this paper is to show the reader the technical/physiological advantages of measuring DLNO compared to DLCO. Furthermore, we will attempt to convince the reader that because of these technical/physiological advantages, the DLNO is a better measure of gas transfer compared to the DLCO and thus can replace the DLCO test in pulmonary function laboratories. We argue that a patient’s management of his/her pulmonary disease can be done equally well with DLNO compared to the DLCO and suggest that the DLNO test is the pulmonary function test of the future. There are a various situations where the DLNO test technically and physiologically differs in a substantial way from DLCO test. We will discuss these situations one by one, pointing out the differences and similarities of DLNO compared to DLCO. 2. Conceptual difference between DLNO and DLCO The uptake of NO molecules from alveolar sacs to the haemoglobin compasses the following steps: passing through the alveolar cell wall, followed by passing through the layer of interstitium, then moving through a layer of endothelial cells, then followed by crossing the plasma, and then crossing the red cell membrane, and finally followed by binding to the Hb molecule near the inner surface of the red cell. As the reactivity of NO with Hb is high, no NO molecule can penetrate the red cell in its depth. The DLNO differs from the DLCO in an important factor, that is, the binding of NO to the Hb is much faster (approximately 1500 times faster) than the binding of CO to Hb (Gibson and Roughton, 1957). Due to this relatively slow binding, the chief barrier to CO uptake is within the red cell (∼70–80%), and ∼25% remaining resistance to CO diffusion is located in the alveolar-capillary membrane (Fig. 1A). In contrast, ∼60% of the resistance for NO diffusion is within the alveolar-capillary membrane, while ∼40% is within the red cell interior (Fig. 1A). Therefore, the DLNO is a better representative of the diffusive properties of the alveolar-capillary membrane than the DLCO (Hughes and van der Lee, 2013). 3. Diffusion dependency on alveolar volume The relationship between the diffusion capacity and lung volume is very complex, in which three distinct associations play an important role: 3.1. DLCO versus alveolar volume The DLCO decreases with voluntary lowering of lung volume at a given breath-hold time (Stam et al., 1991). The DLCO per unit of alveolar volume (DLCO/VA ≈ KCO) increases exponentially when lung volume is lowered, as in voluntary incomplete lung expansion, or in an “extrapulmonary r¨ estrictive disease (neuromuscular disease, chest stiffness or congenital abnormality, Fig. 2). The reason for this phenomenon is the volume to surface area ratio of the lung. We can imagine a simplified model of the lung volume (ignoring the dead space volume and the conducting airways), in which the lung volume equals the total number of alveoli times the mean vol-

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model

ARTICLE IN PRESS

RESPNB-2725; No. of Pages 8

G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

3

Fig. 1. (A) The total resistance to CO and NO uptake has a membrane resistance component and a red cell resistance component. For CO, about 25% of the total resistance to CO uptake in the lung is due to the alveolar-capillary membrane while 70–80% is from the red blood cell. In contrast, the chief barrier to NO uptake in the lung is due to the alveolar-capillary membrane (60%) with the red cell resistance comprising of about 40% of the resistance to NO diffusion (Borland et al., 2010). For NO diffusion, the red cell interior is where the resistance to NO lies (Borland et al., 2014). (B) By simultaneously inspiring about 40–60 ppm NO along with a standard diffusion mixture (0.3% CO, 10% He, 21% O2 , Bal N2 ), DmCO and Vc could be calculated in this one-step NO-CO manoeuvre. Based on endorsements from the recent European Respiratory Society Task Force on DLNO, the ␪NO and ␪CO are provided, ␣ and ␬ are set, and the alveolar oxygen pressure is about 100 mmHg (Zavorsky et al., 2017).

Fig. 2. DLNO is decreased to a larger extent than the DLCO when alveolar volume is reduced similarly. The KCO increases to a larger extent than the KNO when alveolar volume is reduced similarly. Figure obtained with permission from the recent ERS standardization document on DLNO (Zavorsky et al., 2017).

ume per alveolus. In this simplified model, we consider the alveoli as spheres. The surface area of a sphere is 4·␲·r2 and the volume of a sphere 4/3·␲·r3 , which means that the ratio between volume and surface area equals r/3. This leads to the conclusion that the ratio between total alveolar volume and total alveolar surface area increases linearly with lung expansion. In other words, the bigger the inspiration, the bigger the volume per unit surface area is. This means, that when lowering the total volume by 50%, the total surface area of all spheres will be diminished by 40%. The total surface area of all spheres/alveoli is proportional with the diffusion capacity, after all the diffusive area is relative greater, which will, according Fick’s law of diffusion, lead to an increase in diffusion capacity. The DLNO, being more specific for the membrane or ‘true’ diffusing capacity, is more dependent on the VA than the DLCO. Therefore, in subjects with lung restriction (lowered surface area of all alveoli) the changes in DLNO are more easily recognizable than the changes in DLCO. 3.2. KCO and KNO versus alveolar volume The DLCO is calculated as the product of the rate constant for CO uptake and the VA, expressed in [mmol/(min × kPa)]/L, or in tradi-

tional units [mL/(min × mmHg)]/L. The KCO is commonly expressed as DLCO/VA, expressed in mmol/kPa/min/L. Sometimes clinicians use the expression that the KCO is DLCO “corrected” for lung volume. In a physiological sense this is not correct. To fully appreciate this we must first go to the measurement of the single breath DLCO, in which the subjects inhales a mixture consisting of CO, helium (or another poor soluble inert gas) with balanced air, and after a breathhold time (BHT) of 10 s, a fast expiration follows. During the BHT, the alveolar fraction of CO (FACO) decreases exponentially, provided that inspired and expired volumes are simultaneously measured, the helium is needed for the calculation of the inspired CO concentration, the FACO(t = 0) , and for the determination of the VA. The DLCO is calculated as follows: DLCO = VA/[BHT·(Pb − PH2 O)]·ln[FACO(t = 0) /FACO(t = BHT) ] In this formula, Pb stands for barometric pressure and PH2 O is the water vapour pressure. The calculation of KCO, which is DLCO/VA, leads to the value of VA (in ml STPD) in the numerator and also VA (in L BTPS) in the denominator which in fact rule each other out. Therefore, KCO does not need the actual value of VA. KCO is not DLCO corrected to VA even if KCO value is related to

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model

ARTICLE IN PRESS

RESPNB-2725; No. of Pages 8 4

G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

Fig. 3. The association between single breath DLNO and single breath DLCO in nearly 500 healthy subjects from three published studies (Aguilaniu et al., 2008; van der Lee et al., 2007; Zavorsky et al., 2008). DLNO = 4.65·(DLCO) + 3.8, R2 = 0.90, SEE = 11.8 mL/min/mmHg, p < 0.0001, with the 95% CI of the slope = 4.51–4.79. Breath-hold time was about 6 s, Inspired CO was about 0.3%, inspired NO was about 35 ppm, and inspired O2 was about 19.5%. Figure obtained with permission from the recent ERS standardization document on DLNO (Zavorsky et al., 2017). The 95% prediction bands are represented in red, the 95% CI bands are represented by the solid blue lines immediately surrounding the line of best fit.

VA (Fig. 2). Thus, the statement that the KCO is the DLCO corrected to standard lung volume or corrected to actual VA is incorrect. KCO is a rate constant for CO uptake (Hughes and Pride, 2012). The KNO (≈DLNO/VA) is much less dependent on VA than the KCO (Fig. 2), therefore it has the potential of being more useful in restrictive lung disease than the difficult to interpret KCO. Further research is necessary to explore this hypothesis. 3.3. Pulmonary capillary blood volume versus alveolar volume The relationship between the DLCO and VA is not only explained by the surface area versus volume relationship, but by the change in the amount of pulmonary capillaries perfused during the change in lung volume. As lung volume decreases, pulmonary capillary blood volume remains unchanged, therefore the pulmonary capillary blood volume per unit lung volume increases. This has a great impact on the transfer of CO, but much less impact on the transfer of NO because of the very high ␪NO value. 4. DLNO is highly related to DLCO There is a strong association between DLNO and DLCO measured in healthy humans. Based on data from ∼500 subjects that were combined from three different studies (Aguilaniu et al., 2008; van der Lee et al., 2007; Zavorsky et al., 2008), 90% of the variance in DLNO was shared by DLCO. As reported in a European Respiratory Society (ERS) Task Force document on the Standardization of DLNO (Zavorsky et al., 2017), DLNO is about 4.5–4.8 times higher than the DLCO (Fig. 3). As such, because of this tight relationship, any change in DLCO is likely reflected in the change in DLNO (Zavorsky and Lands, 2005). Even in cardiac and pulmonary disease, about 81% of the variance in DLNO was shared by DLCO (Fig. 4). In this case, DLNO is about 3.5 to 4.2 times higher than the DLCO (Fig. 4), which is smaller than in healthy normal subjects (Fig. 3). While the correlation between DLNO and DLCO is statistically different in Fig. 3 compared to Fig. 4 (z-statistic, = 3.30, p = 0.001), an 81–90% shared variance between DLNO and DLCO in diseased and healthy individuals, respectively, demonstrates that DLNO and DLCO can be used interchangeably most of the time. Furthermore, it seems that DLNO can be more sensitive in detecting small changes in gas transfer compared to DLCO

Fig. 4. The association between single breath DLNO and single breath DLCO in 120 patients from three published studies. There were 60 patients with either usual interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) or nonspecific interstitial pneumonia (NSIP) (Barisione et al., 2016), solid black triangles; 50 patients with heart failure (Magini et al., 2015), open white circles; 10 patients with chronic obstructive pulmonary disease (Moinard and Guénard, 1990), solid black circles. DLNO = 3.81·(DLCO) + 6.2, R2 = 0.81 SEE = 10.6 mL/min/mmHg, p < 0.0001, with the 95% CI of the slope = 3.48–4.15. The 95% prediction bands are represented in red, the 95% CI bands are represented by the solid blue lines immediately surrounding the line of best fit.

(Zavorsky et al., 2014). Therefore, replacing DLCO with DLNO should provide the physician with at least similar, if not better interpretation of pulmonary pathophysiology and course of treatment for patient management compared to the previous use of DLCO. 5. Changes in DLNO compared to normal is similar to changes in DLCO compared to normal Table 1 demonstrates that there are several pathophysiological conditions that results in similar decreases in DLNO and DLCO. Some diseases result in a greater percentage decrease in DLNO compared to DLCO, like idiopathic pulmonary fibrosis, cystic fibrosis (CF), sarcoidosis, chronic renal failure (CRF), type II diabetes (T2D) and chronic obstructive pulmonary disease (COPD). Other conditions such as pulmonary arterial hypertension, advanced liver cirrhosis and diffuse parenchymal lung disease (DPLD) demonstrate that there is a greater percentage drop in DLCO compared to DLNO. Nevertheless, Table 1 attests that pulmonary pathophysiology can be tracked similarly when using either DLNO or DLCO since the percentage decrease in DLNO vs DLCO compared to normal is within 14% of each other. To further illustrate this example, a scatter plot is presented in Fig. 5 displaying the tight association between mean percent reductions in DLCO compared to controls versus the mean percent reduction in DLNO compared to controls (r = 0.92, p < 0.01). For example, when there is a 20% reduction in mean DLCO compared to a group of healthy controls, there is an associated 23% reduction in mean DLNO compared to the same control group. The data presented in Fig. 5, along with Fig. 4, confirm that DLNO can replace the DLCO quite well. In order to look at the agreement in patients with cardiopulmonary disease who are below the lower limit of normal (LLN) for diffusing capacity, we examined 60 patients with either usual interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) or nonspecific interstitial pneumonia (NSIP) (Barisione et al., 2016), 50 patients with heart failure (Magini et al., 2015), and 10 patients with chronic obstructive pulmonary disease (Moinard and Guénard, 1990). From those same studies, there were 131 control subjects included. In this cohort it was found that when DLNO was abnormally low, so was DLCO, and when DLNO was normal, then so was

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model

ARTICLE IN PRESS

RESPNB-2725; No. of Pages 8

G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

5

Table 1 Changes in DLNO and DLCO with various cardiopulmonary diseases.

1.

2.

3. 4.

5. 6. 7.

8. 9. 10.

11. 12.

13.

14.

a

Study

% difference in DLNO compared to heathy subjects

% difference in DLCO compared to healthy subjects

% difference in DLNO minus% difference in DLCO

Nonspecific interstitial pneumonia (NSIP) vs healthy subjects (Barisione et al., 2016) Interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) vs healthy subjects (Barisione et al., 2016) Cystic fibrosis (CF) vs healthy subjects (Dressel et al., 2009) Chronic renal failure (CRF) vs healthy subjects (Moinard and Guénard, 1993) Sarcoidosis vs normals (Phansalkar et al., 2004)a Type II diabetes (T2D) vs healthy subjects (Chance et al., 2008)a Chronic obstructive pulmonary disease (COPD) vs healthy subjects (Moinard and Guénard, 1990) Pulmonary arterial hypertension vs healthy subjects (Farha et al., 2013) Early emphysema healthy subjects (van der Lee et al., 2009) Diffuse parenchymal lung disease (DPLD) vs healthy subjects (van der Lee et al., 2006) Cystic Fibrosis (CF) vs healthy subjects (Wheatley et al., 2011)a Advanced liver cirrhosis (HPS) vs healthy subjects (Degano et al., 2009) Pulmonary arterial hypertension vs healthy subjects (van der Lee et al., 2006) Heart failure (HF) vs healthy subjects (Magini et al., 2015) Mean (SD) decrease compared to normal subjects

48% less in NSIP

34% less in NSIP

+14

68% less in UIP-IPF

56% less in UIP-IPF

+12

29% less in CF

17% less in CF

+12

45% less in CRF

35% less in CRF

+10

66% less in sarcoidosis

57% less in sarcoidosis

+9

18% less in T2D, BMI < 30) and 13% less in T2D, BMI < 30) 61% less in COPD

11% less in T2D, BMI < 30 and 7% less in T2D, BMI > 30 56% less in COPD

+7 +6 +5

30% less in PAH

29% less in PAH

+1

13% less in those with early emphysema 42% less in DPLD

13% less in those with early emphysema 45% less in DPLD

0

19% less in CF

22% less in CF

−3

28% less in HPS

34% less in HPS

−6

42% less in PAH

49% less in PAH

−7

18% less in HF

26% less in HF

−8

36% (19%)

33% (17%)

−3

Studies that used rebreathing. All other studies used the single-breath technique.

DLCO (substantial agreement,3 Kappa Statistic = 0.69, 95% bootstrapped CI = 0.61–0.78, p < 0.001, n = 251). Abnormally low was defined as below the LLN (below the 2.5th percentile) determined by prediction equations published in the European Respiratory Society (ERS) Task Force document on the Standardization of DLNO (Zavorsky et al., 2017). The probability that DLNO and DLCO is abnormally low when a cardiopulmonary disease is present (sensitivity) was 79% and 69%, respectively (Tables 2, 3). This implies that a DLNO test detects more patients with a cardiopulmonary disease compared to a DLCO test. The probability that DLNO and DLCO is normal when a cardiopulmonary disease is not present (specificity) was 82 and 97%, respectively (Tables 2, 3). 6. DLNO can better quantify the extent of disease than DLCO There are some preliminary studies that demonstrate that DLNO can be better than the DLCO in quantifying the extent of disease compared to DLCO. In patients with cystic fibrosis (CF), there was a similar shared variance between DLNO and a CF-specific computed tomography (CT) score [R2 = 0.69 between DLNO (in z-scores) and CT score; R2 = 0.62 between DLCO (in z-scores) and CT score] (Dressel et al., 2009). However, when the DLNO per unit of alveolar volume (KNO) was used instead, 40% of the variance in KNO (in z-

3 The following are standards for strength of agreement for the kappa coefficient: ≤0 = poor, 0.01–0.20 = slight, 0.21–0.40 = fair, 0.41–0.60 = moderate, 0.61–0.80 = substantial, and 0.81–1 = almost perfect. Landis and Koch (1977).

scores) was associated with the CF-specific CT score, whereas there was no shared variance between DLCO per unit alveolar volume (KCO, in z-scores) and the CF-specific CT score (Dressel et al., 2009). When comparing these two correlations coefficients (r = −0.63 vs r = 0.01, n = 21), this was statistically significant (z-statistic = −2.25, p = 0.02). Thus KNO follows CF-specific CT changes better than KCO. In those with usual interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) and nonspecific interstitial pneumonia (NSIP), both DLNO (z-scores) better correlated with the visual assessment of fibrosis compared to DLCO (r = −0.81 vs r = −0.69, n = 60, statistically significant between two correlations, zstatistic = 2.01, p = 0.04) (Barisione et al., 2016). Thus, DLNO provides a more sensitive evaluation of fibrotic changes than DLCO because ∼60% of the resistance for NO diffusion is within the alveolar-capillary membrane while ∼40% is within the red cell interior (Borland et al., 2014). In contrast, the chief barrier to CO uptake is within the red cell (∼70–80%), and ∼25% remaining resistance to CO diffusion is located in the alveolar-capillary membrane (Zavorsky et al., 2017). In a cohort of patients with pulmonary arterial hypertension (PAH), there was a better ability to detect decreases in pulmonary diffusing capacity over time when using DLNO compared to DLCO (Farha et al., 2013). One study examined whether DLNO was more sensitive in diagnosing emphysema in heavy smokers compared to DLCO (van der Lee et al., 2009). Out of 750 examinations, emphysema was diagnosed from CT scanning in 36 patients. In those patients, 92% that had the disease tested positive (sensitivity) using KNO, while

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model RESPNB-2725; No. of Pages 8 6

ARTICLE IN PRESS G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx Table 2 Cross tabulation of the number of patients and healthy controls that had normal and abnormal values for single-breath DLNO from three published studies. Cardiopulmonary Disease present? Abnormally low DLNO? Yes

Yes

No

Total

95 (38%)

119 (47%)

No

25 (10%) (false negatives) (Type II error) 120 (48%) Sensitivity (the proportion of positives correctly identified) 95/120 79% (95% CI = 71–86%)

24 (10%) (false positives) (Type 1 error) 107 (43%)

131 (52%) Specificity (the proportion of negatives correctly identified) 107/131 82% (95% CI = 74–88%)

251

Total

Fig. 5. The association between the mean percent decrease in DLCO compared to healthy controls versus the mean percent decrease in DLNO compared to healthy controls across thirteen different studies. There was an approximate 83% shared variance between the two variables (r = 0.92, with the 95% CI of the r using 1000 bootstrapped samples = 0.81–0.97, p < 0.001). The 95% CI bands are represented by the solid blue lines immediately surrounding the line of best fit. The regression equation is: % decrease in DLNO compared to healthy controls = 1.02·(% decrease in DLCO compared to healthy controls) + 2.59, Adjusted R2 = 0.83, SEE = 7.8%, 95% CI of the slope = 0.8%–1.3%, p < 0.01. As one can see, when there is a 20% mean reduction in DLCO compared to healthy controls, there is a concomitant 23% reduction in mean DLNO compared to healthy controls. Nonspecific interstitial pneumonia (NISP) and usual interstitial pneumoniaidiopathic pulmonary fibrosis (IUP-IPF) (Barisione et al., 2016); cystic fibrosis (CF) (Dressel et al., 2009; Wheatley et al., 2011); Chronic renal failure (CRF) (Moinard and Guénard, 1993); Sarcoidosis (Phansalkar et al., 2004); type II diabetes with a body mass index < 30 (T2D, BMI < 30) and > 30 (T2D, BMI > 30) (Chance et al., 2008); chronic obstructive pulmonary disease (COPD) (Moinard and Guénard, 1990); pulmonary arterial hypertension (PAH)(Farha et al., 2013; van der Lee et al., 2006); early emphysema (van der Lee et al., 2009); diffuse parenchymal lung disease (DPLD)(van der Lee et al., 2009); advanced liver cirrhosis (HPS) (Degano et al., 2009); heart failure (HF) (Magini et al., 2015). All studies used the single breath technique except for three of them: (Chance et al., 2008; Phansalkar et al., 2004; Wheatley et al., 2011).

89% that had the disease tested positive when using KCO. In contrast, 73% of patients without disease tested negative (specificity) when using KNO, compared to 57% when using KCO. Finally, 35% of patients who tested positive for emphysema actually had the disease (positive predictive value, or precision rate) when using KNO compared to 25% when using KCO. While the precision rate may not be satisfactory when using KNO, it is still higher than when using KCO in the diagnosis of emphysema. Indeed, the data presented in Tables 2, 3 show that the sensitivity in identifying disease seems to be better when using DLNO compared to DLCO. 7. DLNO, unlike the DLCO, is relatively unaffected by the alveolar oxygen pressure The DLCO is affected by the alveolar oxygen pressure (PA O2 ) due to the inspired oxygen concentration. As CO binds competitively with oxygen on Hb, an increase in inspired oxygen concentration (or alveolar PO2 ) with decrease DLCO and vice versa. The DLCO decreases by about 1.5% per 1% increase in inspired oxygen concentration both humans and animals (Borland and Cox, 1991; Crapo et al., 1988; Guénard et al., 2016; Meyer et al., 1990; Piiper et al., 1988). However, DLNO is relatively unaffected by the PA O2 . Tamhane and colleagues compared the DLNO after breathing 100% oxygen to that after breathing 30% oxygen (Tamhane et al., 2001). There was no appreciable difference in DLNO between those two conditions. Others have also shown that DLNO is not appreciably affected by the inspired oxygen concentration in both human and animal studies (Borland and Cox, 1991; Guénard et al., 2016; Meyer et al., 1990; Piiper et al., 1988). Nevertheless, due to changes in the DLCO with varying inspired oxygen concentrations, the DLNO to

132 (53%)

There were 120 patients with cardiopulmonary disease and 131 individuals that were classified as healthy controls. The patients with cardiopulmonary disease were 60 patients with either usual interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) or nonspecific interstitial pneumonia (NSIP) (Barisione et al., 2016), 50 patients with heart failure (Magini et al., 2015), 10 patients with chronic obstructive pulmonary disease (Moinard and Guénard, 1990). Pearson Chi-Square = 93.0, p < 0.001, Kappa Statistic = 0.61, 95% CI = 0.51-0.70, n = 251). Abnormally low DLNO was defined as a value below the lower limit of normal (<2.5th percentile) according to sex, age, height (Zavorsky et al., 2017). Normal DLNO are values equal to or more than the LLN (≥2.5th percentile). The positive likelihood ratio was 4.3 (95% CI = 3.0–6.3), which is the ratio between the probability of a positive test result given the presence of cardiopulmonary disease and the probability of a positive test result given the absence of cardiopulmonary disease. The negative likelihood ratio was 0.26 (95% CI = 0.18–0.36), which is the ratio between the probability of a negative test result given the presence of cardiopulmonary disease and the probability of a negative test result given the absence of cardiopulmonary disease. Since the sample sizes in the disease present and the disease absent groups do not reflect the real prevalence of cardiopulmonary disease, the positive and negative predicted values cannot be estimated.

DLCO ratio decreases from 0.4% to 2.2% (mean decrease = 1.2%) for every 1% decrease in inspired oxygen concentration (Borland and Cox, 1991; Guénard et al., 2016; Meyer et al., 1990; Piiper et al., 1988). 8. DLNO, unlike the DLCO, is not affected by carboxyhemoglobin concentration Unlike the DLCO, the DLNO remains unchanged with the build-up of carboxyhemoglobin (COHb) (Zavorsky, 2013). After 22 consecutive 5 s breath-hold DLNO-DLCO tests, in which the COHb increased from a baseline value of 1.2% (SD 0.5%) to 11.1% (1.4%), the DLNO remained unaltered. Furthermore, over 22 tests, the DLCO decreased by about 4 mL/min/mmHg, which is about 0.4–0.5% increase in COHb per test and a reduction in DLCO by 0.3–0.4 mL/min/mmHg per test (Zavorsky, 2013). Furthermore, the build-up of methemoglobin is minimal when performing many DLNO tests (Zavorsky, 2013). Other research has demonstrated that several inhalations of 40–60 ppm NO does not affect DLNO (Tamhane et al., 2001; Zavorsky and Murias, 2006). 9. DLNO, unlike the DLCO is unaffected by hemoglobin concentration The DLCO should be corrected to a standard Hb concentration in subjects with abnormal Hb concentration. The DLNO is independent of Hb concentration in vivo (van der Lee et al., 2005), therefore when using the DLNO without accompanying DLCO measurement, one could refrain from measuring Hb concentration. The caveat here is that if one is performing a diffusing capacity test on a patient with the goal of strictly testing for anaemia or poly-

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model

ARTICLE IN PRESS

RESPNB-2725; No. of Pages 8

G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx Table 3 Cross tabulation of the number of patients and healthy controls that had normal and abnormal values for single-breath DLCO from three published studies.

Table 4 Technical, physiological, and diagnostic reasons of why a DLNO test is equal to or better than the traditional DLCO test.

Cardiopulmonary Disease present? Abnormally low DLCO? Yes

Yes

No

Total

81 (32%)

85 (34%)

No

39 (16%) (false negatives) (Type II error) 120 (48%) Sensitivity (the proportion of positives correctly identified) 81/120 68% (95% CI = 58–76%)

4 (2%) (false positives) (Type 1 error) 127 (51%)

131 (52%) Specificity (the proportion of negatives correctly identified) 127/131 97% (95% CI = 92–99%)

251

Total

Reason

Reference

1.

90% of the variance in DLNO is shared by DLCO in healthy subjects

2.

81% of the variance in DLNO is shared by DLCO in patients with cardiopulmonary disease

3.

DLNO can better quantify the extent of cardiopulmonary disease compared to DLCO, especially for cystic fibrosis, pulmonary fibrosis, emphysema, and pulmonary arterial hypertension The probability that DLNO will be abnormally low when a cardiopulmonary disease is present (sensitivity) is 79%. This is a slightly better sensitivity compared to DLCO (68%) When DLNO is abnormally low, so is DLCO, and when DLNO is normal, so is DLCO (substantial agreement, Kappa Statistic = 0.69, n = 251) There is similar agreement between the presence or absence of a cardiopulmonary disease and whether or not DLNO or DLCO is abnormally low (substantial agreement, Kappa Statistic = 0.61–0.65, n = 251) DLNO is more affected by lung volume compared to DLCO, thus DLNO divided by alveolar volume (KNO) is a better measure than KCO in those with restrictive lung disease DLNO, unlike the DLCO, is relatively unaffected by the inspired oxygen concentration (alveolar oxygen pressure)

(Aguilaniu et al., 2008; van der Lee et al., 2007; Zavorsky et al., 2008). See Fig. 3. (Barisione et al., 2016; Magini et al., 2015; Moinard and Guénard, 1990). See Fig. 4. (Barisione et al., 2016; Dressel et al., 2009; Farha et al., 2013; van der Lee et al., 2009)

166 (66%)

There were 120 patients with cardiopulmonary disease and 131 individuals that were classified as healthy controls. The patients with cardiopulmonary disease were 60 patients with either usual interstitial pneumonia-idiopathic pulmonary fibrosis (UIP-IPF) or nonspecific interstitial pneumonia (NSIP) (Barisione et al., 2016), 50 patients with heart failure (Magini et al., 2015), 10 patients with chronic obstructive pulmonary disease (Moinard and Guénard, 1990). Pearson Chi-Square = 116.1, p < 0.001, Kappa Statistic = 0.65, 95% CI = 0.56–0.74, n = 251). Abnormally low DLCO was defined as a value below the lower limit of normal (<2.5th percentile) according to sex, age, height (Zavorsky et al., 2017). Normal DLCO are values equal to or more than the LLN (≥2.5th percentile). The positive likelihood ratio was 22.1 (95% CI = 8.4–58.5), which is the ratio between the probability of a positive test result given the presence of cardiopulmonary disease and the probability of a positive test result given the absence of cardiopulmonary disease. The negative likelihood ratio was 0.34 (95% CI = 0.26–0.43), which is the ratio between the probability of a negative test result given the presence of cardiopulmonary disease and the probability of a negative test result given the absence of cardiopulmonary disease. Since the sample sizes in the disease present and the disease absent groups do not reflect the real prevalence of the disease, the positive and negative predicted values cannot be estimated.

cythaemia, then the DLNO test would not be able to detect anaemia or polycythaemia. This insensivity of DLNO to Hb concentration is due to the high reactivity of NO within the red cell. Owing to this reactivity, NO molecules passing through the red cell membrane are rapidly captured and a few part of the Hb molecules pool participate to this capture unless the total concentration in Hb molecules is severly decreased (about half the normal).

4.

5.

6.

7.

8.

9.

10.

11.

10. The reproducibility of DLNO is tighter than the DLCO The within session repeatability for DLNO is about 17 mL/min/mmHg (Zavorsky and Murias, 2006),4 while the weekto-week reproducibility is about 20 mL/min/mmHg (Murias and Zavorsky, 2007).5 Thus, the difference is about 3 mL/min/mmHg, or ∼15%. In comparison, the within session repeatability for DLCO is about 3.2 mL/min/mmHg (Zavorsky and Murias, 2006), while the week-to-week reproducibility is about 4.9 mL/min/mmHg (Murias and Zavorsky, 2007). Thus, the difference is about 1.7 mL/min/mmHg, or ∼35%. This demonstrates that DLNO is a more stable measure over time compared to the DLCO and that the majority of the variation in DLNO is within sessions and not between sessions. Therefore, the DLNO is likely more precise to detect a change in diffusing capacity compared to the DLCO (Murias and Zavorsky, 2007).

4 Repeatability: The difference between two trials measured on the same subject in the same testing session is expected to be less than 17 mL/min/mmHg for DLNO and 3.2 mL/min/mmHg for DLCO, for 95% of observations. 5 Reproducibility: The difference in DLNO and DLCO measured on the same subject over two different weeks is expected to be less than 20 mL/min/mmHg and 4.9 mL/min/mmHg for DLCO, respectively, for 95% of observations.

7

DLNO, unlike the DLCO, is relatively unaffected by carboxyhaemoglobin concentration DLNO, unlike the DLCO, is relatively unaffected by haemoglobin concentration, thus no adjustment for it is necessary DLNO is a more stable measure over time compared to the DLCO and the majority of the variation in DLNO is within sessions and not between sessions.

(Barisione et al., 2016; Magini et al., 2015; Moinard and Guénard, 1990). See Tables 2, 3.

(Barisione et al., 2016; Magini et al., 2015; Moinard and Guénard, 1990). (Barisione et al., 2016; Magini et al., 2015; Moinard and Guénard, 1990). See Tables 2, 3.

(Zavorsky et al., 2017). See Fig. 2.

(Borland and Cox, 1991; Guénard et al., 2016; Meyer et al., 1990; Piiper et al., 1988; Tamhane et al., 2001). (Zavorsky, 2013)

(van der Lee et al., 2005)

(Murias and Zavorsky, 2007)

11. Conclusion Pulmonary diffusing capacity for CO and DLNO depend on the same components, alveolar-capillary membrane conductance and the blood conductance, but in different proportions. Pulmonary diffusing capacity for CO is mainly sensitive to the pulmonary microcirculation, and DLNO is more sensitive to the alveolarcapillary membrane component of diffusion. Here, we provide evidence that DLNO can provide equal information (if not better information) for the physician in patient management and suggest that DLNO can replace DLCO as the pulmonary function test of the future. The probability that DLNO is abnormally low when a cardiopulmonary disease is present (sensitivity) is 79% and is slightly better than that for DLCO (69%), respectively. There is similar agreement between the presence or absence of a cardiopulmonary disease and whether or not the DLNO or DLCO is abnormally low (Tables 2, 3). Pulmonary diffusing capacity for nitric oxide is a more specific measure of the function of the alveolar-

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008

G Model RESPNB-2725; No. of Pages 8

ARTICLE IN PRESS G.S. Zavorsky, I. van der Lee / Respiratory Physiology & Neurobiology xxx (2016) xxx–xxx

8

capillary membrane, but as presented in Fig. 4, there is a strong association between DLNO and DLCO in disease. Furthermore, there are similar decreases in DLCO and DLNO in many pathophysiological conditions compared to healthy normal subjects (Table 1, Fig. 5). Due to several advantages of DLNO compared to DLCO (Table 4), we suggest that DLNO can replace DLCO for the assessment of gas transfer though the lung. Acknowledgments A synopsis of this paper was presented by G.S. Zavorsky at the 2016 American Association of Respiratory Care Conference in San Antonio, Texas. The title of the talk on page 54 of the Final Program was: “Pulmonary Diffusing Capacity for Nitric Oxide is the Pulmonary Function Test of the Future.” We would like to thank Giovanni Barisione, Alessandra Magini, and their colleagues for generously donating their individual subject data from their previously published papers (Barisione et al., 2016; Magini et al., 2015). Their data was included in Tables 2, 3, and Fig. 4 in this manuscript. References Aguilaniu, B., Maitre, J., Glenet, S., Gegout-Petit, A., Guenard, H., 2008. European reference equations for CO and NO lung transfer. Eur. Respir. J. 31, 1091–1097. Barisione, G., Brusasco, C., Garlaschi, A., Baroffio, M., Brusasco, V., 2016. Lung diffusing capacity for nitric oxide as a marker of fibrotic changes in idiopathic interstitial pneumonias. J. Appl. Physiol. 120, 1029–1038. Borland, C., Cox, Y., 1991. NO and CO transfer. Eur. Respir. J. 4, 766. Borland, C.D., Higenbottam, T.W., 1989. A simultaneous single breath measurement of pulmonary diffusing capacity with nitric oxide and carbon monoxide. Eur. Respir. J. 2, 56–63. Borland, C., Chamberlain, A., Higenbottam, T., 1983. The fate of inhaled nitric oxide [abstract]. Clin. Sci. (Lond.) 65, 37P. Borland, C., Cracknell, N., Higenbottam, T., 1984. Is the measurment of DLNO a true measure of membrane diffusing capacity? [abstract]. Clin. Sci. (Lond.) 67 (S9), 41. Borland, C.D., Dunningham, H., Bottrill, F., Vuylsteke, A., Yilmaz, C., Dane, D.M., Hsia, C.C., 2010. Significant blood resistance to nitric oxide transfer in the lung. J. Appl. Physiol. 108, 1052–1060. Borland, C.D., Bottrill, F., Jones, A., Sparkes, C., Vuylsteke, A., 2014. The significant blood resistance to lung nitric oxide lies within the red cell. J. Appl. Physiol. 116, 32–41. Chance, W.W., Rhee, C., Yilmaz, C., Dane, D.M., Pruneda, M.L., Raskin, P., Hsia, C.C., 2008. Diminished alveolar microvascular reserves in type 2 diabetes reflect systemic microangiopathy. Diabetes Care 31, 1596–1601. Crapo, R.O., Kanner, R.E., Jensen, R.L., Elliott, C.G., 1988. Variability of the single-breath carbon monoxide transfer factor as a function of inspired oxygen pressure. Eur. Respir. J. 1, 573–574. Degano, B., Mittaine, M., Guenard, H., Rami, J., Garcia, G., Kamar, N., Bureau, C., Peron, J.M., Rostaing, L., Riviere, D., 2009. Nitric oxide and carbon monoxide lung transfer in patients with advanced liver cirrhosis. J. Appl. Physiol. 107, 139–143. Dressel, H., Filser, L., Fischer, R., Marten, K., Muller-Lisse, U., de la Motte, D., Nowak, D., Huber, R.M., Jorres, R.A., 2009. Lung diffusing capacity for nitric oxide and carbon monoxide in relation to morphological changes as assessed by computed tomography in patients with cystic fibrosis. BMC Pulm. Med. 9, 30. Farha, S., Laskowski, D., George, D., Park, M.M., Tang, W.H., Dweik, R.A., Erzurum, S.C., 2013. Loss of alveolar membrane diffusing capacity and pulmonary capillary blood volume in pulmonary arterial hypertension. Respir. Res. 14, 6. Gibson, Q.H., Roughton, F.J., 1957. The kinetics and equilibria of the reactions of nitric oxide with sheep haemoglobin. J. Physiol. 136, 507–526. Guénard, H., Varene, N., Vaida, P., 1987. Determination of lung capillary blood volume and membrane diffusing capacity in man by the measurements of NO and CO transfer. Respir. Physiol. 70, 113–120. Guénard, H.J., Martinot, J.B., Martin, S., Maury, B., Lalande, S., Kays, C., 2016. In vivo estimates of NO and CO conductance for haemoglobin and for lung transfer in humans. Respir. Physiol. Neurobiol. 228, 1–8. Hughes, J.M., Pride, N.B., 2012. Examination of the carbon monoxide diffusing capacity (DLCO) in relation to its KCO and VA components. Am. J. Respir. Crit. Care Med. 186, 132–139.

Hughes, J.M., van der Lee, I., 2013. The TL,NO/TL,CO ratio in pulmonary function test interpretation. Eur. Respir. J. 41, 453–461. Johnson, J., Heigenhauser, R.L., Hsia, G.J., Jones, C.C.W., Wagner, N.L., P.D, 1996. Determinants of gas exchange and acid-base balance during exercise. In: Rowell, L.B., Shepherd, J.T. (Eds.), Handbook of Physiology, Section 12: Exercise: Regulation and Integration of Multiple Systems. American Physiological Society, New York, N.Y, pp. 515–584. Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for categorical data. Biometrics 33, 159–174. Magini, A., Apostolo, A., Salvioni, E., Italiano, G., Veglia, F., Agostoni, P., 2015. Alveolar-capillary membrane diffusion measurement by nitric oxide inhalation in heart failure. Eur. J. Prev. Cardiol. 22, 206–212. Martinot, J.B., Guenard, H., Dinh-Xuan, A.T., Gin, H., Dromer, C., 2015. Nitrogen monoxide and carbon monoxide transfer interpretation: state of the art. Clin. Physiol. Funct. Imaging. Meyer, M., Schuster, K.D., Schulz, H., Mohr, M., Piiper, J., 1990. Pulmonary diffusing capacities for nitric oxide and carbon monoxide determined by rebreathing in dogs. J. Appl. Physiol. 68, 2344–2357. Moinard, J., Guénard, H., 1990. Determination of lung capillary blood volume and membrane diffusing capacity in patients with COLD using the NO-CO method. Eur. Respir. J. 3, 318–322. Moinard, J., Guénard, H., 1993. Membrane diffusion of the lungs in patients with chronic renal failure. Eur. Respir. J. 6, 225–230. Murias, J.M., Zavorsky, G.S., 2007. Short-term variability of nitric oxide diffusing capacity and its components. Respir. Physiol. Neurobiol. 157, 316–325. Phansalkar, A.R., Hanson, C.M., Shakir, A.R., Johnson Jr., R.L., Hsia, C.C., 2004. Nitric oxide diffusing capacity and alveolar microvascular recruitment in sarcoidosis. Am. J. Respir. Crit. Care Med. 169, 1034–1040. Piiper, J., Schuster, K.D., Mohr, M., Schulz, H., Meyer, M., 1988. Pulmonary diffusing capacity for carbon monoxide and nitric oxide. Adv. Exp. Med. Biol. 222, 491–495. Roughton, F.J., Forster, R.E., 1957. Relative importance of diffusion and chemical reaction rates in determining rate of exchange of gases in the human lung, with special reference to true diffusing capacity of pulmonary membrane and volume of blood in the lung capillaries. J. Appl. Physiol. 11, 290–302. Stam, H., Kreuzer, F.J., Versprille, A., 1991. Effect of lung volume and positional changes on pulmonary diffusing capacity and its components. J. Appl. Physiol. 71, 1477–1488. Tamhane, R.M., Johnson Jr., R.L., Hsia, C.C., 2001. Pulmonary membrane diffusing capacity and capillary blood volume measured during exercise from nitric oxide uptake. Chest 120, 1850–1856. Wheatley, C.M., Foxx-Lupo, W.T., Cassuto, N.A., Wong, E.C., Daines, C.L., Morgan, W.J., Snyder, E.M., 2011. Impaired lung diffusing capacity for nitric oxide and alveolar-capillary membrane conductance results in oxygen desaturation during exercise in patients with cystic fibrosis. J. Cyst. Fibros. 10, 45–53. Zavorsky, G.S., Lands, L.C., 2005. Lung diffusion capacity for nitric oxide and carbon monoxide is impaired similarly following short-term graded exercise. Nitric Oxide 12, 31–38. Zavorsky, G.S., Murias, J.M., 2006. A small amount of inhaled nitric oxide does not increase lung diffusing capacity. Eur. Respir. J. 27, 1251–1257. Zavorsky, G.S., Cao, J., Murias, J.M., 2008. Reference values of pulmonary diffusing capacity for nitric oxide in an adult population. Nitric Oxide 18, 70–79. Zavorsky, G.S., Milne, E.N., Lavorini, F., Rienzi, J.P., Cutrufello, P.T., Kumar, S.S., Pistolesi, M., 2014. Small changes in lung function in runners with marathon-induced interstitial lung edema. Physiol. Rep. 2, e12056. Zavorsky, G.S., Hsia, C.C.W., Hughes, J.M.B., Borland, C.D.R., Guenard, H.J., Van der Lee, I., Steenbruggen, I., Naeije, R., Cao, J., Dinh-Xuan, A.T., 2017. Standardization and application of the single-breath determination of nitric oxide uptake in the lung. Eur. Respir. J. (in press). Zavorsky, G.S., 2013. The rise in carboxyhemoglobin from repeated pulmonary diffusing capacity tests. Respir. Physiol. Neurobiol. 186, 103–108. van der Lee, I., Zanen, P., Biesma, D.H., van den Bosch, J.M., 2005. The effect of red cell transfusion on nitric oxide diffusing capacity. Respiration 72, 512–516. van der Lee, I., Zanen, P., Grutters, J.C., Snijder, R.J., van den Bosch, J.M., 2006. Diffusing capacity for nitric oxide and carbon monoxide in patients with diffuse parenchymal lung disease and pulmonary arterial hypertension. Chest 129, 378–383. van der Lee, I., Zanen, P., Stigter, N., van den Bosch, J.M., Lammers, J.W., 2007. Diffusing capacity for nitric oxide: reference values and dependence on alveolar volume. Respir. Med. 101, 1579–1584. van der Lee, I., Gietema, H.A., Zanen, P., van Klaveren, R.J., Prokop, M., Lammers, J.W., van den Bosch, J.M., 2009. Nitric oxide diffusing capacity versus spirometry in the early diagnosis of emphysema in smokers. Respir. Med. 103, 1892–1897.

Please cite this article in press as: Zavorsky, G.S., van der Lee, I., Can the measurement of pulmonary diffusing capacity for nitric oxide replace the measurement of pulmonary diffusing capacity for carbon monoxide? Respir. Physiol. Neurobiol. (2016), http://dx.doi.org/10.1016/j.resp.2016.11.008