|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Heart Lung Centre Utrecht, Department of Pulmonary Diseases, St. Antonius Hospital, Nieuwegein, the Netherlands.
Correspondence to: Ivo van der Lee, MD, Department of Pulmonary Diseases, Heart Lung Centre Utrecht, St. Antonius Hospital, PO Box 2500, 3430 EM Nieuwegein, the Netherlands; email: vdlee{at}tiscali.nl
Abstract
Background: The passage of carbon monoxide (CO) through the alveolocapillary membrane and into the plasma and intraerythrocytic compartments determines the diffusing capacity of the lung for CO (DLCO) as defined by the Roughton and Forster equation. On the other hand, the single-breath diffusing capacity of the lung for nitric oxide (DLNO) is thought to represent the true membrane diffusing capacity because of its very high affinity for hemoglobin (Hb) and its independence from pulmonary capillary blood volume. Therefore, the DLNO/DLCO ratio can be used to differentiate between thickened alveolocapillary membranes (both DLNO and DLCO are decreased, and the DLNO/DLCO ratio is normal) and decreased perfusion of ventilated alveoli (the DLNO less decreased than the DLCO; therefore, the DLNO/DLCO ratio is high) in patients with pulmonary disease.
Study design: We measured the combined values of DLCO and DLNO in 41 patients with diffuse parenchymal lung disease (DPLD), 26 patients with pulmonary arterial hypertension (PAH), and 71 healthy subjects.
Results: The DLCO (corrected to the standard Hb value) was lowered in the DPLD group (64% of predicted) and in the PAH group (64% of predicted), and was normal in the control group (105% of predicted). The DLNO/DLCO ratio in patients with PAH (4.98) was significantly higher than that in patients with DPLD (4.56) and in healthy subjects (4.36).
Conclusion: The DLNO/DLCO ratio is significantly higher in patients with PAH than in healthy subjects, although this ratio cannot be applied as a screening test to discriminate between patients with DPLD and PAH as the overlap between these groups is too large.
Key Words: interstitial lung disease pulmonary diffusing capacity pulmonary hypertension
The diffusing capacity of the lung for carbon monoxide (DLCO) estimates the amount of gas uptake by the lungs and is a valuable tool in the assessment of pulmonary diseases. According to the model of Roughton and Forster,1 DLCO is determined by the following several factors: the passage of carbon monoxide (CO) through the alveolocapillary membrane; the transfer of CO into the plasma and the intraerythrocytic compartments; and the reaction rate for the binding of CO on hemoglobin (Hb). Equation 1 enables the estimation of the two components of the DLCO (ie, the diffusing capacity of the alveolocapillary membrane for CO [DMCO] and the pulmonary capillary blood volume [Vcap]) using duplicate measurements of the DLCO with high and low oxygen concentrations.1
![]() | (1) |
Although numerous articles have been published concerning the clinical value of the DMCO and Vcap, this test has not become a standard tool in the pulmonary function laboratory for the following reasons: first, the measurement of the DMCO is complicated and time-consuming; and, second, measurements of DMCO and Vcap are not entirely accurate. The reason for this is that the
CO value in the Roughton and Forster1 equation is an estimate, and the DMCO value is determined by two separate breathholding periods, with high and low oxygen concentrations, in which several factors can influence the measurement.2
Nitric oxide (NO) offers a solution to these problems and can be used as a gas in testing the diffusing capacity of the lung. NO binds 400 times stronger than CO to Hb3; therefore, the diffusing capacity of lung for NO (DLNO) is much less influenced by changes in the Vcap and reflects the properties of the alveolocapillary membrane better than the DLCO. Borland and Higenbottam4 measured DLNO and DLCO in one single-breath maneuver, and calculated the DMCO and Vcap using differences in the
CO and the
NO values.5 Several interesting studies have been conducted in this field. Phansalkar et al6 showed that the DLNO (using the rebreathing technique) closely relates to the DMCO measured by the high/low oxygen method. Tamhane et al3 measured the combined DLNO-DLCO values with a rebreathing technique and found that the DLNO directly correlates to the diffusing capacity of the pulmonary membrane. Thus the DLNO/DLCO ratio should be able to locate the position of the diffusion impairment. Assuming that the DLNO is not affected or is less affected by impaired capillary filling, and thus represents the diffusing capacity of the pulmonary membrane, the DLNO/DLCO ratio has to differ between subjects with a pure alveolocapillary membrane disturbance and subjects with microvascular disease. Subjects with a decreased DLCO due to lowering of the Vcap component will have undisturbed DLNO; therefore, the DLNO/DLCO ratio will increase. Subjects with a decreased DLCO due to thickening of the membrane without disturbing the pulmonary capillaries will also have a lower DLNO, so the DLNO/DLCO ratio will not alter. The study by Harris et al7 demonstrated in sheep that the occlusion of one pulmonary artery increased the DLNO/DLCO ratio by decreasing the DLCO while the DLNO remained constant. This effect is caused by the increase in CO backpressure in stagnant capillaries. The authors concluded that the DLCO has a much greater sensitivity than DLNO in detecting a regional reduction in capillary blood flow. The aim of this study was to test the hypothesis that the DLNO/DLCO ratio significantly differs between patients with diffusion impairment due to fibrotic disease and patients with diffusion impairment due to pulmonary vascular disease.
Materials and Methods
Patients
Subjects were recruited from the pulmonary outpatient clinic of our hospital. This study was approved by the local ethics committee. Inclusion criteria consisted of a definitive diagnosis of diffuse parenchymal lung disease (DPLD)8 or pulmonary arterial hypertension (PAH) according to the Revised Clinical Classification of Pulmonary Hypertension.9 A control group of healthy nonsmoking subjects was recruited from among hospital personnel.
Diagnostic Procedure
All patients were extensively investigated by experienced pulmonologists; the standard procedure consisted of a medical history, a physical examination, laboratory investigations, a chest radiograph, a high-resolution CT scan of the lungs, spirometry, whole-body plethysmography (6200 Autobox DL; SensorMedics; Yorba Linda, CA), the measurement of DLCO (MasterLab Pro; Erich Jaeger GmbH; Wurzburg, Germany), the determination of subdivisions of the DLCO (ie, DMCO and Vcap, as described earlier),10 and ECG.
Patients with (suspected) DPLD underwent bronchoscopy with bronchial lavage and transbronchial biopsies where indicated. Video-assisted thoracoscopic lung biopsy was performed in selected patients when the above-mentioned examinations did not lead to a definite diagnosis. In some patients, cardiologists or rheumatologists were consulted. Immunologic laboratory investigations and echocardiography were performed when indicated. The classification of the DPLD was based on the recommendations of the British Thoracic Society.8
All patients with suspected PAH underwent radionuclide perfusion and ventilation scans, echocardiography with estimation of the pulmonary artery pressure by tricuspid regurgitation measurement (assessed by experienced cardiologists), CT scan of the pulmonary arteries in order to detect central thromboembolic disease, right heart catheterization with measurements of the pulmonary artery pressure with reversibility testing in most patients (epoprostenol), and pulmonary angiography when indicated. Consultation by rheumatologists included the performance of serum immunologic tests in search of collagen vascular disease and scleroderma. A definite diagnosis of PAH was made based on the Revised Clinical Classification of Pulmonary Hypertension.9
DLNO Measurement
A combined single-breath DLNO and DLCO measurement was performed on an adapted instrument (MasterLab Pro; Erich Jaeger GmbH). The test gas contained a mixture of CO 0.25%, He 9.17%, and NO 8 ppm with balance air. The NO was added to the test gas directly before each measurement from a separate tank containing 750 ppm NO in nitrogen (Hoekloos Medical; Schiedam the Netherlands). The single-breath procedure was performed according to American Thoracic Society recommendations11 with an effective breathholding period of 10 s (the Jones and Meade method12), a discard volume of 750 mL, and a sample volume of 750 mL. The device (MasterLab Pro; Erich Jaeger GmbH) sampled the He and CO concentration in the expired air, and a chemoluminescence analyzer (CLD 77 AM; Eco Physics; Zurich, Switzerland [lower detection limit, 0.02 to 0.05 parts per billion (ppb); upper detection limit, 10 ppm; reaction time 0.1 s]) measured the NO concentration. Once a week, the chemoluminescence analyzer was calibrated with 5 ppm NO in nitrogen and NO-free air. All connections between the NO analyzer and the inspiratory and expiratory bags were made of polytetrafluoroethylene (Teflon; DuPont; Wilmington, DE) or stainless steel, which do not interact with NO. The alveolar volume (VA) and DLCO were calculated according to European Respiratory Society recommendations,2 and the DLNO was calculated according to the method described by Borland and Higenbottam,4 which is the same formula for using NO concentrations instead of CO concentrations. Endogenous NO levels and CO backpressure were ignored, and smoking was allowed until 24 h before testing. The DLCO and DLNO were corrected to BTPS conditions, and a minimum of two measurements was performed, in which a variability of
10% for the VA and DLCO was acceptable. All DLCO measurements were corrected to the standard Hb value according to American Thoracic Society recommendations.11 The obtained DLNO/DLCO ratios were compared by means of analysis of variance using a statistical software package (SPSS for Windows, version 11.0; SPSS; Chicago, IL). The relation between the DLNO/DLCO ratio and DMCO was performed with the Pearson correlation coefficient.
Results
In a period of 1 year (April 2003 to April 2004), 71 patients were screened for study inclusion, and 67 patients were included in the study based on eligibility. Four DPLD subjects were excluded due to the presence of secondary pulmonary hypertension. In one patient, this was probably due to left ventricular failure with mitral valve regurgitation, and in the other three patients the cause of the secondary hypertension was associated with the DPLD. In the control group, 71 healthy volunteers were included (36 female volunteers and 35 male volunteers). In the DPLD category (total, 41 subjects; female subjects, 23; male subjects, 18), sarcoidosis was diagnosed in 27 patients, 4 with nonspecific interstitial pneumonia, 5 with idiopathic pulmonary fibrosis (IPF), 2 with pulmonary alveolar proteinosis, 1 with lymphangioleiomyomatosis, 1 with respiratory bronchiolitis-associated interstitial lung disease, and 1 with cryptogenic organizing pneumonia. In the PAH category (total, 26 patients; female patients, 16; male patients, 10), primary pulmonary hypertension (PPH) was diagnosed in 4 patients, chronic thromboembolic pulmonary hypertension (CTEPH) was diagnosed in 20 patients, and pulmonary hypertension associated with scleroderma (without signs of interstitial lung disease) was diagnosed in 2 patients. The demographics of the study population are shown in Table 1 ; the mean age of the control group is lower than that in subjects with DPLD and PAH. The mean DLNO/DLCO ratios in patients with DPLD and PAH were 4.56 and 4.98, respectively; this difference was statistically significant (p = 0.01). The mean DLNO/DLCO ratio in healthy subjects was 4.36, which is significantly different from the DLNO/DLCO ratio of subjects with PAH (p < 0.001) [Table 1] but is not significantly different from that in the DPLD group (p = 0.127). As shown in Figure 1 , the three groups had a high degree of overlap. The DLNO and the DMCO are highly correlated (r2 = 0.81), and the slope of the regression line (DLNO/DMCO) was 2.48 (Fig 2 ).
|
|
|
In this prospective study, we found a difference in the DLNO/DLCO ratio between patients with DPLD and patients with PAH. Although this difference did reach statistical significance, the large overlap between the groups makes the DLNO/DLCO ratio inapplicable as a clinical tool in discriminating between PAH and DPLD.
Although we used a lower inspiratory NO concentration compared to others,341314 we had no reason to expect that this would influence our data as it has been shown that the DLNO is independent of inspiratory NO in rabbits.15 It has been taken into account that the NO concentration in the alveolar sample is well above the natural alveolar output, which is approximately 2 to 3 ppb in healthy subjects,16 4.7 ppb in subjects with scleroderma-associated interstitial lung disease with or without pulmonary hypertension,17 and 4.1 ppb in subjects with hypersensitivity pneumonitis and IPF.18 In our study the NO concentration in the sample volume was well > 200 ppb; therefore, the neglect of natural alveolar NO output can only lead to a very slight underestimation of the DLNO. The NO production by the conducting airways can be neglected because of the very high exhalation flows used.16 We found DLNO/DLCO ratios of 4.36 in healthy subjects, which are comparable to the ratio of 4.3 in 13 healthy subjects assessed by Borland and Higenbottam4 by the single-breath method and of 4.52 in 8 healthy men (single-breath technique) measured by Zavorsky et al.13 In addition, the DLNO strongly correlated with the DMCO. We observed a relation of 2.48, which is similar to the value found by Tamhane et al3 (2.49) using a rebreathing technique3 and that found by Phansalkar et al6 (2.42). DLCO is often, but not always, decreased in patients with PAH.1920 Consequently, DLCO cannot be used as a screening test to exclude pulmonary hypertension in cases in which the pretest probability is high.21 Borland et al22 found the DLNO/DLCO ratio (combined single-breath technique) to be 5.02 in 12 patients with severe PPH vs 4.51 in 10 matched healthy volunteers. This is in accordance with our results. Steenhuis et al19 observed decreased DLCO in subjects with PPH and CTEPH, mainly due to a decreased DMCO (high/low oxygen method). There were no differences in the mean values of DLCO, DMCO, and Vcap between the two groups. Bernstein et al23 measured the DMCO and Vcap (high/low oxygen method) before and 3 weeks after pulmonary thromboendarterectomy in 29 subjects with CTEPH. DMCO and Vcap were decreased prior to the operation, and DMCO decreased further after the operation. However, the short interval after the pulmonary thromboendarterectomy, the fall in VA postoperatively, and the known dependency of DLCO and DMCO on VA24 make it difficult to draw conclusions from this study.
There have been several studies dealing with the subdivision of the DLCO in interstitial lung diseases. In 1976, Saumon et al25 found that in patients with sarcoidosis with radiologic stage I and II disease, the decrease in DLCO was mainly due to decreased DMCO, but that in stage III sarcoidosis the decrease was associated with a decrease in Vcap. The Vcap values in subjects with IPF or due to systemic sclerosis were lower than in the sarcoidosis stage III group. Phansalkar et al6 measured rebreathing DLNO and DMCO values in 25 subjects with stage II-III sarcoidosis compared to 18 healthy nonsmoker subjects. They found a resting DLNO/DLCO ratio of 4.36 in healthy subjects and 3.48 in subjects with sarcoidosis. At 80% of peak workload, the ratios were 3.70 in healthy subjects and 2.97 in subjects with sarcoidosis. Indeed, at rest and during exercise the DLNO/DLCO ratios were lower in subjects with sarcoidosis than in healthy subjects, as expected. The DLNO strongly correlated with the DMCO, indicating that the DLNO closely resembles the true diffusing capacity of the alveolar capillary membrane. The fact that Phansalkar et al6 found lower ratios in subjects with sarcoidosis than we did is troubling. Although Phansalkar et al6 included subjects with stage II-III sarcoidosis compared to our inclusion of subjects with stage I-IV disease, it is unlikely that this explains the difference. The rebreathing technique used by Phansalkar et al6 is performed at the functional residual capacity level, in contrast to our actual measurement at the total lung capacity level as occurred when using the single-breath method. This could explain in part the difference between the findings of Phansalkar et al6 and our own. In 2004, Lamberto et al26 measured the DMCO and the Vcap in patients with stage I-IV sarcoidosis and found that the reduced DLCO was caused mainly by lowered DMCO in all groups. Furthermore, DMCO as well as DLCO are highly predictive of gas exchange abnormalities during exercise.
An interesting study was performed by Bonay et al,27 who investigated whether the Vcap (determined with the single-breath high/low oxygen method) would be lower in subjects with DPLD and associated PAH than in subjects with DPLD without PAH. This appeared not to be the case, thus excluding the Vcap measurement as a screening test for PAH in subjects with DPLD. In this study, the DLNO/DLCO ratios differ between the different diseases, but the overlap is great.
The equation of Roughton and Forster1 assumes that DMCO and Vcap are independent components by assuming that the 1/DLCO resistance is the sum of two resistances. The question is whether this is correct. Hypoxemia due to thickened membranes can lead to pulmonary vasoconstriction. Capillary flow is a prerequisite to measuring the DMCO. Some investigations28 in patients with IPF show that capillary density is significantly decreased in diseased areas, leading to a decrease in the Vcap component of the DLCO in addition to the already lowered DMCO component as a consequence of the diseased-thickened membranes, thus making the Vcap component dependent on the DMCO component. This is of course difficult to assess in vivo, although some research has pointed to the dependency of the DMCO on Hb concentration, and to a relationship between DMCO and Vcap.29 If the DMCO and Vcap components are dependent, the separation of the DLCO in these two components becomes clinically irrelevant. In other words, the lung is defined as a monoalveolar object, with a relative contribution by the Vcap and the DMCO components. The DLCO measurement is not only a function of membrane thickness and surface area, but also (and not in the least) a function of the ventilation and perfusion inhomogeneity. In 1960, Johnson et al30 showed that the increase of the DLCO from rest to exercise is partly based on an increase in the DMCO. This is mainly based on improved matching of ventilation and perfusion than on the recruitment of alveoli. Furthermore, ventilation and perfusion inhomogeneity is the main determinant of the DLCO in patients with asthma.31
If indeed the DLNO is more sensitive than the DLCO in detecting specific disturbances of the alveolocapillary membrane, then the decreased DLCO in patients with PAH and DPLD is probably due to ventilation and perfusion inhomogeneity instead of to decreased passage through the alveolocapillary membrane.3233
In conclusion, although the overlap is large, we observed a statistically significant difference in DLNO/DLCO ratios between patients with PAH and patients with DPLD. The DLNO/DLCO ratio in patients with PAH was significantly higher than that of healthy volunteers.
Acknowledgements
The authors thank Kim Edwards MD for her advice and grammatical corrections of the text.
Footnotes
Abbreviations: CO = carbon monoxide; CTEPH = chronic thromboembolic pulmonary hypertension; DLCO = diffusing capacity of the lung for carbon monoxide; DLNO = diffusing capacity of the lung for nitric oxide; DMCO = diffusing capacity of the alveolocapillary membrane for carbon monoxide; DPLD = diffuse parenchymal lung disease; Hb = hemoglobin; IPF = idiopathic pulmonary fibrosis; PAH = pulmonary arterial hypertension; ppb = parts per billion; PPH = primary pulmonary hypertension; VA = alveolar volume; Vcap = pulmonary capillary blood volume
Received for publication March 2, 2005. Accepted for publication June 24, 2005.
References
This article has been cited by other articles:
![]() |
H. Dressel, L. Filser, R. Fischer, D. de la Motte, W. Steinhaeusser, R. M. Huber, D. Nowak, and R. A. Jorres Lung Diffusing Capacity for Nitric Oxide and Carbon Monoxide: Dependence on Breath-Hold Time Chest, May 1, 2008; 133(5): 1149 - 1154. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. N. Glenet, C. De Bisschop, F. Vargas, and H. J. P. Guenard Deciphering the nitric oxide to carbon monoxide lung transfer ratio: physiological implications J. Physiol., July 15, 2007; 582(2): 767 - 775. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |