- Email: [email protected]
Relationship between lung function, ventilation–perfusion inequality and extent of emphysema as assessed by high-resolution computed tomography
Vol. 96 (2002) 934 ^943
Relationship between lung function, ventilation^ perfusion inequality and extent of emphysema as assessed by high-resolution computed tomography K. SANDEK,T. BRATEL*, L. LAGERSTRANDw AND H. ROSELLz *Division of Respiratory and Allergology, Department of Medicine, wDivision of Clinical Physiology, and zDepartment of Diagnostics Radiology, Karolinska Institutet at Huddinge University Hospital, Stockholm, Sweden Abstract The development of the high-resolution computed tomography (HRCT) has improved the ability to detect and quantify emphysema in various groups of patients with chronic airflow obstruction (COPD). Significant correlations have previously been found betweenindices of air flow obstruction, hyperinflation, reduced diffusing capacity for carbon monoxide (DLCO), and the extent of emphysema (emph.%) assessed by HRCT. However, the relationship between emph.% and ventilation--perfusion (VA/Q) inequality in COPD is unknown.Twenty COPD patients with a mean forced expiratory volume in1s (FEV1) of 38.2 (715.5)% in percent of predicted value (%P), a mean PaO2 value of 9.6 (71.3) kPa, and a mean diffusing capacity of 43.6 (723.0)%P, were subjected to measurements by the multiple elimination inert gas technique (MIGET). The extent of emphysema was determined by HRCTat both full inspiration, emph.I(%) and at full expiration, emph.E(%), with a cut-off limit of 910 Hounsfield Units (HU) using the‘‘Density Mask’’method.The ventilation directed towards high VA/Q areas was 7.3 (710.2)% and the mean ventilation (V-mean) was elevated about three times compared to normal.The mean emph.(I)% and emph.(E) was 45.6 (716.9) and 32.7 (719.0)%, respectively. Significantcorrelations were shown between the emphysema extent and severallung function parameters, but no correlation was found between the emphysema extent and the VA/Q relationships or the blood gas values.Reduced DLCO%P correlated with less high VA/Q ventilation (r=0.73, Po0.05) for the subgroup of COPD patients with DLCO(%P) less than 50% (n=12). Conclusions: In COPD patients, suffering from moderate to severe emphysema without severe blood gas impairment, no correlation was shown between the extent of emphysema, as assessed by HRCT, and the severity of ventilation--perfusion inequality. A substantial collateral ventilation in severe emphysema may be a mechanism that prevents a deterioration in VA/Q relationships and in blood gas levels.r 2002 Published by Elsevier Science Ltd Available online at http://www.sciencedirect.com
Keywords lung function; ventilation ^ perfusion inequality and HRCT in COPD
INTRODUCTION Di¡erent patterns of ventilation ^ perfusion (VA/Q) mismatch in various groups of patients with chronic obstructive pulmonary disease (COPD) have previously been described by Wagner et al. (1). COPD patients with reduced elastic recoil, marked hyperin£ation and a minimum of sputum production have shown substantial ventilation of poorly perfused lung regions (high VA/Q areas) (1). Using xenon gas tracers, measuring the retention of inhaled and infused radioactive xenon gas, it has Received13 November 2001; accepted in revised form18 March 2002 Correspondence should be addressed to: Dr Karin Sandek, Division of Respiratory Medicine and Allergology, Department of Medicine, Karolinska Institutet at Huddinge University Hospital, S-141 86 Stockholm, Sweden Fax: 46 8 7117306. E-mail: [email protected]
been shown that this subgroup of COPD patients has a reduced alveolar blood £ow in relation to lung volumes (2). Compression and destruction of capillaries in the alveolar walls give rise to reduced alveolar blood £ow in emphysema (3,4). Another group of COPD patients with cough and daily sputum production shows a substantial perfusion of low VA/Q regions, probably as a result of reduced ventilation, re£ecting the mechanical airway obstruction and distortion of the small airways (1,2). By the use of the xenon tracer method, it has previously been shown that these bronchitic COPD patients have normal or even increased peripheral perfusion in relation to their lung volumes (2).Thus, not only lung function but also alveolar perfusion and ventilation ^ perfusion ratios are a¡ected by emphysema. However, in COPD patients with severe airway obstruction and with an increase in VA/Q inequality, insu⁄cient equilibration of gas tracers
VENTILATION^PERFUSION INEQUALITY
may lead to an overestimation of ventilation or perfusion to the most severely a¡ected lung regions (5). The multiple inert gas elimination technique (MIGET) has for more than 25 years been used to analyse theVA/Q relationships in COPD (1). Using MIGET, it is possible to accurately re£ect these relationships (at a di¡erence from the xenon tracer method), at least during resting conditions even in severe COPD, irrespective of whether the patients are in a stable condition or if acute respiratory insu⁄ciency has developed (6,7). Mild emphysema is often di⁄cult to be detected by ordinary chest radiograph images (8).The development of a high-resolution computed tomography (HRCT) technique has signi¢cantly improved the possibility of studying the extent and localisation of emphysema in a non-invasive way (8,9). It has previously been shown that the ‘‘Density Mask’’ technique (10) is a reliable method to quantify the extent of emphysema. Microscopic ¢ndings from post-mortem samples and resected lung tissues from COPD patients have shown that a reduced alveolar wall volume/unit lung volume correlates signi¢cantly with mean lung density and to areas of low attenuation shown by computed tomography (CT) (11). The extent of emphysema determined by CT in emphysema patients with varying degrees of air£ow obstruction, has been shown to correlate signi¢cantly with lung function tests such as forced residual capacity (FRC), forced expiratory volume in one second (FEV1), residual volume (RV) and di¡using capacity for carbon monoxide (DLCO) (9,12^14). However, it is not known whether the extent or severity of emphysema assessed by HRCT measurements also relates to the severity of ventilation ^ perfusion inequality. Therefore, we correlated the extent of emphysema assessed by HRCT with ventilation ^ perfusion ratios measured by the multiple inert gas technique in 20 stable COPD patients (with di¡erent levels of airway obstruction). The extent of emphysema assessed by HRCT and VA/Q ratios was then related to spirometry parameters and to the di¡using capacity of carbon monoxide.
PATIENTS Twenty patients (10 smokers and 10 ex-smokers), eight men and 12 women, with a mean duration of COPD of 10 (78) years and a mean age of 60 (78) years were included in our study. Two subjects had homozygotic a-1antitrypsin de¢ciency. None of the patients had a history of atopy or asthma and all su¡ered from irreversible airway obstruction as shown by the FEV1 value which changed less than10% from the baseline value15 min after the inhalation of 2.5 mg nebulised salbutamol. No patient had any other lung disease, other than COPD, and they had all been free from cardio-pulmonary exacerbations for at least 1 month before being included in the study. The patients were selected by consecutive screening
935
when visiting our outpatient clinic.They all received their ordinary medication unchanged for at least 2 months prior to the measurements. Spirometric values remained unchanged for 2 months after the VA/Q tests had been performed, con¢rming that the patients were in stable condition. Arterial oxygen tension (PaO2) was above 8.0 kPa in all the subjects (Table1). Seventeen patients used b2 -inhalers, 10 patients inhaled corticosteroids and four patients used oral diuretics because of mild chronic right-sided heart failure. This study was approved by the Ethical Committee of Huddinge University Hospital and all subjects gave their informed consent.
METHODS Measurement of ventilation^perfusion (VA/Q) ratios Ventilation ^ perfusion ratios were measured using the multiple inert gas elimination technique (15,16). A modi¢ed method was used in which six mixed expired inert gas levels were measured, while arterial levels were estimated from samples of peripheral venous blood taken
TABLE 1. Anthropometric data, spirometric values and extent of emphysema (n=20) Age (years) 60+8
Mean
7SD
EVF (%) BMI (kg/m2) FEV1 (%P) FVC (%P) RV (%P) FRC (%P) TLC (%P) MEF (%P)
45.0 23.6 38.2 74.2 213.6 160.1 121.9 12.2
73.6 74.8 715.5 717.3 747.0 735.3 717.3 7 8.3
PaO2 (kPa) PaCO2 (kPa) DLCO (%P) Mean pulmonary density (HU) Emph. (I) % Emph. (E)%
9.6 5.1 43.6 876.1 45.6 32.7
71.3 70.7 723.0 728.5 716.9 719.0
BMI, body mass index; FEV1, forced expiratory volume in1second; FVC, forced vital capacity;FRC, functional residual capacity; MEF, mean expiratory £ow at 50% of FVC; RV, residual volume;TLC, total lung capacity; DLCO, di¡usion capacity forcarbon monoxide;EVF, erythrocyte volume fraction; (%P), in percentage of predicted; Emph. (I)%, extent of emphysema in percentage of total pulmonary parenchyma at full inspiration; Emph. (E)%, extent of emphysema in percentage of total pulmonary parenchyma at full expiration and HU,Houns¢eld Units.
936
90 min after the commencement of infusion of the inert gases (17,18). The inert gas samples were analysed by gas chromatography (Varian 3300, Varian Associates Inc., CA, U.S.A.). Retention and excretion ratios were computed and the solubility of each inert gas was determined by a two-step procedure, enabling the estimation of the VA/Q distribution. Mean values from three runs were used to achieve a low standard error (18). Oxygen uptake (VO2) and CO2 production were measured by analysing the O2 and CO2 content from a Douglas bag at the same time as the inert gas sampling was carried out (Ametek, Pittsburgh, PA, U.S.A.). Minute ventilation was measured by analysing the volume of the contents of the Douglas bag. The meanVA/Q ratios for ventilation and perfusion distributions are expressed by V-mean and Q-mean, respectively. From the VA/Q distributions, data were derived regarding shunt (perfusion of lung regions with VA/Q ratios o0.005); ‘‘low VA/Q ’’ levels (perfusion of lung regions with 0.005oVA/Qo0.1); ‘‘high VA/Q’’ levels (ventilation of lung regions with 10oVA/Qo100) and dead space (ventilation of lung regions withVA/Q ratios 4100). Cardiac output (CO) (l/min) was assumed to be1/50 of the VO2 (ml/min) (VO2 method). Analysis from two studies in our laboratory (19,20) concerning stable, resting patients with severe COPD and moderate hypoxaemia, showed that the CO estimates by the VO2 method, performed during right heart catheterisation, was not significantly di¡erent from CO assessments using the thermodilution method and that these two methods correlated signi¢cantly (r=0.72, P=0.001). In patients with severe cystic ¢brosis, we have previously shown that changes in CO of 50% are required to signi¢cantly in£uence CO-derived parameters, such as shunting (21). On the basis of these studies, we concluded that the CO estimates (assessed by VO2) and variables derived from CO such as shunting and low VA/Q, in stable, resting COPD patients without severe hypoxaemia, like in the present study, should be accurate enough to result in valid correlations between CO-derived parameters and other variables. The dispersion of ventilation and perfusion for di¡erent VA/Q ratios is expressed as a logarithmic standard deviation of the ventilation distribution (log SDV) and of the perfusion distribution (log SDQ).Thus, the degree of overall ventilation ^ perfusion mismatch is expressed by log SDV and log SDQ. When the peripheral venous sampling technique is used, log SDV and log SDQ are underestimated by about 6% (18) and the upper normal limits regarding log SDVand log SDQ are 0.75 and 0.70, respectively (22).Theoretical work has shown that VA/Q dispersion indices are in£uenced only to a minor degree by estimates of CO (17). Mean values from three runs were used to achieve a low standard error of the method (18). The ¢t of the derived VA/Q distributions to the measured data, expressed as the remaining sum of squares (RSS)
RESPIRATORY MEDICINE
should not exceed 6.0 in more than 50% of the measurements (15,18). The mean RSS from all data in the present study was 1.6 (71.8).
Blood gas analyses Blood gas analyses were performed on radial arterial samples (ABL 520, Radiometer, Copenhagen, Denmark) immediately after VA/Q measurements, while breathing air, as the patients were sitting in a semi-recumbent position.
Spirometry tests Static and dynamic spirometry were performed with the patient in a sitting position. To avoid the in£uence of forced expiration on secretion and thereby gas exchange, the spirometry measurements were always performed after the conclusion of the gas exchange study.The FEV1 (the best value of three was accepted), forced vital capacity (FVC), vital capacity (VC) and mean expiratory £ow at 50% of FVC (MEF 50%) were measured by dynamic spirometry. The functional residual capacity (FRC) was determined by body plethysmography. The residual volume (RV) and the total lung capacity (TLC) were then calculated from the FRC value. Dynamic spirometry and lung volume measurements were carried out with a Pulmonary Function Laboratory 2400 and 6200, respectively (Sensor Medics BV, Bilthoven, The Netherlands). The values from the spirometry tests are presented as percentage of predicted (%P). The normal values were selected from equations of the European Community for Coal and Steel (23). Di¡using capacity for carbon monoxide (DLCO) was measured by the single breath method (24). The mean value of two technically acceptable measurements was used. None of the patients had anaemia or polycythaemia and consequently no correction for the Hb-level was needed.
High-resolution computed tomography (HRCT) CT scanning was performed in the supine position by a Philips Tomoscan SR 7000 equipment (Philips Medical System). The HRCT technique (1.5 mm collimation, a scanning time of 2.0 s, 140 kV and 150 mA) was used and the images were acquired from the apices to the diaphragm at 2 cm intervals. Resting time before CT-scanning was performed was at least 2 minutes and the duration for the whole investigation was 10 min. In each patient, 11^15 layers were scanned at inspiration to TLC and at expiration to RV. The mean normal density (electronic density of the lung parenchyma relative to water) in healthy lungs measured by CTexpressed in Houns¢eld
VENTILATION^PERFUSION INEQUALITY units (HU) values is 722 (744) HU for women and 746 (742) HU for men (25).
Emphysema extent To quantify the extent of emphysema the‘‘Density Mask’’ programme of the CT-scanner was used without contrast medium (10). With a track-ball our radiologist (HR) outlined the lung parenchyma excluding solid bone structure, hilum, mediastinum, air in the trachea and in the main bronchi. In each HRCT-section, the scanner measured the area of pixels (r0.5 mm2) with attenuation values between 400 HU and 1000 HU, corresponding to lung tissue (26). Areas with density levels between 910 HU and 1000 HU were considered emphysematous (27^29). The density of the total pulmonary tissue at full end inspiration was automatically calculated by the scanner’s computer program assessed as the mean value of attenuation (expressed in HU). This was performed at three di¡erent levels; at the top of the aortic arch, at the origin of the right lower lobe bronchus, and at 3 cm above the top of the diaphragm (13). At each of these levels, the ratio of lung areas with low attenuation (o910 HU) to total lung area (400 HU to 1000 HU) was expressed as the percentage of emphysema (emph.%) and was calculated both at full inspiration denominated emph.(I)% and at full expiration denominated emph.(E)%.The cut-o¡ points for separating di¡erent groups of emph.(E)% were modi¢ed from previous studies (29,30). Our radiologist who examined the HRCTscans had no knowledge of the results from the ventilation ^ perfusion ratio measurements or from the pulmonary function tests.
Statistical analysis The data are presented as a mean 7SD Correlations between blood gases and lung function tests was determined by linear regression. Only non-parametric statistical method was used since the extent of emphysema and VA/Q ratios did not show a normal distribution. Correlations between variables were assessed by the Spearman rank test and di¡erences between groups were examined using the Mann ^Whitney U-test. Regarding the di¡erences between the three groups, the Kruskal ^Wallis test was used. The importance of di¡erent variables for the variability of a dependent parameter was assessed by stepwise regression.
RESULTS The chest radiograph images showed hyperin£ation in all patients. HRCTshowed areas with low attenuation levels
937
scattered around the entire lung. No images showed paraseptal or subpleural localised air cysts or ‘‘bullous’’ patterns.The automatically calculated values for the lung density and the extent of emphysema are shown inTable 1. The mean lung density level was 876.1 (728.5)HU, which is lower than normal (722 HU 7 44 for women and 746HU742 for men) (25). Emph.(I)% for the entire group was 45.6 (716.9)% and emph.(E)% was 32.7(719.0)%. Due to technical reasons, measurements of the mean lung density level were performed in only13 patients. Signi¢cant di¡erences between these 13 patients compared to the remaining seven patients were only their mean levels of TLC and FRC (Po0.05, respectively). Two patients had mild airway obstruction [FEV1 60 ^ 70(%P)], seven had moderate [FEV1 40 ^59(%P)], and 11 had severe airway obstruction [FEV1o40(%P)] (Table 1). All patients su¡ered from severe hyperin£ation [RV%P of 213.6 (747.0)%] and a reduced di¡using capacity [DLCO(%P) of 43.6 (723.0)%] compared to normal (23,24). They were all eucapnic with a mean PaCO2 of 5.1 (70.7) kPa and a mean PaO2 close to normal of 9.6 (71.3) kPa (Table 1). High VA/Q value dominated the ventilation ^ perfusion inequality pattern. About 7% of the ventilation was directed towards highVA/Q areas (normal valueo1%) yielding an increase in the V-mean level to 2.9 (71.6) compared to the level of healthy individuals (0.7^1.2) (Table 2).The dispersion of ventilation, log SDV, was also increased compared to normal (normal value r0.75) (22), to a mean value of 0.95 (70.43) and the dispersion of perfusion, log SDQ, was at the upper normal limit of 0.68 (70.20) (normal value r0.70) (22). About 2% of CO was directed to shunt areas and almost no perfusion was directed towards low VA/Q areas (Table 2).
Correlations between lung function and HRCT Close inverse correlations were found between emph.(E)% and FEV1(%P) (Fig. 1) as well as to other indices of air£ow obstruction such as FVC(%P), FEV1/ VC(%) and MEF 50%(%P) (Table 3). A strong inverse correlation was also shown between emph.(E)% and the diffusing capacity (Fig. 2). Emph.(E)% was positively correlated with indices of hyperin£ation such as RV(%P) (Table 3). However, in three patients with rather severe airway obstruction [a median FEV1(%P) of 53 (49^56)% and a median RV(%P) of 165 (153^207)%] a median emph.(E)% of only 5.5 (3.6 ^11.4)% was shown by HRCT. By stepwise regression analysis for all our COPD patients, we noted that 69% of the variability in emph.(E)% was explained by FEV1/VC(%P) and a further 8% by DLCO(%P). The parameters FEV1(%P), RV(%P), FRC(%P), MEF50%(%P), Q-mean and V-mean however, failed to
938
RESPIRATORY MEDICINE
TABLE 2. Ventilation-perfusion relationships (n=20) COPD patients
V-mean Q-mean Log SDV Log SDQ Shunt Low VA/Q HighVA/Q Dead space VE (l/min)
Healthy individuals
Mean
7SD
2.9 1.3 0.95 0.68 1.9 0.1 7.3 30.0 10.2
7 1.6 7 0.4 7 0.43 7 0.20 7 1.4 7 0.4 710.2 7 7.7 7 2.5
Mean 0.9 0.7 r0.75 r0.70 0.08 0 0 39.4
7SD 70.2 7 0.1 F F 70.15 0 0 75.6
The normal values regardinglog SDVand log SD Q are adopted from reference (22) and the other normal values are adopted from reference (30). V-mean, The mean VA/Q ratio for ventilation; Q-mean, The mean VA/Q ratio for perfusion; log SDV, The dispersion of ventilation; log SDQ, The dispersion of perfusion; shunt, VA/Qo0.005 in per cent of cardiac output; low VA/Q, 0.005oVA/Qo0.1 in per cent of cardiac output; high VA/Q, 10oVA/Qo100 in per cent of minute ventilation; dead space, VA/ Q4100 in per cent of minute ventilation and VE,Minute ventilation
Correlations between lung function, blood gas exchange, ventilation^perfusion ratios and extent of emphysema
70
FEV1 (%P)
60 50 40 30 20 10 0 0
10
20
30 40 Emph. (E)%
50
60
70
FIG. 1. Spearmanrankcorrelations betweenthe extentof emphysema atfull expiration emph.(E) (%) as shown by HRCTscans and forced expired volume in one second (FeV1) (%P) in 20 COPD patients, (r=0.83, Po0.001).
explain the remaining variability in emph.(E)%. Similar albeit slightly weaker correlations were found between emph.(1)% and indices of air£ow obstruction, hyperin£ation and DLCO%P (Table 3). No correlations were found between blood gas levels and emph.(E)%. For technical reasons, the mean lung density value at full inspiration [emph.(I)%] was measured in only 13 of our COPD patients. The emph.(I)% for these patients was shown to be inversely and signi¢cantly correlated with emph.(E)% (r=0.70, Po0.05) without any correlation to the air£ow obstruction, hyperin£ation, VA/Q inequality or low DLCO%P (Table 3). However, signi¢cant correlations were found between emph.(E)% and spirometry parameters for the13 patients in whom the mean lung density was measured as for our entire patient group (data not shown).
No signi¢cant correlations were found between air£ow obstruction (measured by spirometry) and ventilation ^ perfusion inequality or between blood gas values and VA/Q ratios. No signi¢cant correlations between reduced DLCO(%P) and any VA/Q parameter (such as high VA/Q ventilation and elevated V-mean) were found. For the entire patient population, no signi¢cant correlations were found between di¡erent VA/Q ratios or blood gas levels vs emph.(I)% or vs emph.(E)%. However, the di¡erence in emph.% at full inspiration and at full expiration deceased with increasing emphysema extent (r=0.47, Po0.05) (Fig. 3). A reduced BMI correlated with emph.(E)% (r=0.53, Po0.05). The PaO2 level related signi¢cantly to the DLCO(%P) level (r=0.73, Po0.001) as well as to the minute ventilation (r=0.48, Po0.05). The PaCO2 level was inversely correlated to FEV1(%P), DLCO(%P) and VC(%P) (r=0.56 to 0.67, Po0.01 for all).
Subgroup analyses Comparisons between groups consisting of ¢ve patients with emph.(E)% of r16% (trivial emphysema), ten patients with emph.(E)% of 17^50 % (moderate emphysema) and ¢ve patients with emph.(E)% of 450% (severe emphysema) (29,31) (Table 4) concerning spirometric values, VA/Q ratios and emph.(%) showed that patients with moderate or severe emphysema had a signi¢cantly more pronounced air£ow obstruction and
VENTILATION^PERFUSION INEQUALITY
939
TABLE 3. Correlations between the extent of emphysema at full inspiration in percentage of healthy pulmonary parenchyma [Emph.(I)%] and at full expiration in percentage of healthy pulmonary parenchyma [Emph.(E)%] and mean lung density atthree levels assessed by HRCTvs spirometric values and VA/Q relationships Emph. (I)% (n=20)
Emph. (E)% (n=20)
Mean density (n=13)
r-Value
P-Value
r-Value
r-Value
NS NS
0.87*** 0.81*** 0.70** 0.77*** 0.57** 0.78** 0.73** 0.15 0.22
0.76** 0.69** 0.69** 0.71*** 0.57** 0.57* 0.60** 0.14 0.25
FEV1/VC (%) FEV1 (%P) FRC (%P) RV (%P) TLC (%P) MEF 50% (%P) DLCO (%P) High VA/Q (%) V-mean (%)
P-Value
NS NS
0.54 0.38 0.37 0.36 0.35 0.28 0.41 0.13 0.49
P-Value NS NS NS NS NS NS NS NS NS
*Po0.05, **Po0.01, ***Po0.001regarding signi¢cant correlations.
100 90 80 DLCO (%P)
70 60 50 40 30 20 10 0 0
10
20
30 40 Emph. (E)%
50
60
70
FIG. 2. Spearman rank correlations between the extent emphysema at full expiration emph.(E) %, as shown by HRCTscans and the di¡using capacity for carbon monoxide (DLCO) (%P) in 20 COPD patients, (r=0.71, Po0.01).
80
Emphysema insp. extent
Emphysema expir. extent
Emphysema extent (%)
70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Number of the patient
FIG. 3. Spearman rankcorrelation between the extent of emphysema as shown by HRCTscans at full inspiration; (upper line) and at full expiration (lower line) emph.(I) (%) and at full expiration; emph.(E) (%) in 20 COPD patients, (r=0.94, P=0.001).
hyperin£ation compared with the trivial emphysema group. Air£ow obstruction and hyperin£ation were also signi¢cantly more pronounced in patients with severe emphysema as compared with patients with moderate emphysema. There was a signi¢cantly lower DLCO(%P) value in patients with moderate or severe emphysema compared with the patients in the trivial emphysema group, but there was no signi¢cant di¡erence with regard to DLCO(%P) value between patients with moderate or severe emphysema. All these di¡erences were con¢rmed by the Kruskal ^Wallis test (Po00.1 for emph.(E)% and for the other variables the P-values varied between 0.01 and 0.05). Ventilation towards high VA/Q areas was signi¢cantly increased in the subgroup with moderate emphysema compared with the trivial emphysema subgroup. In line with this result there was a trend for an increased ventilation distribution (log SDV) including an insigni¢cantly increased shunting in the moderate emphysema subgroup compared with the trivial emphysema subgroup. The mean VA/Q ratio for ventilation (V-mean) was signi¢cantly increased compared with normal values for all subgroups (Table 2) and there were trends towards an increased ventilation of high VA/Q areas and for perfusion of shunting areas in patients with severe emphysema compared with the trivial emphysema subgroup. However, for the 15 patients with an emph.(%E) of 416% no correlation was found between the increasing extent of emphysema and VA/Q inequality. In 13 COPD patients, with a di¡using capacity for carbon monoxide of o50%P, V-mean and high VA/Q ventilation tended to correlate with an increase in DLCO(%P) level, but if one patient with DLCO(%P) 445% was excluded, these correlations were no longer signi¢cant (Figs. 4(a) and (b)).
940
RESPIRATORY MEDICINE
TABLE 4. Di¡erences with regard to spirometric values and VA/Q relationships in COPD patients with di¡erent extent of emphysema as assessed by HRCT Emph. extent r16% (n=5) Mean 7SD FEV1/VC (%) FEV1 (%P) FRC (%P) RV (%P) TLC (%P) DLCO (%P) Log SDV Log SDQ High VA/Q (%) V-mean (%) Q-mean (%) VE (L/min) Three levels: Emph. (I) % Emph. (E)%
Emph. extent17^50% (n=10) Mean 7SD
72.077.8 57.477.1 121.8718.1 159.6735.1 107.079.5 67.0728.4 0.7170.25 0.6070.05 1.172.5 2.170.4 1.470.1 9.373.1 28712 1076
45.777.5** 34.3712.7** 164.1733.7** 217.9733.9** 123.8713.9* 39.4714.8* 1.0370.37 0.7170.21 8.979.7* 2.871.1 1.270.3 10.472.2 44711* 3179**
Emph. extent>50% (n=5) Mean 7SD 36.077.3**w 26.678.1** 186.2713.3** 258.8720.9**w 139.479.4**w 28.4714.1* 1.0370.65 0.7070.29 10.5714.2 3.972.6* 1.470.5 10.772.9 6774** 5974**z
NS NS NS NS NS
NS NS NS NS NS
35
9 8 7 6 5 4 3 2 1 0
30 25 High VA/Q
V-mean
*Po0.05, **Po0.01 regarding signi¢cant di¡erences between COPD patients with extent of emphysema 416% and COPD patients with extent of emphysema r16 %. w Po0.05, zPo0.01 regarding signi¢cant di¡erences between COPD patients with extent of emphysema >50% and COPD patients with extent of emphysema17^50%. Emph. (I)%,Extent of emphysema at full inspiration in percentage of healthy pulmonary parenchyma. Emph. (E)%,Extent of emphysema at full expiration in percentage of healthy pulmonary parenchyma.
20 15 10 5 0
0
5
10
(a)
15
20 25 30 DLCO (%P)
35
40
45
50
0 (b)
5
10
15
20 25 30 DLCO (%P)
35
40
45
50
FIG. 4a. Spearman rank correlations between mean of VA/Q ratios for ventilation (V-mean) and the di¡using capacity for carbon monoxide (DLCO) (%P) in13 COPD patients with DLCO (%P) o50%, (r=0.57, Po0.05).4b Spearman rankcorrelations between ventilation of lungregions with10oVA/Qo100, (highVA/Q) and di¡usingcapacity forcarbon monoxidein13 COPD patients, with DLCO (%P) o50%, (r=0.57, Po0.05).
DISCUSSION Assessment of emphysema by the HRCT method HRCT may detect emphysematous areas which may be as small as 0.5 cm2 (32). The ‘‘Density Mask’’ method has the advantage of being less prone to observer bias compared to visual scorings, since the outlined regions of interest automatically denote the lung areas which are
below the prede¢ned threshold of attenuation (33). However, in healthy individuals and to a lesser extent in emphysematous patients, lung areas of low attenuation level may appear towards the end of the inspiratory manoeuvre to TLC due to hyperin£ation (13,14,29). The low attenuation areas may also appear as a result of obstruction of central or peripheral airways in patients with chronic air£ow limitation, giving rise to air trapping and ¢xed hyperin£ation (13,14). Furthermore, these low
VENTILATION^PERFUSION INEQUALITY
attenuation areas may appear due to the existence of truly emphysematous areas (13,14). In our study, as in previous studies (13,14,34), the di¡erences in the size of areas with low attenuation shown by HRCT, between end inspiration and the end expiration diminished in patients with more widespread emphysema during end expiration. This corroborates the notion that less air is transported to and from the lungs as emphysema develops (13,14). Previous studies have shown that a cut-o¡ limit of 900 or 910 HU, has been found to separate normal from emphysematous lung tissue at full inspiration on CT or HRCT (10,27,28). Recent research, however, implies that a cut-o¡ limit of 950 HU used on CT scans, taken at full inspiration, de¢nes an even more exact limit between hyperin£ated healthy and emphysematous pulmonary tissue (33). It has also been noted previously that similar correlations between the severity of emphysema and spirometry tests are found regardless if the cut-o¡ limit is set to 910 or 950 HU (35).The risk of including pulmonary areas with low attenuation level due to hyperin£ation instead of emphysema is reduced if the CT scans are obtained at full expiration (13,14,27,31) and by this mechanism the in£uence of the supine posture was probably minimised. By the use of standardised coaching, as in our study, the images can be obtained at reproducible levels of expiration (31,34).The extent of emphysema assessed at full expiration by the ‘‘Density Mask’’ technique has previously been shown to be one of the most accurate methods to distinguish patients with no or trivial emphysema from those with moderate and severe emphysema (13,29,30).We therefore chose to assess the extent of emphysema in our patients [emph.(E)%] with a cut-o¡ limit for normal extent of low attenuation areas of r16% (13,27,31). Furthermore, in agreement with earlier studies (13), we found that HRCT scans from three levels correlate closely with images assessed at 10 ^15 levels. Statistical analysis was therefore performed using average values of HRCT scans from only three levels (see methods).
Correlations between lung function and HRCT In the present investigation, as in previous studies, highly signi¢cant correlations were found between indices of air£ow obstruction, hyperin£ation and the extent of emphysema (13,27,29,31,34,35). In agreement with previous results, somewhat more signi¢cant correlations were recorded between lung function tests and emph.(E)% as compared with emph.(I)% (13,27,29,31,34,35). The inverse correlation between the extent of emphysema and FEV1 is primarily dependent on the reduced elastic recoil inherent to emphysema (36). In disagreement with previous reports (8,13,31), indices of air£ow
941
obstruction seemed to be of greater importance for the variability in extent of emphysema than the DLCO(%P) level. A reduction in the di¡usion capacity in COPD may depend on parenchymal and capillary tissue destruction due to the development of emphysema (3,8), which in turn reduces the surface area available for gas exchange. Furthermore, DLCO may be diminished due to small airway disease (37,38) which decreases the in£ux of air into the alveoli and thereby the alveolar ventilation and the gas exchange becomes reduced. Small airway disease may have contributed to the ¢nding that four of our patients had DLCO(%P) below 30% and an emph.(E)% of only 30 ^ 45% (Fig. 2). In disagreement with several other investigations (27,31,34), mean lung density did not correlate to any index of lung function or gas exchange impairment in the present study. The mean lung density measures the proportions of air and liquid in both normal and emphysematous parts of the lung which may explain why this parameter is less sensitive to lung function impairments compared to measurements of the extent of emphysema (27). By measuring the extent of emphysema by HRCT, only the diseased parts of the lung parenchyma are highlighted. Furthermore, in several earlier studies, a wide range of subjects were included, ranging from patients with rather trivial emphysema and normal lung function to patients with large emphysematous areas with severely reduced lung function (27,31,34). Our patients comprised an rather homogenous group, since 80% of our subjects had an FEV1(%P) of less than 50%.This relatively narrow range of forced expiratory volumes may have obscured the plausible correlations between the mean lung density level and impaired lung function.
VA/Q relationships and HRCT The VA/Q ratios assessed by MIGETand the single breath di¡using capacity for carbon monoxide are sensitive indicators of mild emphysema (6,8). In agreement with previous studies of patients with stable COPD and substantial emphysema, theVA/Q relationships in the present study were characterised by substantial ventilation to lung areas with poor perfusion (high VA/Q areas) resulting in an increased V-mean and an increased dispersion of ventilation, whereas shunting and perfusion of low VA/Q areas were scarcely present (2,6,28,39). Moreover, as expected (2,6,28,39), ventilation of poorly perfused areas was higher in patients with substantial emphysema [emph.(E)%P416] than in patients with less severe emphysema [emph.(E)%Pr16] as shown by HRCT. Furthermore, for the entire patient group, no correlations were found between lung function, gas exchange and VA/Q ratios. In previous studies of COPD patients with nocturnal or daytime hypoxaemia, signi¢cant inverse correlation between ventilation of poorly
942
perfused lung areas and di¡usion capacity has been shown (39,40). There are several possible explanations for the absence of correlations betweenVA/Q ratios and the extent of emphysema, di¡using capacity or blood gas levels in the present study. Firstly, a reduced alveolar surface due to emphysema is accompanied by an equal reduction in capillary volume (2,3), preventing the development of severely abnormal VA/Q ratios. Secondly, in most COPD patients an increase in the intra-alveolar pressure occurs due to in£ammation of the bronchioli and/or due to reduced elastic recoil which may decrease the gas £ow when gas is leaving the alveoli (6,28,35,41).The increased intra-alveolar pressure compresses the capillaries and prevents shunting (2,3) and a reduction in the PaO2 level. When air continues to reach these poorly perfused lung regions, an increase in high VA/Q areas will be present as shown in our patients (2,3). Thirdly, as the emphysematous process progresses, the acinar fenestration becomes more extensive and the collateral ventilation will increase (42), preventing a more pronounced VA/Q mismatch despite a reduced gas exchange surface, re£ected by a low DLCO(%P).This may be a possible mechanism for the subgroup of our patients with DLCO(%P) of less than 45%, to have a tendency to an increased ventilation to high VA/Q areas together with augmented di¡using capacity (Figs. 4(a) and (b)). Accordingly, it has been shown that the degree of ventilation towards high VA/Q areas in emphysematous patients with low pulmonary parenchyma density correlated positively with the level of the diffusing capacity (2). Finally, small airway disease may lead to an increasing tendency of atelectases and an increase in the perfusion of poorly ventilated lung areas (28).The patients in the present study had less perfusion of poorly ventilated lung areas compared with the patients in a previously investigated COPD patient group su¡ering from daytime hypoxaemia (40).The patients presently investigated, may thus have been su¡ering from less serious small airway disease than the group of patients with daytime hypoxaemia (40). These mechanisms may explain why 80% of our subjects had normal or almost normal PaO2 (X8.5 kPa) despite that 75% had moderate to severe emphysema measured by HRCT. Severe emphysema is associated with increased basal inspiratory drive (43) and an increase in energy consumption of the respiratory muscles even in patients without daytime hypoxaemia (44). Our patients with severe emphysema thus maintained relatively normal blood gas level presumably at the cost of an increased work of breathing (50% had a BMI below 21kg/m2). Lung function tests and inert gas measurements were performed in the sitting position, whereas the HRCT scans were performed in the supine position. However, the in£uence of the posture was probably small since the HRCTscans were measured at full inspiration and at full expiration.
RESPIRATORY MEDICINE
In conclusion: The extent of emphysema assessed by HRCT is associated with a reduced di¡using capacity for carbon monoxide, hyperin£ation and reduced forced expiratory volumes. In COPD patients without severe blood gas disturbances su¡ering from moderate-to-severe extent emphysema assessed by HRCT images, elevated ventilation of poorly perfused lung areas is found compared to patients with trivial extent of emphysema. However, there is no correlation between the severity of extent of emphysema and the VA/Q inequality. Collateral ventilation and other mechanisms probably prevent VA/Q relationships and blood gas levels to deteriorate at the same pace as the progression of the emphysematous process.
REFERENCES 1. Wagner PD, Dantzker DR, Dueck R, Clausen JL, West JB. Ventilation–perfusion inequality in chronic obstructive pulmonary disease. J Clin Invest 1977; 59: 203–216. 2. Brudin LH, Rhodes CG, Valind S-O, Buckingham PD, Jones TJ, Hughes JMB. Regional structure–function correlations in chronic obstructive lung disease measured with positron emission tomography. Thorax 1992; 47: 914–921. 3. Morrison NJ, Abboud RT, Mu¨ller NL, et al. Pulmonary capillary blood volume in emphysema. Am Rev Respir Dis 1990; 141: 53–61. 4. Barbera` JA, Riverola A, Roca J, et al. Pulmonary vascular abnormalities and ventilation–perfusion relationships in mild chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1994; 149: 423–429. 5. Brudin L., Valind S-O, Rhodes CG. Error analysis of combined measurements of regional ventilation and V/Q ratio using positron emission tomography. Physiol Med Biol 1992; 37: 1077–1093. 6. Barbera` J, Ramirez J, Roca J, et al. Lung structure and gas exchange in mild chronic obstructive pulmonary disease. Am Rev Respir Dis 1990; 141: 895–901. 7. Torres A, Reyes A, Roca J, Wagner PD, Rodriguez-Roisin R. Ventilation–perfusion mismatching in chronic obstructive pulmonary disease during ventilator weaning. Am Rev Respir Dis 1989; 140: 1246–1250. 8. Kuwano K, Matsuba K, Ikeda T, et al. The diagnosis of mild emphysema. Correlation of computed tomography and pathology scores. Am Rev Respir Dis 1990; 141: 169–178. 9. Gurney JW, Jones KK, Robbins RA, et al. Regional distribution of emphysema: correlation of high-resolution CT with pulmonary function tests in unselected smokers. Radiology 1992; 183: 457– 463. 10. Mu¨ller NL, Staples CA, Miller RR, Abboud RT. ‘‘Density Mask’’. An objective method to quantitate emphysema using computed tomography. Chest 1988; 94: 782–787. 11. Gould GA, MacNee W, McLean A, etal. CT measurements of lung density in life can quantitate distal airspace enlargementFan essential defining feature of human emphysema. Am Rev Respir Dis 1988; 137: 380–392. 12. Coxson HO, Rogers RM, Whittall KP, et al. A quantification of the lung surface area in emphysema using computed tomography. Am J Resp Crit Care Med 1999; 159: 851–856. 13. Miniati M, Filippi E, Falaschi F, et al. Radiologic evaluation of emphysema in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995; 151: 1359–1367. 14. Eda S, Kubo K, Fujimoto K, Matsuzawa Y, Sekiguchi M, Sakai F. The relations between expiratory chest CT using helical CT and
VENTILATION^PERFUSION INEQUALITY
15.
16.
17.
18.
19.
20.
21.
22.
23. 24.
25. 26.
27.
28.
29.
pulmonary function tests in emphysema. Am J Respir Crit Care Med 1997; 155: 1290–1294. Wagner PD, Saltzman HA, West JB. Measurement of continuous distributions of ventilation–perfusion ratios: theory. J Appl Physiol 1974a; 36: 588–599. Wagner PD, Nauman PF, Laravuso RB. Simultaneous measurement of eight foreign gases in blood by gas chromatography. J Appl Physiol 1974b; 36: 600–605. Wagner PD, Smith CM, Davies NJH, McEvoy RD, Gale GE. Estimation of ventilation–perfusion inequality by inert gas elimination without arterial sampling. J Appl Physiol 1985; 59: 376–383. Wagner PD, Hedenstierna G, Bylin G, Lagerstrand L. Reproducibility of the multiple inert gas elimination technique. J Appl Physiol 1987; 62: 1740–1746. Bratel T, Hedenstierna G, Nyquist O, Ripe E. The effect of a new calcium antagonist, elodipine, on pulmonary hypertension and gas exchange in chronic obstructive lung disease. Eur J Respir Dis 1985; 67: 244–253. Bratel T, Hedenstierna G, Lundquist H, Nyquist O, Ripe E. Cardiac function and central haemodynamics in severe chronic obstructive lung disease. Acute and long-term effects of felodipine. Eur J Respir Dis 1988; 1: 262–268. Lagerstrand L, Hjelte L, Jorulf H. Pulmonary gas exchange in cystic fibrosis: basal status and the effect of i.v. antibiotics and inhaled amiloride. Eur J Respir Dis 1999; 14: 686–692. Cardu´s J, Burgos F, Diaz O, et al. Increase in pulmonary ventilation–perfusion inequality with age in healthy individuals. Am J Respir Crit Care Med 1997; 156: 648–653. Quanjer PhH. ed. Standardised lung function testing. Bull Eur Physiopath Respir 1983; 9 (Suppl. 5): 45–51. Cotes J. Measurement of transfer factor (diffusing capacity) and its subdivisions. In: Lung FunctionFAssessment on Application in Medicine, 5th edn. Oxford: Blackwell Publishers, 1993: 299–324. Kalef-Ezra J, Karantanas A, Tsekeris P. CT measurement of lung density. Acta Radiol 1999; 40: 333–337. Archer DC, Coblenz CL, de Kemp RA, Nachmias C, Norman G. Automated in vivo quantification of emphysema. Radiology 1993; 188: 835–838. Kinsella M, Mu¨ller NL, Abboud RT, Morrison NJ, DyBuncio A. Quantitation of emphysema by computed tomography using a ‘‘Density Mask’’ program and correlation with pulmonary function tests. Chest 1990; 97: 315–321. Yamaguchi K, Mori M, Kawai A, Takasugi T, Oyamada Y, Koda E. Inhomogeneities of ventilation and the diffusing capacity to perfusion in various chronic lung diseases. Am J Respir Crit Care 1997; 156: 86–93. Dowson LJ, Guest PJ, Hill SL, Holder RL, Stockley RA. Highresolution computed tomography scanning in a1 antitrypsin deficiency: relationships to lung function and health status. Eur Respir J 2001; 17: 1097–1104.
943
30. Wagner PD, Laravuso RB, Uhl RR, West JB. Continuos distributions of ventilation–perfusion ratios in normal subjects breathing air and 100% O2. J Clin Invest 1974c; 54: 54–68. 31. Knudson RJ, Standen JR, Kaltenborn WT, et al. Expiratory computed tomography for assessment of suspected pulmonary emphysema. Chest 1991; 99: 1357–1366. 32. Miller RR, Mu¨ller NL, Vedal S, Morrison NJ, Staples CA. Limitations of computed tomography in the assessment of emphysema. Am Rev Respir Dis 1989; 139: 980–983. 33. Bankier AA, de Maertelaer V, Keyzer C, Genevois PA. Pulmonary emphysema: subjective visual grading versus objective quantification with macroscopic morphometry and thin-section CT densitometry. Radiology 1999; 211: 851–858. 34. Lamers RJ, Thelissen GR, Kessels AG, Wouters EF, van Engelshoven JM. Chronic obstructive pulmonary disease: evaluation with spirometrically controlled CT lung densitometry. Radiology 1994; 193: 109–113. 35. Park KJ, Bergin CJ, Clausen JL. Quantitation of emphysema with three dimensional CT densitometry: comparison with twodimensional analysis, visual emphysema scores and pulmonary function test results. Radiology 1999; 211: 541–547. 36. Nagai A, Yamawaki I, Takizawa T, Thurlbeck WM. Alveolar attachments in emphysema of human lungs. Am Rev Respir Dis 1991; 144: 888–891. 37. Gelb AF, Schein M, Kuei J, et al. Limited contribution of emphysema in advanced chronic obstructive pulmonary disease. Am Rev Respir Dis 1993; 147: 1157–1161. 38. Gelb AF, Zamel N, Hogg JC, Mu¨ller NL, Schein MJ. Pseudophysiologic emphysema resulting from severe small-airways disease. Am J Respir Crit Care 1998; 158: 815–819. 39. Sandek K, Andersson T, Bratel T, Lagerstrand L. Ventilation– perfusion inequality in nocturnal hypoxaemia due to chronic obstructive lung disease (COLD). Clin Physiol 1995; 15: 499–513. 40. Sandek K, Bratel T, Hellstro¨m G, Lagerstrand L. Ventilation– perfusion inequality and carbon dioxide sensitivity in hypoxaemic chronic obstructive pulmonary disease (COPD) and effects of long-term oxygen treatment (LTOT) for six months. Clin Physiol 2001; 5: 584–593. 41. Ranieri VM, Dambrosio M, Brienza N. Intrinsic PEEP and cardiopulmonary interaction in patients with COPD and acute ventilatory failure. Eur Respir J 1996; 9: 1283–1292. 42. Morrell NW, Wignall BK, Biggs T, Seed WA. Collateral ventilation and gas exchange in emphysema. Am J Respir Crit Care Med 1994; 150: 635–641. 43. Sandek K, Andersson T, Bratel T, Hellstro¨m G, Lagerstrand L. Sleep quality, carbon dioxide responsiveness and hypoxemic patterns in nocturnal hypoxaemia due to chronic obstructive pulmonary disease (COPD) without daytime hypoxaemia. Respir Med 1999; 93: 79–87. 44. Shindoh C, Hida W, Kikuchi Y, et al. Oxygen consumption of respiratory muscles in patients with COPD. Chest 1994;105: 790–797.
Relationship between lung function, ventilation^ perfusion inequality and extent of emphysema as assessed by high-resolution computed tomography K. SANDEK,T. BRATEL*, L. LAGERSTRANDw AND H. ROSELLz *Division of Respiratory and Allergology, Department of Medicine, wDivision of Clinical Physiology, and zDepartment of Diagnostics Radiology, Karolinska Institutet at Huddinge University Hospital, Stockholm, Sweden Abstract The development of the high-resolution computed tomography (HRCT) has improved the ability to detect and quantify emphysema in various groups of patients with chronic airflow obstruction (COPD). Significant correlations have previously been found betweenindices of air flow obstruction, hyperinflation, reduced diffusing capacity for carbon monoxide (DLCO), and the extent of emphysema (emph.%) assessed by HRCT. However, the relationship between emph.% and ventilation--perfusion (VA/Q) inequality in COPD is unknown.Twenty COPD patients with a mean forced expiratory volume in1s (FEV1) of 38.2 (715.5)% in percent of predicted value (%P), a mean PaO2 value of 9.6 (71.3) kPa, and a mean diffusing capacity of 43.6 (723.0)%P, were subjected to measurements by the multiple elimination inert gas technique (MIGET). The extent of emphysema was determined by HRCTat both full inspiration, emph.I(%) and at full expiration, emph.E(%), with a cut-off limit of 910 Hounsfield Units (HU) using the‘‘Density Mask’’method.The ventilation directed towards high VA/Q areas was 7.3 (710.2)% and the mean ventilation (V-mean) was elevated about three times compared to normal.The mean emph.(I)% and emph.(E) was 45.6 (716.9) and 32.7 (719.0)%, respectively. Significantcorrelations were shown between the emphysema extent and severallung function parameters, but no correlation was found between the emphysema extent and the VA/Q relationships or the blood gas values.Reduced DLCO%P correlated with less high VA/Q ventilation (r=0.73, Po0.05) for the subgroup of COPD patients with DLCO(%P) less than 50% (n=12). Conclusions: In COPD patients, suffering from moderate to severe emphysema without severe blood gas impairment, no correlation was shown between the extent of emphysema, as assessed by HRCT, and the severity of ventilation--perfusion inequality. A substantial collateral ventilation in severe emphysema may be a mechanism that prevents a deterioration in VA/Q relationships and in blood gas levels.r 2002 Published by Elsevier Science Ltd Available online at http://www.sciencedirect.com
Keywords lung function; ventilation ^ perfusion inequality and HRCT in COPD
INTRODUCTION Di¡erent patterns of ventilation ^ perfusion (VA/Q) mismatch in various groups of patients with chronic obstructive pulmonary disease (COPD) have previously been described by Wagner et al. (1). COPD patients with reduced elastic recoil, marked hyperin£ation and a minimum of sputum production have shown substantial ventilation of poorly perfused lung regions (high VA/Q areas) (1). Using xenon gas tracers, measuring the retention of inhaled and infused radioactive xenon gas, it has Received13 November 2001; accepted in revised form18 March 2002 Correspondence should be addressed to: Dr Karin Sandek, Division of Respiratory Medicine and Allergology, Department of Medicine, Karolinska Institutet at Huddinge University Hospital, S-141 86 Stockholm, Sweden Fax: 46 8 7117306. E-mail: [email protected]
been shown that this subgroup of COPD patients has a reduced alveolar blood £ow in relation to lung volumes (2). Compression and destruction of capillaries in the alveolar walls give rise to reduced alveolar blood £ow in emphysema (3,4). Another group of COPD patients with cough and daily sputum production shows a substantial perfusion of low VA/Q regions, probably as a result of reduced ventilation, re£ecting the mechanical airway obstruction and distortion of the small airways (1,2). By the use of the xenon tracer method, it has previously been shown that these bronchitic COPD patients have normal or even increased peripheral perfusion in relation to their lung volumes (2).Thus, not only lung function but also alveolar perfusion and ventilation ^ perfusion ratios are a¡ected by emphysema. However, in COPD patients with severe airway obstruction and with an increase in VA/Q inequality, insu⁄cient equilibration of gas tracers
VENTILATION^PERFUSION INEQUALITY
may lead to an overestimation of ventilation or perfusion to the most severely a¡ected lung regions (5). The multiple inert gas elimination technique (MIGET) has for more than 25 years been used to analyse theVA/Q relationships in COPD (1). Using MIGET, it is possible to accurately re£ect these relationships (at a di¡erence from the xenon tracer method), at least during resting conditions even in severe COPD, irrespective of whether the patients are in a stable condition or if acute respiratory insu⁄ciency has developed (6,7). Mild emphysema is often di⁄cult to be detected by ordinary chest radiograph images (8).The development of a high-resolution computed tomography (HRCT) technique has signi¢cantly improved the possibility of studying the extent and localisation of emphysema in a non-invasive way (8,9). It has previously been shown that the ‘‘Density Mask’’ technique (10) is a reliable method to quantify the extent of emphysema. Microscopic ¢ndings from post-mortem samples and resected lung tissues from COPD patients have shown that a reduced alveolar wall volume/unit lung volume correlates signi¢cantly with mean lung density and to areas of low attenuation shown by computed tomography (CT) (11). The extent of emphysema determined by CT in emphysema patients with varying degrees of air£ow obstruction, has been shown to correlate signi¢cantly with lung function tests such as forced residual capacity (FRC), forced expiratory volume in one second (FEV1), residual volume (RV) and di¡using capacity for carbon monoxide (DLCO) (9,12^14). However, it is not known whether the extent or severity of emphysema assessed by HRCT measurements also relates to the severity of ventilation ^ perfusion inequality. Therefore, we correlated the extent of emphysema assessed by HRCT with ventilation ^ perfusion ratios measured by the multiple inert gas technique in 20 stable COPD patients (with di¡erent levels of airway obstruction). The extent of emphysema assessed by HRCT and VA/Q ratios was then related to spirometry parameters and to the di¡using capacity of carbon monoxide.
PATIENTS Twenty patients (10 smokers and 10 ex-smokers), eight men and 12 women, with a mean duration of COPD of 10 (78) years and a mean age of 60 (78) years were included in our study. Two subjects had homozygotic a-1antitrypsin de¢ciency. None of the patients had a history of atopy or asthma and all su¡ered from irreversible airway obstruction as shown by the FEV1 value which changed less than10% from the baseline value15 min after the inhalation of 2.5 mg nebulised salbutamol. No patient had any other lung disease, other than COPD, and they had all been free from cardio-pulmonary exacerbations for at least 1 month before being included in the study. The patients were selected by consecutive screening
935
when visiting our outpatient clinic.They all received their ordinary medication unchanged for at least 2 months prior to the measurements. Spirometric values remained unchanged for 2 months after the VA/Q tests had been performed, con¢rming that the patients were in stable condition. Arterial oxygen tension (PaO2) was above 8.0 kPa in all the subjects (Table1). Seventeen patients used b2 -inhalers, 10 patients inhaled corticosteroids and four patients used oral diuretics because of mild chronic right-sided heart failure. This study was approved by the Ethical Committee of Huddinge University Hospital and all subjects gave their informed consent.
METHODS Measurement of ventilation^perfusion (VA/Q) ratios Ventilation ^ perfusion ratios were measured using the multiple inert gas elimination technique (15,16). A modi¢ed method was used in which six mixed expired inert gas levels were measured, while arterial levels were estimated from samples of peripheral venous blood taken
TABLE 1. Anthropometric data, spirometric values and extent of emphysema (n=20) Age (years) 60+8
Mean
7SD
EVF (%) BMI (kg/m2) FEV1 (%P) FVC (%P) RV (%P) FRC (%P) TLC (%P) MEF (%P)
45.0 23.6 38.2 74.2 213.6 160.1 121.9 12.2
73.6 74.8 715.5 717.3 747.0 735.3 717.3 7 8.3
PaO2 (kPa) PaCO2 (kPa) DLCO (%P) Mean pulmonary density (HU) Emph. (I) % Emph. (E)%
9.6 5.1 43.6 876.1 45.6 32.7
71.3 70.7 723.0 728.5 716.9 719.0
BMI, body mass index; FEV1, forced expiratory volume in1second; FVC, forced vital capacity;FRC, functional residual capacity; MEF, mean expiratory £ow at 50% of FVC; RV, residual volume;TLC, total lung capacity; DLCO, di¡usion capacity forcarbon monoxide;EVF, erythrocyte volume fraction; (%P), in percentage of predicted; Emph. (I)%, extent of emphysema in percentage of total pulmonary parenchyma at full inspiration; Emph. (E)%, extent of emphysema in percentage of total pulmonary parenchyma at full expiration and HU,Houns¢eld Units.
936
90 min after the commencement of infusion of the inert gases (17,18). The inert gas samples were analysed by gas chromatography (Varian 3300, Varian Associates Inc., CA, U.S.A.). Retention and excretion ratios were computed and the solubility of each inert gas was determined by a two-step procedure, enabling the estimation of the VA/Q distribution. Mean values from three runs were used to achieve a low standard error (18). Oxygen uptake (VO2) and CO2 production were measured by analysing the O2 and CO2 content from a Douglas bag at the same time as the inert gas sampling was carried out (Ametek, Pittsburgh, PA, U.S.A.). Minute ventilation was measured by analysing the volume of the contents of the Douglas bag. The meanVA/Q ratios for ventilation and perfusion distributions are expressed by V-mean and Q-mean, respectively. From the VA/Q distributions, data were derived regarding shunt (perfusion of lung regions with VA/Q ratios o0.005); ‘‘low VA/Q ’’ levels (perfusion of lung regions with 0.005oVA/Qo0.1); ‘‘high VA/Q’’ levels (ventilation of lung regions with 10oVA/Qo100) and dead space (ventilation of lung regions withVA/Q ratios 4100). Cardiac output (CO) (l/min) was assumed to be1/50 of the VO2 (ml/min) (VO2 method). Analysis from two studies in our laboratory (19,20) concerning stable, resting patients with severe COPD and moderate hypoxaemia, showed that the CO estimates by the VO2 method, performed during right heart catheterisation, was not significantly di¡erent from CO assessments using the thermodilution method and that these two methods correlated signi¢cantly (r=0.72, P=0.001). In patients with severe cystic ¢brosis, we have previously shown that changes in CO of 50% are required to signi¢cantly in£uence CO-derived parameters, such as shunting (21). On the basis of these studies, we concluded that the CO estimates (assessed by VO2) and variables derived from CO such as shunting and low VA/Q, in stable, resting COPD patients without severe hypoxaemia, like in the present study, should be accurate enough to result in valid correlations between CO-derived parameters and other variables. The dispersion of ventilation and perfusion for di¡erent VA/Q ratios is expressed as a logarithmic standard deviation of the ventilation distribution (log SDV) and of the perfusion distribution (log SDQ).Thus, the degree of overall ventilation ^ perfusion mismatch is expressed by log SDV and log SDQ. When the peripheral venous sampling technique is used, log SDV and log SDQ are underestimated by about 6% (18) and the upper normal limits regarding log SDVand log SDQ are 0.75 and 0.70, respectively (22).Theoretical work has shown that VA/Q dispersion indices are in£uenced only to a minor degree by estimates of CO (17). Mean values from three runs were used to achieve a low standard error of the method (18). The ¢t of the derived VA/Q distributions to the measured data, expressed as the remaining sum of squares (RSS)
RESPIRATORY MEDICINE
should not exceed 6.0 in more than 50% of the measurements (15,18). The mean RSS from all data in the present study was 1.6 (71.8).
Blood gas analyses Blood gas analyses were performed on radial arterial samples (ABL 520, Radiometer, Copenhagen, Denmark) immediately after VA/Q measurements, while breathing air, as the patients were sitting in a semi-recumbent position.
Spirometry tests Static and dynamic spirometry were performed with the patient in a sitting position. To avoid the in£uence of forced expiration on secretion and thereby gas exchange, the spirometry measurements were always performed after the conclusion of the gas exchange study.The FEV1 (the best value of three was accepted), forced vital capacity (FVC), vital capacity (VC) and mean expiratory £ow at 50% of FVC (MEF 50%) were measured by dynamic spirometry. The functional residual capacity (FRC) was determined by body plethysmography. The residual volume (RV) and the total lung capacity (TLC) were then calculated from the FRC value. Dynamic spirometry and lung volume measurements were carried out with a Pulmonary Function Laboratory 2400 and 6200, respectively (Sensor Medics BV, Bilthoven, The Netherlands). The values from the spirometry tests are presented as percentage of predicted (%P). The normal values were selected from equations of the European Community for Coal and Steel (23). Di¡using capacity for carbon monoxide (DLCO) was measured by the single breath method (24). The mean value of two technically acceptable measurements was used. None of the patients had anaemia or polycythaemia and consequently no correction for the Hb-level was needed.
High-resolution computed tomography (HRCT) CT scanning was performed in the supine position by a Philips Tomoscan SR 7000 equipment (Philips Medical System). The HRCT technique (1.5 mm collimation, a scanning time of 2.0 s, 140 kV and 150 mA) was used and the images were acquired from the apices to the diaphragm at 2 cm intervals. Resting time before CT-scanning was performed was at least 2 minutes and the duration for the whole investigation was 10 min. In each patient, 11^15 layers were scanned at inspiration to TLC and at expiration to RV. The mean normal density (electronic density of the lung parenchyma relative to water) in healthy lungs measured by CTexpressed in Houns¢eld
VENTILATION^PERFUSION INEQUALITY units (HU) values is 722 (744) HU for women and 746 (742) HU for men (25).
Emphysema extent To quantify the extent of emphysema the‘‘Density Mask’’ programme of the CT-scanner was used without contrast medium (10). With a track-ball our radiologist (HR) outlined the lung parenchyma excluding solid bone structure, hilum, mediastinum, air in the trachea and in the main bronchi. In each HRCT-section, the scanner measured the area of pixels (r0.5 mm2) with attenuation values between 400 HU and 1000 HU, corresponding to lung tissue (26). Areas with density levels between 910 HU and 1000 HU were considered emphysematous (27^29). The density of the total pulmonary tissue at full end inspiration was automatically calculated by the scanner’s computer program assessed as the mean value of attenuation (expressed in HU). This was performed at three di¡erent levels; at the top of the aortic arch, at the origin of the right lower lobe bronchus, and at 3 cm above the top of the diaphragm (13). At each of these levels, the ratio of lung areas with low attenuation (o910 HU) to total lung area (400 HU to 1000 HU) was expressed as the percentage of emphysema (emph.%) and was calculated both at full inspiration denominated emph.(I)% and at full expiration denominated emph.(E)%.The cut-o¡ points for separating di¡erent groups of emph.(E)% were modi¢ed from previous studies (29,30). Our radiologist who examined the HRCTscans had no knowledge of the results from the ventilation ^ perfusion ratio measurements or from the pulmonary function tests.
Statistical analysis The data are presented as a mean 7SD Correlations between blood gases and lung function tests was determined by linear regression. Only non-parametric statistical method was used since the extent of emphysema and VA/Q ratios did not show a normal distribution. Correlations between variables were assessed by the Spearman rank test and di¡erences between groups were examined using the Mann ^Whitney U-test. Regarding the di¡erences between the three groups, the Kruskal ^Wallis test was used. The importance of di¡erent variables for the variability of a dependent parameter was assessed by stepwise regression.
RESULTS The chest radiograph images showed hyperin£ation in all patients. HRCTshowed areas with low attenuation levels
937
scattered around the entire lung. No images showed paraseptal or subpleural localised air cysts or ‘‘bullous’’ patterns.The automatically calculated values for the lung density and the extent of emphysema are shown inTable 1. The mean lung density level was 876.1 (728.5)HU, which is lower than normal (722 HU 7 44 for women and 746HU742 for men) (25). Emph.(I)% for the entire group was 45.6 (716.9)% and emph.(E)% was 32.7(719.0)%. Due to technical reasons, measurements of the mean lung density level were performed in only13 patients. Signi¢cant di¡erences between these 13 patients compared to the remaining seven patients were only their mean levels of TLC and FRC (Po0.05, respectively). Two patients had mild airway obstruction [FEV1 60 ^ 70(%P)], seven had moderate [FEV1 40 ^59(%P)], and 11 had severe airway obstruction [FEV1o40(%P)] (Table 1). All patients su¡ered from severe hyperin£ation [RV%P of 213.6 (747.0)%] and a reduced di¡using capacity [DLCO(%P) of 43.6 (723.0)%] compared to normal (23,24). They were all eucapnic with a mean PaCO2 of 5.1 (70.7) kPa and a mean PaO2 close to normal of 9.6 (71.3) kPa (Table 1). High VA/Q value dominated the ventilation ^ perfusion inequality pattern. About 7% of the ventilation was directed towards highVA/Q areas (normal valueo1%) yielding an increase in the V-mean level to 2.9 (71.6) compared to the level of healthy individuals (0.7^1.2) (Table 2).The dispersion of ventilation, log SDV, was also increased compared to normal (normal value r0.75) (22), to a mean value of 0.95 (70.43) and the dispersion of perfusion, log SDQ, was at the upper normal limit of 0.68 (70.20) (normal value r0.70) (22). About 2% of CO was directed to shunt areas and almost no perfusion was directed towards low VA/Q areas (Table 2).
Correlations between lung function and HRCT Close inverse correlations were found between emph.(E)% and FEV1(%P) (Fig. 1) as well as to other indices of air£ow obstruction such as FVC(%P), FEV1/ VC(%) and MEF 50%(%P) (Table 3). A strong inverse correlation was also shown between emph.(E)% and the diffusing capacity (Fig. 2). Emph.(E)% was positively correlated with indices of hyperin£ation such as RV(%P) (Table 3). However, in three patients with rather severe airway obstruction [a median FEV1(%P) of 53 (49^56)% and a median RV(%P) of 165 (153^207)%] a median emph.(E)% of only 5.5 (3.6 ^11.4)% was shown by HRCT. By stepwise regression analysis for all our COPD patients, we noted that 69% of the variability in emph.(E)% was explained by FEV1/VC(%P) and a further 8% by DLCO(%P). The parameters FEV1(%P), RV(%P), FRC(%P), MEF50%(%P), Q-mean and V-mean however, failed to
938
RESPIRATORY MEDICINE
TABLE 2. Ventilation-perfusion relationships (n=20) COPD patients
V-mean Q-mean Log SDV Log SDQ Shunt Low VA/Q HighVA/Q Dead space VE (l/min)
Healthy individuals
Mean
7SD
2.9 1.3 0.95 0.68 1.9 0.1 7.3 30.0 10.2
7 1.6 7 0.4 7 0.43 7 0.20 7 1.4 7 0.4 710.2 7 7.7 7 2.5
Mean 0.9 0.7 r0.75 r0.70 0.08 0 0 39.4
7SD 70.2 7 0.1 F F 70.15 0 0 75.6
The normal values regardinglog SDVand log SD Q are adopted from reference (22) and the other normal values are adopted from reference (30). V-mean, The mean VA/Q ratio for ventilation; Q-mean, The mean VA/Q ratio for perfusion; log SDV, The dispersion of ventilation; log SDQ, The dispersion of perfusion; shunt, VA/Qo0.005 in per cent of cardiac output; low VA/Q, 0.005oVA/Qo0.1 in per cent of cardiac output; high VA/Q, 10oVA/Qo100 in per cent of minute ventilation; dead space, VA/ Q4100 in per cent of minute ventilation and VE,Minute ventilation
Correlations between lung function, blood gas exchange, ventilation^perfusion ratios and extent of emphysema
70
FEV1 (%P)
60 50 40 30 20 10 0 0
10
20
30 40 Emph. (E)%
50
60
70
FIG. 1. Spearmanrankcorrelations betweenthe extentof emphysema atfull expiration emph.(E) (%) as shown by HRCTscans and forced expired volume in one second (FeV1) (%P) in 20 COPD patients, (r=0.83, Po0.001).
explain the remaining variability in emph.(E)%. Similar albeit slightly weaker correlations were found between emph.(1)% and indices of air£ow obstruction, hyperin£ation and DLCO%P (Table 3). No correlations were found between blood gas levels and emph.(E)%. For technical reasons, the mean lung density value at full inspiration [emph.(I)%] was measured in only 13 of our COPD patients. The emph.(I)% for these patients was shown to be inversely and signi¢cantly correlated with emph.(E)% (r=0.70, Po0.05) without any correlation to the air£ow obstruction, hyperin£ation, VA/Q inequality or low DLCO%P (Table 3). However, signi¢cant correlations were found between emph.(E)% and spirometry parameters for the13 patients in whom the mean lung density was measured as for our entire patient group (data not shown).
No signi¢cant correlations were found between air£ow obstruction (measured by spirometry) and ventilation ^ perfusion inequality or between blood gas values and VA/Q ratios. No signi¢cant correlations between reduced DLCO(%P) and any VA/Q parameter (such as high VA/Q ventilation and elevated V-mean) were found. For the entire patient population, no signi¢cant correlations were found between di¡erent VA/Q ratios or blood gas levels vs emph.(I)% or vs emph.(E)%. However, the di¡erence in emph.% at full inspiration and at full expiration deceased with increasing emphysema extent (r=0.47, Po0.05) (Fig. 3). A reduced BMI correlated with emph.(E)% (r=0.53, Po0.05). The PaO2 level related signi¢cantly to the DLCO(%P) level (r=0.73, Po0.001) as well as to the minute ventilation (r=0.48, Po0.05). The PaCO2 level was inversely correlated to FEV1(%P), DLCO(%P) and VC(%P) (r=0.56 to 0.67, Po0.01 for all).
Subgroup analyses Comparisons between groups consisting of ¢ve patients with emph.(E)% of r16% (trivial emphysema), ten patients with emph.(E)% of 17^50 % (moderate emphysema) and ¢ve patients with emph.(E)% of 450% (severe emphysema) (29,31) (Table 4) concerning spirometric values, VA/Q ratios and emph.(%) showed that patients with moderate or severe emphysema had a signi¢cantly more pronounced air£ow obstruction and
VENTILATION^PERFUSION INEQUALITY
939
TABLE 3. Correlations between the extent of emphysema at full inspiration in percentage of healthy pulmonary parenchyma [Emph.(I)%] and at full expiration in percentage of healthy pulmonary parenchyma [Emph.(E)%] and mean lung density atthree levels assessed by HRCTvs spirometric values and VA/Q relationships Emph. (I)% (n=20)
Emph. (E)% (n=20)
Mean density (n=13)
r-Value
P-Value
r-Value
r-Value
NS NS
0.87*** 0.81*** 0.70** 0.77*** 0.57** 0.78** 0.73** 0.15 0.22
0.76** 0.69** 0.69** 0.71*** 0.57** 0.57* 0.60** 0.14 0.25
FEV1/VC (%) FEV1 (%P) FRC (%P) RV (%P) TLC (%P) MEF 50% (%P) DLCO (%P) High VA/Q (%) V-mean (%)
P-Value
NS NS
0.54 0.38 0.37 0.36 0.35 0.28 0.41 0.13 0.49
P-Value NS NS NS NS NS NS NS NS NS
*Po0.05, **Po0.01, ***Po0.001regarding signi¢cant correlations.
100 90 80 DLCO (%P)
70 60 50 40 30 20 10 0 0
10
20
30 40 Emph. (E)%
50
60
70
FIG. 2. Spearman rank correlations between the extent emphysema at full expiration emph.(E) %, as shown by HRCTscans and the di¡using capacity for carbon monoxide (DLCO) (%P) in 20 COPD patients, (r=0.71, Po0.01).
80
Emphysema insp. extent
Emphysema expir. extent
Emphysema extent (%)
70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Number of the patient
FIG. 3. Spearman rankcorrelation between the extent of emphysema as shown by HRCTscans at full inspiration; (upper line) and at full expiration (lower line) emph.(I) (%) and at full expiration; emph.(E) (%) in 20 COPD patients, (r=0.94, P=0.001).
hyperin£ation compared with the trivial emphysema group. Air£ow obstruction and hyperin£ation were also signi¢cantly more pronounced in patients with severe emphysema as compared with patients with moderate emphysema. There was a signi¢cantly lower DLCO(%P) value in patients with moderate or severe emphysema compared with the patients in the trivial emphysema group, but there was no signi¢cant di¡erence with regard to DLCO(%P) value between patients with moderate or severe emphysema. All these di¡erences were con¢rmed by the Kruskal ^Wallis test (Po00.1 for emph.(E)% and for the other variables the P-values varied between 0.01 and 0.05). Ventilation towards high VA/Q areas was signi¢cantly increased in the subgroup with moderate emphysema compared with the trivial emphysema subgroup. In line with this result there was a trend for an increased ventilation distribution (log SDV) including an insigni¢cantly increased shunting in the moderate emphysema subgroup compared with the trivial emphysema subgroup. The mean VA/Q ratio for ventilation (V-mean) was signi¢cantly increased compared with normal values for all subgroups (Table 2) and there were trends towards an increased ventilation of high VA/Q areas and for perfusion of shunting areas in patients with severe emphysema compared with the trivial emphysema subgroup. However, for the 15 patients with an emph.(%E) of 416% no correlation was found between the increasing extent of emphysema and VA/Q inequality. In 13 COPD patients, with a di¡using capacity for carbon monoxide of o50%P, V-mean and high VA/Q ventilation tended to correlate with an increase in DLCO(%P) level, but if one patient with DLCO(%P) 445% was excluded, these correlations were no longer signi¢cant (Figs. 4(a) and (b)).
940
RESPIRATORY MEDICINE
TABLE 4. Di¡erences with regard to spirometric values and VA/Q relationships in COPD patients with di¡erent extent of emphysema as assessed by HRCT Emph. extent r16% (n=5) Mean 7SD FEV1/VC (%) FEV1 (%P) FRC (%P) RV (%P) TLC (%P) DLCO (%P) Log SDV Log SDQ High VA/Q (%) V-mean (%) Q-mean (%) VE (L/min) Three levels: Emph. (I) % Emph. (E)%
Emph. extent17^50% (n=10) Mean 7SD
72.077.8 57.477.1 121.8718.1 159.6735.1 107.079.5 67.0728.4 0.7170.25 0.6070.05 1.172.5 2.170.4 1.470.1 9.373.1 28712 1076
45.777.5** 34.3712.7** 164.1733.7** 217.9733.9** 123.8713.9* 39.4714.8* 1.0370.37 0.7170.21 8.979.7* 2.871.1 1.270.3 10.472.2 44711* 3179**
Emph. extent>50% (n=5) Mean 7SD 36.077.3**w 26.678.1** 186.2713.3** 258.8720.9**w 139.479.4**w 28.4714.1* 1.0370.65 0.7070.29 10.5714.2 3.972.6* 1.470.5 10.772.9 6774** 5974**z
NS NS NS NS NS
NS NS NS NS NS
35
9 8 7 6 5 4 3 2 1 0
30 25 High VA/Q
V-mean
*Po0.05, **Po0.01 regarding signi¢cant di¡erences between COPD patients with extent of emphysema 416% and COPD patients with extent of emphysema r16 %. w Po0.05, zPo0.01 regarding signi¢cant di¡erences between COPD patients with extent of emphysema >50% and COPD patients with extent of emphysema17^50%. Emph. (I)%,Extent of emphysema at full inspiration in percentage of healthy pulmonary parenchyma. Emph. (E)%,Extent of emphysema at full expiration in percentage of healthy pulmonary parenchyma.
20 15 10 5 0
0
5
10
(a)
15
20 25 30 DLCO (%P)
35
40
45
50
0 (b)
5
10
15
20 25 30 DLCO (%P)
35
40
45
50
FIG. 4a. Spearman rank correlations between mean of VA/Q ratios for ventilation (V-mean) and the di¡using capacity for carbon monoxide (DLCO) (%P) in13 COPD patients with DLCO (%P) o50%, (r=0.57, Po0.05).4b Spearman rankcorrelations between ventilation of lungregions with10oVA/Qo100, (highVA/Q) and di¡usingcapacity forcarbon monoxidein13 COPD patients, with DLCO (%P) o50%, (r=0.57, Po0.05).
DISCUSSION Assessment of emphysema by the HRCT method HRCT may detect emphysematous areas which may be as small as 0.5 cm2 (32). The ‘‘Density Mask’’ method has the advantage of being less prone to observer bias compared to visual scorings, since the outlined regions of interest automatically denote the lung areas which are
below the prede¢ned threshold of attenuation (33). However, in healthy individuals and to a lesser extent in emphysematous patients, lung areas of low attenuation level may appear towards the end of the inspiratory manoeuvre to TLC due to hyperin£ation (13,14,29). The low attenuation areas may also appear as a result of obstruction of central or peripheral airways in patients with chronic air£ow limitation, giving rise to air trapping and ¢xed hyperin£ation (13,14). Furthermore, these low
VENTILATION^PERFUSION INEQUALITY
attenuation areas may appear due to the existence of truly emphysematous areas (13,14). In our study, as in previous studies (13,14,34), the di¡erences in the size of areas with low attenuation shown by HRCT, between end inspiration and the end expiration diminished in patients with more widespread emphysema during end expiration. This corroborates the notion that less air is transported to and from the lungs as emphysema develops (13,14). Previous studies have shown that a cut-o¡ limit of 900 or 910 HU, has been found to separate normal from emphysematous lung tissue at full inspiration on CT or HRCT (10,27,28). Recent research, however, implies that a cut-o¡ limit of 950 HU used on CT scans, taken at full inspiration, de¢nes an even more exact limit between hyperin£ated healthy and emphysematous pulmonary tissue (33). It has also been noted previously that similar correlations between the severity of emphysema and spirometry tests are found regardless if the cut-o¡ limit is set to 910 or 950 HU (35).The risk of including pulmonary areas with low attenuation level due to hyperin£ation instead of emphysema is reduced if the CT scans are obtained at full expiration (13,14,27,31) and by this mechanism the in£uence of the supine posture was probably minimised. By the use of standardised coaching, as in our study, the images can be obtained at reproducible levels of expiration (31,34).The extent of emphysema assessed at full expiration by the ‘‘Density Mask’’ technique has previously been shown to be one of the most accurate methods to distinguish patients with no or trivial emphysema from those with moderate and severe emphysema (13,29,30).We therefore chose to assess the extent of emphysema in our patients [emph.(E)%] with a cut-o¡ limit for normal extent of low attenuation areas of r16% (13,27,31). Furthermore, in agreement with earlier studies (13), we found that HRCT scans from three levels correlate closely with images assessed at 10 ^15 levels. Statistical analysis was therefore performed using average values of HRCT scans from only three levels (see methods).
Correlations between lung function and HRCT In the present investigation, as in previous studies, highly signi¢cant correlations were found between indices of air£ow obstruction, hyperin£ation and the extent of emphysema (13,27,29,31,34,35). In agreement with previous results, somewhat more signi¢cant correlations were recorded between lung function tests and emph.(E)% as compared with emph.(I)% (13,27,29,31,34,35). The inverse correlation between the extent of emphysema and FEV1 is primarily dependent on the reduced elastic recoil inherent to emphysema (36). In disagreement with previous reports (8,13,31), indices of air£ow
941
obstruction seemed to be of greater importance for the variability in extent of emphysema than the DLCO(%P) level. A reduction in the di¡usion capacity in COPD may depend on parenchymal and capillary tissue destruction due to the development of emphysema (3,8), which in turn reduces the surface area available for gas exchange. Furthermore, DLCO may be diminished due to small airway disease (37,38) which decreases the in£ux of air into the alveoli and thereby the alveolar ventilation and the gas exchange becomes reduced. Small airway disease may have contributed to the ¢nding that four of our patients had DLCO(%P) below 30% and an emph.(E)% of only 30 ^ 45% (Fig. 2). In disagreement with several other investigations (27,31,34), mean lung density did not correlate to any index of lung function or gas exchange impairment in the present study. The mean lung density measures the proportions of air and liquid in both normal and emphysematous parts of the lung which may explain why this parameter is less sensitive to lung function impairments compared to measurements of the extent of emphysema (27). By measuring the extent of emphysema by HRCT, only the diseased parts of the lung parenchyma are highlighted. Furthermore, in several earlier studies, a wide range of subjects were included, ranging from patients with rather trivial emphysema and normal lung function to patients with large emphysematous areas with severely reduced lung function (27,31,34). Our patients comprised an rather homogenous group, since 80% of our subjects had an FEV1(%P) of less than 50%.This relatively narrow range of forced expiratory volumes may have obscured the plausible correlations between the mean lung density level and impaired lung function.
VA/Q relationships and HRCT The VA/Q ratios assessed by MIGETand the single breath di¡using capacity for carbon monoxide are sensitive indicators of mild emphysema (6,8). In agreement with previous studies of patients with stable COPD and substantial emphysema, theVA/Q relationships in the present study were characterised by substantial ventilation to lung areas with poor perfusion (high VA/Q areas) resulting in an increased V-mean and an increased dispersion of ventilation, whereas shunting and perfusion of low VA/Q areas were scarcely present (2,6,28,39). Moreover, as expected (2,6,28,39), ventilation of poorly perfused areas was higher in patients with substantial emphysema [emph.(E)%P416] than in patients with less severe emphysema [emph.(E)%Pr16] as shown by HRCT. Furthermore, for the entire patient group, no correlations were found between lung function, gas exchange and VA/Q ratios. In previous studies of COPD patients with nocturnal or daytime hypoxaemia, signi¢cant inverse correlation between ventilation of poorly
942
perfused lung areas and di¡usion capacity has been shown (39,40). There are several possible explanations for the absence of correlations betweenVA/Q ratios and the extent of emphysema, di¡using capacity or blood gas levels in the present study. Firstly, a reduced alveolar surface due to emphysema is accompanied by an equal reduction in capillary volume (2,3), preventing the development of severely abnormal VA/Q ratios. Secondly, in most COPD patients an increase in the intra-alveolar pressure occurs due to in£ammation of the bronchioli and/or due to reduced elastic recoil which may decrease the gas £ow when gas is leaving the alveoli (6,28,35,41).The increased intra-alveolar pressure compresses the capillaries and prevents shunting (2,3) and a reduction in the PaO2 level. When air continues to reach these poorly perfused lung regions, an increase in high VA/Q areas will be present as shown in our patients (2,3). Thirdly, as the emphysematous process progresses, the acinar fenestration becomes more extensive and the collateral ventilation will increase (42), preventing a more pronounced VA/Q mismatch despite a reduced gas exchange surface, re£ected by a low DLCO(%P).This may be a possible mechanism for the subgroup of our patients with DLCO(%P) of less than 45%, to have a tendency to an increased ventilation to high VA/Q areas together with augmented di¡using capacity (Figs. 4(a) and (b)). Accordingly, it has been shown that the degree of ventilation towards high VA/Q areas in emphysematous patients with low pulmonary parenchyma density correlated positively with the level of the diffusing capacity (2). Finally, small airway disease may lead to an increasing tendency of atelectases and an increase in the perfusion of poorly ventilated lung areas (28).The patients in the present study had less perfusion of poorly ventilated lung areas compared with the patients in a previously investigated COPD patient group su¡ering from daytime hypoxaemia (40).The patients presently investigated, may thus have been su¡ering from less serious small airway disease than the group of patients with daytime hypoxaemia (40). These mechanisms may explain why 80% of our subjects had normal or almost normal PaO2 (X8.5 kPa) despite that 75% had moderate to severe emphysema measured by HRCT. Severe emphysema is associated with increased basal inspiratory drive (43) and an increase in energy consumption of the respiratory muscles even in patients without daytime hypoxaemia (44). Our patients with severe emphysema thus maintained relatively normal blood gas level presumably at the cost of an increased work of breathing (50% had a BMI below 21kg/m2). Lung function tests and inert gas measurements were performed in the sitting position, whereas the HRCT scans were performed in the supine position. However, the in£uence of the posture was probably small since the HRCTscans were measured at full inspiration and at full expiration.
RESPIRATORY MEDICINE
In conclusion: The extent of emphysema assessed by HRCT is associated with a reduced di¡using capacity for carbon monoxide, hyperin£ation and reduced forced expiratory volumes. In COPD patients without severe blood gas disturbances su¡ering from moderate-to-severe extent emphysema assessed by HRCT images, elevated ventilation of poorly perfused lung areas is found compared to patients with trivial extent of emphysema. However, there is no correlation between the severity of extent of emphysema and the VA/Q inequality. Collateral ventilation and other mechanisms probably prevent VA/Q relationships and blood gas levels to deteriorate at the same pace as the progression of the emphysematous process.
REFERENCES 1. Wagner PD, Dantzker DR, Dueck R, Clausen JL, West JB. Ventilation–perfusion inequality in chronic obstructive pulmonary disease. J Clin Invest 1977; 59: 203–216. 2. Brudin LH, Rhodes CG, Valind S-O, Buckingham PD, Jones TJ, Hughes JMB. Regional structure–function correlations in chronic obstructive lung disease measured with positron emission tomography. Thorax 1992; 47: 914–921. 3. Morrison NJ, Abboud RT, Mu¨ller NL, et al. Pulmonary capillary blood volume in emphysema. Am Rev Respir Dis 1990; 141: 53–61. 4. Barbera` JA, Riverola A, Roca J, et al. Pulmonary vascular abnormalities and ventilation–perfusion relationships in mild chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1994; 149: 423–429. 5. Brudin L., Valind S-O, Rhodes CG. Error analysis of combined measurements of regional ventilation and V/Q ratio using positron emission tomography. Physiol Med Biol 1992; 37: 1077–1093. 6. Barbera` J, Ramirez J, Roca J, et al. Lung structure and gas exchange in mild chronic obstructive pulmonary disease. Am Rev Respir Dis 1990; 141: 895–901. 7. Torres A, Reyes A, Roca J, Wagner PD, Rodriguez-Roisin R. Ventilation–perfusion mismatching in chronic obstructive pulmonary disease during ventilator weaning. Am Rev Respir Dis 1989; 140: 1246–1250. 8. Kuwano K, Matsuba K, Ikeda T, et al. The diagnosis of mild emphysema. Correlation of computed tomography and pathology scores. Am Rev Respir Dis 1990; 141: 169–178. 9. Gurney JW, Jones KK, Robbins RA, et al. Regional distribution of emphysema: correlation of high-resolution CT with pulmonary function tests in unselected smokers. Radiology 1992; 183: 457– 463. 10. Mu¨ller NL, Staples CA, Miller RR, Abboud RT. ‘‘Density Mask’’. An objective method to quantitate emphysema using computed tomography. Chest 1988; 94: 782–787. 11. Gould GA, MacNee W, McLean A, etal. CT measurements of lung density in life can quantitate distal airspace enlargementFan essential defining feature of human emphysema. Am Rev Respir Dis 1988; 137: 380–392. 12. Coxson HO, Rogers RM, Whittall KP, et al. A quantification of the lung surface area in emphysema using computed tomography. Am J Resp Crit Care Med 1999; 159: 851–856. 13. Miniati M, Filippi E, Falaschi F, et al. Radiologic evaluation of emphysema in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1995; 151: 1359–1367. 14. Eda S, Kubo K, Fujimoto K, Matsuzawa Y, Sekiguchi M, Sakai F. The relations between expiratory chest CT using helical CT and
VENTILATION^PERFUSION INEQUALITY
15.
16.
17.
18.
19.
20.
21.
22.
23. 24.
25. 26.
27.
28.
29.
pulmonary function tests in emphysema. Am J Respir Crit Care Med 1997; 155: 1290–1294. Wagner PD, Saltzman HA, West JB. Measurement of continuous distributions of ventilation–perfusion ratios: theory. J Appl Physiol 1974a; 36: 588–599. Wagner PD, Nauman PF, Laravuso RB. Simultaneous measurement of eight foreign gases in blood by gas chromatography. J Appl Physiol 1974b; 36: 600–605. Wagner PD, Smith CM, Davies NJH, McEvoy RD, Gale GE. Estimation of ventilation–perfusion inequality by inert gas elimination without arterial sampling. J Appl Physiol 1985; 59: 376–383. Wagner PD, Hedenstierna G, Bylin G, Lagerstrand L. Reproducibility of the multiple inert gas elimination technique. J Appl Physiol 1987; 62: 1740–1746. Bratel T, Hedenstierna G, Nyquist O, Ripe E. The effect of a new calcium antagonist, elodipine, on pulmonary hypertension and gas exchange in chronic obstructive lung disease. Eur J Respir Dis 1985; 67: 244–253. Bratel T, Hedenstierna G, Lundquist H, Nyquist O, Ripe E. Cardiac function and central haemodynamics in severe chronic obstructive lung disease. Acute and long-term effects of felodipine. Eur J Respir Dis 1988; 1: 262–268. Lagerstrand L, Hjelte L, Jorulf H. Pulmonary gas exchange in cystic fibrosis: basal status and the effect of i.v. antibiotics and inhaled amiloride. Eur J Respir Dis 1999; 14: 686–692. Cardu´s J, Burgos F, Diaz O, et al. Increase in pulmonary ventilation–perfusion inequality with age in healthy individuals. Am J Respir Crit Care Med 1997; 156: 648–653. Quanjer PhH. ed. Standardised lung function testing. Bull Eur Physiopath Respir 1983; 9 (Suppl. 5): 45–51. Cotes J. Measurement of transfer factor (diffusing capacity) and its subdivisions. In: Lung FunctionFAssessment on Application in Medicine, 5th edn. Oxford: Blackwell Publishers, 1993: 299–324. Kalef-Ezra J, Karantanas A, Tsekeris P. CT measurement of lung density. Acta Radiol 1999; 40: 333–337. Archer DC, Coblenz CL, de Kemp RA, Nachmias C, Norman G. Automated in vivo quantification of emphysema. Radiology 1993; 188: 835–838. Kinsella M, Mu¨ller NL, Abboud RT, Morrison NJ, DyBuncio A. Quantitation of emphysema by computed tomography using a ‘‘Density Mask’’ program and correlation with pulmonary function tests. Chest 1990; 97: 315–321. Yamaguchi K, Mori M, Kawai A, Takasugi T, Oyamada Y, Koda E. Inhomogeneities of ventilation and the diffusing capacity to perfusion in various chronic lung diseases. Am J Respir Crit Care 1997; 156: 86–93. Dowson LJ, Guest PJ, Hill SL, Holder RL, Stockley RA. Highresolution computed tomography scanning in a1 antitrypsin deficiency: relationships to lung function and health status. Eur Respir J 2001; 17: 1097–1104.
943
30. Wagner PD, Laravuso RB, Uhl RR, West JB. Continuos distributions of ventilation–perfusion ratios in normal subjects breathing air and 100% O2. J Clin Invest 1974c; 54: 54–68. 31. Knudson RJ, Standen JR, Kaltenborn WT, et al. Expiratory computed tomography for assessment of suspected pulmonary emphysema. Chest 1991; 99: 1357–1366. 32. Miller RR, Mu¨ller NL, Vedal S, Morrison NJ, Staples CA. Limitations of computed tomography in the assessment of emphysema. Am Rev Respir Dis 1989; 139: 980–983. 33. Bankier AA, de Maertelaer V, Keyzer C, Genevois PA. Pulmonary emphysema: subjective visual grading versus objective quantification with macroscopic morphometry and thin-section CT densitometry. Radiology 1999; 211: 851–858. 34. Lamers RJ, Thelissen GR, Kessels AG, Wouters EF, van Engelshoven JM. Chronic obstructive pulmonary disease: evaluation with spirometrically controlled CT lung densitometry. Radiology 1994; 193: 109–113. 35. Park KJ, Bergin CJ, Clausen JL. Quantitation of emphysema with three dimensional CT densitometry: comparison with twodimensional analysis, visual emphysema scores and pulmonary function test results. Radiology 1999; 211: 541–547. 36. Nagai A, Yamawaki I, Takizawa T, Thurlbeck WM. Alveolar attachments in emphysema of human lungs. Am Rev Respir Dis 1991; 144: 888–891. 37. Gelb AF, Schein M, Kuei J, et al. Limited contribution of emphysema in advanced chronic obstructive pulmonary disease. Am Rev Respir Dis 1993; 147: 1157–1161. 38. Gelb AF, Zamel N, Hogg JC, Mu¨ller NL, Schein MJ. Pseudophysiologic emphysema resulting from severe small-airways disease. Am J Respir Crit Care 1998; 158: 815–819. 39. Sandek K, Andersson T, Bratel T, Lagerstrand L. Ventilation– perfusion inequality in nocturnal hypoxaemia due to chronic obstructive lung disease (COLD). Clin Physiol 1995; 15: 499–513. 40. Sandek K, Bratel T, Hellstro¨m G, Lagerstrand L. Ventilation– perfusion inequality and carbon dioxide sensitivity in hypoxaemic chronic obstructive pulmonary disease (COPD) and effects of long-term oxygen treatment (LTOT) for six months. Clin Physiol 2001; 5: 584–593. 41. Ranieri VM, Dambrosio M, Brienza N. Intrinsic PEEP and cardiopulmonary interaction in patients with COPD and acute ventilatory failure. Eur Respir J 1996; 9: 1283–1292. 42. Morrell NW, Wignall BK, Biggs T, Seed WA. Collateral ventilation and gas exchange in emphysema. Am J Respir Crit Care Med 1994; 150: 635–641. 43. Sandek K, Andersson T, Bratel T, Hellstro¨m G, Lagerstrand L. Sleep quality, carbon dioxide responsiveness and hypoxemic patterns in nocturnal hypoxaemia due to chronic obstructive pulmonary disease (COPD) without daytime hypoxaemia. Respir Med 1999; 93: 79–87. 44. Shindoh C, Hida W, Kikuchi Y, et al. Oxygen consumption of respiratory muscles in patients with COPD. Chest 1994;105: 790–797.