Cardiopulmonar y Imaging • Original Research Chen et al.
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD Cardiopulmonary Imaging Original Research
Correlation of Pulmonary Function Indexes Determined by LowDose MDCT With Spirometric Pulmonary Function Tests in Patients With Chronic Obstructive Pulmonary Disease
Huai Chen1 Rong-chang Chen2 Yu-bao Guan1 Wen Li1 Qin Liu1 Qing-si Zeng1
OBJECTIVE. The objective of our study was to evaluate the correlation between pulmonary function indexes determined by low-dose MDCT and those obtained from routine spirometric pulmonary function tests (PFTs) in patients with chronic obstructive pulmonary disease (COPD). MATERIALS AND METHODS. Lung function of patients with COPD stages 0–III was evaluated by both MDCT and spirometric PFTs. Scanning was performed at maximum end-inspiration and maximum end-expiration. RESULTS. A very strong correlation was found between extrapolated expiratory lung volume (LVex) and COPD stage (r = 0.802, p < 0.001) and between extrapolated LVex and the ratio of forced expiratory volume in 1 second and percentage forced vital capacity (FEV1/ FVC%) (r = –0.831, p < 0.001). Moreover, strong positive correlations were found between inspiratory lung volume (LVin) and total lung capacity (TLC) (r = 0.658, p < 0.001), LVex and residual volume (RV) (r = 0.683, p < 0.001), extrapolated LVex and RV (r = 0.640, p < 0.001), LVex and RV/TLC (r = 0.602, p < 0.001), LVex/LVin and RV/TLC (r = 0.622, p < 0.001), extrapolated LVex and RV/TLC (r = 0.663, p < 0.001), and LVex and COPD stage (r = 0.697, p < 0.001). CONCLUSION. Low-dose MDCT lung function indexes correlate well with spirometric PFT results, and the highest correlation is at end-expiration. Low-dose MDCT may be useful for evaluating lung function in patients with COPD.
Chen H, Chen RC, Guan YB, Li W, Liu Q, Zeng QS
Keywords: chronic obstructive pulmonary disease, low-dose MDCT, pulmonary function, quantification DOI:10.2214/AJR.12.10501 Received December 24, 2012; accepted after revision June 7, 2013. Supported by grant A2011013 from Guangzhou Medical College. 1 Department of Radiology, the First Affiliated Hospital of Guangzhou Medical University, No. 151, Yan Jiang Rd, Yuexiu District, Guangzhou City, Guangdong Province 510120, China. Address correspondence to Q. Zeng ([email protected]). 2
National Key Laboratory of Respiratory Diseases, the First Affiliated Hospital of Guangzhou Medical University, Guangdong Province, China.
AJR 2014; 202:711–718 0361–803X/14/2024–711 © American Roentgen Ray Society
C
hronic obstructive pulmonary disease (COPD) has been the fourth leading cause of death and is projected to rank fifth in 2020 as a worldwide burden of disease according to a study published in the American Journal of Respiratory and Critical Care Medicine [1]. In China, COPD is a major health concern, and the results of a study conducted in seven regions including a total of 20,245 adults showed the prevalence of COPD to be as high as 8.2% in persons 40 years old or older [2]. Thus, early diagnosis and early treatment of COPD are crucial for patients. The diagnosis of COPD depends on pulmonary function tests (PFTs) and spirometry. A ratio of forced expiratory volume in 1 second and percentage forced vital capacity (FEV1/FVC%) of less than 70% predicted in a patient with a postbronchodilator FEV1 of less than 80% of the predicted value is diagnostic for COPD [3]. It is well known that the lung has strong compensatory ability, and injury to more than
30% of lung tissues is typically necessary to result in abnormal lung function. By the time abnormal lung function is detected, irreparable damage has likely occurred. Dual-phase MDCT with a routine radiation dose has also been used for the diagnosis of COPD with favorable results [4–11]. However, the amount of radiation patients are exposed to is high. Standard-dose CT has also been used to objectively evaluate COPD; however, these studies were primarily built on the basis of lung cancer screening by selecting several CT slices obtained during deep inspiration for the evaluation of pulmonary function [12]. Quantitative studies of pulmonary function utilizing CT usually use images of both inspiration and expiration phases, which increases the radiation dose to patients. It is clear that alternatives with a reduced radiation dose are needed. At present, low-dose CT is mainly used for examinations of pediatric patients and screening for early-stage lung cancer [13]. Few studies of low-dose CT have been conducted in
AJR:202, April 2014 711
COPD patients, especially studies with a large sample size using low-dose dual-phase MDCT (i.e., inspiration and expiration phases). Thus, the purpose of the current study was to determine the usefulness of low-dose dual-phase MDCT performed in both inspiration and expiration to quantitatively evaluate pulmonary function in COPD patients. Specifically, the objectives of this study were, first, to examine whether it is feasible to use low-dose CT to evaluate lung function in patients with COPD; second, to determine which indexes obtained from low-dose CT are more correlated with specific measures of COPD lung function; and, third, to determine which phase of scanning is more stable and useful for evaluating COPD. Materials and Methods Subjects This retrospective study was conducted from July 2009 through December 2010. The first step was clinical diagnosis, and the second step was routine PFTs. A total of 200 consecutive subjects diagnosed with COPD according to clinical symptoms and PFT findings were consecutively included in the study. COPD was classified into five stages (stage 0, I, IIA, IIB, III) as described by the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines for COPD [3]. After other lung diseases had been excluded by CT, including lung cancer, tuberculosis, pneumoconiosis, and bronchiectasis, a total of 146 COPD
patients were eligible for the study. All measurements were completed the same day for most patients, with measurements for some patients being completed in 2 days. The study was approved by the institutional review board of our hospital, and all patients provided written informed consent.
CT Techniques and Image Analysis All CT studies were performed using a 16-MDCT unit (Aquilion, Toshiba Medical Systems). The scanning parameters were as follows: 120 kVp, 40 mA, helical pitch of 16, tube rotation rate of 0.5 seconds per rotation, FOV of 400 × 400 mm, 512 × 512 matrix, 1-mm slice thickness, and 2.5-mm-thick reconstructed sections. A lung resolution algorithm was used. Before scanning, we instructed patients on breath-holding at end-inspiration and end-expiration. For the CT examination, patients were in a supine position with both hands holding the head, and contrast material was not injected. Scanning was performed at both end-inspiration and end-expiration. The total lung scanning time was 8 seconds, and scanning was completed within one breath-hold for all patients. Two studies of each patient were obtained: One study was obtained at the end of maximum inspiration and the other at the end of maximum inspiration. Patients were allowed to breathe freely between studies. A computer-aided analysis system (LungCAD, version 2.0.5, Neusoft Medical Systems) was used to collect the DICOM data, which were then automatically separated with 300 HU as the upper limit and 1024 HU as the lower limit. LungCAD is a computer-aided diagnosis system that uses a clas-
0.45
sification algorithm to detect pulmonary lesions and nodules. We chose LungCAD so that we could use its software analysis of pulmonary function. Data about the lung, mediastinum, great vessels, trachea, bronchi, and chest were extracted. The mean CT value of the lung, lung volume, SD of the lung volume, lung area, air volume, tissue volume, tissue weight, and lung density were automatically calculated. Different pixel thresholds were set to calculate the proportion of different pixel intervals and the distribution curve of the pixels (Fig. 1). The trigger threshold and pixel index (PI) for evaluation were 910 and 950 HU, respectively; 910 HU served as a trigger threshold of emphysema, and the air volume of emphysema was automatically calculated. In the current study, techniques for defining the threshold for the lung were similar to those described by Arakawa et al. [14]. The upper and lower thresholds were determined, and the tissues with values outside the thresholds were automatically excluded. The tissues with values within the thresholds were used for further analysis to collect geometric information about the whole lung. In our study, the upper threshold was defined as –300 HU and the lower threshold as –1024 HU. Automatic separation and data analysis were then performed, and the mean attenuation values of the whole lung were obtained. The voxel and its volume were then used to calculate the volume of the whole lung. The values obtained with this method are more precise than those obtained by calculating the density and volume of the whole lung at different slices as was done in previous studies [15, 16].
0.70
0.40 0.60 0.35 0.50 Pixel Index
0.30 Pixel Index
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al.
0.25 0.20
0.40
0.30
0.15 0.20 0.10 0.10
0.05 0.00 −1000
−900
−800
−700
−600
−1000
−900
−800
−700
−600
CT Attenuation Value (HU)
CT Attenuation Value (HU)
A
B
Fig. 1—Term “pixel index” means percentage of pixels under certain threshold. This value can be used to calculate percentage of emphysema. A and B, Pixel index at deep end-expiration (A) and deep end-inspiration (B).
712
AJR:202, April 2014
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD Analysis of data from 10 patients indicated that the average dose-length product (DLP) of lowdose MDCT was 88.5 ± 5.3 (SD) mGy × cm and the corresponding effective dose was 1.50 ± 0.09 mSv. The average DLP of standard-dose chest CT was 256.8 ± 9.4 mGy × cm, and the corresponding average effective dose was 3.87 ± 1.30 mSv. There were significant differences between lowdose MDCT and standard-dose chest CT in DLP and effective dose (p < 0.01).
Routine Pulmonary Function Testing Pulmonary function testing equipment (Quark PFT, Cosmed) met the criteria established by the American Thoracic Society and European Respiratory Society. FVC was measured by having the patient breathe normally for three or four cycles; then, immediately after complete deep inspiration, the patient was asked to blow out as hard and as fast as possible so that expiration was even and complete without interruption (e.g., cough) into a device capable of measuring the amount of air expelled. After complete expiration, deep inspiration was performed again. PFTs were performed at least three times (but fewer than eight times), with the difference between optimal and suboptimal FEV1 and FVC values of less than 150 mL. The maximal FVC and FEV1 were recorded, and the remaining values served as the optimal values based on when FEV1 and FVC were the highest and the curve was smooth. The medical professionals who performed the PFTs are specialists who are highly trained in administering these tests and experienced in judging whether a patient has achieved full inspiration and expiration. In addition, patients would practice the spirometry test three times before the actual measurements to ensure that they could perform both full inspiration and full expiration. The “open nitrogen flushing” method was used to measure lung capacity. At rest, patients were asked to breathe for at least five cycles so that the end-expiration curve was nearly straight. Then, complete deep expiration was performed followed by complete inspiration. A plateau should be present at both end-inspiration and end-expiration. Subsequently, patients would return to normal breathing at rest. When the end-expiration curve was nearly straight, patients would then begin to inhale pure oxygen, and the nitrogen concentration at end-expiration was measured. When the nitrogen concentration reached less than 1.5% or after 7 minutes of oxygen inhalation, the measurements were stopped. The test was performed at least twice with an interval between tests of 20–30 minutes. The average of both tests was obtained and used for analysis.
Statistical Analysis Continuous variables are presented as means ± SD or as medians with the interquartile range
(IQR) depending on the normality of data; the IQR is the range between the 25th and 75th percentiles. Categoric variables are expressed as frequencies (percentages). The differences among five COPD groups (stages 0, I, IIA, IIB, and III as defined by 2002 GOLD guidelines [3]) were detected by the one-way analysis of variance or Kruskal-Wallis test, and the post hoc test for pairwise comparisons was conducted by Bonferroni adjustment to control for type I errors. Differences between smokers and nonsmokers were detected by an independent Student t test or Wilcoxon rank sum test. The relationships between spirometric PFT indexes and MDCT pulmonary function indexes were measured by the Spearman rank correlation coefficient (r). The strength of correlation was defined as follows: very weak, 0–0.19; weak, 0.20–0.39; moderate, 0.40–0.59; strong, 0.60–0.79; and very strong, 0.80–1.00). A Bland-Altman analysis was used to compare the
CT and PFT values that were predicted to correlate with one another. Clinically, LVin (mL), LVex (mL), LVin – LVex (mL), and LVex /LVin (%) seem to be correlated with TLC (L), RV (L), TLC – RV (L), and RV/TLC (%), respectively. Therefore, they were compared with one another. The other two CT parameters—extrapolated LVin (mL) and extrapolated LVex (mL) —did not seem to have specific PFT parameters to reflect; hence, they were compared with all PFT parameters. The statistical analyses were performed using statistics software (SAS, version 9.2, SAS Institute). A two-tailed p < 0.05 was considered statistically significant.
Results The baseline and demographic characteristics of the study subjects are provided in Table 1. A total of 146 patients with COPD were enrolled in this study, with a
TABLE 1: Baseline and Demographic Characteristics of 146 Patients With Chronic Obstructive Pulmonary Disease (COPD) Characteristic
Value
Age (y) Median (IQR)
65.0 (59.0–70.0)
Sex, no. (%) of patients Male
122 (83.6)
Female
24 (16.4)
Weight (kg) Mean ± SD
61.3 ± 10.5
Height (cm) Mean ± SD
163.1 ± 6.5
BMI Mean ± SD
23.0 ± 3.4
Current smokera, no. (%) of patients No
37 (25.3)
Yes
109 (74.7)
Smoking amount (packs/y) Median (IQR)
365.0 (274.0–365.0)
Smoking duration (y)a Median (IQR)
41.0 (33.0–45.0)
COPD stage, no. (%) of patients 0
39 (26.7)
I
40 (27.4)
IIA
30 (20.5)
IIB
22 (15.1)
III
15 (10.3)
Note—Normally distributed data are presented as means ± SD, and nonnormally distributed data are presented as medians (interquartile range [IQR]). BMI = body mass index. aNonsmokers were people who had never smoked. Patients who had quit smoking were included in the smoker category. For current smokers, the analysis is for only the 109 smokers.
AJR:202, April 2014 713
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. median age of 65.0 years (IQR, 59.0–70.0 years). The mean body mass index (BMI) of the study patients was 23.0 ± 3.4. Most patients (74.7%) were smokers, with a median smoking duration of 41.0 years (IQR, 33.0– 45.0 years) and a median smoking amount of 365.0 packs/year (IQR, 274.0–365.0 packs/ year). COPD was classified as stage 0 in 39 (26.7%) patients, stage I in 40 (27.4%), stage IIA in 30 (20.5%), stage IIB in 22 (15.1%), and stage III in 15 (10.3%). Spirometric Pulmonary Function Test Indexes The spirometric PFT indexes are summarized in Table 2 according to COPD stage. Significant differences among the five COPD groups were found in TLC (p = 0.004), RV (p < 0.0001), RV/TLC % (p < 0.0001), the difference between TLC and RV (p < 0.0001), FEV1 (% predicted) (p < 0.0001), FVC (% predicted) (p < 0.0001), and FEV1/FVC % (p < 0.0001). In general, PFT indexes in patients with stage IIB and III COPD were significantly different from those with lower stages (Table 2). We also pooled the data into two groups—patients with less severe COPD (stages 0–IIA) and those with more severe COPD (stages IIB and III)—to evaluate PFT results in these groups. For spirometric PFT results, there were significant differences between patients with less severe COPD stages and those with more severe COPD stages for all the variables except TLC (median [IQR], 5.98 L [5.09–6.84 L] for less severe COPD and 6.04 L [5.49–7.15 L] for more severe COPD) (p = 0.176).
MDCT Pulmonary Function Indexes The MDCT pulmonary function indexes are summarized in Table 3 according to COPD stage. During full inspiration and full expiration, all MDCT pulmonary function indexes differed significantly among patients in the five COPD stages (all, p < 0.0001). Significant differences among the five COPD groups were also found in the difference between LVin and LVex (p < 0.0001) and the LVex/LVin % (p < 0.0001). As with the spirometric PFT indexes, most MDCT pulmonary function indexes, especially during full expiration, were significantly different in patients with stage IIB and III COPD as compared with those with lower stages (Table 3). We also compared MDCT pulmonary function indexes by severity of COPD (less severe = stages 0– IIA and severe = stages IIB and III) by combining COPD stages. Similar to the analysis presented in Table 3, MDCT pulmonary function indexes were lower in patients with less severe COPD compared with patients with severe COPD (all, p < 0.01). Correlation Between Spirometric Pulmonary Function Test Indexes and MDCT Indexes The correlations between PFT indexes and MDCT pulmonary function indexes are shown in Table 4. A very strong correlation was found between extrapolated LVex and COPD stage (r = 0.802, p < 0.001) and between extrapolated LVex and FEV1/FVC% (r = –0.831, p < 0.001). Moreover, strong positive correlations were found between LVin and TLC (r = 0.658, p < 0.001), LVex and RV (r = 0.683, p < 0.001), extrapolat-
ed LVex and RV (r = 0.640, p < 0.001), LVex and RV/TLC (r = 0.602, p < 0.001), LVex / LVin and RV/TLC (r = 0.622, p < 0.001), extrapolated LVex and RV/TLC (r = 0.663, p < 0.001), and LVex and COPD stage (r = 0.697, p < 0.001). Strong negative correlations were found between LVex and FVC (% predicted) (r = –0.649, p < 0.001), extrapolated LVex and FVC (% predicted) (r = –0.745, p < 0.001), LVex and FEV1/FVC% (r = –0.759, p < 0.001), LVex /LVin and FEV1/FVC% (r = –0.600, p < 0.001), and extrapolated LVin and FEV1/ FVC% (r = –0.637, p < 0.001). Other correlations either were weak to moderate or were not significant. Correlation Between Spirometric Pulmonary Function Test Indexes and Pixel Indexes at 910 and 950 HU The correlations between PFT indexes and PI at 910 HU (PI-910) and PI at 950 HU (PI-950) are shown in Table 5. During full inspiration, strong negative correlations were found between FEV1/FVC% and PI-910 (r = –0.613, p < 0.001) and between FEV1/FVC% and PI-950 (r = –0.666, p < 0.001). During full expiration, strong negative correlations were found between FEV1 (% predicted) and PI-910 (r = –0.683, p < 0.001), FEV1 (% predicted) and PI-950 (r = –0.697, p < 0.001), FEV1/FVC% and PI910 (r = –0.776, p < 0.001), and FEV1/FVC % and PI-950 (r = –0.784, p < 0.001). Strong positive correlations were found between RV and PI-910 (r = 0.614, p < 0.001), RV and PI950 (r = 0.616, p < 0.001), RV/TLC and PI910 (r = 0.602, p < 0.001), and RV/TLC and
TABLE 2: Pulmonary Function Test (PFT) Indexes by Chronic Obstructive Pulmonary Disease (COPD) Stage COPD Stage PFT Index TLC (L) RV (L)
Total (n = 146)
0 (n = 39)
I (n = 40)
IIA (n = 30)
IIB (n = 22)
III (n = 15)
pa
6.01 (5.24–6.85)
5.09 (4.62–6.47)
6.44 (5.29–7.15)b
6.36 (5.59–6.79)
5.64 (5.42–6.77)
6.31 (5.81–7.43)b
0.004
2.10 (1.84–2.63)
2.72 (2.37–3.42)b
3.22 (2.70–3.48)b
3.65 (3.22–4.43)b,c
4.24 (3.46–4.84)b,c,d
< 0.0001
64.10 (58.00–70.00)b,c,d
< 0.0001
2.45 (1.78–2.66)b,c
< 0.0001
2.91 (2.29–3.60)
RV/TLC (%)
49.00 (41.40–58.40) 40.20 (37.50–45.70) 46.25 (40.20–51.45) 50.75 (45.74–55.70)b 61.25 (58.40–68.60)b,c,d
TLC – RV (L)
2.21 (2.01–2.64)b,c,d
2.88 (2.41–3.53)
3.01 (2.72–3.80)
3.17 (2.57–3.94)
2.94 (2.49–3.44)
FEV1 (% predicted) 74.40 (43.30–90.40) 93.20 (85.90–108.70) 85.15 (75.65–94.45)b 59.50 (55.20–65.00)b,c 37.15 (33.80–40.80)b,c,d 27.00 (24.20–34.00)b,c,d,e
< 0.0001
FVC (% predicted) 92.10 (78.80–107.20) 101.10 (91.70–113.10) 108.00 (96.50–118.15) 86.40 (79.10–94.70)b,c 71.50 (64.40–77.00)b,c,d 68.70 (60.00–74.30)b,c,d
< 0.0001
FEV1/FVC%
62.30 (47.00–69.60) 76.50 (71.00–80.30)
64.70 (60.95–67.30)b
54.70 (49.30–62.10)b,c
40.85 (35.60–47.80)b,c,d
32.70 (31.30–39.40)b,c,d,e
< 0.0001
Note—Data are nonnormally distributed and are presented as medians (interquartile range). Bonferroni adjustments were applied in pairwise comparisons. TLC = total lung capacity, RV = residual volume, FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, FVC% = percentage forced vital capacity. aDetermined by Kruskal-Wallis test. bp < 0.05 compared with COPD stage 0. cp< 0.05 compared with COPD stage I. dp < 0.05 compared with COPD stage IIA. ep < 0.05 compared with COPD stage IIB.
714
AJR:202, April 2014
Low-Dose MDCT of Patients With COPD TABLE 3: MDCT Pulmonary Function Indexes by Chronic Obstructive Pulmonary Disease (COPD) Stage COPD Stage
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
MDCT Pulmonary Function Index
Total (n = 146)
0 (n = 39)
I (n = 40)
IIA (n = 30)
IIB (n = 22)
III (n = 15)
p
4772 ± 1032
4020 ± 1044
4968 ± 873a
5004 ± 854a
4890 ± 873a
5573 ± 886a
< 0.0001b
2429 (1424–3553)
1133 (161–2025)
2441 (1640–3524)a
2919 (2116–3809)a
2987 (2329–4103)a
4087 (3296–4701)a,c
< 0.0001d
−897 (–916 to –885)a
−904 (–918 to –886)a
−908 (–928 to –892)a
−922 (–936 to –900)a,c
< 0.0001d
0.124 (0.109–0.152) 0.103 (0.084–0.115)a 0.096 (0.082–0.114)a 0.092 (0.072–0.108)a 0.079 (0.064–0.100)a,c
< 0.0001d
Full inspiration LVin (mL) Extrapolated LVin (mL) MLA (HU)
−895 (–918 to –879) −877 (–891 to –848)
MLD (g/cm3)
0.106 (0.082–0.121)
PI-910 (%)
59.32 ± 13.13
47.71 ± 12.12
60.67 ± 10.02a
62.20 ± 10.44a
65.12 ± 11.41a
71.64 ± 8.45a,c
< 0.0001b
PI-950 (%)
42.62 ± 13.89
30.28 ± 8.81
43.10 ± 11.24a
44.93 ± 11.84a
50.53 ± 13.35a
57.19 ± 10.80a,c,e
< 0.0001b
2366 (1740–3025)
1443 (1202–2004)
2256 (1852–2706)a
2391 (2134–2963)a
3225 (2639–3890)a,c,e
3658 (3317–4547)a,c,e
< 0.0001d
992 (366–2104) a,c,e 1722 (892–2820)a,c,e
< 0.0001d
Full expiration LVex (mL) Extrapolated LVex (mL)
51.9 (3.3–547.9)
MLA (HU)
−800 ± 64
MLD (g/cm3)
0.200 ± 0.065
0.320 (0.030–3.060) 30.06 (5.39–189.86)a
207 (32–460)a
−736 ± 43
−793 ± 47a
−807 ± 42a
−858 ± 38a,c,e
−881 ± 37a,c,e
< 0.0001b
0.264 ± 0.043
0.207 ± 0.047a
0.194 ± 0.044a
0.137 ± 0.038a,c,e
0.119 ± 0.037a,c,e
< 0.0001b
PI-910 (%)
27.04 (16.10–42.00) 13.93 (10.62–16.94) 25.72 (19.83–34.73)a 32.84 (20.53–40.09)a 46.36 (37.19–60.83)a,c,e 55.09 (45.12–61.49)a,c,e < 0.0001d
PI-950 (%)
16.65 (9.61–27.30)
7.48 (5.81–10.51)
2266 ± 885
2396 ± 948
LVin – LVex (mL) LVex /LVin (%)
15.00 (10.95–21.51)a 20.32 (12.08–25.58)a 30.62 (23.42–45.77)a,c,e 37.72 (32.77–52.72)a,c,e < 0.0001d
50.82 (40.28–64.24) 38.74 (32.26–45.99)
2607 ± 773
2456 ± 760
1579 ± 638a,c,e
1643 ± 767a,c,e
< 0.0001b
46.33 (40.41–53.73)a
50.22 (43.41–57.11)a
68.55 (60.33–76.81)a,c,e
69.62 (64.50–78.16)a,c,e
< 0.0001d
Note—Normally distributed data are presented as means ± SD, and nonnormally distributed data are presented as medians (interquartile range). Bonferroni adjustments were applied in pairwise comparisons. LV in = inspiratory lung volume, MLA = mean lung attenuation, MLD = mean lung density, PI-910 = pixel index at 910 HU, PI-950 = pixel index at 950 HU, LVex = expiratory lung volume. ap < 0.05 compared with COPD stage 0. bDetermined by one-way analysis of variance. cp < 0.05 compared with COPD stage I. dDetermined by Kruskal-Wallis test. ep < 0.05 compared with COPD stage IIA.
TABLE 4: Correlation Between Spirometric Pulmonary Function Test (PFT) Indexes and MDCT Pulmonary Function Indexes in Patients With Chronic Obstructive Pulmonary Disease (COPD) MDCT Pulmonary Function Index
PFT Index TLC (L)
RV (L)
RV/TLC (%)
TLC – RV (L)
FVC (% Predicted)
LVin (mL)
0.658a
0.490a
0.163b
0.297a
−0.291a
LVex (mL)
0.419a
0.683a
0.602a
−0.245c
−0.649a
LVin – LVex (mL)
0.272a
−0.178b
−0.454a
0.585a
0.622a
−0.452a
FEV1 (% Predicted) FEV1/FVC%
COPD Stage
−0.013
−0.526a
0.405a
−0.411a
−0.759a
0.697a
0.386a
0.451a
0.243c
−0.319a
−0.595a
−0.475a
−0.600a
0.598a
LVex /LVin (%)
0.115
0.526a
Extrapolated LVin (mL)
0.415a
0.465a
0.321a
0.002
−0.450a
−0.159
−0.637a
0.556a
Extrapolated LVex (mL)
0.277a
0.640a
0.663a
−0.384a
−0.745a
−0.500c
−0.831a
0.802a
Note—Correlation was measured by Spearman rank correlation coefficient. TLC = total lung capacity, RV = residual volume, FVC = forced vital capacity, FEV1 = forced expiratory volume in 1 second, FVC% = percentage forced vital capacity, LVin = inspiratory lung volume, LVex = expiratory lung volume. ap < 0.001. bp < 0.05. cp < 0.01.
PI-950 (r = 0.603, p < 0.001). In addition, PI910 and PI-950 were both positively correlated with COPD stage (Table 5) during full inspiration and full expiration.
Comparisons of MDCT Pulmonary Function Indexes Between Smokers and Nonsmokers Each of the MDCT pulmonary function indexes was compared between smokers and
nonsmokers (Table 6). Compared with nonsmokers, smokers exhibited significantly higher LVin, LVex, LVex /LVin, extrapolated LVin, extrapolated LVex, inspiratory PI-910,
AJR:202, April 2014 715
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. expiratory PI-910, inspiratory PI-950, and expiratory PI-950 but significantly lower inspiratory mean lung attenuation, expiratory mean lung attenuation, inspiratory mean lung density, and expiratory mean lung density. Discussion The results of the current study show that certain spirometric and low-dose MDCT pulmonary function indexes differed significantly between various stages of COPD. More important, strong correlations were found between spirometric and MDCT pulmonary function indexes as well as between spirometric indexes and PI-910 and PI950. In particular, key lung parameters that showed strong correlation between MDCT and PFT indexes included LVin (mL) and TLC (L), LVex (mL) and RV (L), extrapolated LVex (mL) and FEV1/FVC%, expiratory PI-950 (%) and FEV1 (% predicted), and expiratory PI-950 (%) and FEV1/FVC% (Tables 4 and 5). These parameters all showed strong correlation between MDCT and PFT (r > 0.5, p < 0.001) and indicated that MDCT has a strong ability to evaluate key COPD-related lung parameters. These results suggest that low-dose MDCT may be useful in the diagnosis and staging of COPD. These results also indicate that imaging patients with low-dose MDCT at deep end-expiration provides better correlation with PFT results than imaging at deep end-inspiration. This finding implies that only end-expiration phase imaging may be necessary, which could further reduce the radiation dose to patients. To our knowledge, this study is the first with a large sample size that investigated the use of low-dose CT to evaluate lung function of patients with COPD. Although still in its infancy, CT is becoming a useful method for the evaluation of large and small airways disease [6, 10, 17, 18]. Routine PFTs have been the reference standard for determining the presence and degree of COPD, but they are not useful for early diagnosis and cannot be used to evaluate unilateral or partial pulmonary tissue function. Additionally, some patients have difficulty performing PFTs. Low-dose MDCT can provide additional useful clinical information: Because many COPD patients required CT evaluation during their treatment course, the possibility that a single low-dose MDCT examination can be used to evaluate both pulmonary structure and function may save both time and cost for patients and physicians.
716
TABLE 5: Correlation Between Spirometric Pulmonary Function Test (PFT) Indexes and Pixel Index at 910 HU (PI-910) and Pixel Index at 950 HU (PI-950) in Patients With Chronic Obstructive Pulmonary Disease (COPD) PFT Index
Full Inspiration
Full Expiration
PI-910 (%)
PI-950 (%)
PI-910 (%)
PI-950 (%)
FEV1 (% predicted)
−0.471a
−0.529a
−0.683a
−0.697a
FVC (% predicted)
−0.222b
−0.271a
−0.462a
−0.476a
FEV1/FVC%
−0.613a
−0.666a
−0.776a
−0.784a
TLC (L)
0.268b
0.275a
0.310a
0.313a
RV (L)
0.413a
0.445a
0.614a
0.616a
RV/TLC (%)
0.360a
0.403a
0.602a
0.603a
−0.318a
−0.316a
0.744a
0.748a
TLC – RV (L)
−0.131
COPD stage
0.560a
−0.162 0.613a
Note—Correlation was measured by Spearman rank correlation coefficient. FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, FVC% = percentage forced vital capacity, TLC = total lung capacity, RV = residual volume. ap < 0.001. bp < 0.01.
TABLE 6: Comparison of MDCT Pulmonary Function Indexes Between Smokers and Nonsmokers With Chronic Obstructive Pulmonary Disease (COPD) MDCT Pulmonary Function Index
Nonsmoker (n = 37)
Smoker (n = 109)
p
LVin (mL)
3766 (3264–4223)
5078 (4462–5724)
< 0.0001a
LVex (mL)
1678 (1314–2283)
2624 (2134–3435)
< 0.0001a
41.55 (36.10–55.97)
52.55 (43.24–64.58)
0.028a
Extrapolated LVin (mL)
1349 (668–2126)
2883 (2004–3908)
< 0.0001a
Extrapolated LVex (mL)
4.65 (0.32–27.47)
181.61 (15.08–892.12)
< 0.0001a
Inspiratory
−876 ± 40
−902 ± 26
0.0006b
Expiratory
−765 ± 56
−811 ± 62
0.0001b
Inspiratory
0.124 ± 0.040
0.098 ± 0.026
0.0006b
Expiratory
0.236 ± 0.056
0.188 ± 0.063
< 0.0001b
Inspiratory
51.62 ± 15.09
61.93 ± 11.33
0.0004b
Expiratory
15.98 (12.86–21.99)
33.20 (21.59–46.77)
< 0.0001a
Inspiratory
34.53 ± 13.59
45.37 ± 12.94
< 0.0001b
Expiratory
8.96 (6.60–13.64)
20.95 (13.05–32.23)
< 0.0001a
LVex /LVin (%)
Mean lung attenuation (HU)
Mean lung density (g/cm3)
PI-910 (%)
PI-950 (%)
Note—Normally distributed data are presented as means ± SD and nonnormally distributed data are presented as medians (interquartile range). LV = lung volume, LVex = expiratory lung volume, LV in = inspiratory lung volume, PI-910 = pixel index at 910 HU, PI-950 = pixel index at 950 HU. aDetermined by Wilcoxon rank sum test. bDetermined by independent Student t test.
In the current study, LVin was 4772 ± 1032 mL (mean ± SD) and median LVex was 2366 mL (IQR, 1740–3025 mL) as determined by MDCT. Spirometric PFTs showed that the median TLC was 6.01 L (IQR, 5.24–6.85 L)
and the median RV was 2.91 L (IQR, 2.29– 3.60 L). The difference between the two detection methods was approximately 1200 and 550 mL, respectively. Routine PFTs are performed with the patient seated, but CT is per-
AJR:202, April 2014
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD formed with the patient in a supine position; this difference in patient positioning may account for the differences in the lung function parameters detected with these two methods. Our results show that the lung volume indexes by MDCT correlated with PFT volume. The correlation coefficient was 0.683 between LVex and RV, 0.658 between LVin and TLC, and 0.622 between LVex /LVin and RV/TLC. Thus, we speculate that values of LVin, LVex, and LVex /LVin detected with lowdose MDCT can be used to evaluate TLC, RV, and RV/TLC, respectively. All of these MDCT indexes reflect the actual air volume in the lung and are thus more precise than the corresponding spirometric PFT values. Kauczor et al. [9] measured lung attenuation at paired high-resolution CT performed at full inspiration and full expiration and correlated the values with PFT results and found that the inspiratory mean lung density and the expiratory attenuation increase were able to differentiate patients with obstructive and restrictive ventilatory impairment from healthy subjects and that scans obtained at full expiration provided the best results. Zaporozhan et al. [4] used 3D high-resolution CT data obtained at inspiration and expiration for the quantitative evaluation of emphysema and reported that emphysema volumes measured from expiratory scans were more consistent with PFT results. In the current study, the lung density was evaluated with two methods. For the first method, we used the CT attenuation value with scanning performed at 40 mA and found that the mean attenuation value of the whole lung was –895 ± 32 HU at end-inspiration and –800 ± 64 HU at end-expiration; these results are consistent with the findings obtained using standarddose dual-phase MDCT [4]. For the second method, we used volume density. Determination of the PI threshold is crucial when evaluating lung disease by CT and is typically determined by the investigator. In this study, we defined the PI threshold as –950 HU. Gevenois et al. [19] compared the area covered by different pixels with pathologic findings and found that –910 HU in the expiration phase and –950 HU in the inspiration phase were significantly related to pathologic findings and that –910 HU in the expiration phase had the most optimal efficiency in determining the area of emphysema. Müller et al. [20] also found that –910 HU was the best threshold for quantifying the area of emphysema. Some investigators have used –950 HU as the trigger threshold for severe emphy-
sema [11]. In the current study, when –910 and –950 HU were used as trigger thresholds, the PI at deep end-inspiration and deep end-expiration was significantly associated with FEV1 (% predicted) and FEV1/FVC% and the relationship in the expiration phase was superior to that in the inspiration phase. The best correlation was observed when PI-910 was used as the trigger threshold (r = –0.78 for expiration phase and –0.69 for inspiration phase). Thus, we speculate that –910 HU as a trigger threshold in the expiration phase can be used to measure emphysema volume. In this study, we compared MDCT pulmonary function indexes between smokers and nonsmokers and found that LVin, LVex, LVex/ LVin, extrapolated LVin, extrapolated LVex, inspiratory PI-950, expiratory PI-950, inspiratory PI-910, and expiratory PI-910 in smokers were markedly increased when compared with nonsmokers. In smokers, inspiratory mean lung attenuation, expiratory mean lung attenuation, inspiratory mean lung density, and expiratory mean lung density were significantly reduced compared with the values in nonsmokers; these results indicate that for patients with clinically proven COPD, lung function in smokers is poorer than that in nonsmokers. We speculate that smoking not only causes inflammation in the small airways but also may cause disruption of the lung anatomic structure in COPD patients. The differences in lung density and emphysema volume show that the emphysema is more severe in smokers than in nonsmokers. There are some limitations of this study that should be considered. We did not compare standard-dose MDCT and low-dose MDCT in the same patients because of concern about increased radiation exposure. In addition, there were no control subjects with COPD who underwent standard-dose MDCT and low-dose MDCT. Conclusion Our results showed that lung function indexes obtained with low-dose dual-phase MDCT correlated well with those obtained from spirometric PFTs in patients with COPD and that the correlation at expiration was superior to that at inspiration. In addition, MDCT indexes varied among the different stages of COPD and thus may be useful for determining the stage of COPD in individual patients. References 1. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS; GOLD Scientific Committee. Glob-
al strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163:1256–1276 2. Group of Chronic Obstructive Pulmonary Disease of Branch of Respiratory Diseases of Chinese Medical Association. Guideline for diagnosis and treatment of chronic obstructive pulmonary disease (2007 revision) [in Chinese]. Chin J Tubercul Respir Dis 2007; 30:8–17 3. Gómez FP, Rodriguez-Roisin R. Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines for chronic obstructive pulmonary disease. Curr Opin Pulm Med 2002; 8:81–86 4. Zaporozhan J, Ley S, Eberhardt R, et al. Paired inspiratory/expiratory volumetric thin-slice CT scan for emphysema analysis: comparison of different quantitative evaluations and pulmonary function test. Chest 2005; 128:3212–3220 5. Bankier AA, Madani A, Gevenois PA. CT quantification of pulmonary emphysema: assessment of lung structure and function. Crit Rev Comput Tomogr 2002; 43:399–417 6. Madani A, Zanen J, de Maertelaer V, Gevenois PA. Pulmonary emphysema: objective quantification at multi-detector row CT—comparison with macroscopic and microscopic morphometry. Radiology 2006; 238:1036–1043 7. Shah E, Gilkeson R, Krishna S, Ciancibello L, Durgan J, Pohlman S. Comparison of computeraided calculation of emphysema volumetry with manual quantification using CT images. Int Congr Ser 2004; 1268:956–960 8. O’Donnell RA, Peebles C, Ward JA, et al. Relationship between peripheral airway dysfunction, airway obstruction, and neutrophilic inflammation in COPD. Thorax 2004; 59:837–842 9. Kauczor HU, Hast J, Heussel CP, Schlegel J, Mildenberger P, Thelen M. CT attenuation of paired HRCT scans obtained at full inspiratory/ expiratory position: comparison with pulmonary function tests. Eur Radiol 2002; 12:2757–2763 10. Park YS, Seo JB, Kim N, et al. Texture-based quantification of pulmonary emphysema on highresolution computed tomography: comparison with density-based quantification and correlation with pulmonary function test. Invest Radiol 2008; 43:395–402 11. Orlandi I, Moroni C, Camiciottoli G, et al. Spirometric-gated computed tomography quantitative evaluation of lung emphysema in chronic obstructive pulmonary disease: a comparison of 3 techniques. J Comput Assist Tomogr 2004; 28:437–442 12. Bastarrika G, Wisnivesky JP, Pueyo JC, et al. Low-dose volumetric computed tomography for quantification of emphysema in asymptomatic
AJR:202, April 2014 717
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. smokers participating in an early lung cancer detection trial. J Thorac Imaging 2009; 24:206–211 13. Wu XH, Ma DQ, Zhang ZJ, et al. The experimental study and clinical application on the detection of pulmonary nodules with low-dose multislice spiral CT [in Chinese]. Zhonghua Fang She Xue Za Zhi 2004; 38:767–770 14. Arakawa A, Yamashita Y, Nakayama Y, et al. Assessment of lung volumes in pulmonary emphysema using multidetector helical CT: comparison with pulmonary function tests. Comput Med Imaging Graph 2001; 25:399–404
718
15. Matsuoka S, Kurihara Y, Yagihashi K, Nakajima Y. Quantitative assessment of peripheral airway obstruction on paired expiratory/inspiratory thin-section computed tomography in chronic obstructive pulmonary disease with emphysema. J Comput Assist Tomogr 2007; 31:384–389 16. Satoh S, Ohdama S, Shibuya H. Sliding thin slab, minimum intensity projection imaging for objective analysis of emphysema. Radiat Med 2006; 24:415–421 17. Kauczor HU, Wielpütz MO, Owsijewitsch M, et al. Computed tomographic imaging of the airways
in COPD and asthma. J Thorac Imaging 2011; 26:290–300 18. Matsuoka S, Yamashiro T, Washko GR, Kurihara Y, Nakajima Y, Hatabu H. Quantitative CT assessment of chronic obstructive pulmonary disease. RadioGraphics 2010; 30:55–66 19. Gevenois PA, De Vuyst P, Sy M, et al. Pulmonary emphysema: quantitative CT during expiration. Radiology 1996; 199:825–829 20. Müller NL, Thurlbeck WM. Thin-section CT, emphysema, air trapping, and airway obstruction. Radiology 1996; 199:621–622
AJR:202, April 2014
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD Cardiopulmonary Imaging Original Research
Correlation of Pulmonary Function Indexes Determined by LowDose MDCT With Spirometric Pulmonary Function Tests in Patients With Chronic Obstructive Pulmonary Disease
Huai Chen1 Rong-chang Chen2 Yu-bao Guan1 Wen Li1 Qin Liu1 Qing-si Zeng1
OBJECTIVE. The objective of our study was to evaluate the correlation between pulmonary function indexes determined by low-dose MDCT and those obtained from routine spirometric pulmonary function tests (PFTs) in patients with chronic obstructive pulmonary disease (COPD). MATERIALS AND METHODS. Lung function of patients with COPD stages 0–III was evaluated by both MDCT and spirometric PFTs. Scanning was performed at maximum end-inspiration and maximum end-expiration. RESULTS. A very strong correlation was found between extrapolated expiratory lung volume (LVex) and COPD stage (r = 0.802, p < 0.001) and between extrapolated LVex and the ratio of forced expiratory volume in 1 second and percentage forced vital capacity (FEV1/ FVC%) (r = –0.831, p < 0.001). Moreover, strong positive correlations were found between inspiratory lung volume (LVin) and total lung capacity (TLC) (r = 0.658, p < 0.001), LVex and residual volume (RV) (r = 0.683, p < 0.001), extrapolated LVex and RV (r = 0.640, p < 0.001), LVex and RV/TLC (r = 0.602, p < 0.001), LVex/LVin and RV/TLC (r = 0.622, p < 0.001), extrapolated LVex and RV/TLC (r = 0.663, p < 0.001), and LVex and COPD stage (r = 0.697, p < 0.001). CONCLUSION. Low-dose MDCT lung function indexes correlate well with spirometric PFT results, and the highest correlation is at end-expiration. Low-dose MDCT may be useful for evaluating lung function in patients with COPD.
Chen H, Chen RC, Guan YB, Li W, Liu Q, Zeng QS
Keywords: chronic obstructive pulmonary disease, low-dose MDCT, pulmonary function, quantification DOI:10.2214/AJR.12.10501 Received December 24, 2012; accepted after revision June 7, 2013. Supported by grant A2011013 from Guangzhou Medical College. 1 Department of Radiology, the First Affiliated Hospital of Guangzhou Medical University, No. 151, Yan Jiang Rd, Yuexiu District, Guangzhou City, Guangdong Province 510120, China. Address correspondence to Q. Zeng ([email protected]). 2
National Key Laboratory of Respiratory Diseases, the First Affiliated Hospital of Guangzhou Medical University, Guangdong Province, China.
AJR 2014; 202:711–718 0361–803X/14/2024–711 © American Roentgen Ray Society
C
hronic obstructive pulmonary disease (COPD) has been the fourth leading cause of death and is projected to rank fifth in 2020 as a worldwide burden of disease according to a study published in the American Journal of Respiratory and Critical Care Medicine [1]. In China, COPD is a major health concern, and the results of a study conducted in seven regions including a total of 20,245 adults showed the prevalence of COPD to be as high as 8.2% in persons 40 years old or older [2]. Thus, early diagnosis and early treatment of COPD are crucial for patients. The diagnosis of COPD depends on pulmonary function tests (PFTs) and spirometry. A ratio of forced expiratory volume in 1 second and percentage forced vital capacity (FEV1/FVC%) of less than 70% predicted in a patient with a postbronchodilator FEV1 of less than 80% of the predicted value is diagnostic for COPD [3]. It is well known that the lung has strong compensatory ability, and injury to more than
30% of lung tissues is typically necessary to result in abnormal lung function. By the time abnormal lung function is detected, irreparable damage has likely occurred. Dual-phase MDCT with a routine radiation dose has also been used for the diagnosis of COPD with favorable results [4–11]. However, the amount of radiation patients are exposed to is high. Standard-dose CT has also been used to objectively evaluate COPD; however, these studies were primarily built on the basis of lung cancer screening by selecting several CT slices obtained during deep inspiration for the evaluation of pulmonary function [12]. Quantitative studies of pulmonary function utilizing CT usually use images of both inspiration and expiration phases, which increases the radiation dose to patients. It is clear that alternatives with a reduced radiation dose are needed. At present, low-dose CT is mainly used for examinations of pediatric patients and screening for early-stage lung cancer [13]. Few studies of low-dose CT have been conducted in
AJR:202, April 2014 711
COPD patients, especially studies with a large sample size using low-dose dual-phase MDCT (i.e., inspiration and expiration phases). Thus, the purpose of the current study was to determine the usefulness of low-dose dual-phase MDCT performed in both inspiration and expiration to quantitatively evaluate pulmonary function in COPD patients. Specifically, the objectives of this study were, first, to examine whether it is feasible to use low-dose CT to evaluate lung function in patients with COPD; second, to determine which indexes obtained from low-dose CT are more correlated with specific measures of COPD lung function; and, third, to determine which phase of scanning is more stable and useful for evaluating COPD. Materials and Methods Subjects This retrospective study was conducted from July 2009 through December 2010. The first step was clinical diagnosis, and the second step was routine PFTs. A total of 200 consecutive subjects diagnosed with COPD according to clinical symptoms and PFT findings were consecutively included in the study. COPD was classified into five stages (stage 0, I, IIA, IIB, III) as described by the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines for COPD [3]. After other lung diseases had been excluded by CT, including lung cancer, tuberculosis, pneumoconiosis, and bronchiectasis, a total of 146 COPD
patients were eligible for the study. All measurements were completed the same day for most patients, with measurements for some patients being completed in 2 days. The study was approved by the institutional review board of our hospital, and all patients provided written informed consent.
CT Techniques and Image Analysis All CT studies were performed using a 16-MDCT unit (Aquilion, Toshiba Medical Systems). The scanning parameters were as follows: 120 kVp, 40 mA, helical pitch of 16, tube rotation rate of 0.5 seconds per rotation, FOV of 400 × 400 mm, 512 × 512 matrix, 1-mm slice thickness, and 2.5-mm-thick reconstructed sections. A lung resolution algorithm was used. Before scanning, we instructed patients on breath-holding at end-inspiration and end-expiration. For the CT examination, patients were in a supine position with both hands holding the head, and contrast material was not injected. Scanning was performed at both end-inspiration and end-expiration. The total lung scanning time was 8 seconds, and scanning was completed within one breath-hold for all patients. Two studies of each patient were obtained: One study was obtained at the end of maximum inspiration and the other at the end of maximum inspiration. Patients were allowed to breathe freely between studies. A computer-aided analysis system (LungCAD, version 2.0.5, Neusoft Medical Systems) was used to collect the DICOM data, which were then automatically separated with 300 HU as the upper limit and 1024 HU as the lower limit. LungCAD is a computer-aided diagnosis system that uses a clas-
0.45
sification algorithm to detect pulmonary lesions and nodules. We chose LungCAD so that we could use its software analysis of pulmonary function. Data about the lung, mediastinum, great vessels, trachea, bronchi, and chest were extracted. The mean CT value of the lung, lung volume, SD of the lung volume, lung area, air volume, tissue volume, tissue weight, and lung density were automatically calculated. Different pixel thresholds were set to calculate the proportion of different pixel intervals and the distribution curve of the pixels (Fig. 1). The trigger threshold and pixel index (PI) for evaluation were 910 and 950 HU, respectively; 910 HU served as a trigger threshold of emphysema, and the air volume of emphysema was automatically calculated. In the current study, techniques for defining the threshold for the lung were similar to those described by Arakawa et al. [14]. The upper and lower thresholds were determined, and the tissues with values outside the thresholds were automatically excluded. The tissues with values within the thresholds were used for further analysis to collect geometric information about the whole lung. In our study, the upper threshold was defined as –300 HU and the lower threshold as –1024 HU. Automatic separation and data analysis were then performed, and the mean attenuation values of the whole lung were obtained. The voxel and its volume were then used to calculate the volume of the whole lung. The values obtained with this method are more precise than those obtained by calculating the density and volume of the whole lung at different slices as was done in previous studies [15, 16].
0.70
0.40 0.60 0.35 0.50 Pixel Index
0.30 Pixel Index
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al.
0.25 0.20
0.40
0.30
0.15 0.20 0.10 0.10
0.05 0.00 −1000
−900
−800
−700
−600
−1000
−900
−800
−700
−600
CT Attenuation Value (HU)
CT Attenuation Value (HU)
A
B
Fig. 1—Term “pixel index” means percentage of pixels under certain threshold. This value can be used to calculate percentage of emphysema. A and B, Pixel index at deep end-expiration (A) and deep end-inspiration (B).
712
AJR:202, April 2014
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD Analysis of data from 10 patients indicated that the average dose-length product (DLP) of lowdose MDCT was 88.5 ± 5.3 (SD) mGy × cm and the corresponding effective dose was 1.50 ± 0.09 mSv. The average DLP of standard-dose chest CT was 256.8 ± 9.4 mGy × cm, and the corresponding average effective dose was 3.87 ± 1.30 mSv. There were significant differences between lowdose MDCT and standard-dose chest CT in DLP and effective dose (p < 0.01).
Routine Pulmonary Function Testing Pulmonary function testing equipment (Quark PFT, Cosmed) met the criteria established by the American Thoracic Society and European Respiratory Society. FVC was measured by having the patient breathe normally for three or four cycles; then, immediately after complete deep inspiration, the patient was asked to blow out as hard and as fast as possible so that expiration was even and complete without interruption (e.g., cough) into a device capable of measuring the amount of air expelled. After complete expiration, deep inspiration was performed again. PFTs were performed at least three times (but fewer than eight times), with the difference between optimal and suboptimal FEV1 and FVC values of less than 150 mL. The maximal FVC and FEV1 were recorded, and the remaining values served as the optimal values based on when FEV1 and FVC were the highest and the curve was smooth. The medical professionals who performed the PFTs are specialists who are highly trained in administering these tests and experienced in judging whether a patient has achieved full inspiration and expiration. In addition, patients would practice the spirometry test three times before the actual measurements to ensure that they could perform both full inspiration and full expiration. The “open nitrogen flushing” method was used to measure lung capacity. At rest, patients were asked to breathe for at least five cycles so that the end-expiration curve was nearly straight. Then, complete deep expiration was performed followed by complete inspiration. A plateau should be present at both end-inspiration and end-expiration. Subsequently, patients would return to normal breathing at rest. When the end-expiration curve was nearly straight, patients would then begin to inhale pure oxygen, and the nitrogen concentration at end-expiration was measured. When the nitrogen concentration reached less than 1.5% or after 7 minutes of oxygen inhalation, the measurements were stopped. The test was performed at least twice with an interval between tests of 20–30 minutes. The average of both tests was obtained and used for analysis.
Statistical Analysis Continuous variables are presented as means ± SD or as medians with the interquartile range
(IQR) depending on the normality of data; the IQR is the range between the 25th and 75th percentiles. Categoric variables are expressed as frequencies (percentages). The differences among five COPD groups (stages 0, I, IIA, IIB, and III as defined by 2002 GOLD guidelines [3]) were detected by the one-way analysis of variance or Kruskal-Wallis test, and the post hoc test for pairwise comparisons was conducted by Bonferroni adjustment to control for type I errors. Differences between smokers and nonsmokers were detected by an independent Student t test or Wilcoxon rank sum test. The relationships between spirometric PFT indexes and MDCT pulmonary function indexes were measured by the Spearman rank correlation coefficient (r). The strength of correlation was defined as follows: very weak, 0–0.19; weak, 0.20–0.39; moderate, 0.40–0.59; strong, 0.60–0.79; and very strong, 0.80–1.00). A Bland-Altman analysis was used to compare the
CT and PFT values that were predicted to correlate with one another. Clinically, LVin (mL), LVex (mL), LVin – LVex (mL), and LVex /LVin (%) seem to be correlated with TLC (L), RV (L), TLC – RV (L), and RV/TLC (%), respectively. Therefore, they were compared with one another. The other two CT parameters—extrapolated LVin (mL) and extrapolated LVex (mL) —did not seem to have specific PFT parameters to reflect; hence, they were compared with all PFT parameters. The statistical analyses were performed using statistics software (SAS, version 9.2, SAS Institute). A two-tailed p < 0.05 was considered statistically significant.
Results The baseline and demographic characteristics of the study subjects are provided in Table 1. A total of 146 patients with COPD were enrolled in this study, with a
TABLE 1: Baseline and Demographic Characteristics of 146 Patients With Chronic Obstructive Pulmonary Disease (COPD) Characteristic
Value
Age (y) Median (IQR)
65.0 (59.0–70.0)
Sex, no. (%) of patients Male
122 (83.6)
Female
24 (16.4)
Weight (kg) Mean ± SD
61.3 ± 10.5
Height (cm) Mean ± SD
163.1 ± 6.5
BMI Mean ± SD
23.0 ± 3.4
Current smokera, no. (%) of patients No
37 (25.3)
Yes
109 (74.7)
Smoking amount (packs/y) Median (IQR)
365.0 (274.0–365.0)
Smoking duration (y)a Median (IQR)
41.0 (33.0–45.0)
COPD stage, no. (%) of patients 0
39 (26.7)
I
40 (27.4)
IIA
30 (20.5)
IIB
22 (15.1)
III
15 (10.3)
Note—Normally distributed data are presented as means ± SD, and nonnormally distributed data are presented as medians (interquartile range [IQR]). BMI = body mass index. aNonsmokers were people who had never smoked. Patients who had quit smoking were included in the smoker category. For current smokers, the analysis is for only the 109 smokers.
AJR:202, April 2014 713
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. median age of 65.0 years (IQR, 59.0–70.0 years). The mean body mass index (BMI) of the study patients was 23.0 ± 3.4. Most patients (74.7%) were smokers, with a median smoking duration of 41.0 years (IQR, 33.0– 45.0 years) and a median smoking amount of 365.0 packs/year (IQR, 274.0–365.0 packs/ year). COPD was classified as stage 0 in 39 (26.7%) patients, stage I in 40 (27.4%), stage IIA in 30 (20.5%), stage IIB in 22 (15.1%), and stage III in 15 (10.3%). Spirometric Pulmonary Function Test Indexes The spirometric PFT indexes are summarized in Table 2 according to COPD stage. Significant differences among the five COPD groups were found in TLC (p = 0.004), RV (p < 0.0001), RV/TLC % (p < 0.0001), the difference between TLC and RV (p < 0.0001), FEV1 (% predicted) (p < 0.0001), FVC (% predicted) (p < 0.0001), and FEV1/FVC % (p < 0.0001). In general, PFT indexes in patients with stage IIB and III COPD were significantly different from those with lower stages (Table 2). We also pooled the data into two groups—patients with less severe COPD (stages 0–IIA) and those with more severe COPD (stages IIB and III)—to evaluate PFT results in these groups. For spirometric PFT results, there were significant differences between patients with less severe COPD stages and those with more severe COPD stages for all the variables except TLC (median [IQR], 5.98 L [5.09–6.84 L] for less severe COPD and 6.04 L [5.49–7.15 L] for more severe COPD) (p = 0.176).
MDCT Pulmonary Function Indexes The MDCT pulmonary function indexes are summarized in Table 3 according to COPD stage. During full inspiration and full expiration, all MDCT pulmonary function indexes differed significantly among patients in the five COPD stages (all, p < 0.0001). Significant differences among the five COPD groups were also found in the difference between LVin and LVex (p < 0.0001) and the LVex/LVin % (p < 0.0001). As with the spirometric PFT indexes, most MDCT pulmonary function indexes, especially during full expiration, were significantly different in patients with stage IIB and III COPD as compared with those with lower stages (Table 3). We also compared MDCT pulmonary function indexes by severity of COPD (less severe = stages 0– IIA and severe = stages IIB and III) by combining COPD stages. Similar to the analysis presented in Table 3, MDCT pulmonary function indexes were lower in patients with less severe COPD compared with patients with severe COPD (all, p < 0.01). Correlation Between Spirometric Pulmonary Function Test Indexes and MDCT Indexes The correlations between PFT indexes and MDCT pulmonary function indexes are shown in Table 4. A very strong correlation was found between extrapolated LVex and COPD stage (r = 0.802, p < 0.001) and between extrapolated LVex and FEV1/FVC% (r = –0.831, p < 0.001). Moreover, strong positive correlations were found between LVin and TLC (r = 0.658, p < 0.001), LVex and RV (r = 0.683, p < 0.001), extrapolat-
ed LVex and RV (r = 0.640, p < 0.001), LVex and RV/TLC (r = 0.602, p < 0.001), LVex / LVin and RV/TLC (r = 0.622, p < 0.001), extrapolated LVex and RV/TLC (r = 0.663, p < 0.001), and LVex and COPD stage (r = 0.697, p < 0.001). Strong negative correlations were found between LVex and FVC (% predicted) (r = –0.649, p < 0.001), extrapolated LVex and FVC (% predicted) (r = –0.745, p < 0.001), LVex and FEV1/FVC% (r = –0.759, p < 0.001), LVex /LVin and FEV1/FVC% (r = –0.600, p < 0.001), and extrapolated LVin and FEV1/ FVC% (r = –0.637, p < 0.001). Other correlations either were weak to moderate or were not significant. Correlation Between Spirometric Pulmonary Function Test Indexes and Pixel Indexes at 910 and 950 HU The correlations between PFT indexes and PI at 910 HU (PI-910) and PI at 950 HU (PI-950) are shown in Table 5. During full inspiration, strong negative correlations were found between FEV1/FVC% and PI-910 (r = –0.613, p < 0.001) and between FEV1/FVC% and PI-950 (r = –0.666, p < 0.001). During full expiration, strong negative correlations were found between FEV1 (% predicted) and PI-910 (r = –0.683, p < 0.001), FEV1 (% predicted) and PI-950 (r = –0.697, p < 0.001), FEV1/FVC% and PI910 (r = –0.776, p < 0.001), and FEV1/FVC % and PI-950 (r = –0.784, p < 0.001). Strong positive correlations were found between RV and PI-910 (r = 0.614, p < 0.001), RV and PI950 (r = 0.616, p < 0.001), RV/TLC and PI910 (r = 0.602, p < 0.001), and RV/TLC and
TABLE 2: Pulmonary Function Test (PFT) Indexes by Chronic Obstructive Pulmonary Disease (COPD) Stage COPD Stage PFT Index TLC (L) RV (L)
Total (n = 146)
0 (n = 39)
I (n = 40)
IIA (n = 30)
IIB (n = 22)
III (n = 15)
pa
6.01 (5.24–6.85)
5.09 (4.62–6.47)
6.44 (5.29–7.15)b
6.36 (5.59–6.79)
5.64 (5.42–6.77)
6.31 (5.81–7.43)b
0.004
2.10 (1.84–2.63)
2.72 (2.37–3.42)b
3.22 (2.70–3.48)b
3.65 (3.22–4.43)b,c
4.24 (3.46–4.84)b,c,d
< 0.0001
64.10 (58.00–70.00)b,c,d
< 0.0001
2.45 (1.78–2.66)b,c
< 0.0001
2.91 (2.29–3.60)
RV/TLC (%)
49.00 (41.40–58.40) 40.20 (37.50–45.70) 46.25 (40.20–51.45) 50.75 (45.74–55.70)b 61.25 (58.40–68.60)b,c,d
TLC – RV (L)
2.21 (2.01–2.64)b,c,d
2.88 (2.41–3.53)
3.01 (2.72–3.80)
3.17 (2.57–3.94)
2.94 (2.49–3.44)
FEV1 (% predicted) 74.40 (43.30–90.40) 93.20 (85.90–108.70) 85.15 (75.65–94.45)b 59.50 (55.20–65.00)b,c 37.15 (33.80–40.80)b,c,d 27.00 (24.20–34.00)b,c,d,e
< 0.0001
FVC (% predicted) 92.10 (78.80–107.20) 101.10 (91.70–113.10) 108.00 (96.50–118.15) 86.40 (79.10–94.70)b,c 71.50 (64.40–77.00)b,c,d 68.70 (60.00–74.30)b,c,d
< 0.0001
FEV1/FVC%
62.30 (47.00–69.60) 76.50 (71.00–80.30)
64.70 (60.95–67.30)b
54.70 (49.30–62.10)b,c
40.85 (35.60–47.80)b,c,d
32.70 (31.30–39.40)b,c,d,e
< 0.0001
Note—Data are nonnormally distributed and are presented as medians (interquartile range). Bonferroni adjustments were applied in pairwise comparisons. TLC = total lung capacity, RV = residual volume, FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, FVC% = percentage forced vital capacity. aDetermined by Kruskal-Wallis test. bp < 0.05 compared with COPD stage 0. cp< 0.05 compared with COPD stage I. dp < 0.05 compared with COPD stage IIA. ep < 0.05 compared with COPD stage IIB.
714
AJR:202, April 2014
Low-Dose MDCT of Patients With COPD TABLE 3: MDCT Pulmonary Function Indexes by Chronic Obstructive Pulmonary Disease (COPD) Stage COPD Stage
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
MDCT Pulmonary Function Index
Total (n = 146)
0 (n = 39)
I (n = 40)
IIA (n = 30)
IIB (n = 22)
III (n = 15)
p
4772 ± 1032
4020 ± 1044
4968 ± 873a
5004 ± 854a
4890 ± 873a
5573 ± 886a
< 0.0001b
2429 (1424–3553)
1133 (161–2025)
2441 (1640–3524)a
2919 (2116–3809)a
2987 (2329–4103)a
4087 (3296–4701)a,c
< 0.0001d
−897 (–916 to –885)a
−904 (–918 to –886)a
−908 (–928 to –892)a
−922 (–936 to –900)a,c
< 0.0001d
0.124 (0.109–0.152) 0.103 (0.084–0.115)a 0.096 (0.082–0.114)a 0.092 (0.072–0.108)a 0.079 (0.064–0.100)a,c
< 0.0001d
Full inspiration LVin (mL) Extrapolated LVin (mL) MLA (HU)
−895 (–918 to –879) −877 (–891 to –848)
MLD (g/cm3)
0.106 (0.082–0.121)
PI-910 (%)
59.32 ± 13.13
47.71 ± 12.12
60.67 ± 10.02a
62.20 ± 10.44a
65.12 ± 11.41a
71.64 ± 8.45a,c
< 0.0001b
PI-950 (%)
42.62 ± 13.89
30.28 ± 8.81
43.10 ± 11.24a
44.93 ± 11.84a
50.53 ± 13.35a
57.19 ± 10.80a,c,e
< 0.0001b
2366 (1740–3025)
1443 (1202–2004)
2256 (1852–2706)a
2391 (2134–2963)a
3225 (2639–3890)a,c,e
3658 (3317–4547)a,c,e
< 0.0001d
992 (366–2104) a,c,e 1722 (892–2820)a,c,e
< 0.0001d
Full expiration LVex (mL) Extrapolated LVex (mL)
51.9 (3.3–547.9)
MLA (HU)
−800 ± 64
MLD (g/cm3)
0.200 ± 0.065
0.320 (0.030–3.060) 30.06 (5.39–189.86)a
207 (32–460)a
−736 ± 43
−793 ± 47a
−807 ± 42a
−858 ± 38a,c,e
−881 ± 37a,c,e
< 0.0001b
0.264 ± 0.043
0.207 ± 0.047a
0.194 ± 0.044a
0.137 ± 0.038a,c,e
0.119 ± 0.037a,c,e
< 0.0001b
PI-910 (%)
27.04 (16.10–42.00) 13.93 (10.62–16.94) 25.72 (19.83–34.73)a 32.84 (20.53–40.09)a 46.36 (37.19–60.83)a,c,e 55.09 (45.12–61.49)a,c,e < 0.0001d
PI-950 (%)
16.65 (9.61–27.30)
7.48 (5.81–10.51)
2266 ± 885
2396 ± 948
LVin – LVex (mL) LVex /LVin (%)
15.00 (10.95–21.51)a 20.32 (12.08–25.58)a 30.62 (23.42–45.77)a,c,e 37.72 (32.77–52.72)a,c,e < 0.0001d
50.82 (40.28–64.24) 38.74 (32.26–45.99)
2607 ± 773
2456 ± 760
1579 ± 638a,c,e
1643 ± 767a,c,e
< 0.0001b
46.33 (40.41–53.73)a
50.22 (43.41–57.11)a
68.55 (60.33–76.81)a,c,e
69.62 (64.50–78.16)a,c,e
< 0.0001d
Note—Normally distributed data are presented as means ± SD, and nonnormally distributed data are presented as medians (interquartile range). Bonferroni adjustments were applied in pairwise comparisons. LV in = inspiratory lung volume, MLA = mean lung attenuation, MLD = mean lung density, PI-910 = pixel index at 910 HU, PI-950 = pixel index at 950 HU, LVex = expiratory lung volume. ap < 0.05 compared with COPD stage 0. bDetermined by one-way analysis of variance. cp < 0.05 compared with COPD stage I. dDetermined by Kruskal-Wallis test. ep < 0.05 compared with COPD stage IIA.
TABLE 4: Correlation Between Spirometric Pulmonary Function Test (PFT) Indexes and MDCT Pulmonary Function Indexes in Patients With Chronic Obstructive Pulmonary Disease (COPD) MDCT Pulmonary Function Index
PFT Index TLC (L)
RV (L)
RV/TLC (%)
TLC – RV (L)
FVC (% Predicted)
LVin (mL)
0.658a
0.490a
0.163b
0.297a
−0.291a
LVex (mL)
0.419a
0.683a
0.602a
−0.245c
−0.649a
LVin – LVex (mL)
0.272a
−0.178b
−0.454a
0.585a
0.622a
−0.452a
FEV1 (% Predicted) FEV1/FVC%
COPD Stage
−0.013
−0.526a
0.405a
−0.411a
−0.759a
0.697a
0.386a
0.451a
0.243c
−0.319a
−0.595a
−0.475a
−0.600a
0.598a
LVex /LVin (%)
0.115
0.526a
Extrapolated LVin (mL)
0.415a
0.465a
0.321a
0.002
−0.450a
−0.159
−0.637a
0.556a
Extrapolated LVex (mL)
0.277a
0.640a
0.663a
−0.384a
−0.745a
−0.500c
−0.831a
0.802a
Note—Correlation was measured by Spearman rank correlation coefficient. TLC = total lung capacity, RV = residual volume, FVC = forced vital capacity, FEV1 = forced expiratory volume in 1 second, FVC% = percentage forced vital capacity, LVin = inspiratory lung volume, LVex = expiratory lung volume. ap < 0.001. bp < 0.05. cp < 0.01.
PI-950 (r = 0.603, p < 0.001). In addition, PI910 and PI-950 were both positively correlated with COPD stage (Table 5) during full inspiration and full expiration.
Comparisons of MDCT Pulmonary Function Indexes Between Smokers and Nonsmokers Each of the MDCT pulmonary function indexes was compared between smokers and
nonsmokers (Table 6). Compared with nonsmokers, smokers exhibited significantly higher LVin, LVex, LVex /LVin, extrapolated LVin, extrapolated LVex, inspiratory PI-910,
AJR:202, April 2014 715
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. expiratory PI-910, inspiratory PI-950, and expiratory PI-950 but significantly lower inspiratory mean lung attenuation, expiratory mean lung attenuation, inspiratory mean lung density, and expiratory mean lung density. Discussion The results of the current study show that certain spirometric and low-dose MDCT pulmonary function indexes differed significantly between various stages of COPD. More important, strong correlations were found between spirometric and MDCT pulmonary function indexes as well as between spirometric indexes and PI-910 and PI950. In particular, key lung parameters that showed strong correlation between MDCT and PFT indexes included LVin (mL) and TLC (L), LVex (mL) and RV (L), extrapolated LVex (mL) and FEV1/FVC%, expiratory PI-950 (%) and FEV1 (% predicted), and expiratory PI-950 (%) and FEV1/FVC% (Tables 4 and 5). These parameters all showed strong correlation between MDCT and PFT (r > 0.5, p < 0.001) and indicated that MDCT has a strong ability to evaluate key COPD-related lung parameters. These results suggest that low-dose MDCT may be useful in the diagnosis and staging of COPD. These results also indicate that imaging patients with low-dose MDCT at deep end-expiration provides better correlation with PFT results than imaging at deep end-inspiration. This finding implies that only end-expiration phase imaging may be necessary, which could further reduce the radiation dose to patients. To our knowledge, this study is the first with a large sample size that investigated the use of low-dose CT to evaluate lung function of patients with COPD. Although still in its infancy, CT is becoming a useful method for the evaluation of large and small airways disease [6, 10, 17, 18]. Routine PFTs have been the reference standard for determining the presence and degree of COPD, but they are not useful for early diagnosis and cannot be used to evaluate unilateral or partial pulmonary tissue function. Additionally, some patients have difficulty performing PFTs. Low-dose MDCT can provide additional useful clinical information: Because many COPD patients required CT evaluation during their treatment course, the possibility that a single low-dose MDCT examination can be used to evaluate both pulmonary structure and function may save both time and cost for patients and physicians.
716
TABLE 5: Correlation Between Spirometric Pulmonary Function Test (PFT) Indexes and Pixel Index at 910 HU (PI-910) and Pixel Index at 950 HU (PI-950) in Patients With Chronic Obstructive Pulmonary Disease (COPD) PFT Index
Full Inspiration
Full Expiration
PI-910 (%)
PI-950 (%)
PI-910 (%)
PI-950 (%)
FEV1 (% predicted)
−0.471a
−0.529a
−0.683a
−0.697a
FVC (% predicted)
−0.222b
−0.271a
−0.462a
−0.476a
FEV1/FVC%
−0.613a
−0.666a
−0.776a
−0.784a
TLC (L)
0.268b
0.275a
0.310a
0.313a
RV (L)
0.413a
0.445a
0.614a
0.616a
RV/TLC (%)
0.360a
0.403a
0.602a
0.603a
−0.318a
−0.316a
0.744a
0.748a
TLC – RV (L)
−0.131
COPD stage
0.560a
−0.162 0.613a
Note—Correlation was measured by Spearman rank correlation coefficient. FEV1 = forced expiratory volume in 1 second, FVC = forced vital capacity, FVC% = percentage forced vital capacity, TLC = total lung capacity, RV = residual volume. ap < 0.001. bp < 0.01.
TABLE 6: Comparison of MDCT Pulmonary Function Indexes Between Smokers and Nonsmokers With Chronic Obstructive Pulmonary Disease (COPD) MDCT Pulmonary Function Index
Nonsmoker (n = 37)
Smoker (n = 109)
p
LVin (mL)
3766 (3264–4223)
5078 (4462–5724)
< 0.0001a
LVex (mL)
1678 (1314–2283)
2624 (2134–3435)
< 0.0001a
41.55 (36.10–55.97)
52.55 (43.24–64.58)
0.028a
Extrapolated LVin (mL)
1349 (668–2126)
2883 (2004–3908)
< 0.0001a
Extrapolated LVex (mL)
4.65 (0.32–27.47)
181.61 (15.08–892.12)
< 0.0001a
Inspiratory
−876 ± 40
−902 ± 26
0.0006b
Expiratory
−765 ± 56
−811 ± 62
0.0001b
Inspiratory
0.124 ± 0.040
0.098 ± 0.026
0.0006b
Expiratory
0.236 ± 0.056
0.188 ± 0.063
< 0.0001b
Inspiratory
51.62 ± 15.09
61.93 ± 11.33
0.0004b
Expiratory
15.98 (12.86–21.99)
33.20 (21.59–46.77)
< 0.0001a
Inspiratory
34.53 ± 13.59
45.37 ± 12.94
< 0.0001b
Expiratory
8.96 (6.60–13.64)
20.95 (13.05–32.23)
< 0.0001a
LVex /LVin (%)
Mean lung attenuation (HU)
Mean lung density (g/cm3)
PI-910 (%)
PI-950 (%)
Note—Normally distributed data are presented as means ± SD and nonnormally distributed data are presented as medians (interquartile range). LV = lung volume, LVex = expiratory lung volume, LV in = inspiratory lung volume, PI-910 = pixel index at 910 HU, PI-950 = pixel index at 950 HU. aDetermined by Wilcoxon rank sum test. bDetermined by independent Student t test.
In the current study, LVin was 4772 ± 1032 mL (mean ± SD) and median LVex was 2366 mL (IQR, 1740–3025 mL) as determined by MDCT. Spirometric PFTs showed that the median TLC was 6.01 L (IQR, 5.24–6.85 L)
and the median RV was 2.91 L (IQR, 2.29– 3.60 L). The difference between the two detection methods was approximately 1200 and 550 mL, respectively. Routine PFTs are performed with the patient seated, but CT is per-
AJR:202, April 2014
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Low-Dose MDCT of Patients With COPD formed with the patient in a supine position; this difference in patient positioning may account for the differences in the lung function parameters detected with these two methods. Our results show that the lung volume indexes by MDCT correlated with PFT volume. The correlation coefficient was 0.683 between LVex and RV, 0.658 between LVin and TLC, and 0.622 between LVex /LVin and RV/TLC. Thus, we speculate that values of LVin, LVex, and LVex /LVin detected with lowdose MDCT can be used to evaluate TLC, RV, and RV/TLC, respectively. All of these MDCT indexes reflect the actual air volume in the lung and are thus more precise than the corresponding spirometric PFT values. Kauczor et al. [9] measured lung attenuation at paired high-resolution CT performed at full inspiration and full expiration and correlated the values with PFT results and found that the inspiratory mean lung density and the expiratory attenuation increase were able to differentiate patients with obstructive and restrictive ventilatory impairment from healthy subjects and that scans obtained at full expiration provided the best results. Zaporozhan et al. [4] used 3D high-resolution CT data obtained at inspiration and expiration for the quantitative evaluation of emphysema and reported that emphysema volumes measured from expiratory scans were more consistent with PFT results. In the current study, the lung density was evaluated with two methods. For the first method, we used the CT attenuation value with scanning performed at 40 mA and found that the mean attenuation value of the whole lung was –895 ± 32 HU at end-inspiration and –800 ± 64 HU at end-expiration; these results are consistent with the findings obtained using standarddose dual-phase MDCT [4]. For the second method, we used volume density. Determination of the PI threshold is crucial when evaluating lung disease by CT and is typically determined by the investigator. In this study, we defined the PI threshold as –950 HU. Gevenois et al. [19] compared the area covered by different pixels with pathologic findings and found that –910 HU in the expiration phase and –950 HU in the inspiration phase were significantly related to pathologic findings and that –910 HU in the expiration phase had the most optimal efficiency in determining the area of emphysema. Müller et al. [20] also found that –910 HU was the best threshold for quantifying the area of emphysema. Some investigators have used –950 HU as the trigger threshold for severe emphy-
sema [11]. In the current study, when –910 and –950 HU were used as trigger thresholds, the PI at deep end-inspiration and deep end-expiration was significantly associated with FEV1 (% predicted) and FEV1/FVC% and the relationship in the expiration phase was superior to that in the inspiration phase. The best correlation was observed when PI-910 was used as the trigger threshold (r = –0.78 for expiration phase and –0.69 for inspiration phase). Thus, we speculate that –910 HU as a trigger threshold in the expiration phase can be used to measure emphysema volume. In this study, we compared MDCT pulmonary function indexes between smokers and nonsmokers and found that LVin, LVex, LVex/ LVin, extrapolated LVin, extrapolated LVex, inspiratory PI-950, expiratory PI-950, inspiratory PI-910, and expiratory PI-910 in smokers were markedly increased when compared with nonsmokers. In smokers, inspiratory mean lung attenuation, expiratory mean lung attenuation, inspiratory mean lung density, and expiratory mean lung density were significantly reduced compared with the values in nonsmokers; these results indicate that for patients with clinically proven COPD, lung function in smokers is poorer than that in nonsmokers. We speculate that smoking not only causes inflammation in the small airways but also may cause disruption of the lung anatomic structure in COPD patients. The differences in lung density and emphysema volume show that the emphysema is more severe in smokers than in nonsmokers. There are some limitations of this study that should be considered. We did not compare standard-dose MDCT and low-dose MDCT in the same patients because of concern about increased radiation exposure. In addition, there were no control subjects with COPD who underwent standard-dose MDCT and low-dose MDCT. Conclusion Our results showed that lung function indexes obtained with low-dose dual-phase MDCT correlated well with those obtained from spirometric PFTs in patients with COPD and that the correlation at expiration was superior to that at inspiration. In addition, MDCT indexes varied among the different stages of COPD and thus may be useful for determining the stage of COPD in individual patients. References 1. Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS; GOLD Scientific Committee. Glob-
al strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO Global Initiative for Chronic Obstructive Lung Disease (GOLD) Workshop summary. Am J Respir Crit Care Med 2001; 163:1256–1276 2. Group of Chronic Obstructive Pulmonary Disease of Branch of Respiratory Diseases of Chinese Medical Association. Guideline for diagnosis and treatment of chronic obstructive pulmonary disease (2007 revision) [in Chinese]. Chin J Tubercul Respir Dis 2007; 30:8–17 3. Gómez FP, Rodriguez-Roisin R. Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines for chronic obstructive pulmonary disease. Curr Opin Pulm Med 2002; 8:81–86 4. Zaporozhan J, Ley S, Eberhardt R, et al. Paired inspiratory/expiratory volumetric thin-slice CT scan for emphysema analysis: comparison of different quantitative evaluations and pulmonary function test. Chest 2005; 128:3212–3220 5. Bankier AA, Madani A, Gevenois PA. CT quantification of pulmonary emphysema: assessment of lung structure and function. Crit Rev Comput Tomogr 2002; 43:399–417 6. Madani A, Zanen J, de Maertelaer V, Gevenois PA. Pulmonary emphysema: objective quantification at multi-detector row CT—comparison with macroscopic and microscopic morphometry. Radiology 2006; 238:1036–1043 7. Shah E, Gilkeson R, Krishna S, Ciancibello L, Durgan J, Pohlman S. Comparison of computeraided calculation of emphysema volumetry with manual quantification using CT images. Int Congr Ser 2004; 1268:956–960 8. O’Donnell RA, Peebles C, Ward JA, et al. Relationship between peripheral airway dysfunction, airway obstruction, and neutrophilic inflammation in COPD. Thorax 2004; 59:837–842 9. Kauczor HU, Hast J, Heussel CP, Schlegel J, Mildenberger P, Thelen M. CT attenuation of paired HRCT scans obtained at full inspiratory/ expiratory position: comparison with pulmonary function tests. Eur Radiol 2002; 12:2757–2763 10. Park YS, Seo JB, Kim N, et al. Texture-based quantification of pulmonary emphysema on highresolution computed tomography: comparison with density-based quantification and correlation with pulmonary function test. Invest Radiol 2008; 43:395–402 11. Orlandi I, Moroni C, Camiciottoli G, et al. Spirometric-gated computed tomography quantitative evaluation of lung emphysema in chronic obstructive pulmonary disease: a comparison of 3 techniques. J Comput Assist Tomogr 2004; 28:437–442 12. Bastarrika G, Wisnivesky JP, Pueyo JC, et al. Low-dose volumetric computed tomography for quantification of emphysema in asymptomatic
AJR:202, April 2014 717
Downloaded from www.ajronline.org by East Carolina University on 06/09/14 from IP address 150.216.68.200. Copyright ARRS. For personal use only; all rights reserved
Chen et al. smokers participating in an early lung cancer detection trial. J Thorac Imaging 2009; 24:206–211 13. Wu XH, Ma DQ, Zhang ZJ, et al. The experimental study and clinical application on the detection of pulmonary nodules with low-dose multislice spiral CT [in Chinese]. Zhonghua Fang She Xue Za Zhi 2004; 38:767–770 14. Arakawa A, Yamashita Y, Nakayama Y, et al. Assessment of lung volumes in pulmonary emphysema using multidetector helical CT: comparison with pulmonary function tests. Comput Med Imaging Graph 2001; 25:399–404
718
15. Matsuoka S, Kurihara Y, Yagihashi K, Nakajima Y. Quantitative assessment of peripheral airway obstruction on paired expiratory/inspiratory thin-section computed tomography in chronic obstructive pulmonary disease with emphysema. J Comput Assist Tomogr 2007; 31:384–389 16. Satoh S, Ohdama S, Shibuya H. Sliding thin slab, minimum intensity projection imaging for objective analysis of emphysema. Radiat Med 2006; 24:415–421 17. Kauczor HU, Wielpütz MO, Owsijewitsch M, et al. Computed tomographic imaging of the airways
in COPD and asthma. J Thorac Imaging 2011; 26:290–300 18. Matsuoka S, Yamashiro T, Washko GR, Kurihara Y, Nakajima Y, Hatabu H. Quantitative CT assessment of chronic obstructive pulmonary disease. RadioGraphics 2010; 30:55–66 19. Gevenois PA, De Vuyst P, Sy M, et al. Pulmonary emphysema: quantitative CT during expiration. Radiology 1996; 199:825–829 20. Müller NL, Thurlbeck WM. Thin-section CT, emphysema, air trapping, and airway obstruction. Radiology 1996; 199:621–622
AJR:202, April 2014
