Morphological And Hemodynamic Study In COPD With HRCT And MR Perfusion Imaging

【Background】Chronic obstructive pulmonary disease (COPD) is a preventable and treatable disease and characterized by airflow limitation that is not fully reversible. The airflow limitation is usually progressive and associated with an abnormal inflammatory response of the lung to noxious particles or gases. COPD is one of the leading causes of morbidity and mortality and one of highest burden in economy worldwide and is projected to be the third leading cause of chronic mortality in2020. The best way to reduce the mortality is early diagnosis and early treatment. Although extremely useful in clinical practice, pulmonary functional testing (PFT) are known to be relatively insensitive to both early stages and small changes of manifest disease. In contrast to spirometry, high-resolution computed tomography (HRCT) might allow for regional assessment of the compartments involved and has been shown to correlate well with pathology. MR perfusion imaging might allow for regional assessment of pulmonary function and has been shown to correlate well with scintigraphy. With reasonable spatial resolution and well anatomy coverage, MR perfusion imaging allows for lobar/segmental analysis in patients with COPD. Intimal enlargement of pulmonary arteries is an early change inCOPD, MR perfusion imaging with high sensitivity can reflect this abnormal change. In this study, we evaluate the Morphological and hemodynamic characteristics of HRCT and MR perfusion imaging in patients with COPD and compare3D dynamic contrast-enhanced (DCE) MR perfusion imaging with PFT and quantitative HRCT using correlation analysis. Based above research, we prospectively explore MR perfusion imaging characteristics in mild COPD for the early diagnosis of COPD. [Purpose] To investigate the characteristics (patient data, parameters of pulmonary functional testing (PFT), HRCT volumetric parameters and MR perfusion values) of various HRCT phenotypes of COPD and to compare HRCT volumetric parameters with body mass index (BMI) and parameters of PFT in patients with Chronic Obstructive Pulmonary Disease (COPD).[Materials and Methods]62consecutive patients with chronic obstructive pulmonary disease (COPD) underwent HRCT, PFT and MRI on the same day. COPD subjects were classified into three phenotypes on the basis of the visual HRCT findings (with or without emphysema and/or bronchial wall thickening). From the volumetric CT data, the total lung volume (TLV) was calculated automatically and the total emphysema volume (TEV) was obtained by applying density thresholds of-950HU, the TEV/TLV was calculated as an emphysema index (EI). The3D MR perfusion data was post processed using Functool, signal intensity of perfusion defects (SIPD), signal intensity of normal lung perfusion (SInormai) were obtained by selecting region of interest(ROI) and the signal intensity ratio (RSI) of the perfusion defects and normal lung perfusion was calculated. The correlation between the parameters from quantitative CT, BMI and PFT was assessed using Spearman correlation analysis. The different stages of COPD subjects were evaluated using Mann-Whitney U testing.[Results]1. According to the visual HRCT findings, the COPD patients were classified into3phenotypes:Phenotype A (n=42), Phenotype E (n=9) and Phenotype M (n=11). COPD patients with phenotype A (23.6±3.5) kg/m2showed a higher body mass index (BMI) vs. phenotype E(20.4±3.3)kg/m2(P<0.05). However, there was no significant difference between phenotype M (21.3±2.7) kg/m2and phenotypes A or E. The FEV1/FVC of phenotype A (64.5±13.2)was higher than phenotype M (51.9±11.6)(P<0.05). But, there was no statistical difference between phenotype E (60.5±13.9) and A or M. TEV and El in phenotype A [(0.4±0.4)1and (8.0±6.3)%]were significantly lower than in phenotype E [(0.9±0.6)1and(16.6±7.2)%] and phenotype M[(0.8±0.4)1and (15.4±6.7)%](P=0.001). TEV and El had no significant difference between phenotype E and M. The TLV in phenotype M (5.3±6.9)1was significantly higher than in phenotype A (4.3±1.3)1(P＜0.05), but was not statistically different from phenotype E (5.1±1.5)1. The RSI in phenotype A (20.3±8.5)was higher than phenotype E (11.8±5.4)(P=0.003), but was not different from that in phenotype M (15.5±5.5), statistically.2. TEV and El of HRCT volumetric parameters separately correlated with BMI and pulmonary function tests [forced expiratory volume in the first minute of expiration (FEV1), FEV1of predicted value (%FEV1), forced vital capacity(FVC), and ratio of FEV1to FVC (FEV1/FVC)(p<0.001)]. When FEVi of50%predicted value was used as threshold, TEV and El moderated correlated with FEV1/FVC (p<0.05). HRCT volumetric parameters for COPD patients at all stages, The TEV and El of Stage IV COPD patients was significantly different from that of Stage I or II The El showed a significant difference between Stage III and Stage I(P＜0.005).[Conclusion] The BMI, FEV1/FVC, Rsi of phenotype A were higher and the TLV was lower, which could relate to the COPD conditions of Phenotype A which were mild. TEV and El in phenotype A lower than that in phenotype E and M, meets the definition of HRCT phenotypes. HRCT volumetric parameters and MR perfusion values had no significant difference between phenotype E and M; which may suggest that the difference tends to associate with the airways. TEV and El separately correlated with BMI and parameters of pulmonary functional test and could reflect the severity of COPD. [Purpose] To prospectively compare values of3D dynamic contrast-enhanced (DCE) MR perfusion imaging with pulmonary function testing (PFT) and quantitative HRCT in patients with COPD for GOLD classification and detecting mild COPD.[Materials and methods]62patients with COPD and17controls (healthy volunteers) underwent PFT, HRCT and3D-DCE MR perfusion imaging. In addition,3patients with COPD underwent PFT and3D-DCE MR perfusion imaging. According to the Global Initiative for Chronic Obstructive Lung Disease (GOLD), the COPD patients were classified. The3D MR perfusion data was post processed using Functool, signal intensity of perfusion defects (SIPD), signal intensity of normal lung perfusion (SInormal) were obtained by selecting region of interest(ROI) and the signal intensity ratio (RSI) of the perfusion defects and normal lung perfusion was calculated. From the volumetric CT data, the total lung volume (TLV) was calculated automatically and the total emphysema volume (TEV) was obtained by applying density threshold of-950HU. The TEV/TLV was calculated as emphysema index (EI). The correlation between the perfusion values of3D-DCE MRI, the PFT values and the HRCT volumetric parameters was assessed using Spearman correlation analysis. Several comparisons were made between the control and all stages of COPD patients by using Bonferroni testing or Mann-Whitney U testing.[Results]1. All3D-DCE MR examinations were successfully performed. Manifestations of MR perfusion images in patients with COPD were generally patchy perfusion defects. The PFT values correlated significantly with almost all MRI perfusion parameters. The strong correlation was detectable between the forced expiratory volume in the first second of expiration of predicted value(％FEV1) and RSI (r=0.659,P<0.001),between FEV1and RSI(r=O.608, p<0.001).In addition,there was a correlation between SIPD and％FEV1(r=0.598,p<0.001), between RsI and ratio of FEVlto forced vital capacity(FEV1/FVC)(r=0.579,P<0.001).RsI showed moderate correlation with FVC(r=0.526,p＜0.001)and SI PD had moderate correlation with FVC.FEV1/FVC、FEV1[r (range0.479~0.541),P≤0.001].MSI、PEI and MSD had signi fi cant correlation with the PFT values[r(range0.255~0.419),P <0.05]. Compared with all patients of COPD,the51smoking-related COPD have stronger correlations between MRI perfusion parameters and PFTs.2.The RSI showed good correlation with the TLV,TEV and EI (r=-0.312,P<0.013；r=-0.466,P<0.001；r=-0.476, P<0.001)for all patients with COPD.In contrast, there was little lower correlation between SI PD and TLV,TEV,EI (r=-0.306,P=0.015；r=-0.400,P=0.001；r=-0.394,P=0.002).For GOLD II patients,the RSI had strong correlation with TEV and EI (r=-0.583,P=0.014； r=-0.623,P=0.008)；the SI PD had moderate correlation with TLV,TEV and EI (r=-0.495, P=0.043；r=-0.586,P=0.013；r=-0.534,P=0.027.The correlation coefficient between the TLV values and the SIPD was r=-0.594(P=0.042),between the TLV values and the RSl was r=-0.713(P=0.009)for GOLD IV patients.However,there was no correlation between the perfusion values of3D-DCE MRI and the HRCT volumetric parameters for patients as GOLD I and III. Correlation between RSI and FVC(r=0.529,P<0.001)and between RSI and％FEVl(r=0.655,P<0.001)was better than that between EI and either FVC(r=-0.372,P<0.001)or％FEVl (r=-0.532,P<0.001).3.All MRI perfusion parameters of the controls were significantly different from those of all stages patients with COPD (P≤0.002).In addition,the RSI of Stage1was significantly different from that of Stage III or IV(P≤0.001). Finally, the RSI showed a significant difference between Stage II and Stage IV(P≤0.001).The EI of the controls showed signi fi cant differences with that of COPD StageII-IV(P<0.001).On the other hand,the EI of stage IV was different from that of Stage II or I(P≤0.002),but there was no significantly difference between the controls and the COPD of Stage I (P=0.504).[Conelusion] The most parameters of3D-DCE MRl was posiitively correlated with the PFT values and was inversely correlated with HRCT volumetric parameters in COPD. The MRI perfusion values and HRCT volumetric parameters could reflect the severity of COPD.The RsI calculated by MRI should be considered a more sensitive value than the HRCT volumetric parameters to distinguish between the controls and COPD(Stage I).