Titre de l’article : L’adiposité et de la composition corporelle évaluées par TACO sont associées aux comorbidités et à un faible état de santé dans la MPOC. Une sous-étude ECLIPSE.
Rationnel : Un indice de masse corporelle élevé a été paradoxalement associé à un meilleur taux de survie dans les maladies chroniques comme la MPOC. Toutefois, l’interprétation de l’impact de l’obésité sur la survie dans la MPOC a pu être influencée par la distribution du tissu gras qui n’as pas été prise en compte dans ces études. Cette sous-étude ECLIPSE teste l’hypothèse selon laquelle l’adiposité intra-abdominale serait augmentée dans la MPOC et serait associée aux comorbidités et à des évènements cliniques néfastes.
Méthodes : Des images de tomodensitométrie axiale contrôlée par ordinateur (TACO) au niveau des vertèbres L2-L3 obtenues à partir de l’étude ECLIPSE ont été utilisées pour quantifier la surface de tissu adipeux viscéral (TAV) et de tissu musculaire (TM) ainsi que de l’atténuation musculaire, un reflet de l’infiltration lipidique du muscle, chez les patients atteints de MPOC et chez les sujets témoins. Des modèles de régression ont été utilisés pour investiguer les possibles relations entre les indices d’adiposité ainsi que la surface de tissu musculaire, avec les comorbidités et les manifestations cliniques de la MPOC.
Résultats : À partir de la cohorte ECLIPSE, nous avons identifié 585 sujets pour qui nous avions des images de TACO à L2-L3 permettant l’analyse de la composition corporelle. La surface de TAV est augmentée (p<0.0001) et la surface musculaire ainsi que l’atténuation réduite (indication d’une infiltration lipidique accrue) chez les patients avec une MPOC comparativement aux témoins. Bien que la sévérité de la maladie ne soit pas associée au TAV augmenté, l’atténuation musculaire diminue en fonction de la progression de la limitation du débit expiratoire. La surface de TAV est associée à une
probabilité accrue de retrouver des comorbidités cardiovasculaires (p = 0.042). Des corrélations ont été établies entre les variables mesurées par TACO et le test six minutes de marche, le score total au St. Georges’ Respiratory Questionnaire et certains biomarqueurs.
Conclusion : Notre étude confirme que l’accumulation de tissu adipeux viscéral, ainsi que l’augmentation de gras intramusculaire et la diminution de la surface de tissu musculaire dans l’abdomen sont associées à plusieurs comorbidités et à des évènements cliniques significatifs dans la MPOC.
Computed tomography-derived adiposity and body composition are associated with comorbidities and poor outcomes in COPD. An
ECLIPSE substudy Authors:
Mickaël Martin1, Natalie Almeras1, Jean-Pierre Després1, Harvey O Coxson2,
George R Washko3, Isabelle Vivodtzev4, Emiel FM Wouters5, Erica Rutten6,
Michelle C Williams7, John T Murchison8, William MacNee7, Don D Sin2,
François Maltais1 for the Evaluation of COPD Longitudinally to Identify
Predictive Surrogate Endpoints (ECLIPSE) investigators.
Affiliations:
1Centre de recherche, Institut universitaire de cardiologie et de pneumologie
de Québec, Université Laval, Québec Canada; 2University of British
Columbia, Vancouver, Canada; 3Division of Pulmonary and Critical Care
Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, United States; Grenoble Alps University, Inserm U1042 Hypoxia Pathophysiology laboratory and Grenoble University Hospital, Grenoble, France; 5Maastricht University Medical Center,
Maastricht, The Netherlands; 6Research and development, CIRO, Horn, The
Netherlands; 7University of Edinburgh, Edinburgh, UK; 8Royal Infirmary of
Edinburgh, Edinburgh, UK. Short title: Adiposity and COPD Corresponding author:
Dr. F. Maltais
Centre de Pneumologie
Institut Universitaire de cardiologie et de pneumologie de Québec, 2725 Chemin Sainte-Foy
Fax: 1 418 656 4762
E-Mail: Francois.Maltais@fmed.ulaval.ca
Key words: metabolic syndrome, obesity, abdominal adiposity, muscle Word count: 3463 (max: 3500)
COMPETING INTERESTS
MM, NA, IV, ER, MCW and JTM have no conflicts of interest to declare. JPD reports personal fees from Abbott Laboratories; AstraZeneca; GlaxoSmithKline; Merck; Pfizer Canada Inc.; Sanofi; Torrent Pharmaceuticals Ltd. outside the submitted work. EFMW reports personal fees from Nycomed; AstraZeneca; GlaxoSmithKline; Novartis outside the submitted work. GRW reports other from GlaxoSmithKline, other from Genentech, outside the submitted work. HOC reports personal fees from GSK, grants from GSK during the conduct of the study; personal fees from Samsung, grants from Spiration Inc outside the submitted work. WM reports personal fees and other from GlaxoSmithKline, personal fees and other from Pfizer, personal fees and other from AstraZeneca, personal fees and other from Boehringer Ingelheim, grants from British Heart Foundation, grants from Chief Scientist Office, grants from MRC, outside the submitted work. DDS reports personal fees from Amgen, grants and personal fees from AstraZeneca, personal fees from Boehringer Ingelheim, grants from Novartis, outside the submitted work. FM
personal fees from GlaxoSmithKline, grants from Nycomed, grants and personal fees from Novartis, outside the submitted work. All fees are pooled with other revenues of the group of pulmonologists to which FM is a member and then shared among members of the group. FM holds a CIHR/GSK research Chair on COPD at Université Laval.
FUNDING
GlaxoSmithKline. MM was recipient of a research training award from the Fonds de la recherche Québec - Santé. FM holds a GSK/CIHR Research Chair on COPD at Université Laval. JPD is the scientific director of the International Chair on Cardiometabolic Risk which is based at the Faculty of medicine, Université Laval.
This work has been presented as an abstract at the 2015 meeting of the American Thoracic society
ABBREVIATIONS LIST
6MWD: 6-min walking distance AT: adipose tissue
BMI: body mass index
CC-16: club cell secretory protein
COPD: chronic obstructive pulmonary disease CSA: muscle tissue cross-sectional area CT: computed tomography
ECLIPSE: Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints
FEV1: Forced expiratory volume in 1 second
GOLD: Global initiative for Chronic Obstructive Lung Disease hs-CRP: high-sensitivity C-reactive protein
HU: Hounsfield units
ICC: intraclass correlation coefficient IL-6: interleukin-6
MT: muscle tissue
PARC/CCL-18: pulmonary and activation-regulated chemokine SCAT: subcutaneous adipose tissue
SGRQ: St. Georges’ Respiratory Questionnaire SP-D: surfactant protein
VAT: visceral adipose tissue WBC: white blood cell count
ABSTRACT
Background : Increased body mass index has been paradoxically associated with prolonged survival in COPD. However, interpreting the impact of obesity on COPD has been confounded by not considering body composition or fat distribution. This ECLIPSE substudy tested the hypothesis that COPD is associated with increased intra-abdominal adiposity and that this would be associated with comorbidities and adverse clinical outcomes. Methods: Computed tomography images of the thorax obtained in the ECLIPSE study were used to quantify, at L2-L3, visceral adipose tissue (VAT) and muscle tissue (MT) cross-sectional areas (CSA) and MT attenuation, a reflection of muscle fat infiltration. Regression models were used to investigate possible relationships between CT body composition indices, comorbidities and COPD-related outcomes. Results: From the entire ECLIPSE cohort, we identified 585 subjects with valid CT images at L2-L3 to assess body composition. VAT CSA was increased (p<0.0001) and muscle CSA and attenuation decreased (indicating more muscle fat accumulation) in COPD compared to controls. Whereas disease severity was not associated with more VAT, muscle attenuation decreased as airflow limitation worsened. VAT CSA was associated with increased probability of diabetes (p = 0.024) and gastroesophageal reflux (p = 0.0048) at baseline while MT attenuation predicted an increased probability of finding cardiovascular comorbidities (p = 0.042). Correlations were found between CT-derived adiposity and body
Respiratory Questionnaire total scores and selected biomarkers. Conclusion: CT-derived adiposity and body composition variables were associated with comorbidities and with relevant clinical outcomes in COPD.
What is the key question?
Is there an increased visceral adipose tissue accumulation in patients with COPD and does this contribute to COPD-related outcomes and comorbidities?
What is the bottom line?
An increased intra-abdominal (visceral) fat accumulation/muscle fat infiltration was observed in COPD and this was associated with diabetes, cardiovascular comorbidities and gastroesophageal reflux and with relevant COPD-related outcomes such as reduced walking capacity and quality of life.
Why read on?
Intra-abdominal (visceral) fat accumulation being a risk factor for cardiovascular diseases, a leading mortality cause in COPD, regional body fat distribution and ectopic fat deposition should be properly assessed to optimally manage patients with COPD in clinical practice.
INTRODUCTION
Obesity is becoming a major health issue in COPD, as it is in the general population. Several studies have reported an increased proportion of patients with COPD who are either overweight or obese.1-4 Based on the well
established link between obesity and hypertension, type 2 diabetes, cancer, and cardiovascular diseases,5 and on the fact that COPD is an independent
risk factor for cardiovascular diseases,6 one would assume that COPD
patients who are obese are characterized by an increased mortality rate. Consistent with these observations, all-cause and cardiovascular mortality are also increased in patients with COPD.7 Depending on COPD severity, cardiovascular diseases are the primary causes of mortality in 20 to 27% of patients with COPD.8,9
In this context, it is somewhat counterintuitive that overweight and obesity paradoxically prolonged survival in COPD10 as well as in other chronic diseases,11 a phenomenon referred to as the “obesity paradox” or “reverse
epidemiology”.12 Many studies have reproduced this finding which appears to
be robust and observed up to very high body mass index (BMI) values (above 40 kg/m2).12
It has been proposed that this paradox may be a “BMI paradox” rather than an “obesity paradox”.13 Indeed, BMI does not provide precise information
consider in the interpretation of the relationship between obesity and COPD. For instance, it is very well documented that the inflammatory, metabolic and cardiovascular consequences of obesity are largely explained by the presence of excess visceral adipose tissue which is also partly a marker of unwanted lipid accumulation in normally lean tissues such as the liver, the heart and the skeletal muscle, this phenomenon being described as ectopic fat accumulation.14 Individuals with a selective accumulation of abdominal
(visceral) fat are at increased risk of developing features of the metabolic syndrome, diabetes, an atherogenic dyslipidemia, inflammation or cardiovascular diseases compared to those who preferentially accumulate lipids in subcutaneous “healthy” adipose tissue.14 We therefore put forward
the hypothesis that the paradoxical relationship between COPD and obesity has been largely confounded by the absence of body composition and regional adipose tissue distribution data.
The Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) was an internationnal multicenter, observational and longitudinal study to identify surrogate markers of COPD progression.15
ECLIPSE enrolled current or former smokers with Global initiative for Chronic Obstructive Lung Disease (GOLD) stage 2 - 4 COPD, smokers with normal lung function and never smokers.16 Interestingly, all ECLIPSE participants
abdominal fat distribution, and muscle tissue area and attenuation, an expression of fat infiltration..17
We therefore took advantage of this unique COPD cohort with a 3-year follow-up to address the following question: i) is COPD associated with increased visceral obesity? ii) does visceral obesity and muscle fat infiltration modify the expression of COPD in being associated with reduced functional and health status, cardiovascular and other comorbidities, as well as with forced expiratory volume in 1 second (FEV1) rate of decline, exacerbations
and mortality? Our overarching hypothesis was that COPD would be associated with increased intra-abdominal fat accumulation compared to controls with normal lung function and that this would be associated with adverse clinical outcomes.
METHODS
Study design and patients
ECLIPSE (Clinicaltrials.gov, NCT00292552) enrolled 2164 current or former smokers (≥ 10 pack-years) with GOLD stage 2-4, as well as 337 smokers with normal lung function and 245 never smokers.16 Patients’ characteristics (age,
race, gender, smoking history), chest CT scan, pulmonary function tests, St. Georges’ Respiratory Questionnaire (SGRQ) scores, 6-min walking distance (6MWD), serum inflammatory biomarkers (white blood cell count [WBC], interleukin-6 [IL-6], high-sensitivity C-reactive protein [hs-CRP], fibrinogen, pulmonary and activation-regulated chemokine [PARC/CCL-18], surfactant protein [SP-D], club cell secretory protein [CC-16]) were obtained at baseline and yearly for the 3-year study duration as previously reported15,18,19
Standardized questionnaires were used to record comorbidity at baseline. COPD exacerbations and mortality were also recorded. All patients provided written informed consent and the study were approved by ethics committees of participating centres. ECLIPSE steering and scientific committees approved the protocol for this substudy.
Imaging assessment
All ECLIPSE participants underwent a low-dose volumetric CT scan of the chest at baseline with the following protocol: 120-kV peak, 40 mA, and 1.00- or 1.25-mm slice thickness, at full inspiration. For the purpose of this
contained images at L2-L3 intervertebral space. Measures of fat and muscle (psoas, paraspinal and abdominal wall muscles) tissues cross-sectional areas and of muscle tissue (MT) attenuation were performed by the same investigator (MM) using the CT imaging Core Lab Unit at the Institut Universitaire de cardiologie et de pneumologie de Québec and a specialized image analysis software (Slice-O-Matic, Tomovision, Montréal, Québec, Canada) according to standardized techniques (Figure 1).17 Adipose and
muscle tissue cross-sectional areas (CSA, cm2) were measured using an
attenuation range of –190 to –30 Hounsfield units (HU) and of -29 to 130 HU, respectively. The mean attenuation value in HU of each structure was generated. To avoid subjectivity in the separation between subcutaneous (SCAT) and visceral (AT) adipose tissues, these two fat tissues were separated from each other midway in the thickness of abdominal wall muscles. Subjects for whom a tissue surface area > 1 cm2 fell outside the
field of view of the scope were excluded from further analysis. A random selection of 10% of all segmented images were redone by the same evaluator to assess for intra-observer variation. Inter-observer variation was evaluated on a random sample of 10% of the segmented images that were also analysed by a second imaging expert.
Statistical analyses
variables. Coronary artery calcification scores were obtained from a previous study in which a random sample of the ECLIPSE cohort was selected for this analysis.20 This variable was available for 394 patients with COPD and 53
controls. Coronary artery calcification (CAC) scores and inflammatory markers were analysed on a log10 scale because they had a log normal
distribution. Patients with COPD were further classified according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD) spirometric classification system.21 The overall test-retest reproducibility of each tissues
compartment parameter was assessed by calculating a Pearson correlation coefficient between evaluation #1 and #2 for each CT body composition parameter. Test-retest reliability was evaluated using Intraclass Correlation Coefficient (ICC) as it reflects both systematic and random differences in test measures.22 An ICC > 0.75 was judged to be excellent.23 CT body
composition variables were compared between COPD and controls, and also across GOLD spirometric grades, using a multivariate analysis of covariance model (MANCOVA), with age, sex, BMI and smoking status included as covariates. When a significant effect of groups or subgroups were identified, univariate analyses of covariance models (ANCOVA) were adjusted to investigate which body composition variables were different, and LSD protected multiple comparisons were performed to differentiate the subgroups. Possible relations between body composition, pulmonary function, coronary artery calcification scores, exacerbation rate and
the effect of body composition on FEV1 rate of decline, as proposed by
Vestbo et al.24. The impact of body composition on comorbidities was
assessed by logistic regression models while its relationship with survival was studied using a Cox proportional hazards regression model. All analyses were controlled for age, sex, BMI, smoking status and FEV1. P values < 0.05
were considered statistically significant. Data analyses were performed using SAS version 9.4 (SAS Institute Inc., CA, USA).
RESULTS
We identified 585 patients (511 COPD and 74 controls) for whom a CT slice at L2-L3 could be analysed for fat distribution. Patient’s characteristics are outlined in Table 1. Patients with COPD and controls had similar BMI but important differences were otherwise noted: COPD were older, had a higher proportion of men and of current/former smokers than controls. They also exhibited higher SGRQ scores (worse health-related quality of life), greater prevalence of cardiovascular comorbidities (angina, heart attack, myocardial infarction, stroke and heart failure), hypertension, greater coronary artery calcification (CAC) score and higher levels of inflammatory biomarkers than controls. During the 3-year follow-up, 50 deaths were recorded (49 patients with COPD and 1 control) resulting in a lower mean survival days in COPD. A total of 1661 moderate to severe exacerbation were reported in COPD, with a mean exacerbation rate of 1.28 ± 0.06 year-1patient-1 (range 0 to 9
exacerbation year-1patient-1).
CT fat distribution data adjusted for age, sex, smoker status and BMI are presented in Figure 2. For a similar BMI (Figure 2A), patients with COPD exhibited higher VAT CSA (Figure 2B), a tendency to have lower MT CSA (Figure 2C) and lower MT attenuation (Figure 2D) than controls. SCAT CSA could be quantified in 384 subjects (327 COPD and 57 controls). SCAT CSA was greater in patients with COPD than in controls (Figure 2E). The same
GOLD 2 patients while GOLD 4 were characterized by greater VAT than GOLD 3 group and SCAT areas and lower muscle tissue attenuation than the remaining groups.
Correlation between visceral fat parameters, MT area, clinical outcomes and biomarkers are presented in Table 3 and Figure 3. 6MWD was negatively associated with VAT CSA (figure 3A) and positively associated with MT CSA (Figure 3B). Reduced MT attenuation was predictive of lower 6MWD (Figure 3C). Increased MT CSA was predictive of better quality of life as indicated by its negative correlation with SGRQ total scores where lower scores indicate better quality of life (Figure 3D). Fat distribution did not predict FEV1 decline
and the rate of exacerbation. VAT CSA predicted WBC and negatively CC- 16, MT CSA was a negative predictor of fibrinogen and MT attenuation was positively associated with CC-16 levels (Table 3).
In COPD, the probability of having diabetes (p = 0.024) and gastroesophageal reflux (p = 0.0048) at baseline increased in parallel with VAT accumulation while MT attenuation predicted increased probability of cardiovascular comorbidities (p = 0.042) (Figure 4). Tissues repartition variables did not correlate with coronary artery scores nor survival.
DISCUSSION
This ECLIPSE sub-study reports altered visceral adipose tissue distribution and MT area findings in a large group of patients with COPD. Patients with COPD were characterized by increased VAT and evidence of muscle fat infiltration as assessed by CT, as well as by an increased prevalence of cardiovascular comorbidities, diabetes and gastroesophageal reflux. Our study is concordant and extends existing knowledge that COPD is associated with increased VAT accumulation compared to controls.7,25 by reporting, for
the first time, alterations in intra-abdominal muscle quantity and quality in COPD and investigating the contribution of such CT body composition and adipose tissue distribution abnormalities to COPD-related outcomes and comorbidities. Beyond a tendency for intra-abdominal muscle CSA to be reduced in COPD, the quality of the muscle, as assessed by its mean tissue density, was also modified, suggesting that COPD is specifically associated with increased muscle fat infiltration. Alterations in CT body composition parameters were particularly notable in patients with GOLD 4 COPD and were clinically relevant as suggested by their association with reduced walking capacity, quality of life and prevalence of cardiovascular diseases, diabetes and gastroesophageal reflux at study entry. Importantly, these associations remained after adjustment for age, sex, BMI, smoking status and FEV1.
paraspinal and abdominal wall muscles, evidence of quadriceps atrophy and fat infiltration has been reported in COPD. 26,27 These authors also reported
inverse relationships between indices of quadriceps fat infiltration and walking capacity in COPD. In this context, it is interesting that lower intra-abdominal muscle tissue density, a reflection of increased fatty content, was also a predictor of reduced walking performance, an association that remained significant after adjustment for muscle tissue CSA. It could be argued that the performance of the psoas and paraspinal muscles that are involved in walking