- Email: [email protected]
Alpha-1 antitrypsin deficiency: From the lung to the heart?
Atherosclerosis 270 (2018) 166e172
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Alpha-1 antitrypsin deficiency: From the lung to the heart? Ivan Curjuric a, b, *, 1, Medea Imboden a, b, Robert Bettschart c, Seraina Caviezel a, b, €ss f, Daiana Stolz g, Julia Dratva a, b, Marco Pons d, Thomas Rothe e, Arno Schmidt-Trucksa Gian Andri Thun a, b, h, i, Arnold von Eckardstein j, Florian Kronenberg k, Ilaria Ferrarotti l, Nicole M. Probst-Hensch a, b a
Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, Basel, Switzerland University of Basel, Basel, Switzerland Lungenpraxis Hirslanden Klinik Aarau, Aarau, Switzerland d Division of Pulmonary Medicine, Regional Hospital of Lugano, Lugano, Switzerland e €henklinik Davos, Davos Clavadel, Switzerland Zürcher Ho f Division of Sports and Exercise Medicine, Department of Sport, Exercise and Health, University of Basel, Basel, Switzerland g Clinic of Respiratory Medicine and Pulmonary Cell Research, University Hospital Basel, Basel, Switzerland h CNAG-CRG, Barcelona Institute of Science and Technology, Barcelona, Spain i Universitat Pompeu Fabra, Barcelona, Spain j Institute for Clinical Chemistry, University Hospital Zürich, Zürich, Switzerland k Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Medical University of Innsbruck, Innsbruck, Austria l Centre for Diagnosis of Inherited Alpha-1 Antitrypsin Deficiency, Department of Internal Medicine and Therapeutics, Pneumology Unit Fondazione IRCCS Policlinico San Matteo, University of Pavia, Pavia, Italy b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 September 2017 Received in revised form 16 January 2018 Accepted 24 January 2018 Available online 31 January 2018
Background and aims: Alpha-1 antitrypsin (A1AT) is the most abundant serine protease inhibitor in human blood and exerts important anti-inflammatory and immune-modulatory effects. In combination with smoking or other long-term noxious exposures such as occupational dust and fumes, genetic A1AT deficiency can cause chronic obstructive pulmonary disease, a condition with elevated cardiovascular risk. The effects of A1AT deficiency on cardiovascular risk have hardly been studied today. Methods: Using data from 2614 adults from the population-based SAPALDIA cohort, we tested associations of serum A1AT and SERPINA1 mutations with carotid intima-media thickness (CIMT, measured by B-mode ultrasonography) or self-reported arterial hypertension or cardiovascular disease in multiple regression models using a Mendelian Randomization like analysis design. Mutations Pi-S and Pi-Z were coded as ordinal genotype score (MM, MS, MZ/SS, SZ and ZZ), according to their progressive biological impact. Results: Serum A1AT concentration presented a u-shaped association with CIMT. At the lower end of the A1AT distribution, an analogous, linear association between SERPINA1 score and higher CIMT was observed, resulting in an estimated 1.2% (95%-confidence interval -0.1-2.5) increase in CIMT per unit (p ¼ 0.060). Genotype score was significantly associated with arterial hypertension with an odds ratio (OR) of 1.2 (1.0e1.5) per unit (p ¼ 0.028). The association with cardiovascular disease was not significant (OR 1.3 (0.9e1.9)). Conclusions: Our results support a possible causal relationship between genetic A1AT deficiency and increased cardiovascular risk, which needs to be better taken into account for the management of affected patients and first-degree relatives. © 2018 Elsevier B.V. All rights reserved.
Keywords: MESH): carotid intima-media thickness Cardiovascular diseases Alpha-1 antitrypsin deficiency Mendelian randomization analysis Cohort study
* Corresponding author. Swiss Tropical and Public Health Institute SwissTPH, Socinstrasse 57, P.O. Box, 4002 Basel, Switzerland. E-mail address: [email protected] (I. Curjuric). 1 Present address: Krebsregister Aargau, P.O. Box 4037, 5001 Aarau, Switzerland. Email: [email protected]. https://doi.org/10.1016/j.atherosclerosis.2018.01.042 0021-9150/© 2018 Elsevier B.V. All rights reserved.
1. Introduction Alpha-1 antitrypsin (A1AT) is an acute phase protein synthesized in the liver, and the most abundant serine protease inhibitor
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
167
by the Swiss Academy of Medical Sciences and respective cantonal ethics committees. All participants gave written informed consent.
in human blood. Its longest known biological function consists of the enzymatic inactivation of elastase, which is expressed on cell surfaces or released by neutrophil leucocytes during inflammation [1]. This function impedes excessive degradation of elastin and collagen fibers in extracellular matrix, and prevents the consecutive activation of matrix metalloproteinases and inflammatory cascades [2] that result in inflammation associated tissue remodeling. In recent years, new functions have been discovered that suggest a broader biological role, including anti-inflammatory and immunemodulatory effects [3], and interactions with serum lipoproteins [4e6]. A1AT is coded by gene SERPINA1 on chromosome 14. In Caucasian populations, the four alleles called Pi-M, Pi-S and Pi-Z are most frequently observed (Pi for “protease inhibitor” and letters given after the positions of the respective protein-bands in electrophoresis). Pi-M is the wildtype allele, while the S- and Z-alleles constitute functional single nucleotide polymorphisms (SNPs) with autosomal recessive inheritance patterns leading to reduced protein stability and impaired liver secretion, respectively [7,8]. The PiZZ genotype is rare (<0.1% in our study population), but highly deleterious, and leads to a 80-90% drop in serum A1AT concentration below the critical threshold of 0.49 g/L defining severe A1ATdeficiency [7,9]. Genotypes SZ, MZ, and SS lead to intermediate deficiency with levels between 0.49 and 1.0 g/L9. A1AT concentration is also influenced by environmental and endogenous biological factors. Likewise, it's highly correlated with C-reactive protein (CRP) level through common upregulation during acute phase reactions. Further, serum A1AT has been positively associated with female sex, age, systolic blood pressure, and tobacco smoking in our cohort, while inverse associations have been observed with body mass index (BMI) and alcohol intake [10]. When combined with smoking or other long-term noxious exposures such as occupational dust and fumes, severe genetic A1AT deficiency highly increases the risk of developing chronic obstructive pulmonary disease (COPD), a relationship constituting the classical paradigm of gene-environment interaction. COPD is a systemic aging-related disease consistently associated with increased cardiovascular risk [11]. The health-related effects of intermediate A1AT deficiency are less clear, but there is evidence showing increased risk of airway obstruction for MZ and SZ genotypes, particularly with concomitant oxidative stress exposure [12e15]. Given the broad biological activity of A1AT, its genetic deficiency could be an independent determinant of cardiovascular effects as observed in COPD. However, only a few studies have investigated this relationship to date, yielding conflicting results [16e18]. Using data from the Swiss population based SAPALDIA cohort, we therefore aimed to study the associations of A1AT serum level and its genetic determinants with CIMT by employing a Mendelian randomization design, to shed further light on the potential role of A1AT deficiency in atherosclerosis and cardiovascular disease.
In SAPALDIA3, spirometry was done with EasyOne handheld spirometers (EasyOne, ndd Medical Technologies, Zürich, Switzerland) following the protocol of the European Community Respiratory Health Survey [25]. Spirometry curves were subjected to ATS/ERS quality standards. SAPALDIA specific lung function equations were used for the calculation of lower limits of normal (representing the 5th percentile of the lung function distribution in healthy, non-smoking adults of our study population) [26]. Spirometric obstruction was defined as having a FEV1/FVC ratio (a marker of airway obstruction) below the lower limit of normal.
2. Materials and methods
2.4. Blood pressure, height, weight and BMI
2.1. Study design and population
Blood pressure was assessed with two measurements 3 min apart in SAPALDIA3 using automatic devices (705CP, Omron, Tokyo, Japan) and cuffs of appropriate size. Arithmetic means of diastolic and systolic values were used for analysis. Participants wore no shoes or heavy clothes when measuring weight and height. BMI was categorized as <25, 25 and < 30, and 30 kg/m2.
The SAPALDIA cohort provides detailed data on lifestyle, lung function, carotid intima-media thickness (CIMT) as marker of early atherosclerosis [19e21], and self-reported doctor's diagnoses of arterial hypertension and cardiovascular disease in participants aged 50 years or older. As depicted in Supplementary Fig. 1, we studied 2614 participants who underwent a CIMT scan at the third assessment (SAPALDIA3), and had A1AT concentration and SERPINA1 genotypes measured 8.4 years earlier (SAPALDIA2). Cohort details were published before [22]. Ethical approval was obtained
2.2. Carotid intima-media thickness In the current study, we used CIMT measurements as marker of early atherosclerotic processes in our general population sample [19]. CIMT was measured at SAPALDIA3 in participants aged 50 years. In a previous study of our cohort, high cardiovascular risk profiles were associated with higher CIMT measures [20], and in international studies, CIMT has repeatedly been evidenced as predictor of future cardiovascular disease [21]. The details of CIMT measurements in SAPALDIA were published previously [23]. Briefly, CIMT was assessed following a standardized protocol using carotid Bmode ultrasound scans on Fukuda Denshi UF-870 equipment. Fieldworkers were trained and supervised by two collaborating vascular labs (the Department of Vascular Medicine at the Academic Medical Center/Imagelab, University of Amsterdam and Erichem, The Netherlands, and the Division of Sports and Exercise Medicine, Department of Sport, Exercise and Health, University of Basel, Switzerland). Bilateral measurements of the distance between the lumen-intima to the media-adventitia layer of the far artery wall were performed for the duration of three heart cycles on the left and right common carotid artery (CCA), 1 cm proximal of the carotid bifurcation in longitudinal ear to ear and horizontal angle. Imaging data was analyzed by the B-mode image analysis program Dynamic Artery Analysis (DYARA). As sonographers were not trained for plaque assessment and following the Mannheim Consensus [24], images of insufficient quality or containing plaque areas were excluded, resulting in the exclusion of N ¼ 174 participants (~5% of the overall sample with available CIMT measurements). For each individual, average CIMT was calculated using the mean values of both CCA sides and scan angles. Duplicate scans were performed on different days within a period of 3 months in 165 randomly chosen participants to assess intra-fieldworker variability. Between-visit coefficient of variation was 3.98% (3.52e4.44) and the intra-class correlation coefficient was 0.89 (0.87e0.93) [20]. 2.3. Spirometry and spirometric obstruction
2.5. Questionnaire data An interview administered questionnaire was used to gather
168
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
data on age, smoking behavior, and cardiovascular disease history. Ever-smoking was defined as having smoked at least 20 packs of cigarettes or 360 g of tobacco during lifetime. Smoking intensity was calculated in packyears (cigarette packs per day times years of exposure). Regular alcohol consumption was defined as at least 1 standard drink per day. Sufficient physical activity was defined by 150 min of moderate or 60 min of vigorous physical activity per week [27]. For the definition of cardiovascular diseases participants were asked “Has a doctor ever told you that you have one of the following diseases?”, followed by a list of different diseases. In the present study, cardiovascular disease comprised self-reports of myocardial infarction or stroke, and arterial hypertension was defined either by a reported diagnosis or a positive history of antihypertensive medication. 2.6. Biomarkers Measurements of A1AT and CRP concentrations were done using latex-enhanced immunoturbidimetric assays (COBAS Integra analyzer, Roche Diagnostics, Indianapolis, Indiana, USA) on serum samples collected at SAPALDIA2. Interassay coefficients of variation were 3.6e4.6%, and lower detection thresholds 0.21 g/L for A1AT and 0.1 mg/L for CRP. Cholesterol and triglycerides were measured using enzymatic tests. SERPINA1 mutations Pi-S (SNP rs-number rs17580) and Pi-Z (rs28929474) were determined by PCR with fluorescently labelled Taq-Man probes on a LightCycler480 (Roche Diagnostics) [15]. Genotype distributions were in Hardy-Weinberg equilibrium. 2.7. Statistical analysis Due to their skewed distribution, mean CIMT measurements were transformed using natural logarithm and the normality of the transformed values was assessed graphically. Then, residuals were derived using sex stratified multiple linear regression models adjusting for age and fieldworker effect at SAPALDIA3. An ordinal genetic score was constructed for SERPINA1 genotypes MM, MS, MZ/SS, SZ and ZZ by assigning them to values 0-4, in accordance to their previously described progressive impact on A1AT serum concentration [9]. Following a Mendelian randomization analysis design, we studied the associations of A1AT serum levels and SERPINA1 genotypes or genotype score with CIMT residuals in multiple linear regression models. Mendelian randomization analysis is an instrumental variable approach based on the idea that the alleles of a genetic locus are passed on to the next generation by chance. This natural randomized experiment allows making causal inferences on biomarker effects, if the genetic mutation strongly determines the biomarker level and is associated with the outcome [28]. In the linear regression models, we controlled for important confounders including mean blood pressure, history of antihypertensive medication, presence of spirometric obstruction, ever smoking, packyears, categories of BMI, physical activity, and regular alcohol consumption at SAPALDIA3, and serum cholesterol, triglycerides and CRP at SAPALDIA2. Regarding CRP, we aimed for a comprehensive adjustment by including restricted cubic spline terms (specified with 5 knots according to Harrel et al. [29]). Also, higher order polynomial terms were included into the models for A1AT concentration to check for non-linear associations. As a quadratic association was observed between A1AT levels and CIMT, we used restricted cubic splines specified with 5 knots [29] to model this relationship graphically with sufficient degrees of freedom. The association of SERPINA1 genotypes with serum A1AT levels in SAPALDIA2 was published in detail before [9] and was thus not
further analyzed in the current study. To assess the clinical significance of the CIMT associations, logistic models were run assessing the risk of arterial hypertension and cardiovascular disease at SAPALDIA3. These models controlled for sex, age, smoking status and packyears, BMI categories, alcohol consumption, physical activity, and study area. As sensitivity analysis, the quadratic serum A1AT and genetic score models were retested with same model specifications in subgroups of never smokers, ever smokers and participants with spirometric obstruction. Under the assumption of an additive genetic model with genetic main effects explaining 0.5e1% in CIMT variability, power calculations yielded 95%e99% statistical power for a sample size of 2614 individuals and allele frequencies between 0.05 and 8%. Analyses were done using STATA IC version 12.1 (StataCorp, College Station, Texas, USA). Two-sided a-values of 0.05 were chosen as significance thresholds. Restricted cubic splines were calculated using STATA module rc_spline. 3. Results 3.1. Study population The characteristics of our study population are described in Table 1. 50.5% of our 2614 participants were female and 56.4% ever smokers. Median age was 63.7 years, and the median CIMT value 0.7 mm. The prevalence of arterial hypertension and cardiovascular disease was 38.9% and 4.4%, respectively. The frequencies of SERPINA1 genotypes MS, SS, MZ, SZ and ZZ were 7.8% (n ¼ 204), 0.2% (n ¼ 5), 2.3% (n ¼ 59), 0.3% (n ¼ 7) and 0.04% (n ¼ 1), respectively, while 89.4% (n ¼ 2338) of the participants carried the wildtype genotype MM. Ever smokers and participants with obstructive lung disease were less frequently female and had a higher prevalence of cardiovascular disease, but were otherwise comparable to never smokers. 3.2. Association of A1AT serum values and SERPINA1 genotypes with CIMT As shown in Table 2, serum A1AT concentration at SAPALDIA2 was significantly associated with pre-adjusted CIMT values at SAPALDIA3 in a non-linear, quadratic manner (p ¼ 0.038 and p ¼ 0.046 for first and second order terms, respectively). A graphical plot of the relationship using restricted cubic splines confirmed the quadratic association with slightly increased CIMT at the low and high ends of the serum A1AT distribution, and a minimum at about 1.3 g/L A1AT (Fig. 1). Regarding the SERPINA1 genetic score, we found a marginally significant linear association with pre-adjusted CIMT (p ¼ 0.060): per score unit, CIMT increased by an estimated 1.2% (95%-confidence interval (95%-CI) -0.1-2.5%). Associations were not statistically significant for single risk genotypes MS, SS/MZ, SZ, and ZZ, but effect estimates generally increased from 1.8 to 11.8%. 3.3. Association of A1AT serum values and SERPINA1 genotypes with arterial hypertension and cardiovascular disease The quadratic association between serum A1AT and arterial hypertension was marginally significant (p ¼ 0.044 for first and p ¼ 0.056 for second order term). A significant association between SERPINA1 score and arterial hypertension was observed (p ¼ 0.028), with each unit increase in score resulting in a 20% higher odds of hypertension (95%-CI 0-50%). Regarding cardiovascular disease, no significant associations were observed with serum A1AT or
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
169
Table 1 Descriptive characteristics of study population. Variable
Whole study population (n ¼ 2614)
Ever smokers (n ¼ 1474)
Female sex Age (years) Ever smoker Pack years Mean blood Pressure (mmHg) CIMT (mm) Doctors diagnosed Arterial hypertension Cardiovascular disease SERPINA1 genotypes MM MS SS MZ SZ ZZ SAPALDIA2 serum values Alpha-1 antitrypsine (g/L) Cholesterol (mmol/L) Triglycerides (mmol/L) C-reactive protein (mg/L)
1320 63.7 1474 0.5 106.8
(50.5) (57.1e69.8) (56.4) (0.0e19.5) (97.8e115.8)
641 63.3 1474 15.7 106.3
0.7
(0.6e0.8)
1016 116
a
Never smokers (n ¼ 1140)
Spirometric obstruction (n ¼ 354)
(43.5) (56.6e69.1) (100) (4.4e35.3) (97.3e115.8)
679 64.0 e e 107
(59.6) (57.8e70.7) e e (98.3e116.0)
197 64.8 228 4.7 106.8
(55.6) (58.3e71.5) (64.4) (0.0e33.5) (96.8e116.5)
0.7
(0.6e0.8)
0.7
(0.6e0.8)
0.7
(0.6e0.8)
(38.9) (4.4)
591 84
(40.1) (5.7)
425 32
(37.3) (2.8)
135 15
(38.1) (4.2)
2338 204 5 59 7 1
(89.4) (7.8) (0.2) (2.3) (0.3) (0.04)
1311 127 4 27 4 1
(88.9) (8.6) (0.3) (1.8) (0.3) (0.1)
1027 77 1 32 3 0
(90.1) (6.8) (0.1) (2.8) (0.3) (0.0)
305 38 2 6 2 1
(86.2) (10.7) (0.6) (1.7) (0.6) (0.3)
1.2 6.1 1.5 1.1
(1.1e1.4) (5.4e6.8) (1.1e2.3) (0.5e2.2)
1.3 6.1 1.6 1.1
(1.1e1.4) (5.4e6.8) (1.1e2.4) (0.6e2.4)
1.2 6.1 1.5 1.0
(1.1e1.3) (5.4e6.8) (1.0e2.1) (0.5e2.0)
1.3 6.1 1.5 1.1
(1.1e1.4) (5.3e6.9) (1.1e2.3) (0.6e2.6)
Estimates are given in absolute numbers (%) for categorical variables, and median (interquartile range) for continuous variables. CIMT: carotid intima-media thickness. a Participants with obstructive lung function overlap with the ever and never smoker groups.
SERPINA1 score. 3.4. Stratification by ever smoking and spirometric obstruction In the stratified sensitivity analysis (Table 3), the association between SERPINA1 score and arterial hypertension remained significant and was stronger in ever smokers (odds ratio (OR) 1.4, 95%CI 1.1e1.8) and participants with obstruction (OR 1.7, 95%-CI 1.0e2.8). For cardiovascular disease, ORs were increased, but not statistically significant in the same two strata. For serum A1AT, significant quadratic associations persisted with CIMT in neversmokers and arterial hypertension in ever-smokers (Supplementary Table 1). 4. Discussion Using data from the population-based SAPALDIA cohort from Switzerland, we report novel associations between serum values of A1AT, variants in SERPINA1 and CIMT. Participants at the outer ends of the observed A1AT serum distribution presented higher CIMT values than those in the middle portion. At the lower end of the A1AT concentration, the inverse association with CIMT was paralleled by an analogous association with known SERPINA1 variants causing progressive A1AT deficiency. In the same population, we also observed a significant, linear association between severity of SERPINA1 deficiency variants and arterial hypertension. Associations with cardiovascular disease (comprising myocardial infarction and stroke) were similar in size, but did not reach statistical significance. The associations with cardiovascular diagnoses only persisted in ever smokers and participants with spirometric obstruction. Our findings suggest that constitutive, genetic deficiency of A1AT might influence early atherosclerotic processes as captured by CIMT, and the study design following Mendelian randomization principles supports a causal relationship in the lower end of the A1AT serum distribution. A1AT possesses a broad spectrum of biological actions besides being a serine-protease. It exerts multiple anti-inflammatory effects including inhibition of TNF-alpha, IL-8
and IL-1beta expression, inhibition of neutrophil superoxide production, and modulation of leucocyte chemotaxis by immune complexes and cytokines [1,30,31]. The anti-inflammatory and immune-modulatory effects persist after the reaction with elastase, and thus appear to be independent of the protease inhibiting action [32]. Moreover, they could also affect allergic pathways [33], which have recently been found to contribute to atherosclerosis development [34]. A1AT has been observed to form complexes with high and low-density lipoproteins in the blood [5,6] and to interact with lipoprotein receptors [4]. Although these interactions are not fully understood, high concentrations of active A1AT appear to have a beneficial effect on the lipid profile [4]. Adding its initially described profound role in the maintenance of normal extracellular matrix, A1AT is thus involved in several key pathways underlying the pathogenesis of atherosclerosis [35e37]. A possible role of A1AT deficiency in atherosclerosis is also supported by the observed associations between SERPINA1 variants and doctor's diagnosed arterial hypertension and cardiovascular endpoints like myocardial infarction and stroke (both etiologically related to atherosclerosis), although the latter were not statistically significant. Regarding arterial hypertension, its definition by self-report or positive medication history rather represents a state of elevated cardiovascular risk than a clear and valid clinical diagnosis. However, the regulation of arterial blood pressure is complex, comprising salt homeostasis, vascular tonus regulation and endocrine signaling, and with its broad biological spectrum, deficiency in A1AT could affect either of these pathways besides altering vascular structures through mechanisms described before. The observed positive relationship of A1AT values with CIMT at the higher end of the serum concentration is difficult to explain, and we can only speculate about underlying mechanisms. A1AT is an acute phase protein whose expression is triggered and maintained by inflammatory pathways. A high A1AT serum concentration could thus be a proxy measure for underlying subclinical inflammation which deteriorates the cardiovascular profile. Likewise, high serum A1AT values have repeatedly been associated with increased arterial blood pressure in previous studies including our own [10,13]. Controlling for CRP in our analysis might not have
170
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
Table 2 Associations with CIMT, arterial hypertension and cardiovascular disease. A1AT/SERPINA1 Variable
Model
a
Mean CIMT value N
%-diff.
95%-CI
p
Serum A1AT values
Linear Linear terma Quadratica
2614 2614
1.0 19.2 8.6
(-4.1-2.2) (34.0 e 1.1) (0.1e17.8)
0.521 0.038 0.046
SERPINA1 genotypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MM ZZ vs. MM
2614 2614
1.2 1.8 0.7 8.5 11.8
(-0.1-2.5) (-0.3-3.9) (-2.8-4.4) (-2.5-20.8) (-16.1-49.0)
0.060 0.095 0.689 0.135 0.446
Estimated effects on arterial hypertension N OR
95%-CI
p
Serum A1AT values
Lineara Linear terma Quadratica
2614 2614
0.8 0.0 3.4
(0.5e1.3) (0.0e0.9) (1.0e11.7)
0.421 0.044 0.056
SERPINA1 gentoypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MM ZZ vs. MMc
2614 2613
1.2 1.1 1.7 4.0 e
(1.0e1.5) (0.8e1.5) (1.0e2.9) (0.8e19.1) e
0.028 0.687 0.072 0.085
Estimated effects on cardiovascular disease N OR
95%-CI
p
Serum A1AT values
Lineara Linear terma Quadratica
2614 2614
0.9 12.4 0.3
(0.3e2.7) (0.0e39763.0) (0.0e9.0)
0.808 0.541 0.513
SERPINA1 genotypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MMc ZZ vs. MMc
2614 2606
1.3 1.2 2.8 e e
(0.9e1.9) (0.6e2.4) (1.1e6.9) e e
0.143 0.576 0.030
A1AT: alpha-1 antitrypsin; CIMT: Carotid intima-media thickness; N: number of cases; OR: Odds ratio; p: p value of association; 95%-CI: 95%-confidence interval; %-diff.: percent difference; in bold: p < 0.05. CIMT was preadjusted for age and fieldworker effect in sex-stratified models. CIMT models adjusted for mean blood pressure, anti-hyptensive medication, spirometric obstruction, ever smoking, packyears, categories of body-mass index (BMI), regular alcohol consumption, physical activity and serum values of cholesterol, triglycerides and Creactive protein (CRP). Models on arterial hypertension and cardiovascular disease adjusted for sex, age, ever smoking, packyears, BMI categories, regular alcohol consumption, physical activity, CRP, and study area. a Estimated effect per g/L increase in serum A1AT. b Estimated effect per unit increase in genetic score. c Effect sizes could not be estimated for genotypes SZ and ZZ due to small sample size and missing outcome variability in this stratum.
Fig. 1. Association of serum A1AT and SERPINA1 genotypes with CIMT. The dark line shows the predicted mean value of CIMT over the observed A1AT serum distribution, while the light line represents the respective 95%-confidence interval (95%-CI). The dots and vertical lines represent the estimated mean CIMT and 95%-CI of the SERPINA1 genotypes. The histogram in the background of the graph represents the observed distribution of serum A1AT.
captured longterm inflammatory effects sufficiently, particularly as A1AT and CIMT were not measured at the same time point. The effects of A1AT on atherosclerosis and cardiovascular events were investigated in a few previous studies. In a nested casecontrol sample enriched with hospital cases from the population based Copenhagen City Heart study, A1AT deficiency genotype MZ was found to be associated with lower risk of ischemic cerebrovascular and heart disease [17]. In the cases with ischemic heart disease, deficiency genotypes were associated with lower blood pressure, a finding in apparent discrepancy to our observations. A recent genome-wide case control study observed associations between large artery stroke and a SERPINA1 variant located in a domain outside the proteolytic loop (likely interacting with serum lipoproteins), but not with the S- and Z-deficiency genotypes [38]. In contrast, functional SERPINA1 SNPs including the Z-deficiency variant were associated with metabolic blood lipid profiles and higher expression in atherosclerotic plaque tissue in another population based genome-wide study [39]. Hospital based studies of patients with overt cardiovascular disease have observed higher severity or progression of atherosclerosis with A1AT deficiency genotypes. Likewise, in a clinical trial on fibrate therapy after bypass-surgery, patients with genetically reduced A1AT expression presented progressive atherosclerotic changes in angiographic
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172 Table 3 Stratum specific effects of SERPINA1 genetic score on CIMT, arterial hypertension and cardiovascular disease. Group
Mean CIMT value n
%-diff.
95%-CI
p
All
2614
1.2
(-0.1-2.5)
0.060
Never smokers
1140
1.2
(-0.6-3.1)
0.204
Ever smokers
1474
1.2
(-0.6-3.0)
0.182
Obstructive spirometrya
354
1.4
(-1.9-4.8)
0.400
Group
Arterial hypertension n OR
95%-CI
p
All
2614
1.2
(1.0e1.5)
0.028
Never smokers
1140
1.1
(0.8e1.5)
0.577
Ever smokers
1474
1.4
(1.1e1.8)
0.017
Obstructive spirometrya
354
1.7
(1.0e2.8)
0.035
Group
Cardiovascular disease n OR
95%-CI
p
All
2614
1.3
(0.9e1.9)
0.143
Never smokers
1140
1.1
(0.5e2.6)
0.816
Ever smokers
1474
1.5
(1.0e2.2)
0.078
Obstructive spirometrya
327
2.4
(0.8e7.1)
0.105
Please refer to the legend in Table 2 for the explanation of abbreviations BMI, CIMT, n, OR, 95-CI and p. Effect estimates are given per increase in one unit of SERPINA1 additive genetic score. CIMT was preadjusted for age and fieldworker effect in sexstratified models. CIMT models adjusted for mean blood pressure, anti-hypertensive medication, spirometric obstruction (except in the stratum itself), packyears, BMI categories, regular alcohol consumption, physical activity and serum values of cholesterol, triglycerides and C-reactive protein. Models on arterial hypertension and cardiovascular disease adjusted for sex, age, packyears, BMI categories, regular alcohol consumption, physical activity, CRP and study area. a Participants with obstructive spirometry overlap with the other strata.
171
Our study also had limitations. CIMT measurements were only available from one time point. However, the temporal relationship with A1AT serum values was clearly defined and SERPINA1 mutations are not modifiable, thus reverse causation is not an issue. Further, for methodological reasons our analysis was limited to CIMT measurements without plaques. As a consequence, a part of our participants with severe atherosclerosis and clinical disease was excluded, and statistical power to assess cardiovascular disease was limited due to the low population frequency. This could explain missing statistical significance in the analysis of cardiovascular disease, particularly for A1AT serum values which are more prone to confounding and less stable over time than genetic determinants. Our results can thus be generalized to middle-aged Caucasian adults of average to good health. Our study did not apply multiple testing corrections because the focus of our study was Mendelian randomization analysis of CIMT which by design requires association-testing of serum A1AT and SERPINA1 genetic score. The additional linear A1AT and genotype-specific SERPINA1 models were included for a better characterization of the observed association patterns. This reasoning does not apply to stratified analyses, and arterial hypertension and cardiovascular disease models. The respective association results can thus only be interpreted with caution. In conclusion, our findings provide evidence from a populationbased observational study for an increased cardiovascular risk profile and potentially elevated cardiovascular disease risk in persons with SERPINA1 variants lowering A1AT concentration. If confirmed by future studies including participants with advanced atherosclerosis, these findings might be relevant for clinical management of moderate to severe A1AT deficiency. In index patients and their direct relatives, a systematic evaluation and active management of the cardiovascular risk profile might be indicated. Conflict of interest DS is advisory board member of Astra AG and Novartis AG, and is supported for lectures by Novartis AG and Roche AG. The other authors have no conflicts of interest to disclose. Financial support
examinations [18]. Several case studies on individuals with severe Pi-ZZ A1AT deficiency observed associations with resistant arterial hypertension, increased arterial stiffness but also structural alterations of the aorta and heart, probably related to the extracellular matrix functions of A1AT [40-43]. The published findings underline the complexity of the relationships between A1AT deficiency and atherosclerosis, which possibly also depend on further factors like pre-existing disease and oxidative stress levels. This is in line with our stratified results suggesting that the cardiovascular risk of A1AT deficiency might depend on concomitant tobacco smoke exposure or presence of pulmonary obstruction. Otherwise, the results of the earlier studies are difficult to compare to our observations, as we primarily assessed CIMT as early marker of atherosclerosis and our definition of clinical disease was based on self-report. Our study population was also in a rather good health status, as reflected by the low frequency of cardiovascular disease. Our study benefitted of a population-based design which permitted studying the effects of A1AT serum levels and SERPINA1 variants on early arteriosclerotic markers. Detailed data on lifestyle and relevant biomarkers enabled comprehensive adjustment for cardiovascular risk factors. Finally, having data on both, serum A1AT values and its genetic determinants allowed causal inferences in the lower end of the A1AT distribution by means of Mendelian randomization analysis.
The Swiss National Science Foundation (grants no 33CS30148470/1&2, 33CSCO-134276/1, 33CSCO-108796, 324730_135673, 3247BO-104283, 3247BO-104288, 3247BO-104284, 3247-065896, 3100-059302, 3200-052720, 3200-042532, 4026-028099, PMPDP3_129021/1, PMPDP3_141671/1), the Federal Office for the Environment, the Federal Office of Public Health, the Federal Office of Roads and Transport, the canton's government of Aargau, BaselStadt, Basel-Land, Geneva, Luzern, Ticino, Valais, and Zürich, the Swiss Lung League, the canton's Lung League of Basel Stadt/Basel Landschaft, Geneva, Ticino, Valais, Graubünden and Zurich, Stiftung €tten, SUVA, Freiwillige Akademische ehemals Bündner Heilsta Gesellschaft, UBS Wealth Foundation, Talecris Biotherapeutics GmbH, Abbott Diagnostics, European Commission 018996 (GABRIEL), Wellcome Trust WT 084703MA. None of the funding bodies had a role in the concept, design, data collection, analysis and interpretation of the study. Autor contributions Author IC wrote the manuscript and carried out the analysis; authors MI, GAT and IF contributed to the design and conception of the work and to data acquisition, authors SC and JD contributed to design of the work and interpretation of data; authors RB, MP, TR, DS contributed to data acquisition; authors AST, AE, FK contributed
172
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
to design of the work and interpretation of data; author NPH contributed to manuscript writing, design and conception of the work, data acquisition and interpretation of the data. All authors have critically revised the manuscript and have accepted the final version for submission. Acknowlegdements The authors acknowledge the contributions of the SAPALDIA team in setting up and running the cohort study assessments. The study could also not have been done without the help of the study participants, technical and administrative support and the medical teams and field workers at the local study sites. A detailed listing of the current SAPALDIA team, local fieldworkers, and technical and administrative staff is given in Supplemental materials. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2018.01.042. References [1] S.M. Janciauskiene, R. Bals, R. Koczulla, et al., The discovery of alpha1antitrypsin and its role in health and disease, Respir. Med. 105 (2011) 1129e1139. [2] T. Fulop, A. Khalil, A. Larbi, The role of elastin peptides in modulating the immune response in aging and age-related diseases, Pathol. Biol. 60 (2012) 28e33. [3] S. Baraldo, E. Balestro, E. Bazzan, et al., Alpha-1 antitrypsin deficiency today: new insights in the immunological pathways, Respiration 91 (2016) 380e385. [4] C.L. Bristow, R. Modarresi, M.A. Babayeva, et al., A feedback regulatory pathway between LDL and alpha-1 proteinase inhibitor in chronic inflammation and infection, Discov. Med. 16 (2013) 201e218. [5] S.M. Gordon, B. McKenzie, G. Kemeh, et al., Rosuvastatin alters the proteome of high density lipoproteins: generation of alpha-1-antitrypsin enriched particles with anti-inflammatory properties, Mol. Cell. Proteomics 14 (2015) 3247e3257. [6] K. Kotani, T. Yamada, N. Taniguchi, The association between adiponectin, HDLcholesterol and alpha1-antitrypsin-LDL in female subjects without metabolic syndrome, Lipids Health Dis. 9 (2010) 147. [7] L. Fregonese, J. Stolk, Hereditary alpha-1-antitrypsin deficiency and its clinical consequences, Orphanet J. Rare Dis. 3 (2008) 16. [8] J.K. Stoller, F.L. Lacbawan, L.S. Aboussouan, Alpha-1 antitrypsin deficiency, in: M.P. Adam, H.H. Ardinger, R.A. Pagon, et al. (Eds.), GeneReviews®, Seattle (WA), University of Washington, 2006 Oct 27, pp. 1993e2018. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1519/ (Updated 2017 Jan 19). [9] I. Ferrarotti, G.A. Thun, M. Zorzetto, et al., Serum levels and genotype distribution of alpha1-antitrypsin in the general population, Thorax 67 (2012) 669e674. [10] O. Senn, E.W. Russi, C. Schindler, et al., Circulating alpha1-antitrypsin in the general population: determinants and association with lung function, Respir. Res. 9 (2008) 35. [11] J.D. Maclay, W. MacNee, Cardiovascular disease in COPD: mechanisms, Chest 143 (2013) 798e807. [12] C.P. Hersh, M. Dahl, N.P. Ly, et al., Chronic obstructive pulmonary disease in alpha1-antitrypsin PI MZ heterozygotes: a meta-analysis, Thorax 59 (2004) 843e849. [13] M. Dahl, C.P. Hersh, N.P. Ly, et al., The protease inhibitor PI*S allele and COPD: a meta-analysis, Eur. Respir. J. 26 (2005) 67e76. [14] I.C. Sorheim, P. Bakke, A. Gulsvik, et al., Alpha(1)-Antitrypsin protease inhibitor MZ heterozygosity is associated with airflow obstruction in two large cohorts, Chest 138 (2010) 1125e1132. [15] G.A. Thun, I. Ferrarotti, M. Imboden, et al., SERPINA1 PiZ and PiS heterozygotes and lung function decline in the SAPALDIA cohort, PLoS One 7 (2012), e42728. [16] L.C. Battes, K.M. Akkerhuis, J.M. Cheng, et al., Circulating acute phase proteins in relation to extent and composition of coronary atherosclerosis and cardiovascular outcome: results from the ATHEROREMO-IVUS study, Int. J. Cardiol. 177 (2014) 847e853. [17] M. Dahl, A. Tybjaerg-Hansen, H. Sillesen, et al., Blood pressure, risk of ischemic cerebrovascular and ischemic heart disease, and longevity in alpha(1)antitrypsin deficiency: the Copenhagen City Heart Study, Circulation 107
(2003) 747e752. [18] P.J. Talmud, S. Martin, G. Steiner, et al., Progression of atherosclerosis is associated with variation in the alpha1-antitrypsin gene, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 644e649. [19] M. Bauer, S. Caviezel, A. Teynor, et al., Carotid intima-media thickness as a biomarker of subclinical atherosclerosis, Swiss Med. Wkly. 142 (2012) w13705. [20] J. Dratva, S. Caviezel, E. Schaffner, et al., Early detection of subjects at risk for vascular remodelling - results from the Swiss population-based study SAPALDIA, Swiss Med. Wkly. 144 (2014) w14052. [21] M. Carpenter, H. Sinclair, V. Kunadian, Carotid intima media thickness and its utility as a predictor of cardiovascular disease: a review of evidence, Cardiol. Rev. 24 (2016) 70e75. [22] U. Ackermann-Liebrich, B. Kuna-Dibbert, N.M. Probst-Hensch, et al., Follow-up of the Swiss cohort study on air pollution and lung diseases in adults (SAPALDIA 2) 1991-2003: methods and characterization of participants, Soz.Praevmed. 50 (2005) 245e263. [23] S. Caviezel, J. Dratva, E. Schaffner, et al., Variability and reproducibility of carotid structural and functional parameters assessed with transcutaneous ultrasound - results from the SAPALDIA Cohort Study, Atherosclerosis 231 (2013) 448e455. [24] P.J. Touboul, M.G. Hennerici, S. Meairs, et al., Mannheim carotid intima-media thickness and plaque consensus (2004-2006-2011), in: An Update on Behalf of the Advisory Board of the 3rd, 4th and 5th Watching the Risk Symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004. Brussels, Belgium, 2006, and Hamburg, Germany, 2011, Cerebrovasc Dis, 2012;34:290e296. [25] P.G. Burney, C. Luczynska, S. Chinn, et al., The European community respiratory health Survey, Eur. Respir. J. 7 (1994) 954e960. [26] P.O. Bridevaux, E. Dupuis-Lozeron, C. Schindler, et al., Spirometer replacement and serial lung function measurements in population studies: results from the SAPALDIA study, Am. J. Epidemiol. 181 (2015) 752e761. [27] S. Endes, E. Schaffner, S. Caviezel, et al., Long-term physical activity is associated with reduced arterial stiffness in older adults: longitudinal results of the SAPALDIA cohort study, Age Ageing 45 (2016) 110e115. [28] M. Bochud, V. Rousson, Usefulness of Mendelian randomization in observational epidemiology, Int. J. Environ. Res. Publ. Health 7 (2010) 711e728. [29] F.E. Harrell, Regression modelling strategies with applications to linear models, logistic regression, and survival analysis, in: Springer Series in Statistics, Springer, New York, 2001, pp. 23e24. [30] M.R. Ehlers, Immune-modulating effects of alpha-1 antitrypsin, Biol. Chem. 395 (2014) 1187e1193. [31] E.C. Lewis, Expanding the clinical indications for alpha(1)-antitrypsin therapy, Mol Med 18 (2012) 957e970. [32] D. Jonigk, M. Al-Omari, L. Maegel, et al., Anti-inflammatory and immunomodulatory properties of alpha1-antitrypsin without inhibition of elastase, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 15007e15012. [33] H. Koga, N. Miyahara, Y. Fuchimoto, et al., Inhibition of neutrophil elastase attenuates airway hyperresponsiveness and inflammation in a mouse model of secondary allergen challenge: neutrophil elastase inhibition attenuates allergic airway responses, Respir. Res. 14 (2013) 8. [34] D.P. Potaczek, Links between allergy and cardiovascular or hemostatic system, Int. J. Cardiol. 170 (2014) 278e285. [35] J. Hartman, W.H. Frishman, Inflammation and atherosclerosis: a review of the role of interleukin-6 in the development of atherosclerosis and the potential for targeted drug therapy, Cardiol. Rev. 22 (2014) 147e151. [36] P.A. Henriksen, J.M. Sallenave, Human neutrophil elastase: mediator and therapeutic target in atherosclerosis, Int. J. Biochem. Cell Biol. 40 (2008) 1095e1100. [37] M. Wang, L. Jiang, R.E. Monticone, et al., Proinflammation: the key to arterial aging, Trends Endocrinol Metab 25 (2014) 72e79. [38] R. Malik, T. Dau, M. Gonik, et al., Common coding variant in SERPINA1 increases the risk for large artery stroke, Proc. Natl. Acad. Sci. U. S. A. 114 (2017) 3613e3618. [39] M. Inouye, S. Ripatti, J. Kettunen, et al., Novel Loci for metabolic networks and multi-tissue expression studies reveal genes for atherosclerosis, PLoS Genet. 8 (2012), e1002907. [40] J.B. Van Leeuw, E. Berra, Y. Gu, et al., Alpha-1 antitrypsin deficiency: a novel cause of isolated systolic resistant hypertension? J. Hypertens. 34 (2016) 1659e1661. [41] J.M. Duckers, D.J. Shale, R.A. Stockley, et al., Cardiovascular and musculskeletal co-morbidities in patients with alpha 1 antitrypsin deficiency, Respir. Res. 11 (2010) 173. [42] E. Vizzardi, L. Corda, N. Pezzali, et al., Elastic properties of the ascending aorta in patients with alpha1-antitrypsin deficiency (Z homozygotes), Heart 98 (2012) 1354e1358. [43] E. Vizzardi, L. Corda, E. Sciatti, et al., Echocardiographic evaluation in subjects with alpha1-Antitrypsin deficiency, Eur. J. Clin. Invest. 45 (2015) 949e954.
Contents lists available at ScienceDirect
Atherosclerosis journal homepage: www.elsevier.com/locate/atherosclerosis
Alpha-1 antitrypsin deficiency: From the lung to the heart? Ivan Curjuric a, b, *, 1, Medea Imboden a, b, Robert Bettschart c, Seraina Caviezel a, b, €ss f, Daiana Stolz g, Julia Dratva a, b, Marco Pons d, Thomas Rothe e, Arno Schmidt-Trucksa Gian Andri Thun a, b, h, i, Arnold von Eckardstein j, Florian Kronenberg k, Ilaria Ferrarotti l, Nicole M. Probst-Hensch a, b a
Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, Basel, Switzerland University of Basel, Basel, Switzerland Lungenpraxis Hirslanden Klinik Aarau, Aarau, Switzerland d Division of Pulmonary Medicine, Regional Hospital of Lugano, Lugano, Switzerland e €henklinik Davos, Davos Clavadel, Switzerland Zürcher Ho f Division of Sports and Exercise Medicine, Department of Sport, Exercise and Health, University of Basel, Basel, Switzerland g Clinic of Respiratory Medicine and Pulmonary Cell Research, University Hospital Basel, Basel, Switzerland h CNAG-CRG, Barcelona Institute of Science and Technology, Barcelona, Spain i Universitat Pompeu Fabra, Barcelona, Spain j Institute for Clinical Chemistry, University Hospital Zürich, Zürich, Switzerland k Division of Genetic Epidemiology, Department of Medical Genetics, Molecular and Clinical Pharmacology, Medical University of Innsbruck, Innsbruck, Austria l Centre for Diagnosis of Inherited Alpha-1 Antitrypsin Deficiency, Department of Internal Medicine and Therapeutics, Pneumology Unit Fondazione IRCCS Policlinico San Matteo, University of Pavia, Pavia, Italy b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 September 2017 Received in revised form 16 January 2018 Accepted 24 January 2018 Available online 31 January 2018
Background and aims: Alpha-1 antitrypsin (A1AT) is the most abundant serine protease inhibitor in human blood and exerts important anti-inflammatory and immune-modulatory effects. In combination with smoking or other long-term noxious exposures such as occupational dust and fumes, genetic A1AT deficiency can cause chronic obstructive pulmonary disease, a condition with elevated cardiovascular risk. The effects of A1AT deficiency on cardiovascular risk have hardly been studied today. Methods: Using data from 2614 adults from the population-based SAPALDIA cohort, we tested associations of serum A1AT and SERPINA1 mutations with carotid intima-media thickness (CIMT, measured by B-mode ultrasonography) or self-reported arterial hypertension or cardiovascular disease in multiple regression models using a Mendelian Randomization like analysis design. Mutations Pi-S and Pi-Z were coded as ordinal genotype score (MM, MS, MZ/SS, SZ and ZZ), according to their progressive biological impact. Results: Serum A1AT concentration presented a u-shaped association with CIMT. At the lower end of the A1AT distribution, an analogous, linear association between SERPINA1 score and higher CIMT was observed, resulting in an estimated 1.2% (95%-confidence interval -0.1-2.5) increase in CIMT per unit (p ¼ 0.060). Genotype score was significantly associated with arterial hypertension with an odds ratio (OR) of 1.2 (1.0e1.5) per unit (p ¼ 0.028). The association with cardiovascular disease was not significant (OR 1.3 (0.9e1.9)). Conclusions: Our results support a possible causal relationship between genetic A1AT deficiency and increased cardiovascular risk, which needs to be better taken into account for the management of affected patients and first-degree relatives. © 2018 Elsevier B.V. All rights reserved.
Keywords: MESH): carotid intima-media thickness Cardiovascular diseases Alpha-1 antitrypsin deficiency Mendelian randomization analysis Cohort study
* Corresponding author. Swiss Tropical and Public Health Institute SwissTPH, Socinstrasse 57, P.O. Box, 4002 Basel, Switzerland. E-mail address: [email protected] (I. Curjuric). 1 Present address: Krebsregister Aargau, P.O. Box 4037, 5001 Aarau, Switzerland. Email: [email protected]. https://doi.org/10.1016/j.atherosclerosis.2018.01.042 0021-9150/© 2018 Elsevier B.V. All rights reserved.
1. Introduction Alpha-1 antitrypsin (A1AT) is an acute phase protein synthesized in the liver, and the most abundant serine protease inhibitor
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
167
by the Swiss Academy of Medical Sciences and respective cantonal ethics committees. All participants gave written informed consent.
in human blood. Its longest known biological function consists of the enzymatic inactivation of elastase, which is expressed on cell surfaces or released by neutrophil leucocytes during inflammation [1]. This function impedes excessive degradation of elastin and collagen fibers in extracellular matrix, and prevents the consecutive activation of matrix metalloproteinases and inflammatory cascades [2] that result in inflammation associated tissue remodeling. In recent years, new functions have been discovered that suggest a broader biological role, including anti-inflammatory and immunemodulatory effects [3], and interactions with serum lipoproteins [4e6]. A1AT is coded by gene SERPINA1 on chromosome 14. In Caucasian populations, the four alleles called Pi-M, Pi-S and Pi-Z are most frequently observed (Pi for “protease inhibitor” and letters given after the positions of the respective protein-bands in electrophoresis). Pi-M is the wildtype allele, while the S- and Z-alleles constitute functional single nucleotide polymorphisms (SNPs) with autosomal recessive inheritance patterns leading to reduced protein stability and impaired liver secretion, respectively [7,8]. The PiZZ genotype is rare (<0.1% in our study population), but highly deleterious, and leads to a 80-90% drop in serum A1AT concentration below the critical threshold of 0.49 g/L defining severe A1ATdeficiency [7,9]. Genotypes SZ, MZ, and SS lead to intermediate deficiency with levels between 0.49 and 1.0 g/L9. A1AT concentration is also influenced by environmental and endogenous biological factors. Likewise, it's highly correlated with C-reactive protein (CRP) level through common upregulation during acute phase reactions. Further, serum A1AT has been positively associated with female sex, age, systolic blood pressure, and tobacco smoking in our cohort, while inverse associations have been observed with body mass index (BMI) and alcohol intake [10]. When combined with smoking or other long-term noxious exposures such as occupational dust and fumes, severe genetic A1AT deficiency highly increases the risk of developing chronic obstructive pulmonary disease (COPD), a relationship constituting the classical paradigm of gene-environment interaction. COPD is a systemic aging-related disease consistently associated with increased cardiovascular risk [11]. The health-related effects of intermediate A1AT deficiency are less clear, but there is evidence showing increased risk of airway obstruction for MZ and SZ genotypes, particularly with concomitant oxidative stress exposure [12e15]. Given the broad biological activity of A1AT, its genetic deficiency could be an independent determinant of cardiovascular effects as observed in COPD. However, only a few studies have investigated this relationship to date, yielding conflicting results [16e18]. Using data from the Swiss population based SAPALDIA cohort, we therefore aimed to study the associations of A1AT serum level and its genetic determinants with CIMT by employing a Mendelian randomization design, to shed further light on the potential role of A1AT deficiency in atherosclerosis and cardiovascular disease.
In SAPALDIA3, spirometry was done with EasyOne handheld spirometers (EasyOne, ndd Medical Technologies, Zürich, Switzerland) following the protocol of the European Community Respiratory Health Survey [25]. Spirometry curves were subjected to ATS/ERS quality standards. SAPALDIA specific lung function equations were used for the calculation of lower limits of normal (representing the 5th percentile of the lung function distribution in healthy, non-smoking adults of our study population) [26]. Spirometric obstruction was defined as having a FEV1/FVC ratio (a marker of airway obstruction) below the lower limit of normal.
2. Materials and methods
2.4. Blood pressure, height, weight and BMI
2.1. Study design and population
Blood pressure was assessed with two measurements 3 min apart in SAPALDIA3 using automatic devices (705CP, Omron, Tokyo, Japan) and cuffs of appropriate size. Arithmetic means of diastolic and systolic values were used for analysis. Participants wore no shoes or heavy clothes when measuring weight and height. BMI was categorized as <25, 25 and < 30, and 30 kg/m2.
The SAPALDIA cohort provides detailed data on lifestyle, lung function, carotid intima-media thickness (CIMT) as marker of early atherosclerosis [19e21], and self-reported doctor's diagnoses of arterial hypertension and cardiovascular disease in participants aged 50 years or older. As depicted in Supplementary Fig. 1, we studied 2614 participants who underwent a CIMT scan at the third assessment (SAPALDIA3), and had A1AT concentration and SERPINA1 genotypes measured 8.4 years earlier (SAPALDIA2). Cohort details were published before [22]. Ethical approval was obtained
2.2. Carotid intima-media thickness In the current study, we used CIMT measurements as marker of early atherosclerotic processes in our general population sample [19]. CIMT was measured at SAPALDIA3 in participants aged 50 years. In a previous study of our cohort, high cardiovascular risk profiles were associated with higher CIMT measures [20], and in international studies, CIMT has repeatedly been evidenced as predictor of future cardiovascular disease [21]. The details of CIMT measurements in SAPALDIA were published previously [23]. Briefly, CIMT was assessed following a standardized protocol using carotid Bmode ultrasound scans on Fukuda Denshi UF-870 equipment. Fieldworkers were trained and supervised by two collaborating vascular labs (the Department of Vascular Medicine at the Academic Medical Center/Imagelab, University of Amsterdam and Erichem, The Netherlands, and the Division of Sports and Exercise Medicine, Department of Sport, Exercise and Health, University of Basel, Switzerland). Bilateral measurements of the distance between the lumen-intima to the media-adventitia layer of the far artery wall were performed for the duration of three heart cycles on the left and right common carotid artery (CCA), 1 cm proximal of the carotid bifurcation in longitudinal ear to ear and horizontal angle. Imaging data was analyzed by the B-mode image analysis program Dynamic Artery Analysis (DYARA). As sonographers were not trained for plaque assessment and following the Mannheim Consensus [24], images of insufficient quality or containing plaque areas were excluded, resulting in the exclusion of N ¼ 174 participants (~5% of the overall sample with available CIMT measurements). For each individual, average CIMT was calculated using the mean values of both CCA sides and scan angles. Duplicate scans were performed on different days within a period of 3 months in 165 randomly chosen participants to assess intra-fieldworker variability. Between-visit coefficient of variation was 3.98% (3.52e4.44) and the intra-class correlation coefficient was 0.89 (0.87e0.93) [20]. 2.3. Spirometry and spirometric obstruction
2.5. Questionnaire data An interview administered questionnaire was used to gather
168
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
data on age, smoking behavior, and cardiovascular disease history. Ever-smoking was defined as having smoked at least 20 packs of cigarettes or 360 g of tobacco during lifetime. Smoking intensity was calculated in packyears (cigarette packs per day times years of exposure). Regular alcohol consumption was defined as at least 1 standard drink per day. Sufficient physical activity was defined by 150 min of moderate or 60 min of vigorous physical activity per week [27]. For the definition of cardiovascular diseases participants were asked “Has a doctor ever told you that you have one of the following diseases?”, followed by a list of different diseases. In the present study, cardiovascular disease comprised self-reports of myocardial infarction or stroke, and arterial hypertension was defined either by a reported diagnosis or a positive history of antihypertensive medication. 2.6. Biomarkers Measurements of A1AT and CRP concentrations were done using latex-enhanced immunoturbidimetric assays (COBAS Integra analyzer, Roche Diagnostics, Indianapolis, Indiana, USA) on serum samples collected at SAPALDIA2. Interassay coefficients of variation were 3.6e4.6%, and lower detection thresholds 0.21 g/L for A1AT and 0.1 mg/L for CRP. Cholesterol and triglycerides were measured using enzymatic tests. SERPINA1 mutations Pi-S (SNP rs-number rs17580) and Pi-Z (rs28929474) were determined by PCR with fluorescently labelled Taq-Man probes on a LightCycler480 (Roche Diagnostics) [15]. Genotype distributions were in Hardy-Weinberg equilibrium. 2.7. Statistical analysis Due to their skewed distribution, mean CIMT measurements were transformed using natural logarithm and the normality of the transformed values was assessed graphically. Then, residuals were derived using sex stratified multiple linear regression models adjusting for age and fieldworker effect at SAPALDIA3. An ordinal genetic score was constructed for SERPINA1 genotypes MM, MS, MZ/SS, SZ and ZZ by assigning them to values 0-4, in accordance to their previously described progressive impact on A1AT serum concentration [9]. Following a Mendelian randomization analysis design, we studied the associations of A1AT serum levels and SERPINA1 genotypes or genotype score with CIMT residuals in multiple linear regression models. Mendelian randomization analysis is an instrumental variable approach based on the idea that the alleles of a genetic locus are passed on to the next generation by chance. This natural randomized experiment allows making causal inferences on biomarker effects, if the genetic mutation strongly determines the biomarker level and is associated with the outcome [28]. In the linear regression models, we controlled for important confounders including mean blood pressure, history of antihypertensive medication, presence of spirometric obstruction, ever smoking, packyears, categories of BMI, physical activity, and regular alcohol consumption at SAPALDIA3, and serum cholesterol, triglycerides and CRP at SAPALDIA2. Regarding CRP, we aimed for a comprehensive adjustment by including restricted cubic spline terms (specified with 5 knots according to Harrel et al. [29]). Also, higher order polynomial terms were included into the models for A1AT concentration to check for non-linear associations. As a quadratic association was observed between A1AT levels and CIMT, we used restricted cubic splines specified with 5 knots [29] to model this relationship graphically with sufficient degrees of freedom. The association of SERPINA1 genotypes with serum A1AT levels in SAPALDIA2 was published in detail before [9] and was thus not
further analyzed in the current study. To assess the clinical significance of the CIMT associations, logistic models were run assessing the risk of arterial hypertension and cardiovascular disease at SAPALDIA3. These models controlled for sex, age, smoking status and packyears, BMI categories, alcohol consumption, physical activity, and study area. As sensitivity analysis, the quadratic serum A1AT and genetic score models were retested with same model specifications in subgroups of never smokers, ever smokers and participants with spirometric obstruction. Under the assumption of an additive genetic model with genetic main effects explaining 0.5e1% in CIMT variability, power calculations yielded 95%e99% statistical power for a sample size of 2614 individuals and allele frequencies between 0.05 and 8%. Analyses were done using STATA IC version 12.1 (StataCorp, College Station, Texas, USA). Two-sided a-values of 0.05 were chosen as significance thresholds. Restricted cubic splines were calculated using STATA module rc_spline. 3. Results 3.1. Study population The characteristics of our study population are described in Table 1. 50.5% of our 2614 participants were female and 56.4% ever smokers. Median age was 63.7 years, and the median CIMT value 0.7 mm. The prevalence of arterial hypertension and cardiovascular disease was 38.9% and 4.4%, respectively. The frequencies of SERPINA1 genotypes MS, SS, MZ, SZ and ZZ were 7.8% (n ¼ 204), 0.2% (n ¼ 5), 2.3% (n ¼ 59), 0.3% (n ¼ 7) and 0.04% (n ¼ 1), respectively, while 89.4% (n ¼ 2338) of the participants carried the wildtype genotype MM. Ever smokers and participants with obstructive lung disease were less frequently female and had a higher prevalence of cardiovascular disease, but were otherwise comparable to never smokers. 3.2. Association of A1AT serum values and SERPINA1 genotypes with CIMT As shown in Table 2, serum A1AT concentration at SAPALDIA2 was significantly associated with pre-adjusted CIMT values at SAPALDIA3 in a non-linear, quadratic manner (p ¼ 0.038 and p ¼ 0.046 for first and second order terms, respectively). A graphical plot of the relationship using restricted cubic splines confirmed the quadratic association with slightly increased CIMT at the low and high ends of the serum A1AT distribution, and a minimum at about 1.3 g/L A1AT (Fig. 1). Regarding the SERPINA1 genetic score, we found a marginally significant linear association with pre-adjusted CIMT (p ¼ 0.060): per score unit, CIMT increased by an estimated 1.2% (95%-confidence interval (95%-CI) -0.1-2.5%). Associations were not statistically significant for single risk genotypes MS, SS/MZ, SZ, and ZZ, but effect estimates generally increased from 1.8 to 11.8%. 3.3. Association of A1AT serum values and SERPINA1 genotypes with arterial hypertension and cardiovascular disease The quadratic association between serum A1AT and arterial hypertension was marginally significant (p ¼ 0.044 for first and p ¼ 0.056 for second order term). A significant association between SERPINA1 score and arterial hypertension was observed (p ¼ 0.028), with each unit increase in score resulting in a 20% higher odds of hypertension (95%-CI 0-50%). Regarding cardiovascular disease, no significant associations were observed with serum A1AT or
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
169
Table 1 Descriptive characteristics of study population. Variable
Whole study population (n ¼ 2614)
Ever smokers (n ¼ 1474)
Female sex Age (years) Ever smoker Pack years Mean blood Pressure (mmHg) CIMT (mm) Doctors diagnosed Arterial hypertension Cardiovascular disease SERPINA1 genotypes MM MS SS MZ SZ ZZ SAPALDIA2 serum values Alpha-1 antitrypsine (g/L) Cholesterol (mmol/L) Triglycerides (mmol/L) C-reactive protein (mg/L)
1320 63.7 1474 0.5 106.8
(50.5) (57.1e69.8) (56.4) (0.0e19.5) (97.8e115.8)
641 63.3 1474 15.7 106.3
0.7
(0.6e0.8)
1016 116
a
Never smokers (n ¼ 1140)
Spirometric obstruction (n ¼ 354)
(43.5) (56.6e69.1) (100) (4.4e35.3) (97.3e115.8)
679 64.0 e e 107
(59.6) (57.8e70.7) e e (98.3e116.0)
197 64.8 228 4.7 106.8
(55.6) (58.3e71.5) (64.4) (0.0e33.5) (96.8e116.5)
0.7
(0.6e0.8)
0.7
(0.6e0.8)
0.7
(0.6e0.8)
(38.9) (4.4)
591 84
(40.1) (5.7)
425 32
(37.3) (2.8)
135 15
(38.1) (4.2)
2338 204 5 59 7 1
(89.4) (7.8) (0.2) (2.3) (0.3) (0.04)
1311 127 4 27 4 1
(88.9) (8.6) (0.3) (1.8) (0.3) (0.1)
1027 77 1 32 3 0
(90.1) (6.8) (0.1) (2.8) (0.3) (0.0)
305 38 2 6 2 1
(86.2) (10.7) (0.6) (1.7) (0.6) (0.3)
1.2 6.1 1.5 1.1
(1.1e1.4) (5.4e6.8) (1.1e2.3) (0.5e2.2)
1.3 6.1 1.6 1.1
(1.1e1.4) (5.4e6.8) (1.1e2.4) (0.6e2.4)
1.2 6.1 1.5 1.0
(1.1e1.3) (5.4e6.8) (1.0e2.1) (0.5e2.0)
1.3 6.1 1.5 1.1
(1.1e1.4) (5.3e6.9) (1.1e2.3) (0.6e2.6)
Estimates are given in absolute numbers (%) for categorical variables, and median (interquartile range) for continuous variables. CIMT: carotid intima-media thickness. a Participants with obstructive lung function overlap with the ever and never smoker groups.
SERPINA1 score. 3.4. Stratification by ever smoking and spirometric obstruction In the stratified sensitivity analysis (Table 3), the association between SERPINA1 score and arterial hypertension remained significant and was stronger in ever smokers (odds ratio (OR) 1.4, 95%CI 1.1e1.8) and participants with obstruction (OR 1.7, 95%-CI 1.0e2.8). For cardiovascular disease, ORs were increased, but not statistically significant in the same two strata. For serum A1AT, significant quadratic associations persisted with CIMT in neversmokers and arterial hypertension in ever-smokers (Supplementary Table 1). 4. Discussion Using data from the population-based SAPALDIA cohort from Switzerland, we report novel associations between serum values of A1AT, variants in SERPINA1 and CIMT. Participants at the outer ends of the observed A1AT serum distribution presented higher CIMT values than those in the middle portion. At the lower end of the A1AT concentration, the inverse association with CIMT was paralleled by an analogous association with known SERPINA1 variants causing progressive A1AT deficiency. In the same population, we also observed a significant, linear association between severity of SERPINA1 deficiency variants and arterial hypertension. Associations with cardiovascular disease (comprising myocardial infarction and stroke) were similar in size, but did not reach statistical significance. The associations with cardiovascular diagnoses only persisted in ever smokers and participants with spirometric obstruction. Our findings suggest that constitutive, genetic deficiency of A1AT might influence early atherosclerotic processes as captured by CIMT, and the study design following Mendelian randomization principles supports a causal relationship in the lower end of the A1AT serum distribution. A1AT possesses a broad spectrum of biological actions besides being a serine-protease. It exerts multiple anti-inflammatory effects including inhibition of TNF-alpha, IL-8
and IL-1beta expression, inhibition of neutrophil superoxide production, and modulation of leucocyte chemotaxis by immune complexes and cytokines [1,30,31]. The anti-inflammatory and immune-modulatory effects persist after the reaction with elastase, and thus appear to be independent of the protease inhibiting action [32]. Moreover, they could also affect allergic pathways [33], which have recently been found to contribute to atherosclerosis development [34]. A1AT has been observed to form complexes with high and low-density lipoproteins in the blood [5,6] and to interact with lipoprotein receptors [4]. Although these interactions are not fully understood, high concentrations of active A1AT appear to have a beneficial effect on the lipid profile [4]. Adding its initially described profound role in the maintenance of normal extracellular matrix, A1AT is thus involved in several key pathways underlying the pathogenesis of atherosclerosis [35e37]. A possible role of A1AT deficiency in atherosclerosis is also supported by the observed associations between SERPINA1 variants and doctor's diagnosed arterial hypertension and cardiovascular endpoints like myocardial infarction and stroke (both etiologically related to atherosclerosis), although the latter were not statistically significant. Regarding arterial hypertension, its definition by self-report or positive medication history rather represents a state of elevated cardiovascular risk than a clear and valid clinical diagnosis. However, the regulation of arterial blood pressure is complex, comprising salt homeostasis, vascular tonus regulation and endocrine signaling, and with its broad biological spectrum, deficiency in A1AT could affect either of these pathways besides altering vascular structures through mechanisms described before. The observed positive relationship of A1AT values with CIMT at the higher end of the serum concentration is difficult to explain, and we can only speculate about underlying mechanisms. A1AT is an acute phase protein whose expression is triggered and maintained by inflammatory pathways. A high A1AT serum concentration could thus be a proxy measure for underlying subclinical inflammation which deteriorates the cardiovascular profile. Likewise, high serum A1AT values have repeatedly been associated with increased arterial blood pressure in previous studies including our own [10,13]. Controlling for CRP in our analysis might not have
170
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
Table 2 Associations with CIMT, arterial hypertension and cardiovascular disease. A1AT/SERPINA1 Variable
Model
a
Mean CIMT value N
%-diff.
95%-CI
p
Serum A1AT values
Linear Linear terma Quadratica
2614 2614
1.0 19.2 8.6
(-4.1-2.2) (34.0 e 1.1) (0.1e17.8)
0.521 0.038 0.046
SERPINA1 genotypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MM ZZ vs. MM
2614 2614
1.2 1.8 0.7 8.5 11.8
(-0.1-2.5) (-0.3-3.9) (-2.8-4.4) (-2.5-20.8) (-16.1-49.0)
0.060 0.095 0.689 0.135 0.446
Estimated effects on arterial hypertension N OR
95%-CI
p
Serum A1AT values
Lineara Linear terma Quadratica
2614 2614
0.8 0.0 3.4
(0.5e1.3) (0.0e0.9) (1.0e11.7)
0.421 0.044 0.056
SERPINA1 gentoypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MM ZZ vs. MMc
2614 2613
1.2 1.1 1.7 4.0 e
(1.0e1.5) (0.8e1.5) (1.0e2.9) (0.8e19.1) e
0.028 0.687 0.072 0.085
Estimated effects on cardiovascular disease N OR
95%-CI
p
Serum A1AT values
Lineara Linear terma Quadratica
2614 2614
0.9 12.4 0.3
(0.3e2.7) (0.0e39763.0) (0.0e9.0)
0.808 0.541 0.513
SERPINA1 genotypes
Additive scoreb MS vs. MM SS/MZ vs. MM SZ vs. MMc ZZ vs. MMc
2614 2606
1.3 1.2 2.8 e e
(0.9e1.9) (0.6e2.4) (1.1e6.9) e e
0.143 0.576 0.030
A1AT: alpha-1 antitrypsin; CIMT: Carotid intima-media thickness; N: number of cases; OR: Odds ratio; p: p value of association; 95%-CI: 95%-confidence interval; %-diff.: percent difference; in bold: p < 0.05. CIMT was preadjusted for age and fieldworker effect in sex-stratified models. CIMT models adjusted for mean blood pressure, anti-hyptensive medication, spirometric obstruction, ever smoking, packyears, categories of body-mass index (BMI), regular alcohol consumption, physical activity and serum values of cholesterol, triglycerides and Creactive protein (CRP). Models on arterial hypertension and cardiovascular disease adjusted for sex, age, ever smoking, packyears, BMI categories, regular alcohol consumption, physical activity, CRP, and study area. a Estimated effect per g/L increase in serum A1AT. b Estimated effect per unit increase in genetic score. c Effect sizes could not be estimated for genotypes SZ and ZZ due to small sample size and missing outcome variability in this stratum.
Fig. 1. Association of serum A1AT and SERPINA1 genotypes with CIMT. The dark line shows the predicted mean value of CIMT over the observed A1AT serum distribution, while the light line represents the respective 95%-confidence interval (95%-CI). The dots and vertical lines represent the estimated mean CIMT and 95%-CI of the SERPINA1 genotypes. The histogram in the background of the graph represents the observed distribution of serum A1AT.
captured longterm inflammatory effects sufficiently, particularly as A1AT and CIMT were not measured at the same time point. The effects of A1AT on atherosclerosis and cardiovascular events were investigated in a few previous studies. In a nested casecontrol sample enriched with hospital cases from the population based Copenhagen City Heart study, A1AT deficiency genotype MZ was found to be associated with lower risk of ischemic cerebrovascular and heart disease [17]. In the cases with ischemic heart disease, deficiency genotypes were associated with lower blood pressure, a finding in apparent discrepancy to our observations. A recent genome-wide case control study observed associations between large artery stroke and a SERPINA1 variant located in a domain outside the proteolytic loop (likely interacting with serum lipoproteins), but not with the S- and Z-deficiency genotypes [38]. In contrast, functional SERPINA1 SNPs including the Z-deficiency variant were associated with metabolic blood lipid profiles and higher expression in atherosclerotic plaque tissue in another population based genome-wide study [39]. Hospital based studies of patients with overt cardiovascular disease have observed higher severity or progression of atherosclerosis with A1AT deficiency genotypes. Likewise, in a clinical trial on fibrate therapy after bypass-surgery, patients with genetically reduced A1AT expression presented progressive atherosclerotic changes in angiographic
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172 Table 3 Stratum specific effects of SERPINA1 genetic score on CIMT, arterial hypertension and cardiovascular disease. Group
Mean CIMT value n
%-diff.
95%-CI
p
All
2614
1.2
(-0.1-2.5)
0.060
Never smokers
1140
1.2
(-0.6-3.1)
0.204
Ever smokers
1474
1.2
(-0.6-3.0)
0.182
Obstructive spirometrya
354
1.4
(-1.9-4.8)
0.400
Group
Arterial hypertension n OR
95%-CI
p
All
2614
1.2
(1.0e1.5)
0.028
Never smokers
1140
1.1
(0.8e1.5)
0.577
Ever smokers
1474
1.4
(1.1e1.8)
0.017
Obstructive spirometrya
354
1.7
(1.0e2.8)
0.035
Group
Cardiovascular disease n OR
95%-CI
p
All
2614
1.3
(0.9e1.9)
0.143
Never smokers
1140
1.1
(0.5e2.6)
0.816
Ever smokers
1474
1.5
(1.0e2.2)
0.078
Obstructive spirometrya
327
2.4
(0.8e7.1)
0.105
Please refer to the legend in Table 2 for the explanation of abbreviations BMI, CIMT, n, OR, 95-CI and p. Effect estimates are given per increase in one unit of SERPINA1 additive genetic score. CIMT was preadjusted for age and fieldworker effect in sexstratified models. CIMT models adjusted for mean blood pressure, anti-hypertensive medication, spirometric obstruction (except in the stratum itself), packyears, BMI categories, regular alcohol consumption, physical activity and serum values of cholesterol, triglycerides and C-reactive protein. Models on arterial hypertension and cardiovascular disease adjusted for sex, age, packyears, BMI categories, regular alcohol consumption, physical activity, CRP and study area. a Participants with obstructive spirometry overlap with the other strata.
171
Our study also had limitations. CIMT measurements were only available from one time point. However, the temporal relationship with A1AT serum values was clearly defined and SERPINA1 mutations are not modifiable, thus reverse causation is not an issue. Further, for methodological reasons our analysis was limited to CIMT measurements without plaques. As a consequence, a part of our participants with severe atherosclerosis and clinical disease was excluded, and statistical power to assess cardiovascular disease was limited due to the low population frequency. This could explain missing statistical significance in the analysis of cardiovascular disease, particularly for A1AT serum values which are more prone to confounding and less stable over time than genetic determinants. Our results can thus be generalized to middle-aged Caucasian adults of average to good health. Our study did not apply multiple testing corrections because the focus of our study was Mendelian randomization analysis of CIMT which by design requires association-testing of serum A1AT and SERPINA1 genetic score. The additional linear A1AT and genotype-specific SERPINA1 models were included for a better characterization of the observed association patterns. This reasoning does not apply to stratified analyses, and arterial hypertension and cardiovascular disease models. The respective association results can thus only be interpreted with caution. In conclusion, our findings provide evidence from a populationbased observational study for an increased cardiovascular risk profile and potentially elevated cardiovascular disease risk in persons with SERPINA1 variants lowering A1AT concentration. If confirmed by future studies including participants with advanced atherosclerosis, these findings might be relevant for clinical management of moderate to severe A1AT deficiency. In index patients and their direct relatives, a systematic evaluation and active management of the cardiovascular risk profile might be indicated. Conflict of interest DS is advisory board member of Astra AG and Novartis AG, and is supported for lectures by Novartis AG and Roche AG. The other authors have no conflicts of interest to disclose. Financial support
examinations [18]. Several case studies on individuals with severe Pi-ZZ A1AT deficiency observed associations with resistant arterial hypertension, increased arterial stiffness but also structural alterations of the aorta and heart, probably related to the extracellular matrix functions of A1AT [40-43]. The published findings underline the complexity of the relationships between A1AT deficiency and atherosclerosis, which possibly also depend on further factors like pre-existing disease and oxidative stress levels. This is in line with our stratified results suggesting that the cardiovascular risk of A1AT deficiency might depend on concomitant tobacco smoke exposure or presence of pulmonary obstruction. Otherwise, the results of the earlier studies are difficult to compare to our observations, as we primarily assessed CIMT as early marker of atherosclerosis and our definition of clinical disease was based on self-report. Our study population was also in a rather good health status, as reflected by the low frequency of cardiovascular disease. Our study benefitted of a population-based design which permitted studying the effects of A1AT serum levels and SERPINA1 variants on early arteriosclerotic markers. Detailed data on lifestyle and relevant biomarkers enabled comprehensive adjustment for cardiovascular risk factors. Finally, having data on both, serum A1AT values and its genetic determinants allowed causal inferences in the lower end of the A1AT distribution by means of Mendelian randomization analysis.
The Swiss National Science Foundation (grants no 33CS30148470/1&2, 33CSCO-134276/1, 33CSCO-108796, 324730_135673, 3247BO-104283, 3247BO-104288, 3247BO-104284, 3247-065896, 3100-059302, 3200-052720, 3200-042532, 4026-028099, PMPDP3_129021/1, PMPDP3_141671/1), the Federal Office for the Environment, the Federal Office of Public Health, the Federal Office of Roads and Transport, the canton's government of Aargau, BaselStadt, Basel-Land, Geneva, Luzern, Ticino, Valais, and Zürich, the Swiss Lung League, the canton's Lung League of Basel Stadt/Basel Landschaft, Geneva, Ticino, Valais, Graubünden and Zurich, Stiftung €tten, SUVA, Freiwillige Akademische ehemals Bündner Heilsta Gesellschaft, UBS Wealth Foundation, Talecris Biotherapeutics GmbH, Abbott Diagnostics, European Commission 018996 (GABRIEL), Wellcome Trust WT 084703MA. None of the funding bodies had a role in the concept, design, data collection, analysis and interpretation of the study. Autor contributions Author IC wrote the manuscript and carried out the analysis; authors MI, GAT and IF contributed to the design and conception of the work and to data acquisition, authors SC and JD contributed to design of the work and interpretation of data; authors RB, MP, TR, DS contributed to data acquisition; authors AST, AE, FK contributed
172
I. Curjuric et al. / Atherosclerosis 270 (2018) 166e172
to design of the work and interpretation of data; author NPH contributed to manuscript writing, design and conception of the work, data acquisition and interpretation of the data. All authors have critically revised the manuscript and have accepted the final version for submission. Acknowlegdements The authors acknowledge the contributions of the SAPALDIA team in setting up and running the cohort study assessments. The study could also not have been done without the help of the study participants, technical and administrative support and the medical teams and field workers at the local study sites. A detailed listing of the current SAPALDIA team, local fieldworkers, and technical and administrative staff is given in Supplemental materials. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.atherosclerosis.2018.01.042. References [1] S.M. Janciauskiene, R. Bals, R. Koczulla, et al., The discovery of alpha1antitrypsin and its role in health and disease, Respir. Med. 105 (2011) 1129e1139. [2] T. Fulop, A. Khalil, A. Larbi, The role of elastin peptides in modulating the immune response in aging and age-related diseases, Pathol. Biol. 60 (2012) 28e33. [3] S. Baraldo, E. Balestro, E. Bazzan, et al., Alpha-1 antitrypsin deficiency today: new insights in the immunological pathways, Respiration 91 (2016) 380e385. [4] C.L. Bristow, R. Modarresi, M.A. Babayeva, et al., A feedback regulatory pathway between LDL and alpha-1 proteinase inhibitor in chronic inflammation and infection, Discov. Med. 16 (2013) 201e218. [5] S.M. Gordon, B. McKenzie, G. Kemeh, et al., Rosuvastatin alters the proteome of high density lipoproteins: generation of alpha-1-antitrypsin enriched particles with anti-inflammatory properties, Mol. Cell. Proteomics 14 (2015) 3247e3257. [6] K. Kotani, T. Yamada, N. Taniguchi, The association between adiponectin, HDLcholesterol and alpha1-antitrypsin-LDL in female subjects without metabolic syndrome, Lipids Health Dis. 9 (2010) 147. [7] L. Fregonese, J. Stolk, Hereditary alpha-1-antitrypsin deficiency and its clinical consequences, Orphanet J. Rare Dis. 3 (2008) 16. [8] J.K. Stoller, F.L. Lacbawan, L.S. Aboussouan, Alpha-1 antitrypsin deficiency, in: M.P. Adam, H.H. Ardinger, R.A. Pagon, et al. (Eds.), GeneReviews®, Seattle (WA), University of Washington, 2006 Oct 27, pp. 1993e2018. Available from: https://www.ncbi.nlm.nih.gov/books/NBK1519/ (Updated 2017 Jan 19). [9] I. Ferrarotti, G.A. Thun, M. Zorzetto, et al., Serum levels and genotype distribution of alpha1-antitrypsin in the general population, Thorax 67 (2012) 669e674. [10] O. Senn, E.W. Russi, C. Schindler, et al., Circulating alpha1-antitrypsin in the general population: determinants and association with lung function, Respir. Res. 9 (2008) 35. [11] J.D. Maclay, W. MacNee, Cardiovascular disease in COPD: mechanisms, Chest 143 (2013) 798e807. [12] C.P. Hersh, M. Dahl, N.P. Ly, et al., Chronic obstructive pulmonary disease in alpha1-antitrypsin PI MZ heterozygotes: a meta-analysis, Thorax 59 (2004) 843e849. [13] M. Dahl, C.P. Hersh, N.P. Ly, et al., The protease inhibitor PI*S allele and COPD: a meta-analysis, Eur. Respir. J. 26 (2005) 67e76. [14] I.C. Sorheim, P. Bakke, A. Gulsvik, et al., Alpha(1)-Antitrypsin protease inhibitor MZ heterozygosity is associated with airflow obstruction in two large cohorts, Chest 138 (2010) 1125e1132. [15] G.A. Thun, I. Ferrarotti, M. Imboden, et al., SERPINA1 PiZ and PiS heterozygotes and lung function decline in the SAPALDIA cohort, PLoS One 7 (2012), e42728. [16] L.C. Battes, K.M. Akkerhuis, J.M. Cheng, et al., Circulating acute phase proteins in relation to extent and composition of coronary atherosclerosis and cardiovascular outcome: results from the ATHEROREMO-IVUS study, Int. J. Cardiol. 177 (2014) 847e853. [17] M. Dahl, A. Tybjaerg-Hansen, H. Sillesen, et al., Blood pressure, risk of ischemic cerebrovascular and ischemic heart disease, and longevity in alpha(1)antitrypsin deficiency: the Copenhagen City Heart Study, Circulation 107
(2003) 747e752. [18] P.J. Talmud, S. Martin, G. Steiner, et al., Progression of atherosclerosis is associated with variation in the alpha1-antitrypsin gene, Arterioscler. Thromb. Vasc. Biol. 23 (2003) 644e649. [19] M. Bauer, S. Caviezel, A. Teynor, et al., Carotid intima-media thickness as a biomarker of subclinical atherosclerosis, Swiss Med. Wkly. 142 (2012) w13705. [20] J. Dratva, S. Caviezel, E. Schaffner, et al., Early detection of subjects at risk for vascular remodelling - results from the Swiss population-based study SAPALDIA, Swiss Med. Wkly. 144 (2014) w14052. [21] M. Carpenter, H. Sinclair, V. Kunadian, Carotid intima media thickness and its utility as a predictor of cardiovascular disease: a review of evidence, Cardiol. Rev. 24 (2016) 70e75. [22] U. Ackermann-Liebrich, B. Kuna-Dibbert, N.M. Probst-Hensch, et al., Follow-up of the Swiss cohort study on air pollution and lung diseases in adults (SAPALDIA 2) 1991-2003: methods and characterization of participants, Soz.Praevmed. 50 (2005) 245e263. [23] S. Caviezel, J. Dratva, E. Schaffner, et al., Variability and reproducibility of carotid structural and functional parameters assessed with transcutaneous ultrasound - results from the SAPALDIA Cohort Study, Atherosclerosis 231 (2013) 448e455. [24] P.J. Touboul, M.G. Hennerici, S. Meairs, et al., Mannheim carotid intima-media thickness and plaque consensus (2004-2006-2011), in: An Update on Behalf of the Advisory Board of the 3rd, 4th and 5th Watching the Risk Symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004. Brussels, Belgium, 2006, and Hamburg, Germany, 2011, Cerebrovasc Dis, 2012;34:290e296. [25] P.G. Burney, C. Luczynska, S. Chinn, et al., The European community respiratory health Survey, Eur. Respir. J. 7 (1994) 954e960. [26] P.O. Bridevaux, E. Dupuis-Lozeron, C. Schindler, et al., Spirometer replacement and serial lung function measurements in population studies: results from the SAPALDIA study, Am. J. Epidemiol. 181 (2015) 752e761. [27] S. Endes, E. Schaffner, S. Caviezel, et al., Long-term physical activity is associated with reduced arterial stiffness in older adults: longitudinal results of the SAPALDIA cohort study, Age Ageing 45 (2016) 110e115. [28] M. Bochud, V. Rousson, Usefulness of Mendelian randomization in observational epidemiology, Int. J. Environ. Res. Publ. Health 7 (2010) 711e728. [29] F.E. Harrell, Regression modelling strategies with applications to linear models, logistic regression, and survival analysis, in: Springer Series in Statistics, Springer, New York, 2001, pp. 23e24. [30] M.R. Ehlers, Immune-modulating effects of alpha-1 antitrypsin, Biol. Chem. 395 (2014) 1187e1193. [31] E.C. Lewis, Expanding the clinical indications for alpha(1)-antitrypsin therapy, Mol Med 18 (2012) 957e970. [32] D. Jonigk, M. Al-Omari, L. Maegel, et al., Anti-inflammatory and immunomodulatory properties of alpha1-antitrypsin without inhibition of elastase, Proc. Natl. Acad. Sci. U. S. A. 110 (2013) 15007e15012. [33] H. Koga, N. Miyahara, Y. Fuchimoto, et al., Inhibition of neutrophil elastase attenuates airway hyperresponsiveness and inflammation in a mouse model of secondary allergen challenge: neutrophil elastase inhibition attenuates allergic airway responses, Respir. Res. 14 (2013) 8. [34] D.P. Potaczek, Links between allergy and cardiovascular or hemostatic system, Int. J. Cardiol. 170 (2014) 278e285. [35] J. Hartman, W.H. Frishman, Inflammation and atherosclerosis: a review of the role of interleukin-6 in the development of atherosclerosis and the potential for targeted drug therapy, Cardiol. Rev. 22 (2014) 147e151. [36] P.A. Henriksen, J.M. Sallenave, Human neutrophil elastase: mediator and therapeutic target in atherosclerosis, Int. J. Biochem. Cell Biol. 40 (2008) 1095e1100. [37] M. Wang, L. Jiang, R.E. Monticone, et al., Proinflammation: the key to arterial aging, Trends Endocrinol Metab 25 (2014) 72e79. [38] R. Malik, T. Dau, M. Gonik, et al., Common coding variant in SERPINA1 increases the risk for large artery stroke, Proc. Natl. Acad. Sci. U. S. A. 114 (2017) 3613e3618. [39] M. Inouye, S. Ripatti, J. Kettunen, et al., Novel Loci for metabolic networks and multi-tissue expression studies reveal genes for atherosclerosis, PLoS Genet. 8 (2012), e1002907. [40] J.B. Van Leeuw, E. Berra, Y. Gu, et al., Alpha-1 antitrypsin deficiency: a novel cause of isolated systolic resistant hypertension? J. Hypertens. 34 (2016) 1659e1661. [41] J.M. Duckers, D.J. Shale, R.A. Stockley, et al., Cardiovascular and musculskeletal co-morbidities in patients with alpha 1 antitrypsin deficiency, Respir. Res. 11 (2010) 173. [42] E. Vizzardi, L. Corda, N. Pezzali, et al., Elastic properties of the ascending aorta in patients with alpha1-antitrypsin deficiency (Z homozygotes), Heart 98 (2012) 1354e1358. [43] E. Vizzardi, L. Corda, E. Sciatti, et al., Echocardiographic evaluation in subjects with alpha1-Antitrypsin deficiency, Eur. J. Clin. Invest. 45 (2015) 949e954.