The high coronary artery disease (CAD) prevalence in patients with abdominal aortic aneurysm (AAA) is well known. However, the inverse relation has been little explored. We present, based on a systematic review and meta-analysis of the published evidence, a critical appraisal of the issue of AAA prevalence and also AAA predictive risk factors in patients with CAD, comparing it with AAA prevalence in subjects without CAD. A total of 22 studies involving 13,388 patients with CAD met the inclusion criteria. Overall, AAA prevalence in patients with CAD was 8.4% (95% confidence interval [CI] 6.9 to 10.3), significantly higher than in subjects without CAD (odds ratio [OR] 2.42, 95% CI 2.08 to 2.81). Pooled analysis revealed that smoking, arterial hypertension, and concomitant carotid artery stenosis were significantly associated with AAA in patients with CAD (OR 1.72, 95% CI 1.14 to 2.61; OR 1.57, 95% CI 1.06 to 2.35; OR 2.14, 95% CI 1.20 to 3.79, respectively). In patients with CAD, AAA prevalence tended to be higher with concomitant peripheral artery disease (OR 2.66, 95% CI 0.82 to 8.61, p = 0.08). In conclusion, AAA prevalence was significantly higher in patients with CAD versus subjects without CAD.
Abdominal aortic aneurysm (AAA), a focal dilation of the abdominal aorta, is defined by current guidelines as an abdominal aortic diameter of ≥30 mm, in either anterior-posterior or transverse planes. AAA prevalence in general population ranges from 3.9% to 7.2% in men and 1.0% to 1.3% in women. Health organizations recommend a 1-time ultrasonographic AAA screening for men aged 65 to 75 years, with a smoking history, thereby reducing AAA-related mortality rates by 50%. AAA and atherosclerosis share common risk factors, such as male gender, age, and smoking. A high coronary artery disease (CAD) prevalence in patients with AAA is well known, affecting on short-term survival after AAA repair. Indeed, coronary investigation is often required before aortic surgery, finding concomitant 31% to 90% CAD prevalence. In contrast, the opposite relation, namely AAA prevalence in patients with CAD, was little explored. Accordingly, no specific recommendation on AAA screening in patients with advanced atherosclerosis is available. We present, based on a systematic review and meta-analysis of the published evidence, a critical appraisal of the issue of AAA prevalence, and also AAA predictive risk factors, in patients with CAD comparing it with AAA prevalence in subjects without CAD when data are available.
Methods
This analysis was planned in accordance with current MOOSE (Meta-analysis Of Observational Studies in Epidemiology) and PRISMA (Preferred reporting items for systematic review and meta-analysis) guidelines for performing comprehensive systematic reviews and meta-analysis. We searched in MEDLINE, EMBASE, and Cochrane databases up to April 15, 2015, using a combination of “abdominal aortic aneurysm” with the following terms: “coronary artery disease” OR “angina” OR “ischemic heart disease” OR “heart attack” OR “myocardial infarction” OR “coronary revascularization” with no language, study size, or publication year restriction. Reference lists from all retrieved original articles were also hand-searched to identify further relevant studies, and this process was repeated until no additional articles could be identified. Study inclusion criteria were all types of studies reporting AAA prevalence in Caucasian patients with CAD or data allowing calculating this prevalence. Included studies had to use a recognized definition of CAD and the use of imaging by ultrasound scan or computed tomography to establish the presence or absence of AAA. Studies were excluded if AAA was not identified by abdominal imaging. Only published studies in peer-review journals were finally selected.
Two independent investigators (AE and AD) reviewed all relevant articles and identified eligible studies. The following data were extracted from each publication using standardized data extraction forms and were transcribed into a predefined excel document: inclusion period, publication year, characteristics of the patients with CAD (number, age, men proportion, and geographic origin), CAD definition, CAD severity (need for coronary artery bypass graft surgery [CABG]), AAA definition, prevalence, and size and associated risk factors. When the included studies were primarily screening studies in the general population (not in preselected cohorts), we collected AAA prevalence in the non-CAD groups. Discrepancies between the 2 investigators were resolved by consensus with third-party adjudication.
For the prevalence of AAA and for each risk factor, studies’ heterogeneity was assessed with Cochrane’s chi-square test, with p <0.10 indicating evidence of heterogeneity. Data were pooled using fixed-effects model, or random-effects model when statistically significant heterogeneity was present, to calculate overall AAA prevalence, overall odds ratios (ORs), and 95% confidence intervals (CIs). For age, we generated pooled standardized differences and 95% CI. Publication bias was assessed graphically using a funnel plot. Statistical significance was assumed for p <0.05. Statistical analyses were performed using SAS software 9.2 (SAS Institute, Cary, North Carolina).
Results
A total of 22 observational studies, involving 13,388 Caucasian patients with CAD, were eligible, meeting predetermined inclusion criteria ( Table 1 ). A schematic representation of studies’ selection and exclusion is shown in Figure 1 . Three studies were excluded from the meta-analysis because they enrolled non-Caucasian patients (of Iranian, Chinese, and Korean origin). They reported very low AAA prevalence (2.09%, 1.76%, and 2.4%, respectively). Studied CAD populations ranged from 65 to 3,160 patients and AAA prevalence in patients with CAD oscillated between 2.22% and 15.28%. A fixed-effects model revealed studies’ heterogeneity (chi-square = 95.95, I 2 = 91.66, p-heterogeneity <0.0001). Overall AAA prevalence, determined by random-effects model, was 8.4% (95% CI 6.9 to 10.3). Interestingly, the overall prevalence was lower in studies published after 2000 compared with before 2000 (random-effects model 7.1%, 95% CI 5.5 to 9.3, vs 11.6%, 95% CI 9.6 to 13.8, p = 0.026). The prevalence of >50 mm AAA, reported in 9 studies (2,533 patients), ranged from 0% to 5.21%, and the overall prevalence was estimated at 0.69% (95% CI 0.11 to 1.2, p-heterogeneity = 0.024).
| First author (Year) Ref number | Inclusion period | Country | Age (Years) | Post-CABG | AP | Prior MI | Men | CAD patients | Non-CAD patients | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | AAA prevalence | >50 mm AAA prevalence | n | AAA prevalence | ||||||||
| Nevelsteen (1991) | – | Belgium | 60 | + | 0 | 0 | 100% | 65 | 10.76% | 4% | – | — |
| Smith (1993) | 1989 | UK | 65-75 | 0 | + | 0 | 100% | 325 | 14.76% | – | 2272 | 7.52% |
| Simoni (1995) | 1991-1994 | Italy | 65-75 | 0 | + | + | – | 279 | 9.31% | – | 1322 | 3.32% |
| Alcorn (1996) | 1989-1993 | USA | >65 | + | + | + | 43% | 1244 | 13.83% | – | – | – |
| Kanagasabay (1996) | – | UK | 65-80 | 0 | + | + | – | 418 | 9.56% | – | 4974 | 3.57% |
| Bergersen (1998) | – | USA | >50 | + | 0 | 0 | 65,6% | 192 | 13.02% | 5.21% | – | – |
| Lindholt (1998) | 1994 | Denmark | 65-73 | 0 | + | + | 100% | 377 | 7.69% | – | 2967 | 3.77% |
| Jaussi (1999) | 1995-1996 | Switzerland | – | 0 | + | 0 | – | 72 | 15.28% | – | – | – |
| Benzaquen (2001) | – | Canada | – | 0 | 0 | 0 | – | 99 | 8.08% | 1.01% | – | – |
| Bonamigo (2003) | 1987-1993 | Brazil | >54 | 0 | 0 | 0 | 100% | 501 | 6.78% | – | 1012 | 1.67% |
| Monney (2004) | – | Switzerland | 70 | + | 0 | 0 | 100% | 395 | 10.13% | 1.01% | – | – |
| Calderwood (2004) | – | UK | 65 | + | 0 | 0 | – | 118 | 15.25% | 4.23% | – | – |
| Madaric (2005) | – | Slovakia | 67 | 0 | 0 | 0 | 83% | 109 | 14.68% | 4.59% | – | – |
| Goessens (2006) | 1996-2004 | Netherlands | 59 | + | + | + | 83% | 1034 | 2.22% | 0.09% | – | – |
| Golledge (2007) | 1996-1999 | Australia | 65-83 | + | + | + | 100% | 3160 | 11.51% | – | 9043 | 5.65% |
| Dupont (2010) | 2002-2006 | France | 64 | + | 0 | 0 | 87% | 217 | 5.99% | 0.92% | – | – |
| Long (2010) | 2008-2009 | France | 64 | 0 | + | + | 77% | 304 | 6.58% | 0% | – | – |
| Johnsen (2010) | 1994-1995 | Norway | 25-84 | 0 | + | + | – | 779 | 7.95% | – | 5667 | 2.96% |
| Svenjo (2011) | 2006-2010 | Sweden | >65 | 0 | + | + | 100% | 1642 | 3.65% | – | 12,969 | 1.33% |
| Duncan (2012) | 2001-2004 | UK | 65-74 | 0 | + | + | 100% | 1132 | 7.77% | – | 7014 | 4.64% |
| Cueff (2012) | 2008-2009 | France | 65 | 0 | 0 | + | 77% | 193 | 4.66% | – | – | – |
| Durieux (2014) | 2009-2010 | Belgium | 64 | 0 | 0 | 0 | 69.9% | 733 | 5.18% | – | – | – |
Nine of 22 studies included subjects without CAD (non-CAD group). In these subjects without CAD, AAA prevalence ranged from 1.33% to 7.52%, with studies’ heterogeneity by fixed-effects model (chi-square = 1,800, I 2 = 99, p-heterogeneity <0.0001) and an overall AAA prevalence, by random-effects model, of 4.6% (95% CI 2.2 to 9.7). For these 9 studies, overall AAA prevalence was significantly higher in patients with CAD versus subjects without CAD (random-effects model: OR 2.42; 95% CI 2.08 to 2.81, p <0.0001; Figure 2 ). Interestingly, overall prevalence was still significantly higher in patients with CAD versus subjects without CAD for studies including >500 patients with CAD or for studies published after 2000 (random-effects model: OR 2.44; 95% CI 1.95 to 3.057, p <0.0001, for both). Regarding these 9 studies, there was no clear asymmetry of the funnel plot ( Figure 3 ).
Whereas 8 studies exclusively included male patients, the others recruited a female proportion varying from 13% to 57%. Among the 22 studies, the overall association of male gender with AAA in patients with CAD could be studied out of only 4 studies (798 patients) and was found to be not significant (OR 1.52, 95% CI 0.80 to 2.89, p = 0.19, p-heterogeneity = 0.59; Figure 4 ).
Of the 22 studies, the overall association of smoking with AAA in patients with CAD could be studied out of only 5 studies (1,502 patients) and was significant (OR 1.72, 95% CI 1.14 to 2.61, p = 0.010, p-heterogeneity = 0.24, Figure 4 ).
Six of the 22 publications evaluated the association between AAA and diabetes (mainly defined as fasting glucose ≥7 mmol/L and/or treatment). Overall analysis revealed no association between diabetes and the presence of AAA in patients with CAD (1,719 patients; OR 0.85, 95% CI 0.53 to 1.36, p = 0.51, p-heterogeneity = 0.08, Figure 5 ).
