Author + information
- Received July 19, 2003
- Revision received August 26, 2003
- Accepted October 6, 2003
- Published online March 17, 2004.
- Ignatios Ikonomidis, MD, FESC*,* (, )
- John Lekakis, MD, FESC*,
- Kimon Stamatelopoulos, MD*,
- Nikolaos Markomihelakis, MD†,
- Phaedon G Kaklamanis, MD† and
- Myron Mavrikakis, MD*
- ↵*Reprint requests and correspondence:
Dr. Ignatios Ikonomidis, University of Athens, Department of Clinical Therapeutics, Alexandra Hospital, Perikleous 19, Nea Chalkidona, Athens, 14343, Greece.
Objectives We investigated whether Adamantiades-Behcet's disease (ABD) is related to impaired aortic (Ao) elastic properties and left ventricular (LV) function.
Background Adamantiades-Behcet's disease is an inflammatory disorder characterized by vasculitis leading to vascular complications and, rarely, myocarditis.
Methods We studied 82 patients with ABD (age: 40 ± 12 years) and 24 normal control subjects by echocardiography. Abdominal Ao diameter (mm/m2) and Ao elastic indexes—namely, Ao strain (%), distensibility (cm2× dyn−1× 10−6), stiffness index, and pressure strain modulus (Ep) (cm2× dyn−1× 10−6)—were calculated from the echocardiographically derived thoracic Ao diameters (mm/m2), and the measurement of pulse pressure obtained by cuff sphygmomanometry. Isovolumic relaxation time (IVRT) (ms), deceleration time (DT) (ms), and flow propagation velocity (FPV) (cm/s) were measured by Doppler echocardiography to assess diastolic LV function. The duration of disease and presence of vascular complications were noted.
Results Patients versus control subjects had increased Ao diameters (p < 0.01), lower mean Ao strain and distensibility (4 vs. 9 and 1.4 vs. 3.4, respectively, p < 0.01), higher mean aortic stiffness index and Ep (15.6 vs. 6 and 1.17 vs. 0.44, respectively, p < 0.01), and impaired IVRT and FPV (p < 0.01). Aortic function indexes were related to the duration of disease (p < 0.01) and increased DT (p < 0.01). Deceleration time >190 ms predicted vascular complications with 80% sensitivity and 71% specificity (odds ratio 6.52 [confidence interval: 2.23 to 19.03]).
Conclusion Aortic elastic properties and diastolic LV function are impaired in patients with ABD and are interrelated. The link between diastolic LV dysfunction and vascular complications suggests the presence of a common pathophysiologic pathway and provides a possible marker of risk for vascular disease.
Adamantiades-Behcet's disease (ABD) is a chronic, relapsing, inflammatory disease characterized by recurrent oral and genital ulceration, skin and ocular lesions (1,2). Approximately one-third of these patients develop vascular manifestations, which include aneurysm formation, arterial or venous occlusive disease, and neurological disorders due to vasculitis of large and small vessels (3,4). Aortic distensibility is an important determinant of left ventricular (LV) function and coronary blood flow (5–7). Additionally, aortic distensibility has been related to increased cardiovascular mortality in different populations (8,9)and increased risk of aortic aneurysm rupture (10). However, the elastic properties of the aortic wall in patients with ABD have not been investigated yet.
Cardiomyopathy (1,11,12)linked to vascular and intracardiac thrombosis (11)has been frequently described in patients with ABD. Coronary vasculitis or focal myocardial inflammation (12)causing ischemia and fibrosis (11)may lead to significant diastolic dysfunction of the LV (13)even in the presence of normal systolic function. However, there are conflicting data about the presence of LV diastolic dysfunction in patients with ABD (14,15).
Two-dimensional guided mode echocardiography of the aortic root combined with simultaneous sphygmomanometric measurements of the arterial pressure at the brachial artery has been used to assess aortic function indexes (9,16). Additionally, two-dimensional and Doppler echocardiography has extensively been used for assessing systolic and diastolic LV function (13,17).
In the present study, we hypothesized that aortic elastic properties are impaired in patients with ABD, deteriorate as disease progresses in time, and are linked to significant LV diastolic dysfunction. We also investigated whether the presence of vascular complications in these patients is linked to a compromised cardiac function by means of two-dimensional and Doppler echocardiography.
A total of 82 consecutive patients (age: 40 ± 12 years; 54 males, 28 females) who fulfilled the International Study Group (ISG) criteria for diagnosis of Behcet disease (recurrent oral ulcerations and two of the following features: recurrent genital ulceration, eye lesions, skin lesions, positive pathergy skin test ) were prospectively studied as outpatients in the hospital's vascular laboratory. Exclusion criteria were: 1) severe valve disease; 2) atrial fibrillation; and 3) the need for use of calcium channel blockers and/or beta-blockers during the echocardiographic study.
All patients were in functional class I of the New York Heart Association classification and had a history of recurrent oral ulcerations; 28 had signs of active disease (>2 diagnostic criteria of ISG for Behcet's disease), and none had signs of active vascular disease during the study. All patients had an erythrocyte sedimentation rate <50 mm at the time of examination. All patients were receiving treatment (Table 1). Because aortic and LV diastolic function indexes are associated with age and presence of atherosclerotic risk factors, patients with ABD were compared with 24 control subjects with similar age, gender, cholesterol, and triglyceride plasma levels, as well as incidence of smoking, hypertension, and diabetes mellitus. Subjects of the control group had no history of cardiovascular disease, normal physical examination, and normal resting electrocardiogram.
The study was approved by the local ethics committee, and patients gave informed written consent. Each patient underwent standard transthoracic two-dimensional and Doppler echocardiography. The duration of disease was calculated from the time the patient fulfilled the ISG criteria to the time of examination. History of vascular complications (vascular aneurysms, stroke, arterial or venous occlusive disease) were also noted for each patient. Vascular complications were documented by conventional angiography, digital subtraction angiography, ultrasonography, and computed tomography of the involved organ.
Echocardiography and Doppler
All studies were performed using a Toshiba Sonolayer SSH 140A phased array ultrasound system (Toshiba Corp., Japan). A 2.5-MHz duplex transducer for combined cross-sectional imaging and Doppler echocardiography was used.
Echocardiography of the aorta (Ao)
The following parameters were measured from cross-sectional echocardiographic images of the Ao:
1. Thoracic Ao diameters (mm/m2) were measured 3 cm above the aortic valve by two-dimensional guided M-mode transthoracic echocardiography of the aortic root at left parasternal long-axis view. Aortic systolic diameter (AoS) was measured at the time of full opening of the aortic valve, and diastolic (AoD) diameter at the peak of the QRS complex at the simultaneous electrocardiogram recording. The following indexes of aortic function were calculated: aortic strain (AS) = 100 × (AoS − AoD)/AoD, aortic root distensibility = 2 × (AoS − AoD)/(AoD × PP) (cm2× dyn(−1)× 10(−6)), aortic stiffness index = ln(SBP/DBP) × (AoS − AoD)/AoD, and aortic pressure strain modulus (Ep) = PP/[(AoS − AoD)/AoD] (cm2× dyn(−1)× 10(−6)).
Pulse pressure (PP) was obtained simultaneously by cuff sphygmomanometry of the left brachial artery as systolic (SBP) minus diastolic blood pressure (DBP) (9,16).
Central aortic PP was also measured by radial artery tonometry and pulse wave analysis using the Spygmocor apparatus (Sphygmocor, ATCOR Medical, Sydney, Australia) (18). The above indexes of aortic function were then calculated by inserting central aortic PP in the place of the brachial-artery-derived PP in their formulas to correct for possible overestimation of the peripherally measured PP (19).
Central PP and the derived Ao distensibility, stiffness index, and Ep were related to the corresponding left brachial artery PP and the derived Ao distensibility, stiffness index, and Ep (r = 0.75, r = 0.89, r = 0.88, r = 0.89, p = 0.01, respectively).
2. Maximal Ao root diameter (mm/m2) was obtained from the parasternal short-axis view of the aortic valve and left atrium.
3. Maximal abdominal Ao diameter (mm/m2) was measured from the subcostal window.
Echocardiography of the LV
The following parameters were measured from cross-sectional echocardiographic images of the LV: 1) end-diastolic and end-systolic diameter (mm/m2); 2) fractional shortening (%); 3) LV outflow tract diameter (mm); 4) left atrial dimensions (mm/m2); and 5) interventricular septal (IVS), posterior wall (PW) thickness (mm/m2), and LV mass using the formula proposed by the Penn Convention (20).
Flow velocity profiles were obtained using color, pulsed, and continuous wave Doppler from apical projections, and measurements were made utilizing the software of the ultrasound equipment. Mitral diastolic flow was obtained after the pulsed Doppler sample volume was positioned perpendicular to the tips of the mitral valve leaflets and 7 to 11 cardiac cycles were recorded. The Doppler cursor was then moved toward the LV outflow position, and the sample volume was placed approximately 1 cm proximal to the aortic valve so that it would come in contact with the anterior mitral valve leaflet. Isovolumic relaxation time (IVRT) (ms) was measured as the interval between the end of the aortic click artefact and the onset of mitral inflow waveform.
The following indexes were measured from the mitral valve diastolic wave form: peak early (E) and atrial (A) flow velocities (cm/s), E/A ratio, deceleration time (DT) (ms) of the LV diastolic filling using standard methodology as previously described (13,21). Similarly, velocity time integral (cm) was measured by tracing the perimetry of the LV outflow tract wave form, and the corresponding stroke volume was calculated utilizing the equipment's software. Heart rate (beats/min) was measured from simultaneous electrocardiogram recordings.
Color M-mode echocardiography was performed in the apical four-chamber view with the cursor aligned parallel with the LV inflow. The longest column of flow from the mitral annulus to the apex of the LV was recorded. The M-mode cursor was placed at the center of the LV inflow. Flow propagation velocity (FPV) was measured as the slope of the first color aliasing velocity (45 cm/s) from mitral annulus in early diastole to 40 mm distally in the LV cavity (22). The ratio E/FPV was calculated, and patients were divided into those with <1.7 and those with ≥1.7 corresponding to significantly elevated LV diastolic pressures (pulmonary capillary wedge pressure >12 mm Hg) (22).
All variables are expressed as mean ± SD. Statistical analysis used SPSS (10.0 version) software. Unpaired comparisons between patient subgroups or between patients and control subjects were performed using the standard chi-square for categorical and the unpaired ttest for continuous data. Linear relations were checked using simple linear regression analysis. When a dependent variable was related to more than one independent variable, backward stepwise regression analysis model was used to exclude interrelation between the examined variables. Interactions between various independent variables had been excluded using simple linear regression analysis before these variables were used in the stepwise regression analysis. Atherosclerotic risk factors (smoking status, systolic and diastolic blood pressure, cholesterol, triglyceride, and glucose plasma levels) were forced in the regression analysis model, and each model was weighted for heart rate or age to take into account any possible relation with the examined variables. Differences between patients with vascular complications, patients without, and control subjects were assessed using factorial analysis of variance. Logistic regression analysis was performed to examine the relation of the incidence of vascular complications with the measured echocardiographic indexes. Receiver operator curve analysis was used afterwards to determine the values of the echocardiographic indexes, which would best predict the incidence of vascular complications. The odds ratio of the examined index was then calculated using crosstabs analysis. A p < 0.05 was considered to be the level of statistical significance.
Assessment of reproducibility
In 30 randomly selected patients, measurements were repeated in a blinded manner by a second observer. Differences between the two sets of measurements were examined using the paired ttest and were found not statistically significant.
Patients versus control subjects
Patients and control subjects had similar age, gender, and atherosclerotic risk factor distribution (Table 1). The mean duration of the disease was 10 ± 7.7 (range, 1 to 38) years.
Aortic wall characteristics
Maximal thoracic and abdominal aortic diameters were increased in patients compared with control subjects (p < 0.05). Aortic strain and distensibility were approximately twofold lower and Ep and aortic stiffness index approximately threefold higher in patients compared with control subjects (p < 0.01) (Table 2).
All patients had normal fraction shortening. Isovolumic relaxation time was prolonged, and FPV was reduced in all patients compared with control subjects (p < 0.01). Deceleration time was similar between patients and control subjects (p = NS) (Table 3). However, DT values were significantly higher in patients with vascular complications compared with control subjects (p < 0.01) (Table 4). Moreover, the ratio E/FPV was significantly higher in patients than in control subjects. A total of 41 patients (50%) had an E/FPV ≥1.7, indicating elevated LV diastolic pressures (p < 0.05) (Table 3).
There was no association between the measured echocardiographic indexes and the presence of any of the diagnostic criteria for Bechet's disease. There were also no differences in the indexes of aortic elastic properties or of LV diastolic function between patients treated with corticosteroids or other anti-inflammatory regimens (Table 1) and those receiving no anti-inflammatory medication at the time of examination.
Relations with duration of disease
The duration of disease was related to low Ao strain and distensibility as well as increased aortic stiffness index and Ep (r = −0.30, r = −0.38, and r = 0.26, r = 0.22, respectively, weighted for age; p < 0.01).
Interrelation of echocardiographic indexes
Low Ao distensibility, increased Ep, and aortic stiffness index were associated with prolonged DT (r = −0.26, r = 0.24, and r = 0.22, p < 0.05, respectively). As expected (13,20,21), LV mass, IVS, and PW thickness were related to increased brachial PP (r = 0.26, r = 0.23, and r = 0.26, p < 0.05, respectively), maximal thoracic Ao diameters (r = 0.26, r = 0.23, and r = 0.26, p < 0.05, respectively), prolonged IVRT (r = 0.25, r = 0.31, and r = 0.25, p < 0.05, respectively), and DT (r = 0.22, r = 0.24, and r = 0.23, p < 0.05, respectively). Age was also related to Doppler diastolic indexes (p < 0.05) (21).
Characteristics of patients with vascular complications
Patients with vascular complications (n = 29) were older (43 ± 12 vs. 37 ± 12 years, p < 0.05), had higher LV mass (g/m2), IVS, PW thickness, IVRT, and DT than patients without (Table 4). Logistic regression analysis showed that, of the above indexes, only DT was significantly associated with the incidence of vascular complications independently of age and atherosclerotic risk factors (exp [B]: 1.02 [confidence interval: 1.004 to 1.030], p = 0.01). A DT ≥190 ms predicted the incidence of vascular complications with 80% sensitivity and 71% specificity (receiver operator curve area: 75% [confidence interval: 60% to 87%]). The odds ratio of a patient with DT ≥190 ms to present vascular complications was calculated to be 6.52 (confidence interval: 2.23 to 19.03) (p < 0.01). Interestingly, patients with vascular complications had a lower aortic distensibility than those without (Table 4), though the difference was of borderline statistical significance (p = 0.07).
To our knowledge, this is the first prospective study demonstrating that aortic elastic properties and diameters are severely impaired in patients with ABD compared with control subjects with similar atherosclerotic risk factor profiles. We have also shown that the extent of the changes in aortic elastic properties is associated with the duration of the disease and with LV diastolic dysfunction. Furthermore, we demonstrated that a DT ≥190 ms indicates patients with vascular complications with an odds ratio of 6.52 (confidence interval: 2.23 to 19.03) (p < 0.01) and, thus, may provide a useful marker of increased risk for vascular disease.
Aortic elastic properties
Approximately one-third of patients with ABD develop vascular manifestations, which include aneurysm formation, arterial or venous occlusive disease, and stroke due to vasculitis of all size vessels (1,3,4). In this study, we have shown that patients with ABD develop abnormal aortic elastic properties in addition to dilation of thoracic and abdominal Ao compared with control subjects with similar age and atherosclerotic risk factor distribution. Decreased aortic distensibility has been related to increased risk of aneurysm rupture (10). Thus, our findings provide a possible mechanism of aneurysm formation and rupture in patients with ABD (3,4).
Vasculitis of the vasa vasorum (1,23)causing fibrosis and degradation of elastin within the aortic wall may explain the presence of significant functional abnormalities of the ascending Ao and the dilation of the proximal and abdominal Ao observed in our study.
Increased circulating levels of tumor necrosis factor-alpha, soluble interleukin 2 receptor, interleukin-1b, interleukin-6, and interleukin-8 have been reported in patients with ABD (24). Cytokines promote production of metalloproteinases (25,26)that degrade collagen and elastin content of the aortic intima (27)and may, thus, contribute to impaired function and dilation of the Ao (28). In the present study, we have also shown that the longer the duration of the disease the greater the impairment of aortic wall elastic properties. During the course of the disease, the constant exposure of the aortic wall to increased circulating cytokine levels (24)causes significant endothelial dysfunction (29). Reduced nitric oxide production by an injured aortic endothelium may sustain cytokine-induced matrix metalloproteinase release within the aortic layers over time (30). This mechanism may partly explain the relation of abnormal aortic function indexes with the duration of the disease observed in our study.
Patients treated with steroids had aortic elastic properties similar to those receiving other anti-inflammatory treatment or no treatment at the time of examination. It is possible that treatment with steroids may have reduced aortic wall inflammation and, thus, eliminated differences in aortic elastic properties between those patients with more severe disease in need of steroids compared with those with less severe disease. However, this study was not designed to compare aortic wall characteristics before and after treatment with steroids.
In the present study, we found that Doppler indexes of LV diastolic function were impaired in patients with ABD compared with control subjects with similar atherosclerotic risk factors including hypertension. Furthermore, 50% of the patients had an E/FPV ≥1.7 suggesting significantly elevated LV diastolic pressures (22). Our findings are in agreement with those of Gemici et al. (14)who also found impaired LV diastolic function associated with repolarization abnormalities in patients with ABD.
In our study, indexes of aortic function were associated with a prolonged DT suggesting a link between increased aortic stiffness and impaired myocardial relaxation. Experimental studies have shown that increased aortic stiffness augments LV afterload and impairs myocardial perfusion (6,7)causing significant LV diastolic dysfunction (13,21). Clinical studies have also linked increased aortic stiffness to overt diastolic heart failure (5). Thus, abnormal aortic elastic properties may represent one of the mechanisms of LV diastolic dysfunction in our study population. Alternatively, coronary vasculitis (1,3,12)or cytokine-induced myocardial fibrosis (11,24,26,31)may have contributed to LV diastolic dysfunction (13,21)in these patients.
Diastolic indexes and vascular complications
In our study, patients with vascular complications were older and had higher LV mass, wall thickness, and prolonged IVRT and DT compared with patients without. Studies have shown that older age, increased LV mass, increased wall thickness, and prolonged indexes of LV diastolic function are interrelated (13,21). However, in a logistic regression model including age, of all measured echocardiographic indexes and atherogenic risk factors, only DT was associated with presence of vascular complications. Moreover, a DT ≥190 ms predicted the incidence of vascular complications with 80% sensitivity and 71% specificity (odds ratio, 6.52 [confidence interval: 2.23 to 19.03]) and may, thus, provide a useful marker of risk for vascular disease.
The relationship between DT and vascular complications suggests the presence of a common pathophysiologic pathway leading to thrombosis and myocardial and aortic wall dysfunction. Interleukin-2, interleukin-6, interleukin-1b, and tumor necrosis factor-alpha have been reported elevated in patients with ABD (24). Interleukin-2 has been related with reduced peak diastolic filling rate at radionuclide ventriculography in patients with cancer (31)as well as with vascular thrombosis (32,33). Interleukin-6 has been found to initiate coagulation (34)and to cause myocardial dysfunction in experimental studies (35). Finally, the procoagulant cytokines interleukin-1b and tumor necrosis factor-alpha (34)may induce metalloproteinase expression in the aortic wall (25)and myocardium (26)and, thus, promote fibrosis (26–28). Through these actions, cytokines may link vascular thrombosis, stroke, or formation of arterial aneurysms (3,4)to impairment of LV diastolic function. Therefore, circulation of increased levels of inflammatory cytokines may explain the link between DT and incidence of vascular complications observed in our study.
Pulse pressure was measured by cuff sphygmomanometry of the brachial artery and not invasively in the ascending Ao. However, several reports have demonstrated the excellent correlation of the non-invasively calculated aortic function indexes with indexes derived from aortography (9,16). Additionally, calculation of the same indexes using the central PP as assessed by radial artery tonometry and pulse wave analysis instead of the brachial artery PP (19)confirmed the results of our study. Another potential limitation is that we have not excluded subclinical coronary artery involvement. However, given the patients' young age, the lack of clinical findings, and the low atherosclerotic risk factor profile of our study population, the likelihood for coronary artery disease is relatively low, though coronary involvement due to the Behcet syndrome per se cannot be excluded.
In this study, we have shown that patients with ABD present abnormal aortic elastic properties and diameters in addition to impaired LV diastolic function compared with control subjects with similar atherosclerotic risk factors. In these patients aortic stiffness was associated with the duration of the disease and with LV diastolic dysfunction. Finally, a DT ≥190 ms was a marker of the incidence of vascular complications, suggesting the presence of a common pathophysiologic pathway leading to thrombosis and aortic and myocardial wall dysfunction.
Improvement of aortic elastic properties by medical treatment such as calcium channel blockers (36), angiotensin-converting enzyme inhibitors, or angiotensin type 1 receptor blockers (37)may be an important therapeutic goal in patients with ABD in order to decrease the risk of aneurysm formation (3)and rupture (4,10), to improve LV diastolic function (5), and, consequently, improve prognosis (4,8,9)in these patients. Additionally, measurement of DT by Doppler echocardiography may provide a simple, reproducible, and non-invasive technique to identify patients at risk for vascular disease.
- Adamantiades-Behcet's disease
- aortic diastolic diameter
- aortic systolic diameter
- deceleration time
- pressure strain modulus
- flow propagation velocity
- International Study Group
- isovolumic relaxation time
- interventricular septum
- left ventricle/left ventricular
- pulse pressure
- posterior wall
- Received July 19, 2003.
- Revision received August 26, 2003.
- Accepted October 6, 2003.
- American College of Cardiology Foundation
- Hundley W.G,
- Kitzman D.W,
- Morgan T.M,
- et al.
- Watanabe H,
- Ohtsuka S,
- Kakihana M,
- Sugishita Y
- Kingwell A.B,
- Waddell T.K,
- Medley T.L,
- et al.
- Stefanadis C,
- Dernellis J,
- Tsiamis E,
- et al.
- Le Thi Huong D,
- Wechsler B,
- Papo T,
- et al.
- Wechsler B,
- Le Thi Huong D,
- Kieffer E
- Gemici K,
- Baran I,
- Gullullu S,
- et al.
- Ozkan M,
- Emel O,
- Ozdemir M,
- et al.
- Quinones M,
- Waggoner A.D,
- Redutto L.A,
- et al.
- Chen C.H,
- Nevis E,
- Fectis B,
- et al.
- Devereux R,
- Reichek N
- Ikonomidis I,
- Tsoukas A,
- Parthenakis F,
- et al.
- Garcia J.M,
- Ares A.M,
- Asher C,
- et al.
- Stefanadis C,
- Vlachopoulos C.J,
- Karayannakos P,
- et al.
- Yoo H.G,
- Shin B.A,
- Park J.S,
- et al.
- Peng J,
- Gurantz D,
- Tran V,
- et al.
- Katsuda S,
- Okada Y,
- Imai K,
- et al.
- Chambers J,
- Haskard D,
- Kooner J
- Fragasso G,
- Tresoldi M,
- Benti R,
- et al.
- Giannattasio C,
- Achilli F,
- Failla M,
- et al.