Author + information
- Received January 30, 1998
- Revision received June 29, 1998
- Accepted July 17, 1998
- Published online November 1, 1998.
- Carlos A Roldan, MD, FACCa,* (, )
- Joe Chavez, MD, FACCa,
- Philip W Wiest, MDa,
- Clifford R Qualls, PhDa and
- Michael H Crawford, MD, FACCa
- ↵*Address for correspondence: Dr. Carlos A. Roldan, Cardiology Division 5B-111, Veterans Affairs Medical Center, 2100 Ridgecrest Drive SE, Albuquerque, New Mexico 87108
Objectives. This study sought to determine the prevalence, characteristics, relation to clinical features and evolution of aortic root disease and valve disease associated with ankylosing spondylitis (AKS).
Background. Aortic root disease and valve disease are common in patients with AKS, but their clinical and prognostic implications have not been well defined.
Methods. Forty-four outpatients with AKS and 30 age- and gender-matched healthy volunteers underwent initial transesophageal echocardiography and rheumatologic evaluations. Twenty-five patients underwent clinical and echocardiographic follow-up 39 ± 10 months later.
Results. Aortic root disease and valve disease were common in patients (82%) as compared with controls (27%; p < 0.001). Aortic root thickening, increased stiffness and dilatation were seen in 61%, 61% and 25% of patients, respectively. Valve thickening (41% for the aortic and 34% for the mitral valve) manifested predominantly (74%) as nodularities of the aortic cusps and basal thickening of the anterior mitral leaflet, forming the characteristic subaortic bump. Valve regurgitation was seen in almost half of patients, and 40% had moderate lesions. Except for the duration of AKS, aortic root disease and valve disease were unrelated to the activity, severity or therapy of AKS. During follow-up of 25 patients, in up to 24% new aortic root or valve abnormalities developed, in 12% existing valve regurgitation worsened significantly and in 20% abnormalities resolved. Twenty percent of patients developed heart failure, underwent valve replacement, had a stroke or died, as compared with 3% of control subjects.
Conclusions. Aortic root disease and valve disease are common in patients with AKS, are unrelated to clinical features of AKS, can resolve or progress over time and are associated with clinically important cardiovascular morbidity.
Aortic root disease and valve disease associated with ankylosing spondylitis (AKS) include aortic root thickening and dilatation, aortic cusp thickening and retraction, subaortic bump and aortic and mitral regurgitation (1,2). Their prevalence ranges from 24% to 100% in postmortem series and 8% to 31% by transthoracic echocardiography (3–7). However, their clinical characterization, relation to clinical features of AKS, evolution and prognostic implications are unclear (1,3,5–7). Thus, in this study we sought to determine: (1) the prevalence and characteristics of aortic root disease and valve disease associated with AKS by transesophageal echocardiography (TEE), (2) the association of heart disease with clinical features of AKS, and (3) the evolution of aortic root disease and valve disease during 3 years of follow-up.
From 1993 to 1997, 102 patients with AKS were identified at two affiliated hospitals. We excluded 34 patients >60 years old because of the increased prevalence of degenerative heart disease (8,9), 2 with history of rheumatic fever or intravenous drug abuse, 4 with other spondyloarthropathies, and 18 because of refusal to participate or inability to establish communication. The remaining 44 patients constituted our study group (71% of those eligible) and prospectively underwent TEE and rheumatologic evaluations. They all denied cardiac symptoms, but four (9%) were known to have valve disease or a heart murmur (three had had valve surgery). Thirty-nine patients tested for syphilis had negative results.
The results of TEE and clinical evaluations were compared with those of 30 gender- and age-matched healthy volunteers (Table 1). The Human Research Committee of the University of New Mexico approved the study, and all subjects signed informed consent.
The activity, functional and global indexes of AKS disease for the 6 months preceding TEE were determined and each was scored from 0 to 10 (10–12). Spinal disease was assessed by the metrology index of cervical, lumbar and sacroiliac joint motion and scored from 0 to 10, the Stoke’s lumbar spine radiographic score ranged from 0 to 72, and the Dale’s sacroiliitis radiographic staging ranged from 0 (normal joints) to V (bilateral bony ankylosis) (13–15). An observer unaware of clinical and TEE data determined the radiographic scores.
A laboratory activity score of 0 to 5 was derived from five abnormalities: anemia, leukopenia or leukocytosis, thrombocytopenia, high erythrocyte sedimentation rate and high C-reactive protein. This score resulted from at least three tests performed within 12 months preceding TEE.
To assess the relation of immunoglobulin G and immunoglobulin M anticardiolipin antibodies to cardiovascular disease, these antibodies were measured in 38 patients and considered present when >22 or >10 IU, respectively, was detected (16). Finally, the anti-inflammatory therapy for ≥12 months preceding the TEE was determined.
All but four subjects underwent multiplane color Doppler TEE with a Hewlett-Packard 2500 imaging system (Andover, Massachusetts). Four subjects had transthoracic echocardiography because of TEE refusal, pregnancy or unsuccessful esophageal intubation. The aortic root and heart valves were imaged at 4 to 8-cm depth settings and with a narrow sector scan to improve image resolution. All studies were coded by Social Security number, randomly intermixed with those of controls and interpreted by an observer unaware of clinical data.
Electronic calipers were used to measure by M-mode: (1) the end diastolic thickness of the aortic root walls at the sinuses and tubular portions from the longitudinal or short axis view; (2) the aortic root diameters at the annulus, sinuses and tubular levels (17); (3) the mitral leaflet thickness at the basal, mid and tip portions from the four-chamber view and the aortic cusps at their mid and thickest portions from the longitudinal or short axis view; and (4) the end diastolic diameter of the mid descending thoracic aorta. The aorto-mitral junction length and height at the end diastole were measured using two-dimensional images from the longitudinal view. All measurements were averaged during three cardiac cycles.
From four to six blood pressure measurements during TEE and the root diameters, aortic elasticity at the sinuses and tubular portions were calculated by the Peterson’s pressure–strain elastic modulus as 106dynes cm−2and the pressure-independent stiffness index (B) (18,19).
Criteria for echocardiographic interpretation
Aortic root abnormalities were defined as 2 SD above the mean values obtained for the controls (Table 2). Aortic root thickening was defined as an anterior or posterior wall thickness of >2.2 mm. Sinus or tubular aortic root dilatation was defined as >41 or >35 mm, respectively. These diameters are above normal for healthy subjects <60 years old (17,20). Abnormal aortic root pressure–strain elastic modulus and stiffness were defined as >2.1 and >4.1 at the sinuses or >3.1 and >3.9 at the tubular level, respectively. Subaortic bump was defined as an aorto-mitral junction length and height of >7.7 and >3.2 mm, respectively (Fig. 1A).
Valve thickening was defined as a thickness >3 mm for the mitral valve or >2 mm for the aortic valve in at least two leaflets, or in one leaflet if associated with at least mild regurgitation of the respective valve (21)or if a subaortic bump was present (for the mitral valve). Valve regurgitation assessed by semiquantitive methods (22,23)was present if it was greater than mild or if it was mild with thickening of the respective valve.
All subjects had 12-lead electrocardiography.
Evolution of aortic root disease and valve disease
To assess the evolution of cardiovascular disease, 25 patients had repeat transthoracic echocardiography and electrocardiography 39 ± 10 months later. These evaluations were performed >12 months after the initial TEE or at the time of complications. An observer unaware of initial results interpreted these studies.
Morbidity and mortality
During follow-up of the same 25 patients, the incidences of heart failure, valve replacement, infective endocarditis, stroke and death were determined. Findings were compared with those of all controls followed for 50 ± 16 months.
Student and paired ttests were used for comparison of continuous variables, and analysis of variance was used to compare characteristics of the aortic root and heart valves between groups after adjusting for smoking, hypertension, diabetes or coronary artery disease (24). A subset analysis of patients was done by analysis of variance with Fisher’s least significant differences as post hoc tests. Fisher’s exact and McNemar’s tests were used for comparison and associations of categorical variables (25). To assess interrater agreement for detecting cardiovascular disease, a second observer interpreted 35 randomly selected TEE studies (20 patients and 15 controls). Interrater agreement was determined by the percentage of agreement as well as by Cohen’s kappa test (26). A p value <0.05 was considered significant.
Characteristics of the aortic root and heart valves
In patients the aortic root thickness, diameter and elasticity; the aorto-mitral junction length and height; and the aortic and mitral valve thickness were significantly different from those of controls even after adjusting for smoking, hypertension, diabetes mellitus or coronary artery disease (Table 2).
Aortic root disease and valve disease
Aortic root disease and valve disease were detected in 36 patients (82%) and 8 controls (27%) (p < 0.001) (Table 3).
Aortic root thickening and decreased elasticity, but not root dilatation, were common in patients (61%) as compared with controls (<10%) (p < 0.001). Most patients (78%) with root thickening had decreased root elasticity, and most (73%) with root dilatation had root thickening.
Valve thickening was frequent (48%) in patients and similar for the mitral and aortic valves. Aortic valve thickening was predominantly (72%) manifested as nodularities (up to 6 mm in diameter), generally single, homogeneously echoreflectant, with well-defined borders, located on any portion of the three cusps (Fig. 1C)and unrelated to the patient’s age (logistic regression, p = 0.19). Diffuse thickening was uncommon, was noted on the cusp margins and rarely caused decreased cusp mobility. Mitral valve thickening was also common and predominantly localized (73%) to the basal anterior mitral leaflet, forming a subaortic bump in most cases. The decreased diastolic mobility of the basal anterior mitral leaflet, in contrast to normally moving mid and distal portions, gave the leaflet the appearance of an elbow (elbow sign) (Fig. 1B)(p < 0.001 for the association of subaortic bump with elbow sign).
Valve regurgitation was also frequent (45%) in patients and similar for the mitral and aortic valves. Eight (40%) of 20 patients with valve regurgitation had moderate lesions (aortic in three and mitral in five). Seven patients (16%) had valve prolapse: aortic in two and mitral in five (posterior leaflet in two, anterior in one and both in two).
Aortic root disease and valve disease were interrelated (73% to 100%, p < 0.02 for all associations). Most patients with aortic valve thickening had root and mitral valve thickening, most with mitral valve thickening had root thickening and subaortic bump, and all with subaortic bump had root thickening. All patients with aortic regurgitation had root and valve thickening, most with mitral regurgitation had valve thickening, and all with moderate aortic or mitral regurgitation had subaortic bump (Figs. 1 and 2). ⇓
Of 35 TEE studies evaluated by a second observer, 66% had aortic root or valve disease. There was an overall 85% agreement for detecting aortic root disease (kappa = 0.70) and an overall 89% agreement for valve disease (kappa = 0.77): 83% for root thickening, 91% for dilatation, 82% for stiffness, 89% for valve thickening, 93% for regurgitation and 85% for subaortic bump.
Other cardiac findings
Sixteen patients (36%) had other cardiac disease: 10 (23%) had conduction disturbances (left bundle branch block in 3, right bundle branch block in 5, atrioventricular, nonspecific intraventricular or fascicular block in 2 of them; and 8 had root thickening and subaortic bump; five (11%) had coronary artery disease; two (5%) had nonischemic cardiomyopathy; and 2 had left ventricular hypertrophy. Four controls (13%) had other cardiac disease: 3 had conduction disturbances and 1 had left ventricular hypertrophy (p = 0.02 for patients vs. controls).
Relation of cardiovascular disease to clinical features of AKS
To define an association with clinical features of AKS, three groups were defined: those with aortic root disease and valve disease (24 patients), those with aortic root disease or valve disease (12 patients) and those with neither (8 patients). All parameters of AKS activity and severity or therapy did not differ according to the presence or severity of cardiovascular disease. However, most patients >45 years old with AKS duration of >15 years had aortic root or valve disease (Table 4).
Evolution of cardiovascular disease
Six of 25 patients (24%) developed new or other aortic root or valve abnormalities. In three other patients (12%) valve regurgitation progressed from moderate or less to severe regurgitation in two (one with mitral and one with aortic) and from mild to moderate aortic regurgitation in the third. In five patients (20%) aortic root or valve abnormalities resolved during follow-up. One patient developed a new conduction disturbance.
Morbidity and mortality
Before entry into the study, one patient had had aortic valve replacement for severe regurgitation. At surgery the aortic root was moderately dilated and the aortic cusps were thickened and retracted. Pathology showed cusp fibrosis and mucinoid degeneration. A second patient had had mitral valve repair and annuloplasty for severe regurgitation due to a flail myxomatous posterior leaflet. A third patient had had mitral valve replacement for severe regurgitation due to chronic valvulitis. In these three patients surgery confirmed preoperative TEE findings.
During follow-up five patients (20%) developed seven complications. A 34-year-old man underwent aortic valve replacement for severe regurgitation and heart failure. The aortic root and annulus were moderately dilated and the three cusps were thickened and retracted, forming a central regurgitant jet. These findings matched the preoperative TEE (Fig. 2). Pathology showed cusps nodular thickening and fibrosis but no inflammation. A 50-year-old man with a previously replaced aortic valve and residual diffuse thickening of the anterior mitral leaflet developed severe mitral regurgitation and heart failure, but refused valve surgery. A 52-year-old woman with a small mass of low echoreflectance on the aortic noncoronary cusp developed two transient ischemic attacks. A 41-year-old man with root thickening and dilatation but no valve disease had a stroke. These last two patients had no aortic atheroma, atrial pseudocontrast or anticardiolipin antibodies. Finally, a 55-year-old man with a normal TEE died from a subarachnoid hemorrhage. A 55-year-old woman control subject (3%) with a normal TEE developed hypertension and a transient ischemic attack (p = 0.08, patients vs. controls).
Aortic root disease and valve disease in patients with AKS: (1) are common, are interrelated and have unique TEE characteristics; (2) are unrelated to clinical features of AKS with the exception of patients’ age and duration of AKS; and (3) show resolution or progression over time and are associated with clinically important cardiovascular morbidity.
Pathogenesis and comparison with postmortem studies
The cellular inflammation of the first 3 cm of the aortic root results in a marked fibroblastic reparative response, adventitial thickening, focal destruction of the medial elastic tissue and intimal proliferation. The surrounding inflammation and intimal proliferation of the vasa vasora lead to obliterative endarteritis and then to root dilatation. This process extends to the aortic annulus and produces basal thickening and downward retraction of the cusps. These chronic or acute changes of the aortic root and valve lead to aortic regurgitation (1,2,27). Extension of the fibrosis into the aorto-mitral junction forms a subvalvular fibrous ridge or subaortic bump. Mitral regurgitation also results from chronic or acute valvulitis, but its association with a subaortic bump was not well defined. Further extension of fibrosis into the proximal septum and atrioventricular node leads to conduction disturbances. However, the antigenic trigger, the pathogenesis of inflammation and the selective involvement of the left heart skeleton and proximal aorta are unknown.
The aortic root disease and valve disease described in this study correspond to those described in pathology (1,2). However, in this study aortic root thickening was more common than dilatation (61% vs. 25%, respectively) and dilatation was usually mild. Also, aortic valve thickening was usually localized instead of diffuse. In addition, an association of structural with functional aortic root disease and valve disease was established. Root thickening was associated with increased stiffness, and these abnormalities in addition to aortic valve thickening or prolapse, but not root dilatation, were associated with aortic regurgitation. Also, all patients with at least moderate regurgitation had subaortic bump. The subaortic bump decreased the mobility and may have caused anterior mitral leaflet retraction, asymmetric or incomplete coaptation and a relative or true leaflet prolapse. Finally, conduction disturbances were associated with aortic root thickening or subaortic bump.
Comparison with previous echocardiographic studies
Previous transthoracic echocardiographic series (3–7,28)reported a lower prevalence of aortic root disease and valve disease (8% to 31%). The abnormalities described were subjectively assessed and were frequently nonspecific, and an interrelation between aortic root disease and valve disease was not well defined. In addition, aortic root stiffness was not reported and the mechanisms of valve regurgitation were not described. The use of a less sensitive echocardiographic technique, the inclusion of patients >60 years old, the qualitative definition of root and valve abnormalities and the lack of control for other conditions associated with aortic root disease and valve disease may explain these differences.
Relation of cardiovascular disease to clinical features of AKS
The presence and severity of aortic root disease and valve disease were temporally dissociated from the activity, severity or therapy of AKS, but were associated with the patients’ age or duration of AKS. Previous series (3,5,7,29)using less objective clinical evaluations also suggested a dissociation. Similar dissociation appears to exist with pulmonary or eye disease associated with AKS (30,31).
The dissociation between cardiovascular disease and the clinical features of AKS is probably explained by several factors.
1. A highly sensitive technique for detecting mild cardiovascular abnormalities in outpatients with a wide spectrum of duration, activity and severity of AKS is applied.
2. The currently utilized clinical and radiographic parameters of activity and severity of AKS may not be sensitive or specific (32,33). Elevation of the erythrocyte sedimentation rate or C-reactive protein occurs in only 30% to 60% of patients with severe AKS (32,34).
3. The cardiovascular and skeletal inflammation may be temporally dissociated or may be triggered by different target antigens.
4. In contrast to other connective tissue diseases, AKS is frequently subclinically active or symptomatically regressive (32,34).
Thus, the traditional parameters of skeletal inflammation of AKS cannot be utilized as markers of cardiovascular disease.
Evolution of aortic root disease and valve disease
During follow-up, up to 24% of patients developed new or other aortic root or valve abnormalities, in 12% valve regurgitation progressed significantly and in 20% abnormalities resolved. These changes may be related to intermittent, recurrent or persistent aortitis or valvulitis. Similar changes have been observed in the immune-mediated valve disease of systemic lupus erythematosus and rheumatic fever (21,35).
Seven patients had complications related to aortic root disease or valve disease. Five patients required valve surgery for severe valve dysfunction, but one declined it; two patients developed cerebral ischemic events. In the four patients who had valve surgery, surgical findings confirmed those of preoperative TEE. This increased cardiovascular morbidity in patients with AKS may account for their higher than expected mortality at 10 years of follow-up (36).
The study of 71% of eligible patients may have skewed the study results toward an overestimation of the prevalence and morbidity of the cardiovascular disease associated with AKS. The study of a larger population with longer follow-up is necessary to better assess our findings. The comparison of TEE with follow-up precordial echocardiography may decrease the validity of the observed changes of aortic root disease and valve disease.
Aortic root stiffness may cause hypertension or increased left ventricular afterload, leading to ventricular hypertrophy and diastolic dysfunction. In this study patients had higher systolic and diastolic blood pressures than controls. Previous series have reported a high prevalence (20%) of systemic hypertension (4)and of left ventricular diastolic dysfunction (20% to 43%) in nonhypertensive patients with AKS <45 years old (29,37).
The most feared consequence of valve disease is progression to severe valve dysfunction and the need for valve replacement. In this study four of five patients with severe valve regurgitation underwent valve replacement. However, also in this study no clinical predictors other than duration of AKS were predictive of the presence, severity and progression of cardiovascular disease.
Aortic root or valve lesions may be a substrate for cardioembolism. In two patients with cerebral ischemic events, an aortic valve mass in one and aortic root thickening and dilatation in the other were the only possible substrates. Although a causal relation with embolism could not be established, an association is possible. Studies of the natural history of AKS have demonstrated that cerebrovascular disease is a common cause of mortality in patients <60 years old (36). Thus, while awaiting the study of larger populations with a longer follow-up, prophylactic antiplatelet therapy may be indicated in these patients. Infective endocarditis is another reported complication of asymptomatic valve disease associated with AKS (38).
In this study most patients had asymptomatic aortic root disease and valve disease, but 40% of those with valve regurgitation had moderate lesions and in another 12% valve regurgitation progressed significantly. Also, patients ≥45 years old with an AKS duration of ≥15 years had the highest prevalence of cardiovascular disease. Therefore, these patients should be approached with high clinical suspicion, and a careful cardiovascular history and physical examination should be the preferred screening method for detecting significant valve disease (39). Finally, serial clinical or echocardiographic follow-up of patients with AKS and aortic root disease or valve disease appears warranted.
We thank Frank T. Gurule, Aggie M. Schaeffer and Kristine E. Councilman for echocardiographic technical assistance, and Jennifer Jones, RN, for collecting clinical data.
- ankylosing spondylitis
- transesophageal echocardiography
- Received January 30, 1998.
- Revision received June 29, 1998.
- Accepted July 17, 1998.
- American College of Cardiology
- Bulkley B.H,
- Roberts W.C
- Davidson P,
- Baggenstoss A.H,
- Slocumb C.H,
- Daugherty G.W
- O’Neil T.W,
- King G,
- Graham I.H
- Alves M.G,
- Espirito-Santo J,
- Queiroz M.V,
- Madeira H,
- Macieira-Coelho E
- Jones S.D,
- Steiner A,
- Garrett S.L,
- Calin A
- Averns H.L,
- Oxtoby J,
- Taylor H.G,
- Jones P.W,
- Dziedzic K,
- Dawes P.T
- Gardin J.M,
- Henry W.L,
- Savage D.D,
- Ware J.H,
- Burn C,
- Borer J.S
- Paterson L.H,
- Jensen R.E,
- Parnell J
- Kawasaki T,
- Sasayama S,
- Yagi S,
- Asakawa T,
- Hirai T
- Gerstenblith G,
- Frederiksen J,
- Yin F.C.P,
- Fortuin N.J,
- Lakatta E.G,
- Weisfeldt M.L
- Castello R,
- Lenzen P,
- Aguirre F,
- Labovitz A.J
- Perry G.J,
- Helmcke F,
- Nanda N.C,
- Byard C,
- Soto B
- ↵StatXact 3 for Windows User Manual. Cambridge, MA: CYTEL Software, 1995 (Software).
- ↵(1990) SAS/STAT User’s Guide, Version 6 (SAS Institute, Cary, NC).
- Fleiss J.L
- Vasan R.S,
- Shrivastava S,
- Vijayakumar M,
- Narang R,
- Lister B.C,
- Narula J
- Hoppmann R.A,
- Wose C.M,
- Challa R,
- Peacock J