Author + information
- Received July 30, 1998
- Revision received November 11, 1998
- Accepted December 24, 1998
- Published online April 1, 1999.
- ↵*Reprint requests and correspondence: Dr. Emile Ferrari, 30 Avenue de la Voie Romaine, B.P. 69 06002 Nice, France
Our aim was to confirm the poor prognosis related to thoracic aortic plaques, in particular aortic debris, diagnosed by transesophageal echocardiography (TEE) and to evaluate patients’ prognosis as a function of the antithrombotic treatment.
Aortic atheroma (AA) has been widely studied. However, it is still not known which antithrombotic treatment should be adopted in this disease.
Patients referred for TEE and diagnosed with AA were followed. All thromboembolic events and deaths were recorded during a follow-up of 22 ± 10 months. The antithrombotic treatment to be adopted was left to the discretion of the practitioner in charge of the patient.
Aortic atheroma was found in 12% of all TEE performed and in 27.5% of TEE performed for stroke. This prevalence was higher when no other etiology existed to explain the stroke (p < 0.001). During follow-up, an end point occurred in 22.5% of patients. The more severe the AA the greater the incidence of events (p = 0.007). A higher mortality rate is shown in patients with aortic debris (p = 0.049). Compared with those treated with oral anticoagulants, patients with aortic plaques >4 mm thick treated with antiplatelets had more embolic events and combined events (p = 0.01 and p = 0.007, relative risk [RR] = 5.9, 95% confidence interval [CI] = 1.4 to 15, respectively); patients with aortic debris had more combined events and a higher mortality rate (p = 0.001, RR = 7.1, 95% CI = 1.2 to 19 and p = 0.019, RR = 9.1, 95% CI = 1.2 to 25, respectively).
We confirm the high incidence of vascular events and deaths in patients with AA. We have demonstrated, for the first time in this condition, a better outcome among patients treated with oral anticoagulants versus antiplatelets.
Atherosclerotic disease of the thoracic aorta as a cause of embolic events has been well documented (1–18). However, although the prevalence of aortic atherosclerosis and its value as an indicator of poor prognosis have been studied, the prognostic value of “aortic debris,” a particular form of aortic atheroma, is less well documented. Furthermore, the optimal antithrombotic treatment in the presence of such disease remains unknown.
The first aim of our study was to evaluate the risk of future vascular events as a function of the antithrombotic treatment prescribed in patients with aortic atheroma (AA) diagnosed by transesophageal echocardiography (TEE). Furthermore, we sought to confirm the poor prognosis related to AA of the thoracic aorta and, in particular, to assess the prognosis related to aortic debris.
Patients referred to our echocardiographic laboratory for TEE from January 1993 to December 1995 and presenting with AA were followed. At the time of inclusion, information was recorded relative to arterial hypertension, hypercholesterolemia, smoking, diabetes mellitus, peripheral vascular disease and atrial fibrillation.
The TEE procedure was performed by two practitioners in awake patients with local premedication (lidocaine spray). The investigation was performed using a Hewlett-Packard Sonos 1000 (biplane probe: 5 MHz). The thoracic aorta was examined with special care in all patients after the usual examination of the cardiac structures. The location and thickness of all aortic plaques were carefully noted. All examinations were recorded on VHS allowing a second physician, with no knowledge of the patient status, to view the procedure and make an independent assessment of the AA classification.
Aortic atheroma was classified using modified FAP study criteria (14). As one of our aims was to individualize the prognosis of patients with aortic debris, we divided the patients into three groups according to the thickness of the aortic wall and the presence of an atherosclerotic mobile component in the lumen of the aorta. As a result, grade I was defined as a plaque with a thickness ranging from 1 to 3.9 mm. Grade II was defined as a plaque of more than 4 mm in thickness. Grade III defined any plaque, whatever its thickness, with an obvious mobile component (aortic debris). As the prognosis of patients presenting with aortic plaque <1 mm in thickness has been demonstrated to be no different from the prognosis of patients with aortic plaques of 1 to 3.9 mm (14), aortic plaques less than 1 mm thick were not taken into account.
No interobserver variability was observed concerning the classification of AA.
During follow-up, information was obtained from all patients or their family (when patients were unable to answer). Forty-five percent of patients were examined at least once at the end of follow-up. The remaining 55% were interviewed by telephone. In particular, we sought to collect all clinical events possibly caused by an embolic mechanism such as stroke, transient ischemic attack, acute ischemia of a limb and death. Whatever the event, the latter was always confirmed by the physician in charge of the patient or from hospital records.
It is noteworthy that, when no other embolic potential etiology apart from AA was found, patients were referred back to their physician with no particular recommendation concerning the treatment to be adopted. The antithrombotic treatment followed was divided into two groups: oral anticoagulants (OAC) and antiplatelet agents (APA) (aspirin or ticlopidine hydrochloride). Neither the data on anticoagulation intensity (i.e., the International Normalized Ratio) nor the reasons for the choice of either APA or OAC (i.e., contraindication to warfarin) were prospectively collected.
Data concerning patients’ characteristics are expressed as mean ± SD. For patient outcome, comparisons are made using the three-grade classification of aortic atheroma defined by the echocardiographic criteria and the antithrombotic treatment followed.
Univariate comparisons of characteristics between patients were performed with the chi-square test for categorial variables (gender, history of diabetes, hypertension, current smoking status, atrial fibrillation) and with two-sample ttests for continuous variables (age and left ventricular ejection fraction).
End points were stroke or peripheral embolic event on the one hand and death, of whatever etiology, on the other. Kaplan–Meier estimates of the distribution of times from inclusion to end point were computed. Log rank analysis was performed to compare the event curves for different groups. Relative risks (RRs) with 95% confidence interval (CI) (Taylor method) are given. In all analyses, a p value of 0.05 or less was considered to indicate statistical significance. The data were analyzed with SAS (version 6.11).
Patients and prevalence of aortic plaques
During the study period, 1,116 TEEs were performed and examined by two independent practitioners. Transesophageal echocardiography was prescribed for the following reasons: to seek an embolic cause after an embolic event (n = 429), to check for valvular disease (n = 397), to look for clots in the left atrium before cardioversion (n = 149), to confirm or rule out endocarditis or aortic dissection (n = 86) or for miscellaneous reasons (n = 55). Protruding atheroma was found in 139 patients, that is, 12% of all cases. Among these, TEE was prescribed in 118 cases (85%) for an embolic event (stroke in 107 cases [77%], an embolic event in a limb in 11 cases [8%]) and for other reasons in 21 cases (15%).
As a result, a significant atheroma, as defined above, was found in 25.6% of patients referred for TEE after a stroke (107/418), and in 27.5% of patients referred for TEE after a stroke or peripheral embolic event (118/429).
It is worth noting that, among the patients referred for an embolic event, and despite a complete etiologic search (carotid Doppler, ECG Holter monitoring, transthoracic and TEE), in all, prevalence of aortic plaques was statistically more frequent in patients in whom no other etiology for this embolic event was found, that is, 47.7% of cases (56/118) versus 16.7% (52/311) in the group of patients in whom a potential cause of emboli was discovered (p < 0.0001).
Ten of the 139 patients were lost during follow-up. As a result, our study group comprised 129 patients. Mean age was 68 ± 8.7 years, 71% were male, 32% were diabetics, 39% were current smokers and 24% had arterial hypertension. Their classification according to the severity of the AA is shown in Table 1. It is worth noting that 33 of 34 patients presenting with a complex atherosclerotic plaque with a mobile component (aortic debris: grade III) presented with at least one plaque >4 mm in thickness. Mean follow-up was 22 ± 10 months.
In 92 out of the 129 cases (71.3%), aortic plaques involved a single portion of thoracic aorta: the ascending aorta and/or aortic arch (to the ostium of the left subclavian artery) in half the cases, and the descending thoracic aorta in the other half. In 37 cases (28.7%), aortic plaques were localized in both aortic segments. This diffused localization was more frequent in grades II and III (Table 1).
Table 2shows the treatment followed according to the aortic atheroma. Fifty-four percent of all patients were treated with APA (aspirin [250 mg] or ticlopidine [500 mg]); 46% were treated with OAC. None was treated with both OAC and APA.
It is worth noting that the great majority (71%) of group I patients were treated with APA (p < 0.01), whereas the distribution between the two treatments was identical for grade II and III patients.
Vascular events and death during follow-up
During follow-up, 12 thromboembolic events (eight strokes, four peripheral embolic events) and 17 deaths occurred. All thromboembolic events were documented by computed tomography scan (no hemorrhagic stroke occurred) or angiography. As a result, during follow-up, 22.5% of the 129 patients presented with one of the serious events being collected; the global mortality rate was 13.2% (Table 3).
Death was related to documented vascular events in five cases (three acute myocardial infarctions, one type A aortic dissection, one massive stroke). Sudden death (defined as witnessed sudden unexpected death within 6 h of the onset of symptoms from a stable medical condition, or death occurring <24 h after an individual had been seen in a stable condition) occurred in six other patients. Other patients died of causes having no clear relation to their aortic atheromas, including cancer (n = 2), cardiac failure (n = 2), pulmonary embolism (n = 1) and renal failure of unknown cause (n = 1).
Mean age of patients who died or who experienced a vascular event was 69 ± 6 years (range 50 to 80) and was not statistically different from the rest of the study group (p = 0.16).
Prognostic value of aortic atheroma severity
Considering deaths from any cause, mortality rate was 4%, 14% and 23.5% respectively, for grades I, II and III.
When comparing the three groups, although the threshold of statistical significance is not reached, log rank analysis shows a trend in favor of higher mortality in patients with more severe AA (p = 0.069). Comparison between two groups (grade II vs. I, grade III vs. I and grade III vs. II) shows a gradual increase in mortality risk (Table 3).
However, if one compares more severely diseased patients (grade III) with the others (grades I and II), a statistically higher mortality rate is observed in the grade of more severely diseased patients (nine deaths out of 95 patients in grades I and II vs. eight deaths out of 34 patients in grade III) (log rank = 3.82, RR = 2.51, 95% CI = 1.7 to 10) (Table 3).
Of grade I, 4.4%, 10% of grade II and 14.7% of grade III patients suffered a documented embolic event (stroke for two grade I patients, three grade II patients and three grade III patients; a peripheral embolic event for two grade II patients and two grade III patients). The occurrence of an embolic event in the three groups, although showing a trend, is not statistically different (p = 0.26). However, although not statistically significant, comparisons between two groups show a gradual increase for embolic event risk (Table 3).
Combined end points
The occurrence of combined events (embolic event or death) increased with the severity of the disease. The more severe the AA, the worse the patients’ prognosis. In effect, when comparisons are made between the three grades, Kaplan–Meir curves show a significant difference with a higher incidence of events in the more severe grades (p = 0.007) (Fig. 1).
If one analyzes the occurrence of embolic events alone, there is no difference between grade I patients whatever the treatment used. In grade III a trend is shown for fewer embolic events in patients treated with OAC (p = 0.08). However, in grade II a real benefit is highlighted with OAC. Indeed, the risk of an embolic event is higher, in this group, when patients are not treated with OAC (p = 0.01) (Table 4).
As regards the mortality rate, although no statistical difference exists concerning mortality in grade I or II patients according to the treatment, a significant benefit is observed in group III patients taking OAC as compared with those patients of the same grade classification taking APA (p = 0.019, RR = 9.1, 95% CI = 1.2 to 50) (Table 4) (Fig. 2). ⇓
As a result, although in grade I patients the incidence of combined end points is not different according to the treatment used (p = 0.7), the occurrence of these combined end points is significantly lowered in patients treated with OAC versus those treated with APA in grade II and III patients (p = 0.007, RR = 5.9, 95% CI = 1.4 to 15; and p = 0.001, RR = 7.1, 95% CI = 1.2 to 19, respectively) (Table 4) (Fig. 3 and 4). ⇓⇓
Prevalence and prognosis value of aortic atheroma
Among the 1,112 TEEs performed, we found a 12% prevalence of AA. This prevalence rose to 27.5% when considering TEE performed to look for an emboligenic source.
These figures are very similar to those reported in published reports (5,7,10). In our study, the great majority of patients in whom AA was found were checked for an embolic event (108/129: 84%). It is worth noting that when an AA is found, the result of an etiologic search to find a cause for the embolic event is more frequently unproductive (p < 0.001). This supports a possible causal link between AA and embolic events.
We confirm the poor prognosis of patients with AA, in particular those with plaques greater than 4 mm in thickness. The high incidence of vascular events or death over a period of almost two years is consistent with previous data (5,7,8,14). Our result means that one patient out of four presenting with an aortic plaque >4 mm and more than one patient out of three with aortic debris will experience a vascular event or die within two years.
One of our major findings is that, during follow-up, vascular events or deaths increased proportionately to the severity of AA. The more severe the AA, the more frequent the vascular events (p = 0.007). In particular, when compared with patients with aortic plaques <4 mm thick with no mobile component (grade I), patients with aortic debris had a higher risk for combined events and death (RR = 4.3, 95% CI = 1.5 to 12; and RR = 5.3, 95% CI = 1.2 to 23, respectively).
The same analysis tends to show a worse prognosis for these patients as compared with those with AA plaques >4 mm thick (but no aortic debris), a condition already associated with a poor prognosis.
Finally, mortality rate alone is higher when patients with aortic debris are compared with those with significant AA but no aortic debris (grade I+II) (p = 0.049, RR = 2.51, 95% CI = 1.7 to 10).
These aortic debris were first described by Karalis et al. (5). They have been associated with an increased incidence of embolization but, to our knowledge, not with a higher mortality (5,7,11,16,17).
Many published data support the hypothesis that complex aortic plaques are valuable “markers” of severe and widespread atheroma. In particular, the link between AA diagnosed with TEE and coronary artery disease has been demonstrated (19,20). We believe this may explain why severe AA, though not always the direct cause of death, is correlated with higher global mortality.
Benefit of oral anticoagulants
The other major result in our study is the benefit of oral anticoagulation as opposed to antiplatelet agents. We demonstrate this benefit for embolic events and combined end points in patients with aortic plaques >4 mm thick, and for combined end points and also mortality in patients with aortic debris.
In particular, our results show that patients with aortic debris treated with APA have a nine-fold higher mortality risk as compared with the same patients treated with OAC. On the other hand, patients with aortic plaque >4 mm thick have almost a six-fold higher risk for combined events when treated with APA as compared with OAC.
In a recent study, dealing with 31 patients followed for about 10 months, Dressler et al. reported a lower incidence of vascular events in those patients treated with warfarin as compared with those treated with antiplatelets (21). In the FAP study (14), among patients with plaques >4 mm thick, no significant difference in the event rate between patients receiving warfarin and those receiving aspirin was noted. However, only nine patients (20%) received warfarin in this group. Furthermore, patients with a mobile lesion received warfarin for one to three months and aspirin thereafter. We believe this may have minimized efficacy of OAC.
In our study population, the incidence of aortic debris, which obviously represents the most severe form of AA, was frequent, involving 26% of patients. This may have increased the incidence of vascular events or deaths and probably made it easier to demonstrate the beneficial effect of OAC.
The main limitation of our study could be that treatment was not randomly assigned. Fortunately, however, the distribution of the two antithrombotic regimens was homogeneous in the most severe groups (Table 2).
Furthermore, in clinical practice, physicians usually use oral anticoagulation instead of antiplatelet agents only when the “global vascular risk” is higher. Indeed, Table 5details a comparison between the main patient characteristics in the two treatment groups and shows that patients treated with OAC more frequently presented with atrial fibrillation or left ventricular dysfunction. This may have minimized the beneficial effects of OAC we found.
We did not use a control group without atherosclerotic disease of the aorta. As a result, the comparisons shown were made between three severe forms of AA. However, some published data suggest that a patient with aortic plaque between 1 and 3.9 mm in thickness (our grade I) runs the same risk as a patient with no aortic plaque (14).
Many treatments may have a favorable effect on AA (e.g., hypocholesterolemic drugs, antithrombotics). In addition to a comparison of medical and surgical strategies, further studies are needed to investigate these medications that may have a favorable impact on this condition. In particular, randomized studies are warranted. These should compare not only different antithrombotic treatments but also investigate the need for such a treatment (placebo).
Recent studies have emphasized the possible changes in the echocardiographic appearance of aortic debris and even the poor correlation between echocardiographic data and pathologic findings (22,23). They suggest that what is described as aortic debris may often constitute a superimposed thrombus on a classic aortic plaque. Whatever the pathologic findings, the link between “aortic debris” and poorer prognosis signifies at very least that these ulcerated aortic plaques present a greater propensity to evolve with all inherent complications. Moreover, the favorable evolution (decrease in size) of these aortic debris described by serial TEE examinations in patients treated with OAC is an argument for OAC treatment (22).
We confirm that patients with protruding AA seen by TEE are at high risk for the development of vascular events, especially embolic events and death. In particular, the presence of aortic debris is correlated with a higher mortality rate whatever the cause.
Furthermore, we demonstrated, for the first time, a better outcome among patients treated with OAC compared with those treated with antiplatelet agents. This beneficial effect is highlighted in patients with aortic plaques >4 mm in thickness for embolic events alone or combined embolic events and mortality. The same benefit is observed in patients with aortic debris for combined embolic events and mortality but also mortality alone. This finding should justify further randomized studies.
- aortic atheroma
- antiplatelet agents
- confidence interval
- oral anticoagulants
- relative risk
- transesophageal echocardiography
- Received July 30, 1998.
- Revision received November 11, 1998.
- Accepted December 24, 1998.
- American College of Cardiology
- Katz E.S,
- Tunick P.A,
- Rusinek H,
- et al.
- Montgomery D.H,
- Ververis J.J,
- McGorisk G,
- et al.
- Karalis D.G,
- Chandrasekaran K,
- Victor M.F,
- Ross J.J Jr.,
- Mintz G.S
- Amarenco P,
- Cohen A,
- Baudrimont M,
- Bousser M.G
- Tunick P.A,
- Rosenzweig B.P,
- Katz E.S,
- et al.
- Khatibzadeh M,
- Mitusch R,
- Stierle U,
- et al.
- Ferrari E,
- Taillan B,
- Drai E,
- et al.
- Filiale echocardiographie de la societé Française de cardiologie,
- Laperche T,
- Laurian C,
- Roudaut R,
- Steg P.G
- Fazio G.P,
- Redberg R.F,
- Winslow T,
- Schiller N.G
- Dressler F.A,
- Craig W.R,
- Castello R,
- Labovitz A.J
- Craig W.R,
- Dressler F.A,
- Vaughn L.M,
- et al.
- Vaduganathan P,
- Ewton A,
- Nagheh S.F,
- et al.