Author + information
- Received November 26, 1997
- Revision received May 6, 1998
- Accepted May 14, 1998
- Published online September 1, 1998.
- ↵*Address for correspondence: Dr. Craig Frances, Veterans Affairs Medical Center, Room 111A1, 4150 Clement Street, San Francisco, California 94121
Left ventricular (LV) pseudoaneurysms form when cardiac rupture is contained by adherent pericardium or scar tissue. Although LV pseudoaneurysms are not common, the diagnosis is difficult and they are prone to rupture. We evaluated the clinical presentation, diagnostic accuracy of imaging modalities, results of therapy and prognosis of 290 patients with LV pseudoaneurysms. Most cases of LV pseudoaneurysm were related to myocardial infarction (particularly inferior wall myocardial infarction) and cardiac surgery. Congestive heart failure, chest pain and dyspnea were the most frequently reported symptoms, but >10% of patients were asymptomatic. Physical examination revealed a murmur in 70% of patients. Almost all patients had electrocardiographic abnormalities, but these were usually nonspecific ST segment changes; only 20% of patients had ST segment elevation. Although radiographic findings were also usually nonspecific, the appearance of a mass was present in more than one half of patients and may be an important clue to the correct diagnosis. Left ventricular angiography was the most definitive test and can be useful in planning surgery since concomitant coronary angiography can be performed. Regardless of treatment, patients with LV pseudoaneurysms had a high mortality rate, especially those who did not undergo surgery. Because the symptoms, signs, electrocardiographic abnormalities and radiographic findings seen in patients with LV pseudoaneurysms can be indistinguishable from those in patients with coronary disease alone, a high clinical index of suspicion is needed to avoid missing the diagnosis.
Left ventricular (LV) pseudoaneurysms form when cardiac rupture is contained by adherent pericardium or scar tissue (1). Thus, unlike a true LV aneurysm, a LV pseudoaneurysm contains no endocardium or myocardium (2). Wall stress, which is related to LV pressure and radius, and loss of myocardial integrity (e.g., from myocardial infarction) are the likely determinants of cardiac rupture (3). Free intrapericardial rupture usually results in cardiac tamponade and death (4). Less frequently, cardiac rupture is contained and LV pseudoaneurysm formation occurs. Although LV pseudoaneurysm is clinically uncommon, diagnosis is difficult and rupture often leads to death. The most recent quantitative review of this subject was published in 1977 (5)and did not include information on the accuracy of various diagnostic tests. Therefore, we evaluated the clinical presentation, diagnostic accuracy of imaging modalities, results of therapy and prognosis of 290 patients with LV pseudoaneurysms.
We performed an English language search of MEDLINE from 1966 through November 1997 for articles with the key word “pseudoaneurysm” or the subject “false aneurysm,” “infected aneurysm” or “ruptured aneurysm.” We also reviewed references from all identified case reports. Articles not pertaining to pseudoaneurysms of the LV free wall and those without individual case data were excluded. We extracted data from 201 articles that contained 290 case reports (references available from the investigators).
Two reviewers (CF and AR) collected data using a standardized data abstraction form. All disagreements were resolved by discussion and consensus. We collected clinical information including patient age, sex, race, symptoms, signs and electrocardiographic and radiographic findings. We assumed that the pseudoaneurysm was due to the most proximate cardiac event. For example, a patient with a myocardial infarction who later underwent LV aneurysmectomy and was subsequently diagnosed with a LV pseudoaneurysm was placed in the “surgical etiology” category. We determined whether transthoracic and transesophageal echocardiography, radionuclide scan, computed tomography (CT) scan, magnetic resonance imaging (MRI) and angiography were helpful in diagnosing LV pseudoaneurysms. We categorized a study’s diagnostic usefulness as “inadequate” if no abnormality was detected, “helpful” if an abnormality was detected but further assessment was necessary or “diagnostic” if a definitive diagnosis was made.
We abstracted the location of the pseudoaneurysm, its diameter and the diameter of its orifice. When pseudoaneurysm or orifice size was reported in two or three dimensions, the diameter was estimated assuming a circular area or a spherical volume. If more than one imaging study reported pseudoaneurysm or orifice dimensions, the results from the more definitive study were used. The average orifice/pseudoaneurysm diameter ratio was calculated for cases that reported both values and, because pseudoaneurysms are reported to have ratios of <0.5, we calculated the proportion of patients who met this criterion.
We also determined the percentage of patients who received a particular treatment including observation, anticoagulation and surgery. When surgery was performed, we recorded all surgical complications. When data were available, we recorded the time of survival from the diagnosis of LV pseudoaneurysm.
The median patient age was 60 years (range 2 to 88) (Table 1). More than two thirds of patients were men and three quarters were white. The most common reported symptoms or conditions were congestive heart failure (36%), chest pain (30%) and dyspnea (25%). Sudden death was the presenting symptom in 3% of cases. Approximately 12% of patients were asymptomatic at the time of diagnosis.
Murmurs and abnormalities on the electrocardiogram and chest X-ray were common. Murmurs were found in more than two thirds of patients and more than 95% of patients had electrocardiographic and chest X-ray abnormalities. ST segment elevation was noted in 20% of patients, whereas nonspecific ST segment changes were found in 75% of patients. More than half of the patients had evidence of a mass on chest X-ray and 65% had cardiomegaly.
The most common etiology of LV pseudoaneurysm was myocardial infarction (Table 2); inferior infarcts were approximately twice as common as anterior infarcts. One third of pseudoaneurysms resulted from a surgical procedure, most often mitral valve replacement.
Angiography was the more frequently used imaging modality followed by two-dimensional (2D) echocardiography (Table 3). Angiography almost always revealed some abnormality (98%) and provided a definitive diagnosis in 87% of patients. Two-dimensional, color Doppler and pulsed Doppler echocardiography showed some abnormality in approximately 85% to 90% of patients, and a definitive diagnosis was made in about one quarter to one third of cases. The etiology of the pseudoaneurysm (myocardial infarction v. surgery) did not affect whether a definitive diagnosis was provided by angiography (93% v. 90%, p = 0.57) or 2D echocardiography (28% v. 31%, p = 0.71). M-mode echocardiography was inadequate in 40% of patients and rarely provided a definitive diagnosis (7%). Transesophageal echocardiography (TEE) was diagnostic in 75% of patients, but only 16 patients underwent this test. Radionuclide and CT scans usually produced equivocal results with only 12% and 7% definitive diagnoses. In the 17 patients who underwent MRI, an abnormality was detected in all and a definitive diagnosis was provided in 9 cases (53%).
For patients who underwent an alternative imaging study and angiography, similar or improved diagnostic accuracy was obtained in 18% of patients (6 of 34) who also had radionuclide scintigraphy, 28% of patients (21 of 74) who also had 2D echocardiography, 42% (5 of 12) who also had a CT scan, 75% (6 of 8) who also had TEE and 80% (8 of 10) who also had MRI. Compared with TEE, 2D echocardiography had similar or improved accuracy in only 23% of patients (3 of 13). TEE produced results equivalent to MRI in 1 of the 2 patients who underwent both studies.
Posterior pseudoaneurysms were more than twice as common as anterior pseudoaneurysms (Table 4). Other common locations were the lateral wall, apex and inferior wall. The median LV pseudoaneurysm diameter was 6.0 cm (range 1.5 to 20.1) and the median orifice diameter was 2.0 cm (range 0.1 to 9.0). In the 92 patients with data, the median orifice/pseudoaneurysm diameter ratio was 0.23 (range 0.02 to 1.04) and was <0.5 in 75 patients (82%).
Treatment of LV pseudoaneurysms could be evaluated in 244 cases. Surgery was performed in 193 patients (80%) and the presence or absence of surgical complications was reported in 107. Death (9%), hemorrhage (7%) and arrhythmia (6%) were the complications most often reported. The recurrence of pseudoaneurysm after surgery was reported in 6 patients (5%). Stroke and infection were each reported in 3% of cases. Five patients (2%) received anticoagulation and 46 patients (19%) received only conservative therapy.
Follow-up was reported in a total of 140 patients. Of 107 patients who underwent surgery, 25 (23%) died at a median of 3 days after operation. All 82 patients who survived surgery were alive at a median of 46 weeks. Of the 31 patients who were treated conservatively, 15 (48%) died at a median of less than one week. All 16 surviving patients who were treated conservatively were alive at a median of 156 weeks. Of patients who did not undergo surgery, 12 lived at least 1 year, 5 lived at least 5 years and 2 patients lived at least 10 years. Two patients who had been treated with anticoagulation died within 2 weeks of follow-up.
We found that myocardial infarction accounted for most LV pseudoaneurysms followed by cardiac surgery, trauma, and infection. In contrast to a previous review of 67 cases in which anterior myocardial infarctions accounted for most LV pseudoaneurysms (5), we found that inferior myocardial infarctions account for approximately twice as many cases as anterior myocardial infarctions. The predilection for inferior myocardial infarction to result in LV pseudoaneurysm is consistent with the location of pseudoaneurysms on the posterior, lateral, apical or inferior surface of the left ventricle and not on the anterior surface.
In contrast to LV pseudoaneurysms, only about 4% of true LV aneurysms are located at the posterolateral or diaphragmatic surface (6). One proposed explanation for the relative lack of anterior LV pseudoaneurysms is that anterior rupture may be more likely to result in hemopericardium and death than posterior rupture (7). Because hospitalized patients usually are in the recumbent position, an inflammatory reaction of the posterior pericardium may result in pericardial adhesions and the formation of a posterior LV pseudoaneurysm rather than cardiac tamponade.
Left ventricular pseudoaneurysms usually present with symptoms, but we found that >10% of patients can be asymptomatic. Diagnosis is complicated because the most frequently reported symptoms are heart failure, chest pain and dyspnea, all of which are common in patients with coronary artery disease. In addition, patients also have nonspecific complaints such as cough, altered mental status, and dizziness that rarely elicit a concern for a LV pseudoaneurysm. Although the classic finding on physical examination is a new to-and-fro murmur (8), previous reports have noted that the murmur may be indistinguishable from mitral regurgitation (9), or absent (3). Thirty percent of patients had no detectable murmur. Electrocardiographic and chest X-ray abnormalities are present in >95% of patients; however, electrocardiographic changes are usually nonspecific and the most common chest X-ray finding is an enlarged heart; therefore, these tests are unlikely to suggest the diagnosis.
Angiography of the left ventricle and coronary arteries is considered to be the best available test for the diagnosis of LV pseudoaneurysm. Angiographic findings that help distinguish false aneurysms include a narrow orifice leading into a saccular aneurysm and the lack of surrounding coronary arteries (10). In addition, coronary angiography is usually necessary before surgery to evaluate the need for concomitant bypass grafting and may also prevent inappropriate surgical ligation of the left circumflex artery (11,12). We found that LV angiography was normal in only 3 of 197 patients (2%) who underwent the procedure and that a definitive diagnosis was made in >85% of patients. Cases that are missed may occur when the radiograph is not perpendicular to the pseudoaneurysm (resulting in overlap with the left ventricle) or when insufficient contrast is used (13–15).
Transthoracic 2D echocardiography is the only alternative imaging modality that was used in an adequate number of patients to evaluate usefulness. Although the test was usually not diagnostic, a definitive diagnosis was made in 26% of patients. Because angiography is invasive and more costly, transthoracic echocardiography is a reasonable first test. Although TEE and MRI appear to have diagnostic accuracies of ≥75% compared with angiography, the small numbers of patients who were evaluated did not allow accurate assessment of the value of these tests. In two studies with small numbers of patients, an orifice-to-pseudoaneurysm diameter ratio of <0.5 by 2D echocardiography was used to separate patients with pseudoaneurysm from those with true aneurysms (14,16). Using data from a number of imaging modalities, we found that almost 20% of patients with LV pseudoaneurysms had an orifice-to-pseudoaneurysm diameter ratio ≥0.5. This measurement alone is therefore unable to accurately distinguish true from false LV aneurysms.
MRI was first reported to diagnose LV pseudoaneurysm after myocardial infarction in 1991 (17). Loss of epicardial fat at the orifice of the pseudoaneurysm is seen on MRI (18), but distinction from true aneurysms with low signal myocardium resulting from a previous infarction may be difficult (19). Cine MRI may provide additional diagnostic capability by evaluating blood flow turbulence in the cardiac chambers, one of the hemodynamic features of the pseudoaneurysm (20). We found that in the 17 patients who underwent any MRI, an abnormality was noted in all cases, but 8 patients (47%) required further testing for definitive diagnosis.
Treatment and prognosis
Most investigators have supported surgery as the appropriate treatment for LV pseudoaneurysm (21–23)since untreated pseudoaneurysms have an approximately 30% to 45% risk of rupture (2,4,5). We found high mortality rates in patients who underwent surgery (23%), as well as in those who were treated medically (48%). Improvements in surgical technique may have decreased the perioperative mortality rate for most patients with pseudoaneurysms to ≤10% (24,25); however, patients who also require a mitral valve replacement for severe mitral regurgitation probably have a much higher risk of death (26). We found high mortality rates for patients who do not undergo surgery; however, our study also demonstrates prolonged survival in some patients who were treated conservatively. Because publication bias may have affected observed mortality rates, the true natural history of LV pseudoaneurysm cannot be determined from our study.
LV pseudoaneurysms are most often the result of myocardial infarction (particularly inferior wall myocardial infarction) and cardiac surgery. Because patients frequently present with nonspecific symptoms, a high index of suspicion is needed to make the diagnosis. Whereas electrocardiography and chest X-ray abnormalities are almost always present, they are also usually nonspecific; however, the appearance of a mass on chest X-ray film may be found in more than one half of patients and should raise the consideration of this diagnosis. Although transthoracic echocardiography is a reasonable first step, further imaging should be performed when the diagnosis is suspected despite a negative evaluation. Transesophageal echocardiography or MRI likely have a greater sensitivity than transthoracic echocardiography for detecting LV pseudoaneurysms. LV angiography remains the best test and is also useful in planning surgery. For appropriate candidates, surgery is considered the treatment of choice. Although patients with LV pseudoaneurysms have high mortality rates regardless of treatment, prolonged survival has been observed even in a few patients who do not undergo surgery.
We would like to acknowledge Joel S. Karliner, MD for his editorial assistance.
- computed tomography
- left ventricular
- magnetic-resonance imaging
- transesophageal echocardiography
- Received November 26, 1997.
- Revision received May 6, 1998.
- Accepted May 14, 1998.
- American College of Cardiology
- Vlodaver Z,
- Coe J.I,
- Edwards J.E
- Kupari M,
- Verkkala K,
- Maamies T,
- Hartel G
- Gueron M,
- Hirsch M,
- Venderman K,
- Freund H,
- Borman J