Author + information
- Received February 16, 2000
- Revision received June 7, 2000
- Accepted June 13, 2000
- Published online September 1, 2000.
- James Slater, MD, FACC∗,* (, )
- Robert J Brown, MD∗,
- Tracy A Antonelli, MPH†,
- Venu Menon, MD, FACC∗,
- Jean Boland, MD‡,
- Jacques Col, MD§,
- Vladimir Dzavik, MD∥,
- Mark Greenberg, MD, FACC¶,
- Mark Menegus, MD, FACC¶,
- Cliff Connery, MD∗,
- Judith S Hochman, MD, FACC∗,
- for the SHOCK Investigators
- ↵*Reprint requests and correspondence:
Dr. James Slater, Division of Cardiology, St. Luke’s/Roosevelt Hospital, 1111 Amsterdam Ave., New York, NY 10025
We sought to compare the characteristics and outcomes of patients with acute myocardial infarction (MI) and cardiogenic shock (CS) caused by rupture of the ventricular free wall or tamponade versus shock from other causes.
Free-wall rupture is a recognized cause of mortality in patients with acute MI. Some of these patients present subacutely, which provides an opportunity for intervention. Recognition of factors that distinguish them from the overall shock cohort would be beneficial.
The international SHOCK Trial Registry enrolled patients concurrently with the randomized SHOCK Trial. Thirty-six centers consecutively enrolled all patients with suspected CS after MI, regardless of trial eligibility.
Of the 1,048 patients studied, 28 (2.7%) had free-wall rupture or tamponade. These patients had less pulmonary edema, less diabetes, less prior MI, and less prior congestive heart failure (all p < 0.05). They more often had new Q waves in two or more leads (51.9% vs. 31.5%, p < 0.04), but MI location and time to shock onset after MI did not differ. Of patients with rupture or tamponade, 75% had pericardial effusions. No hemodynamic characteristics identified patients with rupture/tamponade. Most patients with rupture/tamponade had surgery and/or pericardiocentesis (27/28); their in-hospital survival rate was identical to that of the group overall (39.3%). Women and older patients with rupture/tamponade tended to survive intervention less often.
Free-wall rupture and tamponade may present as CS after MI, and survival after intervention is similar to that of the overall shock cohort. All patients with CS after MI should have echocardiography in order to detect subacute rupture or tamponade and initiate appropriate interventions.
Rupture of the ventricular free wall, a dramatic clinical event, is an uncommon but not rare cause of death in patients hospitalized with acute myocardial infarction (MI) (1,2). Its overall incidence is estimated to be about 6%, but it accounts for as much as 15% of the in-hospital mortality after MI (3,4). Late administration of thrombolytic therapy, large infarctions, advanced age and female gender are known risk factors for rupture (5–8).
Many patients succumb almost instantaneously with rapid, irreversible, electromechanical dissociation, but others present with a less acute clinical course, which, when recognized, allows for potentially life-saving therapeutic intervention (9). These patients often present with hypotension and other signs of cardiogenic shock (CS) (10,11). It is therefore important to attempt to distinguish free-wall rupture and tamponade from the spectrum of patients developing CS after MI.
The recent SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK? (SHOCK) Trial (12) concurrently compiled an international registry of patients who developed CS after MI. The SHOCK Trial Registry allowed us to identify patients with CS after free-wall rupture and compare their presentation, course and outcomes with those of the range of patients who develop CS in the setting of acute MI.
The SHOCK Trial Registry was initiated to ensure that all potentially eligible patients were considered for the randomized SHOCK Trial and to reduce the possibility that an enrolling center was systematically excluding a particular subgroup of patients from the Trial. The Registry was maintained at all enrolling centers and included all patients with suspected CS complicating acute MI, regardless of trial eligibility. The Registry enrolled 1,190 patients between April 1993 and September 1997. Detailed Registry methodology is being reported in this supplementary issue of the Journal(13). Institutional Review Board approval was obtained at all enrolling centers prior to commencing the recruitment.
Failure to meet all trial inclusion criteria, presentation outside of the specified time period or inability to give informed consent were reasons for enrollment in the Registry rather than the trial. Cardiogenic shock was considered to be present if all the following conditions were met: 1) systolic blood pressure (BP) persistently was <90 mm Hg, or vasopressors were required to maintain BP ≥90 mm Hg; 2) there was evidence of end-organ hypoperfusion, such as altered mental status, cold or diaphoretic extremities, or low urine output; 3) there was evidence of elevated filling pressures (for example, pulmonary congestion at physical examination or in chest radiograph or, if right-heart catheterization had been performed, a capillary wedge pressure of ≥15 mm Hg). Causes of CS other than predominant left ventricular (LV) failure were recorded and they included isolated right ventricular (RV) shock; acute, severe mitral regurgitation (MR); ventricular septal rupture; free-wall rupture or tamponade; and shock related to noncardiac causes, such as hemorrhage or sepsis.
Patients locally diagnosed with free-wall rupture or tamponade are the focus of this analysis. Patients with CS complicated by rupture of the interventricular septum or acute, severe MR (n = 142) were excluded from the dataset and are reported separately (14,15). Patients with free-wall rupture or tamponade were compared with the larger cohort of patients with CS from other causes (n = 1,020), of which primary LV failure (n = 884) formed the largest group.
The Fisher exact test was used to examine the association between categorical variables and diagnosis (presence or absence of rupture/tamponade). The means of normally distributed variables were compared by using the Student t-test. The distributions of skewed variables were compared by using the Wilcoxon rank-sum test. Multivariate modeling of mortality was not performed, because of the small sample size of the rupture/tamponade group. Descriptive statistics are expressed as mean ± SD. All p values are two-sided and considered statistically significant at p < 0.05. No adjustments were made for multiple univariate comparisons.
After excluding patients with CS secondary to septal rupture or acute, severe MR, 28 patients in the remaining cohort (n = 1,048) had rupture or tamponade (2.7%). The overall prevalence was 2.3% if all 1,190 patients in the Registry were considered. The diagnosis was based on echocardiographic findings or clinical presentation and confirmed at surgery or pericardiocentesis in 96% of the patients. Of these 28 patients, 6 were characterized as rupture alone, 9 as tamponade alone, and 13 as both rupture and tamponade.
Patients with rupture or tamponade tended to be older and more commonly were female, but these differences did not reach statistical significance (Table 1). Patients with rupture or tamponade had significantly less prior MI, congestive heart failure, diabetes and peripheral vascular disease. At physical examination, patients with rupture or tamponade less often had pulmonary edema (16.7% vs. 55.3%, p < 0.001). Patients with rupture or tamponade more often showed new Q waves in two or more leads (51.9% vs. 31.5%, p = 0.035), but there was no difference in MI location or development of ST-segment elevation after the appearance of Q waves. The time from MI onset to CS onset did not differ between the two groups (median 12.0 h [interquartile range, 3.6 to 21.0 h] for the rupture/tamponade group vs. 6.0 h [1.7 to 20.1 h] for the other patients), and similar proportions of patients received thrombolytic therapy (39.3% vs. 34.2%, p = ns). Administration of a thrombolytic agent did not appear to accelerate the time from MI onset to rupture or tamponade. In all, 75% of the patients who received thrombolytic therapy developed rupture or tamponade within 47 h after MI.
No hemodynamic variables appeared to distinguish between patients with rupture or tamponade and those without (Table 2). Systolic and diastolic BPs, heart rate, right-heart pressures and cardiac index values did not differ significantly between the two groups, but patients with rupture or tamponade tended to have higher a right atrial mean pressure and heart rate and a lower pulmonary capillary wedge pressure (p < 0.20).
Of the 28 patients with rupture or tamponade, 20 had echocardiograms, of which 15 (75%) showed a pericardial effusion. Most of these were generalized effusions (67%), and the rest localized. The location of the effusion did not correlate with MI location. Right atrial or RV collapse was described in 39% of patients, and a myocardial tear was seen in 39%.
There was no difference in the distributions of the number of diseased vessels between patients with rupture or tamponade (n = 18) and those without (n = 578) (Table 3). There were significant differences in the culprit vessel between patients with and without rupture (p = 0.033); the left anterior descending or circumflex artery was the culprit vessel more often in those with rupture or tamponade. It is interesting that no patients with rupture had the right coronary artery as the culprit vessel versus 29.5% of the patients in the no-rupture group. The proportion with TIMI grade 2 or 3 flow in the culprit vessel did not differ significantly between groups.
Overall in-hospital survival in the rupture or tamponade patients was 39.3%, which was identical to that in the 1,020 patients without this complication. One patient died 1.8 h after admission and did not undergo either pericardiocentesis or surgery. Of the six patients who had pericardiocentesis alone, 50% survived, and 38% of the 21 patients whom had surgical repair (with or without bypass surgery) survived. Survival in men was 54% (7/13) compared with 27% (4/15) in women (p = 0.246), the average age of those surviving was 66 ± 9.8 years compared with 74.6 ± 5.3 years for patients not surviving (p = 0.005). Data on the timing of pericardiocentesis or surgical repair are not available.
Rupture of the free wall of the LV after acute MI often is a catastrophic event. Its prevalence in the SHOCK Trial Registry was 2.7%, but it is impossible to know the true prevalence because many patients die immediately and the cause of death is not confirmed. Lopez-Sendon et al. (16) estimated the overall incidence of rupture at 6.2%, of which about 30% presented subacutely. In their study, 29 of 1,453 patients (2%) had subacute rupture confirmed at operation, and 94% of those patients presented with hypotension. Considering that almost all patients with subacute rupture present with CS, the Registry’s cohort of patients with this condition represents an important addition to the literature. Although rupture and tamponade were grouped for purposes of our analysis, tamponade alone may represent instances of spontaneously sealed or unrecognized rupture. Serous or hemorrhagic effusions, however, can occur without documented free-wall rupture; in the Lopez-Sendon series, 6 of 1,214 patients presented this way, which is in general agreement with the 9 of 1,048 patients seen in our analysis.
Demographics of patients with rupture
The National Registry of Myocardial Infarction found a higher mortality from cardiac rupture in women than in men (6.8% vs. 3%) after adjusting for age. In multivariate analysis of that series, thrombolytic therapy and prior MI also were found to be independent predictors of myocardial rupture (17). We found a trend toward increased myocardial rupture in women and older patients, but only age—not gender—is associated with mortality in the SHOCK Trial Registry. Administration of thrombolytics was not associated with a higher rate of rupture in our series, and prior MI was found less often in our patients with rupture than in those with CS from other causes. It is interesting to speculate whether a previous MI with scar formation or pericardial inflammation offers some degree of protection against rupture during a later infarction. Diabetes and peripheral vascular disease were also less prevalent in our patients with rupture or tamponade; it is unknown whether these conditions predispose patients to LV pump failure after MI (reflecting more extensive coronary artery disease) or whether they protect against rupture (via increased myocardial fibrosis). In a large series of patients dying of rupture after undergoing thrombolysis for MI, Becker et al. (18) also found prior MI and diabetes to be less frequent in patients with rupture than in those succumbing primarily to left ventricular failure.
Hemodynamic, echocardiographic and angiographic features
The hemodynamic profile of patients with rupture or tamponade in our series did not differ significantly from that of the larger group of patients in CS after MI. We did not collect detailed information about whether typical hemodynamic findings, such as a blunted Y descent or pulsus paradoxus, were present, but other series have shown these findings to be neither sensitive nor specific for the diagnosis of tamponade in this setting (4,19). Pulmonary edema, however, was noted in only 16.7% of our patients with rupture or tamponade, compared with 55.3% of patients without this complication. This may be a useful new clinical indicator of rupture or tamponade.
The echocardiogram has obvious utility in the diagnosis of rupture or tamponade. In the series of Lopez-Sendon, the presence of pericardial effusion >5 mm was 100% sensitive for the diagnosis of subacute ventricular-wall rupture (16). It is difficult to explain why only 75% of our patients with rupture or tamponade showed an effusion on echocardiography. Perhaps the timing of the rupture or severity of the clinical condition prevented the acquisition of technically adequate images. Alternatively, perhaps some patients who are prone to develop CS after rupture or tamponade have small effusions, which are less easily detected by echocardiography but may produce substantial hemodynamic effects.
The most interesting angiographic finding in our series was that the right coronary artery was less often the culprit vessel in patients with rupture or tamponade. Although subacute free-wall rupture has been reported after right coronary-artery occlusion, and inferior-wall location by ECG is well represented in patients with rupture, right coronary-artery occlusion more often may lead to rupture of the lower ventricular septum, with development of ventricular septal defect (20,21). The number of patients with rupture or tamponade in our series whose culprit artery was known, however, was small (12/28); this observation may represent random variation.
Clinical outcomes after rupture
The overall survival of patients with rupture or tamponade was identical to that of patients with CS secondary to primary pump failure (39.3%). All but one of the patients with rupture or tamponade had either pericardiocentesis alone or surgical evacuation and repair, which should be considered the standard of care for this condition (22–25). This overall survival rate compares favorably with the 48.5% long-term survival reported by Lopez-Sendon et al. (16). The slightly better long-term results in that series may reflect the lower mean age of their patients (67.8 vs. 71.1 years). In any event, a sizable proportion of patients with rupture or tamponade present subacutely, which provides an opportunity for diagnosis and effective treatment.
The small number of patients with cardiac rupture or tamponade prevents detailed analysis of treatment approaches or predictors of survival. Furthermore, due to the low prevalence of this condition, the tamponade cohort in this Registry is small, and thus, the primary intent of this report is descriptive rather than comparative. Accordingly, no adjustments were made for multiple univariate comparisons. The multicenter organization of the Registry precluded a uniform approach to diagnosis and treatment and, therefore, restricts our ability to determine the sensitivity and specificity of clinical findings and diagnostic methods. Because echocardiography was not performed uniformly and serially in all patients, we may have underestimated the true incidence of rupture or tamponade. Finally, we grouped rupture and tamponade for purposes of analysis. Although tamponade alone may reflect patients with spontaneously sealed rupture or hemorrhagic inflammation, it also could reflect a different, potentially more benign, etiology (26,27).
Cardiac rupture or tamponade was found in 2.7% of the patients developing CS after MI. Survival after appropriate diagnosis and treatment was no different from that of the overall cohort of patients in CS after MI; thus, rapid diagnosis is crucial if there is to be any opportunity for life-saving intervention. Diabetes, prior MI, and peripheral vascular disease were significantly less prevalent in patients developing rupture or tamponade, and pulmonary edema was found less frequently upon physical examination compared with shock patients without rupture or tamponade. Hemodynamic variables did not distinguish the two groups, but echocardiography was useful in making the diagnosis and should be obtained quickly in all patients developing hypotension after MI.
☆ Supported by RO1 grants HL50020, HL49979, 1994-99, from the National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland.
- blood pressure
- cardiogenic shock
- left ventricular, left ventricle
- myocardial infarction
- mitral regurgitation
- right ventricular, right ventricle
- SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK? (trial)
- Received February 16, 2000.
- Revision received June 7, 2000.
- Accepted June 13, 2000.
- American College of Cardiology
- Becker R.C,
- Gore J.M,
- Lambrew C,
- et al.
- Honan M.B,
- Harrell F.E Jr.,
- Reimer K.A,
- et al.
- Investigators of the Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico (GISSI-2),
- Maggioni A.P,
- Maseri A,
- Fresco C,
- et al.
- Mahoney P.G,
- Slesser B.V,
- Baignent D.F,
- Humber P.J,
- Slade P.R,
- Lawson C.W
- Balakumaran K,
- Verbaan C.J,
- Essed C.E,
- et al.
- ↵Hochman JS, Christopher BE, Sleeper LA, et al., for the SHOCK Investigators. Cardiogenic shock complicating acute myocardial infarction—etiologies, management and outcome: a report from the SHOCK Trial Registry. J Am Coll Cardiol 2000;36:1063–70.
- ↵Thompson CR, Buller CE, Sleeper LA, et al., for the SHOCK Investigators. Cardiogenic shock due to acute severe MR complicating acute myocardial infarction: a report from the SHOCK Trial Registry. J Am Coll Cardiol 2000;36:1104–9.
- Menon V, Webb JG, Hillis LD, et al., for the SHOCK Investigators. Outcome and profile of ventricular septal rupture with cardiogenic shock after myocardial infarction: a report from the SHOCK Trial Registry. J Am Coll Cardiol 2000;36:1110–6.
- Lopez-Sendon J,
- Gonzalez A,
- Lopez de Sa E,
- et al.
- Becker R.C,
- Hochman J.S,
- Cannon C.P,
- et al.
- Oliva P.B,
- Hammill S.C,
- Edwards W.D
- Galve E,
- Garcia del Castillo H,
- Evangelista A,
- Batlle J,
- Permanyer-Miralda G,
- Soler-Soler J
- Renkin J,
- de Bruyne B,
- Benit E,
- Joris J.M,
- Carlier M,
- Col J