ARTICLE IN PRESS doi:10.1510/icvts.2006.129718
Interactive CardioVascular and Thoracic Surgery 6 (2007) 30–34 www.icvts.org
Institutional report - Cardiac general
Value of echocardiography for stroke and mortality prediction following coronary artery bypass grafting Orit Klucka,1, Marius Bermanb,1,*, Alon Stamlerb, Gideon Saharb, Alexander Koganb, Eyal Poratb, Alexander Sagiea Department of Cardiology, Echocardiographic Unit, Rabin Medical Center, Beilinson Campus, Petach Tikva, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel b Department of Cardiothoracic Surgery, Rabin Medical Center, Beilinson Campus, 49100 Petach Tikva, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
a
Received 30 January 2006; received in revised form 2 October 2006; accepted 10 October 2006
Abstract There are known clinical and laboratory predictors for stroke and death following CABG. The aim of this study was to determine if transthoracic echocardiographic findings prior to CABG have additional predictive value for occurrence of perioperative cerebrovascular accident (CVA) and death. The files of patients who underwent CABG between January 2002 and November 2004, with perioperative echocardiographic assessment were reviewed. Echocardiographic variables examined included LV size, function and hypertrophy, mitral annulus calcification (MAC) and aortic valve calcification (AVC). Patients in whom post-CABG stroke or death was documented were compared with those without these endpoints. Of the 572 patients who met the study criteria, 33 (5.8%) had a neurological event and 26 (4.5%) died, four after a major stroke. One hundred and sixty-seven patients had MAC and 228 AVC. On multivariate analysis, risk factors for stroke were previous stroke (ORs2.91 CI 1.179–7.24; P-0.005), renal failure (ORs2.48 CI 1.039–5.95; P-0.001) and older age (ORs1.60 CI 0.971–2.63; P-0.001); risk factors for death were perioperative insertion of intra-aortic balloon pump (ORs33.7 CI 11.38– 100; P-0.001) and peripheral vascular disease (ORs3.89 CI 1.32–11.45; P-0.001). Medically treated dyslipidemia was protective factor. LV hypertrophy significantly predicts stroke post-CABG by univariate analysis (Ps0.02). There was no significant correlation between AVC and MAC with stroke, although death was slightly increased in patients with MAC (44% vs. 29.2%, Ps0.114). 䊚 2007 Published by European Association for Cardio-Thoracic Surgery. All rights reserved. Keywords: Risk stratification; Mitral annulus calcification; Aortic valve calcification; Stroke; Mortality; LV mass; CABG
1. Introduction Coronary artery bypass grafting (CABG) is associated with neurological complications. Nevertheless, stroke today has an incidence of 1–5% in post-CABG patients and accounts for 21% of the postoperative mortality rate. It also has an important influence on morbidity and recovery and significantly prolongs the length of in-hospital stay w1x. Since the 1960s there has been a notable decrease in the postoperative stroke rate, probably because of improvements in surgical techniques, advancement in anesthesia and oxygenation and new methods of cardioplegia w1x. Yet the number with postoperative stroke remains high, and researchers assume that these high figures are explained by the current tendency to operate even in older and complicated cases, more severe heart diseases and more comorbidity w2x. Therefore, recognizing the risk factors for stroke after CABG will help lower its frequency or prevent it altogether. The most significant predictors found in previous studies were chronic renal failure, recent myocardial infarction, hypertension, previous stroke, type 2 dia1 Orit Kluck and Marius Berman contributed equally to this work. *Corresponding author. Tel.: q972-3-9376710; fax: q972-3-9240762. E-mail address:
[email protected] (M. Berman).
䊚 2007 Published by European Association for Cardio-Thoracic Surgery
betes mellitus, age over 75 years, peripheral vascular disease, moderate to severe left ventricular dysfunction, atrial fibrillation, low ejection fraction, and operation in an emergency setting w1,3x. Another important predictor of postoperative stroke is atherosclerotic aorta. Although off-pump CABG lowers the risk of atheroemboli associated with aortic cannulation, cross-clamping and pump-related emboli, it is still associated with 0–3.3% risk of stroke w4x. Some reports demonstrated no consistent differences in the occurrence of cognitive dysfunction and stroke at several time points after on-pump versus off-pump CABG w5x. A correlation between an atherosclerotic aorta and valve calcification has been previously reported w6,7x. The aim of the present study was to determine if mitral annulus calcification (MAC) and aortic valve calcification (AVC) are predictors of post-CABG stroke. 2. Material and methods A retrospective case–control design was used. Between January 2002 and November 2004 the computerized database of the departments of cardiothoracic surgery department and echocardiograhic laboratory of a single tertiary medical center were searched for all patients who under-
ARTICLE IN PRESS O. Kluck et al. / Interactive CardioVascular and Thoracic Surgery 6 (2007) 30–34
went first-time CABG. Only patients in whom preoperative echocardiographic data were available were included. Patients who had concomitant procedures were excluded. Cases were defined as those patients who had a postoperative stroke or transient ischemic attack (TIA), or, who died. Controls were defined as all other patients in the sample. 2.1. Echocardiography Routine standard 2-D transthoracic echocardiography was done using Sonos 5000 ultrasound machines (Philips). MAC was defined as a dense, localized, highly reflective area at the base of the posterior mitral leaflet. AVC was defined by TTE as focal areas of increased echogenicity and thickening of the aortic-valve leaflets without restriction in motion. To isolate valve calcification as a risk factor, all other known predictors for post-CABG stroke were examined. 2.1.1. Statistical analysis Data were analyzed using BMDP software (1993). Continuous variables were compared between the two groups using analysis of variance and discrete variables were compared using Pearson’s x2 test or Fisher’s exact test, as appropriate. A P-value of F0.05 was considered significant. 3. Results The patient characteristics are shown in Table 1. A total of 572 patients met the study criteria, of which 33 (5.8%) had a postoperative neurological event: 21 (3.7%) CVA and 12 (2.1%) TIA. Twenty-six patients (4.5%) died in hospital, four after a massive neurological event. Death in a later course occurred 3–24 days (mean 9) after surgery, and in the other 22 cases who deceased, death occurred within 0–45 days (mean 10.85) after surgery. There were 167 (29.9%) cases of MAC and 228 (40.8%) of AVC. The mean length of in-hospital stay was 13.323"9.393 days for patients with a neurological event, and 6.609"5.312 days for the control group. 3.1. Predictors of stroke Several factors were associated with a significant increase in the risk of stroke (Table 2): older age, chronic renal failure, previous stroke, and postoperative paroxysmal atrial fibrillation (PAF). Among the echocardiographic factors (Table 3), there was no association of MAC and AVC with stroke. Left ventricular hypertrophy, calculated as the mean of the sum of the intraventricular septum width and posterior wall thickness (Ps0.03), was associated with postoperative stroke, and low left ventricular function showed a tendency for predicting stroke. 3.2. Predictors of mortality Factors found to be predictors of death (Table 4) were older age, chronic renal failure, peripheral vascular disease and PAF. Some surgery-related factors were also highly predictive, including high Parsonnet score, postoperative PAF, insertion of an intra-aortic balloon pump, and emergency operation (P-0.0001) for all.
31
Table 1 Patient characteristics of 572 patients undergoing CABG Preoperative characteristics: Age (year) mean"S.D. Sex (male) Diabetes Hypertension Chronic renal failure Dyslipidemia Dyslipidemic – medically treated No dyslipidemia but medically treated Smoker Perioperative smoker Past smokers Previous myocardial infarction Carotid artery disease Previous CVA Peripheral vascular disease Preoperative arrhythmia NSR PAFyCAF Parsonnet risk score, mean Emergency operation Mitral annulus calcification Aortic valve calcification Aortic atheroma LVFS, mean LV dysfunction Normal-mild Mild-moderate Moderate-severe Regional wall abnormality LA thrombus Inter-ventricular septum, mmHg (mean) Posterior wall thickness mmHg (mean) Pulmonary pressure mmHg (mean) Operative characteristics: Intraaortic balloon pump Preoperative Intraoperative Postoperative On-pump CABG Postoperative characteristics: Arrhythmia NSR PAFyCAF VF Neurological events CVA TIA Mortality
66.72"10.35 393 (68.7%) 245 (43.1%) 373 (74.7%) 75 (13.3%) 271 (55.1%) 73 (14.8%) 96 92 329 102 60 137
(26.5%) (25.4%) (62.1%) (68.4%) (11.7%) (26.9%)
399 (90.5%) 24 (5.4%) 15.88 66 (11.5%) 167 (29.9%) 228 (40.8%) 39 (78%) 34.049 357 (64%) 125 (22.4%) 76 (13.7%) 329 (58.4%) 10 (1.8%) 11.02 10.48 39.70
43 8 10 525
(7.5%) (1.4%) (1.7%) (92.8%)
409 (77.9%) 102 (19.5%) 7 (1.3%) 3.7% (21) 2.1% (12) 26 (4.5%)
NSR – normal sinus rhythm, PAF – paroxysmal atrial fibrillation, CAF – chronic atrial fibrillation, LVFS – left ventricular fraction shortening, VF – ventricular fibrillation, CVA – cerebro-vascular accident, TIA – transient ischemic attack.
Among the echocardiographic factors (Table 5), moderate to severe left ventricular dysfunction was a strong predictor of mortality (P-0.0001). The presence of MAC or AVC had no significant influence on postoperative mortality. Multivariate analysis risk factors for stroke and mortality are represented in Table 6. 4. Discussion Neurological complications remain a major cause of postCABG morbidity and mortality. In our sample, 3.7% of the patients had a postoperative CVA and 2.1%, a TIA. Although inclusion of the TIAs may have led to an overestimation of
ARTICLE IN PRESS 32
O. Kluck et al. / Interactive CardioVascular and Thoracic Surgery 6 (2007) 30–34
Table 2 Clinical predictors for postoperative stroke by univariate analysis
Preoperative characteristics: Age (year) mean Male sex Diabetes Hypertension Chronic renal failure Dyslipidemia Dyslipidemia-medically treated No dyslipidemia-statin treatment Smoker Perioperative smoker Past smoker Previous myocardial infarction Carotid artery disease Previous CVA Peripheral vascular disease Preoperative arrhythmia NSR PAFyCAF Parsonnet risk score, mean Emergency operation Intraoperative characteristics: On-pump operation Intraaortic balloon pump Preoperative Intraoperative Postoperative Postoperative characteristics: Length of stay, mean Postoperative arrhythmia NSR PAFyCAF VF
P-value
Neurological events (ns33)
No neurological events (ns539)
n
n
%
72.7 54.5 81.8 33.3
66.152 369 227 346 64
68.5 42.4 74.2 12.1
9 4
31 13.8
262 69
56.6 14.9
0.0078
4 8 22 6 9 7
20 40 71 66.6 27.3 21.9
92 84 307 96 53 130
26.9 24.6 61.5 68.5 11.1 27.2
0.3015
24 2 18.726 2
88.9 7.4
375 22 15.697 64
90.6 5.3 – 11.9
0.6200 0.0870 0.3103
492
92.3
0.0981
7.4 1.1 1.7
0.1065
%
70.364 24 18 27 11
33
– 6.1 100
3 2 1 13.323 19 11 2
9.1 6.1 3.0
40 6 9 6.609
– 59.4 34.4 6.3
390 91 5
–
0.0238 0.6078 0.1726 0.3334 0.0005
0.2930 0.9823 0.0059 0.5108
-0.0001
79.1 18.4 1.0
0.0156
NSR – normal sinus rhythm, PAF – paroxysmal atrial fibrillation, CAF – chronic atrial fibrillation, LVFS – left ventricular fraction shortening, VF – ventricular fibrillation, CVA – cerebro-vascular accident, TIA – transient ischemic attack.
examination and to determine, in particular, whether aortic and mitral valve calcifications are major risk factors. We found that older age, chronic renal failure, previous stroke and PAF were predictors of postoperative stroke, and that
the neurological events, we were limited by a relatively small number of patients who had a full stroke. The aim of our study was to identify possible predictors of stroke and death, mainly from the echocardiographic Table 3 Echocardiographic predictors for postoperative stroke by univariate analysis
Mitral annulus calcification Aortic valve calcification Aortic atheroma LV dysfunction Normal-mild Mild-moderate Moderate-severe Regional wall abnormality LA thrombus LVFS (%) Interventricular septum (mm) mean Posterior wall thickness (mm) mean (PWTqIVS)y2(mm)* Pulmonary pressure mean**
P-value
Neurological events (ns33)
No neurological events (ns539)
n
%
n
%
10 17 3
33.3 54.8 100
157 211 36
29.7 40.0 76.6
339 118 71 306 10 34.231 10.985 10.441 10.672 39.498
64.2 22.3 13.4 57.6 1.9 – – – – –
18 7 5 23 0 29.850 11.739 11.304 11.522 44.273
60.0 23.3 16.7 71.9 0 – – – – –
*(PWTqIVS)y2 is an indicator of left ventricular hypertrophy. **Pulmonary pressuresJVPq10 mmHg.
0.6705 0.1014 0.3427
0.8578 0.1122 0.6178 0.0567 0.0681 0.0227 0.0320 0.1349
ARTICLE IN PRESS O. Kluck et al. / Interactive CardioVascular and Thoracic Surgery 6 (2007) 30–34 Table 4 Clinical predictors for postoperative mortality by univariate analysis
Preoperative characteristics: Age (year) mean Male sex Diabetes mellitus Hypertension Chronic renal failure Dyslipidemia Dyslipidemia-medically treated No dyslipidemia-statin treatment Smoker Perioperative smoker Past smoker Previous myocardial infarction Carotid artery disease Previous CVA Peripheral vascular disease Preoperative arrhythmia NSR PAFyCAF Parsonnet risk score (mean) Emergency operation Intraoperative characteristics: On-pump operation Intraaortic balloon pump Preoperative Intraoperative Postoperative Postoperative characteristics: Length of stay, mean Postoperative arrhythmia NSR PAFyCAF VF
Mortality (ns26)
Control group (ns546)
n
%
n
73.1 52.0 87.5 28.0
66.185 374 232 359 68
70.808 19 13 14 7
Table 5 Echocardiographic predictors of postoperative mortality by univariate analysis
P-value
%
68.5 42.7 74.3 12.6
0.0267 0.6228 0.3599 0.2328 0.0268
9 1
64.3 262 7.1 72
54.8 15.1
0.6678
4 2 18 1 5 10
44.4 92 22.2 90 81.8 311 33.3 101 25.0 57 50.0 127
26.1 25.5 61.2 69.2 11.6 25.9
0.4557 0.0512 0.3153 0.1661 0.0172
8 61.5 391 91.4 0.0011 2 15.4 22 5.1 30.420 – 15.133 – -0.0001 11 42.3 55 10.1 -0.0001 22
88.0 503
11 3 1
42.3 11.5 3.0
10.560 2 5 4
–
93.0
0.3480
32 5 9
5.9 0.9 -0.0001 1.7
6.822
–
16.7 407 41.6 97 33.3 3
33
0.0016
79.3 18.9 -0.0001 0.6
NSR – normal sinus rhythm, PAF – paroxysmal atrial fibrillation, CAF – chronic atrial fibrillation, VF – ventricular fibrillation, CVA – cerebro-vascular accident, TIA – transient ischemic attack.
previous stroke, chronic renal failure, and older age were predictors of death. Valve calcification did not seem to have any influence on either stroke or death, although left ventricular hypertrophy, easily measured by echocardiography, was considered as a strong predictor of stroke. Previous studies have described the relationship between MAC and AVC, and between valve calcification and aortic atheromas w6x. An association was also reported between MAC and cerebral stroke w8x, and between aortic atheromas and postoperative stroke w9x. These findings led to the assumption that left-sided valve or annular calcification may be a predictor for postoperative stroke. However, no statistical evidence to support this theorem was noted in the present study. Our results may be explained by a true absence of a relationship between MACyAVC and post-CABG stroke or, a too low incidence of stroke in the study population to achieve a statistical significance. The risk factors identified in our analysis are in line with earlier studies. Previous cerebral stroke (ORs2.91) has also been identified by several other groups as a strong predictor of postoperative stroke w1,10x. Twelve of our patients had a carotid endarterectomy in the past; none of them had a postoperative stroke. Therefore, we recom-
Mitral annulus calcification Aortic valve calcification Aortic atheroma LV dysfunction Normal-mild Mild-moderate Moderate-severe Regional wall abnormality LA thrombus LVFS (%) Inter-ventricular septum (mm) mean Posterior wall thickness (mm) mean (PWTqIVS)y2 (mean)* Pulmonary pressure (mean)**
Mortality (ns26)
Control group (ns546)
n
%
n
%
11 11 3
44.0 44.0 75.0
156 217 36
29.2 40.6 78.3
8 3 13 13 0 29.61 11.38 10.61 11.00 43.63
33.0 349 12.5 122 54.16 63 56.5 316 0 10 – 34.17 – 11.01 – 10.47 – 10.70 – 39.52
P-value
0.114 0.738 0.880
65.3 22.8 -0.05 11.8 58.5 0.849 1.9 0.692 – 0.107 – 0.492 – 0.783 – 0.571 – 0.198
*(PWTqIVS)y2 is an indicator of left ventricular hypertrophy. **Pulmonary pressuresJVPq10 mmHg.
mend that prior to CABG, patients undergo carotid artery evaluation according to the guidelines. In the case of diseased arteries, physicians should consider carotid treatment first or combined surgery. The strong correlation shown here between chronic renal failure and postoperative stroke (ORs2.48) agrees with previous studies showing that an elevated creatinine level is a strong predictor for cerebral stroke. In addition, left ventricular hypertrophy was found in previous studies to predict cerebral stroke and asymptomatic brain damage (not postoperative) w11x. We found that it also predicts stroke after surgery (Ps0.032). The size of the older population continues to grow, and CABG is being applied more and more in older patients. Although CABG can provide them with a good quality of life, clinicians must be aware that older age is also a major risk factor for postoperative complications and stroke (ORs1.60 for each 10 years of age). Therefore, careful patient selection is imperative. PAF is a common complication of a CABG operation, affecting about 30% of patients. Its incidence rises with an increasing rate of myocardial ischemia during the operation and depends on the method of cardioplegia. In our group, 34.4% of patients with stroke had PAF, compared with 18.4% Table 6 Significant risk factors for postoperative stroke and mortality in multivariate analysis Risk factor Stroke Past stroke Renal failure Age)70 Mortality IABP PVD Urgent operation Parsonnet)25
Odds ratio
P-value
CI
2.91 2.48 1.60
1.179–7.24 1.039–5.95 0.971–2.63
-0.005 -0.001 -0.001
33.7 3.89 3.18 1.10
11.38–100 1.32–11.45 1.03–9.80 1.05–1.16
-0.001 -0.001 -0.001 -0.001
ARTICLE IN PRESS 34
O. Kluck et al. / Interactive CardioVascular and Thoracic Surgery 6 (2007) 30–34
in the control group (Ps0.0156). Although PAF is usually not considered lethal, it can cause hemodynamic instability, thromboembolic events, and patient anxiety. The finding that pharmacologically treated dyslipidemia was a protective factor against postoperative stroke (ORs2.71), emphasizes the importance of treating this condition. It also agrees with many clinical trials that have shown the effectiveness of treating high cholesterol levels in lowering the number of cardiovascular events, including stroke. Regarding the mortality risk factors, older age, CRF, PVD, preoperative PAF and postoperative arrhythmias were found as factors that predict death, although most of the patients who died had PAF (41.6% compared to 18.9% in the control group), ventricular fibrillation was much more significant, leading to death in four patients (33.3%) in the study group compared to only three controls (0.6%). The Parsonnet score is based on the calculation of many risk factors for death w12x. We found that a high Parsonnet score was correlated with high mortality (P-0.0001), supporting the use of this score to calculate individual patient operative mortality risk. Insertion of an intra-aortic balloon pump (IABP) and emergency operation were also associated with a higher probability of death. It is very likely that these are not direct causes of death, but rather factors that signify poor medical status. The finding of moderate to severe left ventricular dysfunction as a very important predictor of death (P-0.0001), agrees with earlier studies and raises the question of whether these patients require different management before, during and after surgery. Another echocardiographic parameter, the LV mass and hypertrophy, was previously associated with increased risk for sudden death w13x. The correlation between heart weight and severity of coronary artery disease in sudden death victims is debatable. Left ventricular hypertrophy reduces coronary flow reserve while increasing myocardial oxygen consumption. This imbalance may predispose to ischemia, arrhythmias and sudden death w14x. Coronary blood supply also may be impaired by atherosclerosis in persons with LV hypertrophy because some factors associated with myocardial hypertrophy are atherogenic. On the contrary, Cooper et al. w15x found that increased LV mass predicts subsequent mortality more strongly in patients without angiographic evidence of obstructive coronary artery disease than in those with stenosis of the epicardial coronary arteries. These findings have been interpreted to suggest that LV mass reflects the integrated adverse effects on the heart of increased hemodynamic load and vascular damage. In our study, we documented a significant association between LH hypertrophy parameters and stroke. It may suggest that hypertrophic hearts adapted to hypertensive pattern, reflects diffuse vascular alterations, intra- and extracranial. Although our study is limited by its retrospective design and relatively few cases of stroke, the importance of
recognizing the predictors for stroke, which has important effects on morbidity and mortality, cannot be overemphasized. Surgeons must take patient characteristics and risk factors into consideration, and decide accordingly, whether to operate or not and what operative technique to use. 5. Conclusion The independent risk factors for post-CABG stroke are previous CVA, RF and older age. MAC and AVC are apparently not independent risk predictors associated with either a postoperative neurological event or death, though the mortality rate slightly increased by the presence of MAC. LV hypertrophy suggests increase risk for post-CABG stroke. Further investigation is needed to determine other possible echocardiographic risk factors for stroke in this setting. References w1x Ascione R, Reeves BC, Chamberlain MH, Ghosh AK, Lima Kh, Angelini GD. Predictors of stroke in the modern era of coronary artery bypass grafting: a case control study. Ann Thorac Surg 2002;74:474–480. w2x Newman SP, Harrison MJ. Coronary artery bypass surgery and the brain: persisting concerns. Lancet Neurol 2002 Jun;1:119–125. w3x Stamou SC, Hill PC, Dangas G, Pfister AS, Boyce SW, Dullum MK, Bafi AS, Corso PJ. Stroke after coronary artery bypass – incidence, predictors and clinical outcome. Stroke 2001;32:1508–1513. w4x Lev-Ran O, Braunstein R, Sharony R, Kramer A, Paz Y, Mohr R, Uretzky G. No-touch aorta off-pump coronary surgery: the effect on stroke. J Thorac Cardiovasc Surg 2005;129:307–313. w5x Floyd T, Fleisher LA. Off-pump coronary artery bypass and the hypothesis from which it grew: is it yet to be tested? What are the downsides of the lingering questions? Anesthesiology 2005 Jan;102:3–5. w6x Adler Y, Fink N, Spector D, Wiser I, Sagie A. Mitral annulus calcification – a window to diffuse atherosclerosis of the vascular system. Atherosclerosis 2001 Mar;155:3–8. w7x Weisenberg D, Sahar Y, Sahar G, Shapira Y, Iakobishvili Z, Vidne BA, Sagie A. Atherosclerosis of the aorta is common in patients with severe aortic stenosis: an intraoperative transesophageal echocardiographic study. J Thorac Cardiovasc Surg 2005 July;130:29–32. w8x Adler Y, Koren A, Fink N, Tanne D, Fusman R, Assali A, Yahav J, Zelikovski A, Sagie A. Association between mitral annulus calcification and carotid atherosclerotic disease. Stroke 1998;29:1833–1837. w9x Trehan N, Mishra M, Kasliwal RR, Mishra A. Reduced neurological injury during CABG in patients with mobile aortic atheromas: a five-year follow up study. Ann Thorac Surg 2000 Nov;70:1558–1564. w10x Bucerius J, Gummert JF, Borger MA, Walther T, Doll M, Onnasch JF, Metz S, Falk V, Mohr FW. Stroke after cardiac surgery: a risk factor analysis of 16,184 consecutive adult patients. Ann Thorac Surg 2003;75:472–478. w11x Selvetella G, Notte A, Maffei A, Calistri V, Scamardella V, Frati G, Trimarco B, Colonnese C, Lembo G. Left ventricular hypertrophy is associated with asymptomatic cerebral damage in hypertensive patients. Stroke 2003;34:1766–1770. w12x Bernstein AD, Parsonnet V. Bedside estimation of risk as an aid for decision-making in cardiac surgery. Ann Thorac Surg 2000;69:823–828. w13x Haider AW, Larson MG, Benjamin EJ, Levy D. Increased left ventricular mass and hypertrophy are associated with increased risk for sudden death. J Am Coll Cardiol 1998;32:1454–1459. w14x Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WB. Prognostic implication of echocardiographically determined left ventricular mass in the Farmingham Heart Study. N Engl J Med 1990;322:1561–1566. w15x Cooper RS, Simmons BE, Castaner A, Santhanam V, Ghali J, Mar M. Left ventricular hypertrophy is associated with worse survival independent of ventricular function and number of coronary arteries severely narrowed. Am J Cardiol 1990;65:441–445.
