Supplemental Digital Content

Cardiac and central vascular functional alterations in the acute phase of Aneurysmal Subarachnoid Hemorrhage

JohnPapanikolaou1, DemosthenesMakris1, DimitriosKarakitsos2, Theodosios Saranteas3, AndreasKarabinis4, Georgia Kostopanagiotou3, Epaminondas Zakynthinos1

1Department of Critical Care, School of Medicine, University of Thessaly, University Hospital of Larissa, Thessaly, Greece

2Department of Critical Care, General State Hospital of Athens, Athens, Greece

32nd Department of Anaesthesiology, School of Medicine, University of Athens, University Hospital of Athens ‘Attikon’, Athens, Greece

4Onassis Cardiac Surgery Centre, Athens, Greece

Materials and Methods


Neurological functional outcome was evaluated according to the Glasgow Outcome Scale (GOS) at discharge from ICU(E1); GOS 1-3 (death or severe dependence) was considered as poor neurological outcome (PNO) and GOS 4-5 (independence) as good outcome (E2).

Cardiovascular assessment – Ultrasound studies

Cardiac and vascular ultrasound examinations were performed by specialized doctors (JP, MD, Cardiologist-Intensivist- University Hospital of Thessaly, Greece, official national qualification, six-year experience in transthoracic and transesophageal echocardiography) and TS (MD, PhD, Anaesthesiologist-Intensivist, anesthetic consultant, General State Hospital of Athens, Greece, accredited by the European Society of Echocardiography). Ultrasound examinations were performed by same investigators at each centre.Echocardiographywas performed from the apical 4-chamber view with an ultrasound device (Philips, XD11 XE, Andover, MA, U.S.A), equipped with the tissue Doppler imaging program and a phased array multifrequency transducer (E3). Echocardiographic studies were performed while patients were sedated and mechanically ventilated (tidal volume 8ml/kg) on supine position. Patients wereeuvolemic during examination based on clinical criteria and pulse pressure variation (ΔPP) monitoring, as previously described (E4-E5).Transesophageaelechocardiography (TEE)including transgastric views was used in patients who were not applicable for transthoracic echocardiography (TTE) (mainly due to poor acoustic window).

Two dimensional (2-D) images and Doppler echocardiographic signals were recorded along with the electrocardiogram and respiratory cycle waveform at a sweep speed of 100 mm/s and were stored digitally in the hardware for later analysis. Off-line echocardiographic analysiswas carried out by an experienced cardiologist (EZ, MD, PhD Cardiologist-Intensivist, professor of Critical Care Medicine of University Hospital of Thessaly, accredited and awarded by the British Society of Echocardiography) blinded to patients’ identity. Doppler echocardiographic variables were measured at end-expiratory phase, and the average of 3 end-expiratory cycles was used (E6).

LV 2-D analysis

The LV ejection fraction (LVEF) was measured using the Simpson’s biplane method of disks (E3)and a LVEF≤50% was considered abnormal (E7). The American Society of Echocardiography 16-segment model and regional wall motion score index (WMSI) (E3) were used for the analysis of regional wall motion abnormalities (RWMA). Each LV segment was graded as normal (score = 1), hypokinetic (score = 2), or akinetic/dyskinetic (score = 3), and WMSI was calculated by averaging the sum of the 16 segments.

Doppler echocardiographic LV function assessment

Mitral inflow curveswere recorded from the apical 4-chamber view, as previously described[E3].The peak Doppler velocitiesof early (E) and late diastolic flow (A), the E-wave deceleration time(DTE) and the E/A ratio were measured. Several mitral inflow velocity variables (short DTE, increased E/A ratios) have been reported to correlate well with advanced LV diastolic dysfunction and increased LV filling pressures; this association has been outlined in patients with impaired LV systolic function. However, in patients with normal LV systolic function mitral variables are poorly associated with central hemodynamics (E6).

Tissue Doppler Imaging (TDI) myocardial velocities were obtained from the apical 4-chamber view, as suggested [E3, E6].A 1.5-mm sample volume was placed at the lateral corner of mitral annulus and peak systolic (Sm), early diastolic (Em) and late diastolic (Am) velocities were recorded. Analysis was also performed for Em/Am as well as for E/Em ratios. TDI variables also reflect LV diastolic function. Accordingly, decreased Em velocities and Em/Am ratios are also associated with LV diastolic dysfunction.

An increased E/Em ratio, which is a combined mitral inflow and TDI Doppler index, has been reported to be a strong indicator of elevated LV filling pressures [E6], even in mechanically ventilated patients (E8-E9). However, the clinical utility of TDI signals (E/Em and Em/Am) in predicting LV diastolic dysfunction may be hampered by several factors, such as young age, mitral regurgitation, mitral annular calcification, structural myocardial or pericardial disease. Nevertheless, TDI variables correlate better with LV filling pressures and invasive indices of LV stiffness than mitral inflow variables (E-wave DT, E/A velocity ratio), as the latter are associated poorly with central hemodynamics in patients with normal LV systolic function. [E6].

Peak systolic velocity (Sm) has been reported to correlate well and to be a surrogate of LVEF(E10), especially in those patients with poor image quality, where quantitative assessment of LVEF is technically difficult; it is also significantly decreased in subendocardial ischemia as well as in diastolic heart failure even in patients with normal LV ejection fraction (E11-E12).

Assessment of aortic stiffness

Aortic stiffness was assessed by determining the carotid-femoral PWV by using transcutaneous Doppler flow recordings and the foot-to-foot method, as previously described (E13). An Acuson Aspen (Siemens, Mountainview, CA, USA) with a 7 to 10 MHz linear transducer was used for this purpose and the pulse wave sample gate was positioned at 1 to 2cm proximally to the bifurcation of the right common carotid and the right femoral artery. Electrocardiogram (ECG) and Doppler signals were recorded simultaneously at a sweep speed of 100mm/s. The R-wave on ECG was used as a timing reference to determine the time intervals to the upstroke (foot) of the Doppler waveforms. The time intervals were averaged over five cardiac cycles for both carotid and femoral signals. The "carotid-femoral pulse transit time (T)" was calculated by the subtraction of carotid from the femoral mean time interval. The distance between the sternal notch and femoral artery scanning site was measured by a tape measure and was used as the path length (D)travelled by the flow wave in this above mentioned time interval T. PWV was calculated from D and T(PWV= D/T) (E13). Echocardiography and vascular ultrasound were performed during the same examination session for each individual subject.

PWV/LVEF ratio

PWV/LVEF ratio is a combined sonographic index we introduced to incorporate both aortic stiffness and LV systolic dysfunction in the same parameter (e.g. a patient with PWV=10m/sec and LVEF=0.5 has a PWV/LVEF ratio of 20m/sec).

Inter-observer reproducibility in ultrasound measurements

Reproducibility of measurements was evaluated in 10 randomly selected ICU patients at each centre in a preliminary phase. We compared evaluations of different echocardiographic items by two independent observers or by the same observer by using both Bland-Altman analyses [%difference vs. average] and Pearson bivariate two-tailed correlations. Interobserver variability [bias of agreement (95% limits of agreement)] between two pairs of observers (EZ/TS and EZ/JP) in the two centers were: for LV ejection fraction (LVEF), 0.35 (-19.4,20), r=0.94, P<0.001 and -0.43 (-14.7,13.8), r=0.94, P<0.001, respectively; for pulsed-Doppler mitral inflow velocities, -1.28 (-25,23), r=0.83, P<0.001 and 0.97 (-17.8,19.8), r=0.85, P<0.001; for tissue Doppler Imaging (TDI) velocities at the lateral mitral annulus -4.7 (-26,17), r=0.95, P<0.001 and -2.7 (-30,24.5), r=0.91, P<0.001, respectively. Interobserver variability for aortic pulse wave velocity was 1.32 (-8.1, 10.7), r=0.94, P<0.001; Intraobserver variability was overall -1.39 (-13.2,10.4), r=0.98, P< 0.001.


There were no statistically significant differences between the two centres in terms of baseline clinical data according to age (42.6±2.9 vs. 42.8±3.3, P=0.97), female sex (57% vs. 69%, P=0.49), APACHE II score (20.5±0.9 vs. 20±1, P=0.72), SOFA score (6.9±0.4 vs. 6.3±0.3, P=0.35) , aneurysm’s location (P=0.73-0.96), Hunt-Hess grades>III(62% vs. 63%, P=1.0), Fisher CT grades>II (52% vs. 75%, P=0.19).

There were also no differences either in risk factors for ASAH, such as hypertension (24% vs. 19%, P=0.72), smoking (52% vs. 31%, P=0.21) and alcohol abuse (10% vs. 19%, P=0.43), or in interventional therapies performed, such as surgical clipping (33% vs. 50%, P=0.32), endovascular coiling (24% vs. 13%, P=0.4) and decompressive craniectomy (67% vs. 63%, P=0.8).In addition, there were no significant differences between the two centres according to ASAH outcomes, such as intra-ICU death (29% vs. 25%, P=0.82), severe dependence (GOS 2-3) (30% vs. 50%, P=0.19), and delayed cerebral infarctions (25% vs. 44%, P=0.25) at ICU discharge.

ECHO indices assessed in this study were similar between the two centers (P=0.79), either measured during the acute ASAH (P level between 0.15-0.99) or at stable state (P level between 0.17-0.91); the only exception was the measurements of E-wave DT at stable stage, P=0.04.

Cardiovascular ultrasonography and serum biomarkers in the acute phase of ASAH

i) LV function

Among patients with LVEF≤50%, a) two patients (2/12, 17%) demonstrated the classic phenotypic form of Tako Tsubo Cardiomyopathy (apical ballooning syndrome) (E14), b) seven patients (7/12, 58%) manifested the apex-sparing pattern of LV systolic dysfunction (where RWMAs were predominately observed at the basal and mid-ventricular LV walls) (E7, E15), and c) three patients (3/12, 25%) showed extensive LV circumferential dyskinesia-akinesia with a hyperkinetic apex and severe mitral regurgitation (inverted Tako Tsubo Cardiomyopathy) (E16); these three patients demonstrated severe hemodynamic compromise (fatal in one case, in which post-mortem analysis revealed normal coronary arteries and no evidence of myocarditis).

Left ventricular systolic function was found significantly impaired in ASAH patients with more severe initial bleeding. LVEF and TDI-derived peak systolic velocity (Sm) at the lateral border of the mitral annulus were significantly decreased in patients with Hunt and Hess grading>III (53.1±2.92 vs. 63.57±2.67%, P=0.02 and 7.78±0.47 vs. 11.52±0.71cm/sec, P<0.001, respectively), as well as in patients with Fisher grade>II (51.9±2.8 vs. 65.5±2.2%, P=0.002 and 8.19±0.6 vs. 10.84±0.7cm/sec, P=0.008, respectively) compared to the remainders (Hunt and Hess grading≤ III, Fisher grade≤ II, respectively). In addition, statistically significant differences between the distribution of ruptured aneurysms in terms of baseline LVEF values were found (ANOVA, P= 0.001); notably, ruptured aneurysms of the ACoA were characterized by lower LVEFs comparing to patients with MCA (-14.2±4, P= 0.008) or ACA (-24.1±7.3, P= 0.007) aneurysms (Bonferoni post hoc test). Finally, patients presenting impaired LV wall motion (WMSI>1) demonstrated higher SOFA scores upon admission (7.64±0.5 vs. 6±0.27, P=0.003).

ii) Aortic stiffness

PWV values on admission were 8.1±0.37 m/sec. Twenty three patients (62%) were found to present supra-normal PWV value compared to normal expected value (E17). PWV was significantly associated with both Fisher and Hunt and Hess severity grading. Patients with severe ASAH (Hunt-Hess grade>III or Fisher grade of >II) presented higher PWV(8.76±0.44 vs. 7.14±0.57m/sec, P=0.03 and 8.7±0.43 vs. 7.23±0.61m/sec, P=0.05, respectively). PWV values were significantly different in respect of the site of the ruptured aneurysm (ANOVA, P=0.017); patients with anterior communicating artery aneurysms presented with depressed aortic elasticity and significantly higher PWV (compared to patients with middle cerebral artery aneurysms (2.1±0.72, P= 0.018, Bonferoni post hoc test). Finally, patients who presented increased PWV values (N=23) demonstrated higher SOFA scores upon admission (7.1±0.36 vs. 5.86±0.4, P=0.033).



In our cohort, BNP levelson admissioncorrelated with LV systolic dysfunction indices (LVEF, TDI-derived Sm and WMSI) and aortic stiffness (PWV values) but not either with diastolic function surrogates (E/Em ratio)or the severity of ASAH (Fisher scale). This might indicate that BNP at this stage is primarily of cardiac than cerebral origin (E18). Another hypothesis may be that the trigger for BNP cardiac production in acute ASAH is increased LV systolic wall stress (LV afterload) due to acute aortic stiffeningrather than elevated LV end-diastolic pressure(E19). BNP levels on admission were found to be related with ASAH outcomein our series (E2,E20-21) (see Figure 4 on main text). Despite that BNP has been reported to lack specificity in ICU, as it may be influenced by several clinical conditions, particularly sepsis (E22), BNP>164pg/mL early after ASAH manifested a specificity of 92% and a positive predicting value of 95% for predicting poor neurological outcome.

PWV/LVEF index

In addition to established echocardiographic markers of ASAH outcome (such as LVEF and WMSI) (E2, E23-24), our study introduces novel early ultrasound indicators (PWV, Sm, Em/Am), which may possibly have a role in the prediction of DCI or poor neurological functional status (Table 5). Furthermore, PWV/LVEF ratio (a combined sonographic index we introduced in order to incorporate in the same parameter both aortic stiffness and LV systolic dysfunction)>16.3m/sec manifested the greatest diagnostic performance in predicting DCI (AUC=0.9) and positive predictive value of 100% in predicting death or severe dependency.

Cardiovascular alterations and acute brain injury early in ASAH

There is increasing evidence that acute brain injury following aneurysm rupture may significantly contribute to patient outcomes, by causing acute development of cerebral edema, oxidative stress, inflammation, apoptosis, and infarction (E25). Acute cardiovascular alterations and BNP over-secretion may reduce cardiac output and plasma volume, impairing cerebral blood flow; however, whether these changes are associated with DCI and neurological outcomes through the pathway of the development of ASAH-induced acute brain injury or not remains to be elucidated in the future.


E1.Jennett B, Bond M: Assessment of outcome after severe brain damage. Lancet. 1975;1(7905):480-484.

E2.van der Bilt IA, Hasan D, Vandertop WP, et al: Impact of cardiac complications on outcome after aneurysmal subarachnoid hemorrhage: a meta-analysis.Neurology. 2009;72(7):635-642.

E3.Schiller NB, Shah PM, Crawford M, et al: Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr 1989; 2(5):358-367

E4.Michard F, Chemla D, Richard C, et al:Clinical use of respiratory changes in arterial pulse pressure to monitor the hemodynamic effects of PEEP. Am J Respir Crit Care Med. 1999;159:935–939.

E5.De Backer D, Heenen S, Piagnerelli M, et al: Pulse pressure variations to predict fluid responsiveness: influence of tidal volume. Intensive Care Med. 2005;31(4):517-523.

E6.NaguehSF, AppletonCP, GillebertTC, et al:Recommendations for the evaluation of left ventricular diastolic function by echocardiography. Eur J Echocardiogr 2009; 10(2):165-193

E7. Banki N, Kopelnik A, Tung P, et al:Prospective analysis of prevalence, distribution, and rate of recovery of left ventricular systolic dysfunction in patients with subarachnoid hemorrhage. J Neurosurg. 2006;105:15–20.)

E8.Vignon P, AitHssain A, Franηois B, et al:Echocardiographic assessment of pulmonary artery occlusion pressure in ventilated patients: a transoesophageal study. Crit Care 2008;12(1):R18.

E9.Combes A, Arnoult F, Trouillet JL:Tissue Doppler imaging estimation of pulmonary artery occlusion pressure in ICU patients. Intensive Care Med. 2004;30(1):75-81.

E10. Yuda S, Inaba Y, Fujii S, et al: Assessment of left ventricular ejection fraction using long-axis systolic function is independent of image quality: a study of tissue Doppler imaging and m-mode echocardiography. Echocardiography. 2006; 23(10):846-852.

E11.Vinereanu D, Nicolaides E, Tweddel AC, et al:"Pure" diastolic dysfunction is associated with long-axis systolic dysfunction. Implications for the diagnosis and classification of heart failure. Eur J Heart Fail 2005; 7(5):820-828.

E12.YipG, WangM, ZhangY,et al:Left ventricular long axis function in diastolic heart failure is reduced in both diastole and systole: time for a redefinition? Heart2002; 87(2):121-125

E13.Jiang B, Liu B, McNeill KL, et al:Measurement of Pulse Wave Velocity Using Pulse Wave Doppler Ultrasound: Comparison with Arterial Tonometry. Ultrasound Med Biol 2008; 34:509-512.

E14.Lee VH, Connolly HM, Fulgham JR, et al: Tako-tsubo cardiomyopathy in aneurysmal subarachnoid hemorrhage: an underappreciated ventricular dysfunction. J Neurosurg. 2006;105(2):264-270.

E15.Zaroff JG, Rordorf GA, Ogilvy CS, et al. Regional patterns of left ventricular systolic dysfunction after subarachnoid hemorrhage: evidence for neurally mediated cardiac injury. J Am Soc Echocardiogr. 2000;13(8):774-779.

E16.Maréchaux S, Fornes P, Petit S, et al:Pathology of inverted Takotsubo cardiomyopathy. Cardiovasc Pathol.2008;17(4):241-243.

E17.Reference Values for Arterial Stiffness' Collaboration. Determinants of pulse wave velocity in healthy people and in the presence of cardiovascular risk factors: 'establishing normal and reference values'. Eur Heart J. 2010;31(19):2338-2350

E18.Tung PP, Olmsted E, Kopelnik A, et al:Plasma B-type natriuretic peptide levels are associated with early cardiac dysfunction after subarachnoid hemorrhage. Stroke. 2005;36(7):1567-1569.

E19.Meaudre E, Jego C, Kenane N, et al: B-type natriuretic peptide release and left ventricular filling pressure assessed by echocardiographic study after subarachnoid haemorrhage: a prospective study in non-cardiac patients.Crit Care. 2009;13(3):R76.

E20.Sviri GE, Shik V, Raz B, et al: Role of brain natriuretic peptide in cerebral vasospasm. Acta Neurochir (Wien).2003;145(10):851-860

E21.McGirt MJ, Blessing R, Nimjee SM, et al: Correlation of serum brain natriuretic peptide with hyponatremia and delayed ischemic neurological deficits after subarachnoid hemorrhage. Neurosurgery.2004;54(6):1369-1373

E22.Christenson RH: What is the value of B-type natriuretic peptide testing for diagnosis, prognosis or monitoring of critically ill adult patients in intensive care?Clin Chem Lab Med. 2008;46(11):1524-1532.

E23.Sugimoto K, Watanabe E, Yamada A, et al. Prognostic implications of left ventricular wall motion abnormalities associated with subarachnoid hemorrhage. Int Heart J. 2008;49(1):75-85.

E24.Vannemreddy P, Venkatesh P, Dinesh K, et al: Myocardial dysfunction in subarachnoid hemorrhage: prognostication by echo cardiography and cardiac enzymes. A prospective study. Acta Neurochir Suppl. 2010;106:151-154

E25.Ayer RE, Zhang JH:The clinical significance of acute brain injury in subarachnoid hemorrhage and opportunity for intervention. Acta Neurochir Suppl.2008;105:179-184.

Supplemental Figure 1. Changes in left ventricular ejection fraction (LVEF) and aortic pulse wave velocity (PWV) between acute ASAH and stable condition. In order to view the Supplemental Figure 1, follow this link (Supplemental Digital Content 2,