Intensive Care Med (2008) 34:917–922 DOI 10.1007/s00134-008-1022-y
Alain Cariou Michael R. Pinsky Mehran Monchi Ivan Laurent Christophe Vinsonneau Jean-Daniel Chiche Julien Charpentier Jean-François Dhainaut
BRIEF REPORT
Is myocardial adrenergic responsiveness depressed in human septic shock?
Abstract Objective: To assess left ventricular (LV) contractile function and adrenergic responsiveness in septic patients. Methods: We used echocardiographically defined Electronic supplementary material fractional area of contraction (FAC), The online version of this article and LV area to end-systolic arterial (doi:10.1007/s00134-008-1022-y) contains supplementary material, which is available pressure estimates of end-systolic to authorized users. elastance (E’es) and its change in response to dobutamine (5 µg/kg/min) Dr. Pinsky was a Professeur Associé at in 10 subjects in septic shock admitCochin Hospital and Paris Descartes University. ted to an intensive care unit of an academic medical center. Subjects A. Cariou · M. R. Pinsky · M. Monchi · were studied on admission and again I. Laurent · C. Vinsonneau · J.-D. Chiche · at both 5 days and 8–10 days after J. Charpentier · J.-F. Dhainaut admission. Results: Three of the Cochin-Saint Vincent de Paul Hospital 10 subjects died as a result of their and Paris Descartes University, acute process, while the others were Paris, France discharged from hospital. Nine out of 10 subjects required intravenous vasoM. R. Pinsky (u) pressor therapy on day 1, while only University of Pittsburgh, Department 1 of 9 subjects required vasopressor of Critical Care Medicine, 606 Scaife Hall, 3550 Terrace St., Pittsburgh support at day 5. LV end-diastolic area (EDA) increased from day 1 to 15261, PA, USA day 5 and days 8–10 (p < 0.05), but e-mail:
[email protected] Received: 11 May 2007 Accepted: 16 January 2008 Published online: 8 February 2008 © Springer-Verlag 2008
Introduction Left ventricular (LV) dysfunction occurs in septic shock despite increases in cardiac output [1]. Previous data showed decreased diastolic compliance and LV ejection fraction in non-survivors but reversible myocardial dilation in survivors [2]. However, LV ejection fraction is a load-dependent parameter and may not reflect contractility. Myocardial performance may be a pivotal aspect of survival from sepsis because survivorship is proportional
neither FAC nor E’es was altered by time (EDA 15.7 ± 5.8, 21.4 ± 5.1, and 19.4 ± 5.6 cm2 ; FAC 0.46 ± 0.19, 0.50 ± 0.20, and 0.48 ± 0.15%; E’es 21.6 ± 12.6, 23.2 ± 8.5, and 19.2 ± 6.3 mmHg/cm2 , mean ± SD, for days 1, 5 and 8–10 respectively). Although dobutamine did not alter E’es on day 1 or day 5, E’es increased in all of the 5 subjects studied on days 8–10 (p < 0.05). Conclusions: Adrenergic hyporesponsiveness is present in septic shock and persists for at least 5 days into recovery, resolving by days 8–10 in survivors. Keywords Contractility · Critical care · Echocardiography · Elastance · Human · Left ventricle · Nitroprusside · Sepsis
to the ability to increase maximal O2 delivery [3] even if done spontaneously [4]. We validated a clinical bedside tool that accurately estimates end-systolic elastance (E’es) a primary measure of contractility, using echocardiographic estimates of LV volume and arterial pressure measures of LV ejection pressure [5]. Thus, we sequentially measured over time baseline contractility and its change in response to a dobutamine challenge in critically ill patients with septic shock.
918
Methods Subject recruitment Following approval from the Cochin Hospital and the University of Pittsburgh Institutional Review Board we enrolled 14 patients over 4 months. All patients or their families gave informed consent. Entry criteria included the presence of septic shock for < 48 h requiring invasive monitoring and mechanical ventilation. Septic shock was defined using standard criteria [6]. Exclusion criteria included the inability to acquire adequate LV area signals by transesophageal echocardiography. Four of the 14 patients were excluded because of inadequate LV area signals or because they were not stable enough to have their initial studies. No patient had evidence of isolated regional LV dilation or marked regional wall motion abnormalities or was receiving beta-adrenergic blocking agents, and all patients were assumed to be adequately resuscitated as quantified by a pulse pressure variation of < 10%. Severity of illness, scored using SAPS2 [7], and organ system failure (OSF), assessed using the McCabe score of pre-morbid functional status by [8], were recorded for all subjects.
automated border detection ultrasound system (VingMed CFM 800, General Electrics Inc., Les Ulis, France) assessed LV end-diastolic area (EDA) and end-systolic area (ESA) as previously described [5]. Protocol Subjects were studied three times: on initial presentation, 5 days later and then at 8–10 days. Each time, baseline apneic measures were made over 20 s and then repeated after 15 min infusion of dobutamine at 5 µg/kg/min. Transient reduction in venous return and arterial pressure was induced by bolus intravenous infusion of nitroprusside (50 µg) at each interval. Data analysis
From the apneic steady state data prior to nitroprusside infusions, LV EDA and ESA were measured. Fractional area of contraction (FAC) was calculated as the ratio of ESA–EDA to EDA. Arterial pressure at LV ESA was taken to reflect end-systolic LV pressure. End-systolic elastance (E’es) was calculated from the end-systolic pressure–area relationship as arterial pressure and LV volumes decreased Instrumentation in response to nitroprusside infusion using the iterative technique as previously described by us [5]. We also All subjects received radial arterial catheterization, stand- calculated the ESA at 100 mmHg from the E’es curves ard two-lead ECG and endotracheal intubation with mech- to assess pressure-independent LV volume shifts in reanical ventilation. A 5.0-MHz transesophageal probe with sponse to dobutamine, as a second measure of contractility
Table 1 Subject characteristics Age Sex OSF Type (years)
SAPS2 McCabe a Sepsis history
Blood culture positive
Outcome and VasopresICU LOS (days) sor at D1 b
Vasopressor at D5 b
1
78
F
3
Med
85
I
N
Death 12
0.8 NE
0
2
68
M
3
Med
42
I
Y
Survived 31
0.23 Epi
0.23 Epi
3
49
M
3
Med
88
I
N
Survived 22
0.5 Epi
0
4
78
M
1
Med
41
II
N
Death 32
0.2 Epi
0
5
62
F
0
Med
30
I
N
Death 11
0.07 Epi
0
6
64
M
3
Med
55
I
N
Survived 20
0.14 Epi
0
7
38
M
1
Med
23
I
Y
Survived 11
0
0
8
46
F
1
Med
28
II
N
Survived 25
8 Dopa
0
9 10
36 71
M M
2 2
Med Surg
76 69
I II
Y Y
Survived 85 Survived 34
0 0.7 NE
0 0
Pneumococcal meningitis/pneumonia Pneumococcal pneumonia E. coli urosepsis Haemophilus pneumonia E. coli urosepsis Klebsiella pneumonia MRSA pneumonia and mediastinitis Haemophilus pneumonia MRSA endocarditis Aeromonas hydrophilia peritoneal abscess
OSF, organ system failure (no. of systems); Med, medical; Surg, surgical; SAPS, Simplified Acute Physiological Scale; MRSA, methicillinresistant Staphylococcus aureus; LOS, length of stay; NE, norepinephrine; Epi, epinephrine; D, dopamine a McCabe scoring of pre-morbid functional status: I, normal, II some impairment by self-sufficient b Vasopressor use: maximum for each day in µg/kg/min
9.5 6.0 9.5 4.4 0.42 0.55 0.47 0.10
change. Data were compared using a repeated-measures analysis of variance with a post hoc Newman–Keuls analysis with significance corresponding to p < 0.05.
Results
SAP, systolic arterial pressure (mmHg); EDA, end-diastolic area (cm2 ); ESA, end-systolic area (cm2 ), FAC, fractional area of contraction (%) ∗ p < 0.05 compared to day 1
0.36 0.81 0.51 0.22 0.51 0.5 0.2 17.7 30.2 13.5 19.3 23.8 21.3∗ 5.1
11.3 5.7 6.6 15.0 11.7 10.9 4.4
20.5 21.9 15.2 13.7 18.1 21.1 6.1
0.42 0.65 0.61 0.35 0.48 0.53 0.20
11.9 7.7 5.9 8.9 9.4 9.2 3.5
17.8 16.5 19.4 5.6
0.36 0.63 0.48 0.15
11.4 6.1 9.8 2.8
16.3 13.3 17.3 5.3
16.2 5.1 10.8 0.38 0.61 0.4 26.0 12.9 18.0 11.7 7.5 12.3 0.6 0.53 0.29 29.4 15.9 17.3 2.5 11.5 11.3 14.0 0.91 0.45 0.66 0.25 27.7 20.9 33.3 18.7 4.0 15.6 14.8 13.5 0.8 0.35 0.43 0.23 20.2 24.1 26.0 17.5
5.2 17.8 6.7 7.4 3.0 5.8 9.0 5.8 10.3 13.5 8.5 4.4 0.62 0.39 0.69 0.42 0.59 0.42 0.55 0.55 0.14 0.38 0.48 0.16 13.6 29.2 21.5 12.7 7.4 9.9 20.1 12.9 12.0 21.8 16.1 6.7 5.8 16.1 7.6 7.2 2.7 6.8 7.8 5.6 11.6 11.5 8.3 3.8 123 104 122 106 137 134 154 106 136 117 124 16.9 1 2 3 4 5 6 7 8 9 10 Mean SD
0.59 0.44 0.61 0.31 0.83 0.37 0.45 0.42 0.14 0.42 0.46 0.19 14.2 28.8 19.5 10.4 15.8 10.7 14.2 9.59 13.5 19.8 15.7 5.8
Day 1 dobutamine EDA FAC ESA ESA Day 1 EDA FAC SAP Subject
Table 2 Echocardiographic apneic steady-state left ventricular area data
Day 5 EDA FAC
ESA
Day 5 dobutamine EDA FAC ESA
Day 8–10 EDA FAC
ESA
Day 8–10 dobutamine EDA FAC ESA
919
Ten subjects were studied on day 1, nine on day 5 and five on day 8–10. One subject was not studied on day 5 due to hemodynamic instability, and five on day 8–10 due to early recovery and discharge. All subjects had been previously healthy, as assessed by McCabe score, but became profoundly ill during this event, as assessed by SAPS2 and OSF at time of admission (Table 1). All but one subject were receiving inotropic drugs on day 1. All but one subject were off vasopressors by day 5, and all were deemed hemodynamically stable. All five subjects studied on days 8–10 were off vasopressors. Neither heart rate (103 ± 17, 96 ± 16, 101 ± 14 beats/min), mean arterial pressure (77 ± 22, 82 ± 10, 97 ± 9 mmHg), CVP (17 ± 7, 16 ± 6, 14 ± 4 mmHg) nor temperature (37.7 ± 1.2, 37.5 ± 0.6, 37.6 ± 0.2°C) differed significantly between days 1, 5 and 8–10, respectively. LV EDA increased from day 1 to days 5 and 8–10 (p = 0.028), whereas neither ESA nor FAC changed over time or during dobutamine infusions (Table 2). Marked inter-subject variability in baseline E’es and LV area at 100 mmHg was seen (Table 3). Baseline E’es was unchanged over the study period and its values were in a range similar to those of subjects with normal LV contractility prior to coronary artery bypass grafting [12] (E’es 16–24 mmHg/cm2 ). Although baseline E’es did not change for the group as a whole over the study period, 1 of the 10 subjects had a marked improvement in E’es from day 1 to day 5. These data are shown in comparison to a representative of the remaining 9 subjects (Fig. 1) and for all subjects (Fig. E1 in ESM). Individual subject E’es data is displayed in Figs. 4–6. Dobutamine challenge did not alter E’es at either day 1 (Fig. 2A) or day 5 (Fig. 2B). Baseline E’es did not change from day 1 to day 5 (Fig. 2D). All 5 subjects at days 8–10 increased E’es in response to dobutamine, and these increases were to greater E’es values than in the same subjects earlier (Fig. 2C) (p < 0.05). LV area at 100 mmHg did not change with dobutamine challenge on any of the days (Table 3).
Discussion Ognibene et al. [9] showed that preload-recruitable stroke work was depressed more in septic than in non-septic critically ill patients. However, they did not challenge their patients with inotropes. Dobutamine challenge identifies septic shock survivors, presumably because the associated increasing cardiac output reflects less cardiovascular impairment [10]. Our data show that impaired baseline car-
920
Table 3 End-systolic elastance (E’es, mmHg/cm2 ) and left ventricular area (LVA, cm2 ) at 100 mmHg Subject 1 2 3 4 5 6 7 8 9 10 Mean SD Median
Day 1 E’es
LVA
Day 1 + dobuta E’es LVA
Day 5 E’es
LVA
Day 5 + dobuta E’es LVA
Day 8–10 E’es LVA
Day 8–10 + dobuta E’es LVA
15.7 13.1 22.0 19.7 31.0 54.2 13.8 15.6 15.8 14.8 21.6 12.6 15.7
4.6 11.9 4.7 6.4 1.6 6.1 5.5 5.2 10.2 10.2 6.6 3.1 5.8
23.1 4.1 24.2 24.7 27.0 35.1 12.6 24.6 11.5 13.8 20.1 9.2 23.7
24.2 16.0 16.6 15.8
1.8 8.9 11.3 12.0
32.0 19.0 16.0 12.5
1.8 9.1 5.9 9.3
24.3 27.0 12.0
9.3 3.6 10.9
37.5 56.0 16.6
10.8 1.8 7.8
25.8 21.0 25.8 43.1 20.0 23.2 8.5 21.0
9.8 4.4 6.8 14.8 9.8 8.8 4.0 9.8
4.1 20.5 49.9 34.9 14.0 22.5 14.0 19.0
9.4 4.3 3.9 8.6 7.6 6.7 2.8 7.6
14.8 18.0 19.2 6.3 18.0
8.6 5.5 7.6 3.0 8.6
21.5 24.0 31.1 15.9 24.0
8.7 5.3 14.0 4.1 7.8
4.5 15.5 3.7 6.2 2.7 5.1 4.1 3.9 8.8 10.8 6.5 4.0 4.8
End-systolic elastance was calculated from the end-systolic arterial pressure to end-systolic left ventricular area relation during a transient decrease in arterial pressure Dobuta, Dobutamine infusions (5 µg/kg/min)
Fig. 1 End-systolic arterial pressure and left ventricular (LV) area data during baseline nitroprusside infusion on days 1, 5 and 8 for two subjects. Subject one reflects the typical behavior of 10 of the 11 subjects, with minimal change in end-systolic elastance (E’es) over time, whereas subject 2 demonstrated a marked increase in performance between days 1 and 5. E’es is defined as slope of the lines through the individual data points, increased from day 1 to 5. E’es values in mmHg/cm2 . Note, however, that with time absolute LV areas decrease
diac contractility is uncommon whereas depressed adrenergic responsiveness universally occurs. Whether this reflects transient impaired LV contractility or receptor downregulation can not be accessed by our data. Interestingly, we only saw minimal LV dilation and then only by day 5. Since we did not measure pulmonary artery occlusion pressure, we are unable to assess LV diastolic compliance. However, LV dilation was not a feature of our surviving patients. LV contractility can be measured by several parameters. E’es is only one such measure. Parallel E’es shifts may occur such that the heart is working at smaller volumes while developing similar ejection pressures. Such changes would reflect increased contractility. This appears to be the case for these two subjects (Fig. 1). However, dobutamine did not increase baseline stroke work or decrease ESA for a common pressure on days 1 and 5 (Fig. 2 and Fig. E1). If anything ESA increased from day 1 to day 8–10, probably reflecting volume resuscitation. Thus, we also measured FAC and LV area at 100 mmHg as an estimate of both contractility and potential LV volume shifts (Table 3). Neither FAC nor LV area at 100 mmHg demonstrated changes with dobutamine or over time. Therefore, our data suggest that baseline contractility remains constant over the first week of severe sepsis with the only change being a recovery of adrenergic responsiveness by day 8–10. Finally, we used a bolus infusion of nitroprusside to alter loading conditions. Nitroprusside induced a short (< 2 min) reduction in afterload, allowing calculation of E’es. Although this challenge was safe, with no subject demonstrating sustained hypotension, we recently showed that an inspiratory hold maneuver of 10 cmH2 O can also be used to estimate E’es [11]. Previous workers have documented non-specific adrenergic hyporesponsiveness in septic patients. Silverman et al. [12] showed that intrinsic myocardial performance of isolated cardiac myocytes was unaltered by being bathed
921
Fig. 2 Effect of dobutamine infusion (5 µg/kg/min) on left ventricular (LV) end-systolic elastance (E’es) on day 1 (a), day 5 (b), and 8–10 days after the initial presentation of septic shock (c); and (d) the change in LV E’es for individual subjects from day 1 to 5, during which time all subjects were hemodynamically stable and off all vasoactive drug therapy. E’es values in mmHg/cm2
in human sepsis serum, but β-adrenergic responsiveness was blunted. Human sepsis also blunts vasculature to β-adrenergic responsiveness [13]. Nitric oxide-induced depression of adrenergic responsiveness appears to be a key mechanism [14]. Our data support the position that in severe sepsis myocardial dysfunction reflects more impaired adrenergic responsiveness than a reduction in intrinsic myocardial contractility. However, we used only one low dose of dobutamine as our adrenergic challenge. Furthermore, day 1 baseline contractility could reflect increased inotropic stimulation. All patients were already receiving significant vasoactive drug infusions and may not have responded to the addition of dobutamine. However, by day 5 all but one subject were off exogenous adrenergic drug therapy and all still showed no increase in E’es in response to dobutamine. These data are consistent with down-regulation of β-adrenergic receptors’ responsiveness. Our study does not allow us to quantify the degree of adrenergic hyporesponsiveness present or its resolution over time because we did not titrate dobutamine to effect. Many recent studies implicate cytokines as important mediators in both myocardial depression of sepsis and adrenergic hyporesponsiveness. Cytokines are present in the sera of humans with sepsis, and both TNF-α and
IL-1β appear responsible for the myocardial depression in isolated heart cell preparations in vitro [15] and in intact animal models [16]. Unfortunately, animal models of sepsis have yielded conflicting results when analyzed for myocardial performance. Eichenholz et al. demonstrated that bolus infusion of relatively high doses of TNF-α induced a variable time-dependent depression of canine LV performance, as assessed by LV ejection fraction measurements at times from 2 h to 240 h post infusion [17]. However, Pinsky et al. were unable to demonstrate any measurable impairment in LV systolic performance in response to bolus endotoxin infusion (1 mg/kg) [18]. Furthermore, adrenergic responsiveness to low-dose dobutamine infusion was heightened rather than depressed in their acute endotoxemic model. Thus, animal models of septic shock reflect poorly the clinical state because of species differences in cardiovascular response and the chronic and complex nature of clinical sepsis compared to a single-process animal model of sepsis. Thus, impaired myocardial performance in human septic shock reflects adrenergic hyporesponsiveness more than depressed myocardial contractility, and this hyporesponsiveness persists after clinical resolution of circulatory shock.
922
References 1. Parrillo JE (1993) Pathogenetic mechanisms of septic shock. N Engl J Med 328:1471–1477 2. Parker MM, Shelhamer JH, Bacharach SL, Green MV, Natanson C, Frederick TM, Damske BA, Parrillo JE (1984) Profound but reversible myocardial depression in patients with septic shock. Ann Intern Med 100:483–490 3. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M (2001) Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 345:1368–1377 4. Tuchschmidt J, Fired J, Astiz M, Rackow E (1992) Elevation of cardiac output and oxygen delivery improves outcome in septic shock. Chest 102:216–220 5. Gorcsan J, Denault AY, Gasior TA, Mandarino WA, Kancel MJ, Deneault LG, Hattler BG, Pinsky MR (1994) Rapid estimation of left ventricular contractility from end-systolic relations by echocardiographic automated border detection and femoral arterial pressure. Anesthesiology 81:553–562 6. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, Schein RM, Sibbald WJ (1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. Chest 101:1644–1652
7. Le Gall JR, Lemeshow S, Saulnier F (1993) A new simplified acute physiology score (SAPS II) based on a European /North American multicenter study. J Am Med Assoc 270:2957–2963 8. McCabe WR, Jackson GG (1962) Gram-negative bacteremia. I. Etiology and ecology. Arch Intern Med 110:847–855 9. Ognibene FP, Parker MM, Natanson C, Shelhamer JH, Parrillo JE (1988) Depressed left ventricular performance. Response to volume infusion in patients with sepsis and septic shock. Chest 93:903–910 10. DeBacker D, Moraine JJ, Barre J, Kahn RJ, Vincent JL (1994) Effects of dobutamine on oxygen consumption in septic patients. Direct versus indirect determinants. Am J Respir Crit Care Med 150:95–100 11. Kim HK, Alhammouri MT, Mokhtar YM, Pinsky MR (2007) Estimating left ventricular contractility using inspiratory-hold maneuvers. Intensive Care Med 33:181–189 12. Silverman HJ, Penaranda R, Orens JB, Lee NH (1993) Impaired betaadrenergic receptor stimulation of cyclic adenosine monophosphate in human septic shock: association with myocardial hyporesponsiveness to catecholamines. Crit Care Med 21:31–39
13. Bersten AD, Hersch M, Cheung H, Rutledge FS, Sibbald WJ (1992) The effect of various sympathomimetics on the regional circulations in hyperdynamic sepsis. Surgery 112:549–561 14. Hare JM, Loh E, Creager MA, Colucci WS (1995) Nitric oxide inhibits the positive inotropic response to β-adrenergic stimulation in humans with left ventricular dysfunction. Circ 92:2198–2203 15. Kumar A, Brar R, Wang P, Dee L, Skorupa G, Khadour F, Schulz R, Parrillo JE (1999) Role of nitric oxide and cGMP in human septic serum-induced depression of cardiac myocyte contractility. Am J Physiol 276:R265–R276 16. Walley KR, Herbert PC, Wakai Y, Wilcox PG, Road JD, Cooper DJ (1994) Decrease in left ventricular contractility after tumor necrosis factor-α infusion in dogs. J Appl Physiol 76:1060–1067 17. Eichenholz PW, Eichacker PQ, Hoffman WD, Banks SM, Parrillo JE, Danner RL, Natanson C (1992) Tumor necrosis factor challenges in canines: patterns of cardiovascular dysfunction. Am J Physiol 263:H668–H675 18. Pinsky MR, Rico P (2000) Cardiac contractility is not depressed in early canine endotoxic shock. Am J Respir Crit Care Med 161:1087–1093
