Preload optimisation in severe sepsis and septic shock

Preload optimisation in severe sepsis and septic shock Prof. Jean-Louis TEBOUL Medical ICU Bicetre hospital University Paris South France

Conflicts of interest Member of themedical Advisory Board ofpulsion

Decision of starting fluid administration presence of hemodynamic instability/peripheral hypoperfusion (mottled skin, hypotension, oliguria, hyperlactatemia ) and presence of preload responsiveness and limited risks of fluid overload

Fluid Challenge Predictors of fluid responsiveness/unresponsiveness

Crit Care Med 2006; 34:1333-1337 Rate of infusion: 500-1000 ml crystalloids or 300-500 ml colloids over 30 mins Goal:reversal of themarker of perfusion failurethat prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) Safety limits:cvpof 15 mmhg measured every 10 mins

Crit Care Med 2006; 34:1333-1337 Limitations Fluid challenge cannot serve as a test to predict fluid responsiveness First, it isnot a test but areal therapy 500-1000 ml crystalloids or 300-500 ml colloids/30 mins Not negligible amounts! Second, by definition, itcannot predictfluid responsiveness Fluid challenge is successful in only 50% cases

CHEST 2002, 121:2000-8

Tento obrázek nyní nelze zobrazit. Tento obrázek nyní nelze zobrazit.

Fluid Challenge Predictors of fluid responsiveness/unresponsiveness Can help to choose the best fluid strategy by avoiding to fluid overload patients who would be fluid unresponsive

Fluid infusion will increase LV stroke volume only ifboth ventricles arepreload responsive preload unresponsiveness Stroke Volume Fluid responsiveness equivalent to preload responsiveness biventricular preload responsiveness Ventricular preload

Fluid Challenge Predictors of fluid responsiveness/unresponsiveness Static markers of preload

normal heart Stroke volume preload responsiveness «static» measures ofpreload cannot reliablypredict fluid responsiveness. preload unresponsiveness failing heart Ventricular preload

Tento obrázek nyní nelze zobrazit. neither baseline PAOP nor baseline CVP predicted volume responsiveness Responders Nonresponders Responders Nonresponders

Tento obrázek nyní nelze zobrazit. Tento obrázek nyní nelze zobrazit. Tento obrázek nyní nelze zobrazit. Tento obrázek nyní nelze zobrazit. Crit Care Med 2013; 41: 1474-81 1802 pts Summary AUC 0.56 Crit Care Med 2013; 41:1774-1781

Tento obrázek nyní nelze zobrazit.???

Dynamic indices of preload responsiveness normal heart Stroke volume preload responsiveness failing heart. preload unresponsiveness Ventricular preload

MV induces cyclic changes in SV only in pts with biventricular preload responsiveness fluid responsiveness occurs only in pts with biventricular preload responsiveness correlates with the magnitude of the induced by

Fluid Challenge Predictors of fluid responsiveness/unresponsiveness Static markers of preload Dynamic markers of preload responsiveness o heart-lung interaction tests o variability of stroke volume and of its surrogates Invasive indices

Tento obrázek nyní nelze zobrazit. Stroke volume preload responsiveness preload unresponsiveness A B Ventricular preload

Tento obrázek nyní nelze zobrazit. Am J Respir Crit Care Med 2000; 162:134-8 PPV = PPmax- PPmin (PPmax +PPmin) /2 Arterial catheter PPmax PPmin

PP (%) before fluid Infusion 13 % 45 40 35 30 25 20 15 10 non-responders n = 24 5 0 responders n = 16

PPV Sensitivity CVP PAOP 1 -Specificity

Chest 2005;128;848-854 Anesth Analg 2011; 113:523-8 Chest 2004, 126:1563-1568 PPV Crit Care Med 2005;33:2534-9 Tento obrázek ny ní nelze zobrazit. M. Cannesson, J. Slieker, O. Desebbe, F. Fahdi,O. Bastien, JJ. Lehot Tento obrázek ny ní nelze zobrazit. Tento obrázek ny ní nelze zobrazit. X. Monnet1,2*, L. Guerin1,2, M. Jozwiak1,2, A. Bataille1,2, F. Julien1,2, C. Richard1,2, J-L. Teboul1,2 Tento obrázek ny ní nelze zobrazit.

The larger the PPbeforefluid infusion, the larger the increase in Am CO J Respir Crit after Care Med 2000; fluid 162:134-8 infusion 50 40 Fluid-induced changes in cardiac index (%) 30 20 10 r 2 = 0. 85 0 0 10 20 30 40 50 The smallerthe PPVbefore fluid infusion, PP (%) before fluid infusion the smaller the increase in CO after fluid infusion

Pulse Pressure Variation Calculated automatically and displayed in real-time by usual hemodynamic monitors All thesemonitors aresuitable todisplay PPV in real-time

Arterial pressure waveform analysis Stroke volume Arterial Pressure Stroke Volume Variation Calculated automatically anddisplayed inreal-time by new hemodynamic monitors

X. Monnet1,2*, L. Guerin1,2, M. Jozwiak1,2, A. Bataille1,2,F. Julien1,2, C. Richard1,2, J-L. Teboul1,2 Assessing fluid responsiveness by stroke volume variation in mechanically ventilated patients with severe sepsis G. Marx, T. Cope, L. McCrossan, S. Swaraj, C. Cowan, SM. Mostafa, R. Wenstone, M. Leuwer European Journal of Anaesthesiology 2004; 21:132-138 Chest 2005;128;848-854

PPV SVV 685 pts

Average cut-off: 12.5%

PPVi PPVni Non-invasive finger blood pressure monitoring device

Esophageal Doppler 18% ABF % = ABF max-abf min (ABF max+ ABF min)/2 R NR

Doppler-echo Vpeak Vpeak (%) before fluid 36 32 Vpeak max Vpeak min 24 28 20 16 12 8 Vpeak = Vpeak max Vpeak min (Vpeak max + Vpeak min) / 2 4 responders non responders

Pulse oximeter

PPV SVV PVI GEDV

NE (-) NE (+)

Subcostal view divcmin divcmax divc% = divcmax-divcmin (divcmax + divcmin)/2

66 pts with MV Systematic fluid loading with 10 ml/kg HES

36 % 100 SVC collapsibility (AUC: 0.99) 80 PPV 12% (AUC: 0.94) Specificity (%) 60 40 20 PPV and SVC collapsibility perform equally for predicting fluid responsiveness Transesophageal approach is required 0 0 20 40 60 80 100 100 -Specificity (%)

Limitations of respiratory variability indices impossible to interpret in pts with spontaneous breathing activity

PPV (threshold: 12 %) fully adapted to ventilator patients with spontaneous breathing sensitivity PPV 1 -specificity

Limitations of respiratory variability indices impossible to interpret in pts with spontaneous breathing activity impossible to interpret in patients with arrhythmias mmhg 110 PPmax 90 70 50 PPmin

PPV 12.8 % PPV 8 % Sensitivity Normal TV Low TV 1- Specificity

Ability of PPV to predict fluid responsiveness in function of lung compliance 100 80 Crs 40 ml/cmh 2 O Sensitivity 60 40 20 0 30 Crs < 40 ml/cmh 2 O Crs < 30 ml/cmh 2 O 0 20 40 60 80 100 100-Specificity (%)

Limitations of respiratory variability indices impossible to interpret in pts with spontaneous breathing activity impossible to interpret in patients with arrhythmias difficult to interpret if tidal volume is too low difficult to interpret if lung compliance is too low difficult to interpret in case of high frequency ventilation PPV can be not reliable when the heart rate/respiratory rate is> 3.6 De Backer et al Anesthesiology 2009

Limitations of respiratory variability indices impossible to interpret in pts with spontaneous breathing activity impossible to interpret in patients with arrhythmias difficult to interpret if tidal volume is too low difficult to interpret if lung compliance is too low difficult to interpret in case of high frequency ventilation difficult to interpret under open-chest conditions difficult to interpret in case of severe RV failure Mahjoub et al Crit Care Med 2009, Wyler von Ballmoos et al Crit Care 2010

Limitations of respiratory variability indices impossible to interpret in pts with spontaneous breathing activity impossible to interpret in patients with arrhythmias difficult to interpret if tidal volume is too low difficult to interpret if lung compliance is too low difficult to interpret in case of high frequency ventilation difficult to interpret under open-chest conditions In all these situations and in case of any doubt about interpretation other reliable dynamic tests are required difficult to interpret in case of severe RV failure and are now available

End-expiratory occlusion test Cyclic decrease in preload Transient increase in preload and hence in CO in case of preload-dependency Fluid responders should be identified by anincrease of theirco during theend-expiration occlusion test Systemic venous return

% 50 effects ofend-expiratory occlusion on Pulse contour CO % 50 effects ofend-expiratory occlusion on pulse pressure 40 40 30 Any real-time CO monitor could be suitable 20 10 30 A simple arterial catheter 20 10 could be suitable 0 0-10 NR R -10 NR R

Passive Leg Raising: the advantages PLR provides a good prediction of fluid responsiveness Unlike fluid challenge, effects of PLR are rapidly reversible PLR may well assess fluid responsiveness in situations where PPVfails to do it

RAP (mmhg) PAOP (mmhg) 20 30 25 15 20 10 15 5 10 5 0 PLR 0 PLR Base post-plr Base post-plr

Stroke Volume PLR mimics fluid challenge preload responsiveness b' a' Unlike fluid challenge, no fluid is infused, preload unresponsiveness and, the effects are reversible and transient b a The hemodynamic response to PLR A PLR B Ventricular preload can predict the hemodynamic response to volume infusion

The hemodynamic response to PLR can predict the hemodynamic response to fluid infusion Real-time CO response to PLR

VTI VTI +35%

24 pts with circulatory failure and SB TTE before and after 500 ml saline PLR-induced change in Velocity-Time Integral PLR-induced changes in VTIAo (%) goodprediction of volume responsiveness 40 35 30 VTI 25 20 15 10 5 0 * 12% -5 R NR

reversible hemodynamic effects ABF PLR ABF max 90 sec No risk of pulmonary edema

Pts with spontaneous breathing PLR-induced changes in ABF sensitivity PPV 1 -specificity

Passive Leg Raising: The advantages PLR provides a good prediction of fluid responsiveness Unlike fluid challenge, effects of PLR are rapidly reversible PLR may well assess fluid responsiveness in situations where PPVfails to do it Spontaneous Breathing activity Low lung compliance

Lung compliance 30 ml/cmh 2 O Lung compliance < 30 ml/cmh 2 O 100 PPV effects of PLR on CO 100 effects of PLR on CO 80 80 Sensitivity 60 40 Sensitivity 60 40 PPV 20 20 0 0 20 40 60 80 100 0 0 20 40 60 80 100 100 - Specificity 100-Specificity

Passive Leg Raising: the limits PLRshould notstart from a horizontalpatient s position but from a semi-recumbent position

45 45 semi-recumbent position Passive Leg Raising horizontal position 45 Passive Leg Raising

37.5 % change in CO from baseline 25 12.5 0

35 patients (all responders to fluid administration) 35 pts correctly classified as responders 20 pts correctly classified as responders 15 pts classified as nonresponders

45 45 semi-recumbent position Passive Leg Raising rather than horizontal position 45 Passive Leg Raising

Passive Leg Raising: the limits PLRshould notstart from a horizontalpatient s position but from a semi-recumbent position The hemodynamic response to PLR should not be monitored with arterial pressure but with CO measurements

PLR-induced changes in CO PLR-induced changes in AP Study name sample size AUC Study name sample size AUC Monnet CCM 2006 71 0.96 Lafanéchère CC 2006 22 0.95 Lamia ICM 2007 24 0.96 Maizel ICM 2007 34 0.89 Monnet CCM 2009 34 0.94 Thiel CC 2009 102 0.89 Biais CC 2009 30 0.96 Preau CCM 2010 34 0.94 Monnet CCM 2006 71 0.75 Monnet CCM 2009 34 0.68 Preau CCM 2010 34 0.86 139 0.76 351 0.95

80 PLR-induced changes in aortic blood flow PLR-induced changes in pulse pressure 60 % change from Baseline 40 Following thechanges in arterial pressure 20 10 0-20 -40 * * during PLR is not suitable (false negative cases) Falses negative cases A real-time CO monitor is necessary NR R NR R

Decision of stopping fluid administration disappearance of hemodynamic instability/peripheral hypoperfusion or presence of preload unresponsiveness or high risks of fluid overload or severe hypoxemic lung injury Value ofevlw andpvpi

Conclusion Predictors of fluid responsiveness/unresponsiveness Pulse pressure variation orstroke volume variation PLR or end-expiratory occlusion tests Can help to choose the best fluid strategy by identifying patients eligible for fluid infusion and by avoiding to fluid overload patients who would be fluid unresponsive Thank you