Prof. Jean-Louis TEBOUL Medical ICU Bicetre hospital University Paris XI France What is the best way to assess What is the best way to assess fluid responsiveness in a spontaneously breathing patient ? Member of the Medical Advisory Board of Pulsion

Three different scenarios 2- Patients in the ER for high suspicion of septic shock 3- Patients in the ICU, already resuscitated for several hours or days 1- Patients in the ER for acute blood losses or body fluid losses

Diagnosis of hypovolemia is almost certain clinical signs of hemodynamic instability Presence of clinical signs of hemodynamic instability No therapeutic dilemma prediction of volume responsiveness good prediction of volume responsiveness lacking of sensitivity although lacking of sensitivity

2- Patients in the ER for high suspicion of septic shock no needs predictors Most often, no needs for searching sophisticated predictors of volume responsiveness volume resuscitation is mandatory of volume responsiveness since volume resuscitation is mandatory in the first hours in the first hours (see Rivers et al NEJM 2001)

- with hemodynamic instability requiring therapy - without certainty of volume responsiveness - with potential risks of pulmonary edema and/or excessive cumulative fluid balance and/or excessive cumulative fluid balance 3- Patients in the ICU, already resuscitated for several hours or days

How to deal with this therapeutic dilemma? Prediction of volume responsiveness ? Fluid challenge ? - with hemodynamic instability requiring therapy - without certainty of volume responsiveness - with potential risks of pulmonary edema and/or excessive cumulative fluid balance and/or excessive cumulative fluid balance

Rate of infusion: mL crystalloids mL colloids 30 mins Rate of infusion: mL crystalloids or mL colloids over 30 mins Goal: reversal of the marker of perfusion failure that prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) Safety limits: CVP Crit Care Med 2006; 34: of 15 mmHg measured every 10 mins

Question: benefit/risk ratio ? Fluid challenge successful in only 50% cases Crit Care Med 2006; 34: Rate of infusion: mL crystalloids or mL colloids over 30 mins Goal: reversal of the marker of perfusion failure that prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) (ex: hypotension, tachycardia, oliguria, etc) Safety limits: CVP of 15 mmHg measured every 10 mins

CHEST 2002, 121:2000-8

Question: benefit/risk ratio ? Fluid challenge successful in only 50% cases Crit Care Med 2006; 34: Rate of infusion: mL crystalloids or mL colloids over 30 mins Goal: reversal of the marker of perfusion failure that prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) (ex: hypotension, tachycardia, oliguria, etc) Safety limits: CVP of 15 mmHg measured every 10 mins Fluid challenge potentially risky Is a CVP of 15 mmHg a reasonable safety limit? 1) PAOP often > CVP 2) Pulmonary capillary pressure (Pcap) > PAOP

Collee et al. Anesthesiology 1987 Radermacher et al. Anesthesiology 1989 Radermacher et al. Anesthesiology 1990 Teboul et al J. Appl Physiol 1992 Benzing et al. Acta Anaesthesiol Scand.1994 Rossetti et al. Am J Respir Crit Care Med 1996 Benzing et al. Br J Anaesth Nunes et al. Intensive Care Med Her et al. Anesthesiology 2005 Pcap-PAOP difference is high in ALI/ARDS

Question: benefit/risk ratio ? Fluid challenge successful in only 50% cases Crit Care Med 2006; 34: Rate of infusion: mL crystalloids or mL colloids over 30 mins Goal: reversal of the marker of perfusion failure that prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) (ex: hypotension, tachycardia, oliguria, etc) Safety limits: CVP of 15 mmHg measured every 10 mins Fluid challenge potentially risky Is a CVP of 15 mmHg a reasonable safety limit? 1) PAOP often > CVP 2) Pulmonary capillary pressure (Pcap) > PAOP 3) Degree of pulmonary edema poorly evaluated by Pcap since lung capillary permeability is often altered in ICU pts since lung capillary permeability is often altered in ICU pts

EVLW Pulmonary capillary hydrostatic pressure Pcrit normal lung capillary permeability Increased permeability mmHg PVC: 15, PAOP:19

Fluid challenge successful in only 50% cases Crit Care Med 2006; 34: Rate of infusion: mL crystalloids or mL colloids over 30 mins Goal: reversal of the marker of perfusion failure that prompted the fluid challenge (ex: hypotension, tachycardia, oliguria, etc) (ex: hypotension, tachycardia, oliguria, etc) Safety limits: CVP of 15 mmHg measured every 10 mins Fluid challenge potentially risky Question: benefit/risk ratio ?

3- Patients in the ICU, already resuscitated for several hours or days How to deal with this therapeutic dilemma? Prediction of volume responsiveness ? Fluid challenge ? - with hemodynamic instability requiring therapy - without certainty of volume responsiveness - with potential risks of pulmonary edema and/or excessive cumulative fluid balance and/or excessive cumulative fluid balance

Stroke Volume Ventricular preload preload-dependence preload-independence Volume expansion will increase stroke volume only if ventricles are preload-dependent

How to predict preload-dependence and hence volume responsiveness? 1- By estimating cardiac preload - using filling pressures: RAP, PAOP

Stroke Volume Ventricular preload preload-dependence preload-independence The lower the ventricular preload, the more likely the preload-dependency

Calvin Schneider Reuse Wagner responders nonresponders * mmHg RAP before volume expansion in responders (R) and non-responders (NR) RAP before volume expansion in responders (R) and non-responders (NR) number of pts SB pts (%) *

Wagner et al. Chest 1998 Baseline P RA (mmHg) Changes in stroke volume (%) r = 0.45

responders nonresponders mmHg PAOP before volume expansion in responders (R) and non-responders (NR) PAOP before volume expansion in responders (R) and non-responders (NR) Calvin Schneider Reuse Diebel number of pts SB pts (%)

How to predict preload-dependence and hence volume responsiveness? 1- By estimating cardiac preload - using filling pressures: RAP, PAOP - using dimensions: RVEDVi, LVEDVi

* Calvin Schneider Reuse Diebel Diebel Wagner Calvin Schneider Reuse Diebel Diebel Wagner * responders nonresponders RVEDVi before volume expansion in responders (R) and non-responders (NR) RVEDVi before volume expansion in responders (R) and non-responders (NR) number of pts SB pts (%)

0 120 LVEDVi before volume expansion in responders (R) and non-responders (NR) LVEDVi before volume expansion in responders (R) and non-responders (NR) responders nonresponders Calvin Schneider Calvin Schneider mL/m number of pts SB pts (%) 54 33

How to predict preload-dependence and hence volume responsiveness? 1- By estimating cardiac preload - using filling pressures: RAP, PAOP - using dimensions: RVEDVi, LVEDVi markers of preload: poor markers of volume responsiveness Why ?

Why do ventricular preload indicators not predict fluid responsiveness ? 2- Because RAP, PAOP, RVEDVi, LVEDVi are not always accurate indicators of preload of preload 1- In the available studies, pts were already resuscitated so that values of markers of preload were rarely low. of markers of preload were rarely low. It cannot be excluded that low values predict volume responsiveness, whereas high values well predict the absence of hemodynamic response to volume On the other hand, values were rarely high before fluid challenges 3- Because assessment of preload is not assessment of preload-dependence

. Stroke volume Ventricular preload normal heart normal heart failing heart failing heart preload-dependence preload-independence

How to detect fluid responsiveness ? 1- By estimating cardiac preload ? 2- By using dynamic tests detecting cardiac preload reserve ? cardiac preload reserve ? 2.1- using heart-lung interaction - SPV, PPV?. for physiological reasons, these indices must not work. as confirmed in clinical studies

Patients with MV Rooke et al Anesth & Analg 1995 Systolic pressure variation (mmHg) Patients with SB Patients with SB Systolic pressure variation (mmHg)

Soubrier et al. ATS 2005 nonresponders responders 3 % 23 % 12 % PPV before volume infusion Patients breathing without mechanical support

1 - specificity PPV sensitivity patients with SB patients with SB PPV (threshold: 12 %) sedated patients sedated patients

How to detect fluid responsiveness ? 1- By estimating cardiac preload ? 2- By using dynamic tests detecting cardiac preload reserve ? cardiac preload reserve ? 2.1- using heart-lung interaction - SPV, PPV? NO - Inspiratory decrease in RAP?

RAP 20 0 mmHg Inspiration J Crit Care 1992, 7:76-85 RAP 20 0 mmHg Inspiration RAP decrease by ≥ 1 mmHg at inspiration at inspiration positive respiratory response no hemodynamic response to volume challenge hemodynamic response to volume challenge RAP decrease by < 1 mmHg at inspiration at inspiration negative respiratory response

Changes in CO after volume loading (L/min) negative respiratory response positive respiratory response Magder et al J Crit Care 1992 Limitation : to be sure that the inspiratory effort is sufficient

How to detect fluid responsiveness ? 1- By estimating cardiac preload ? 2- By using dynamic tests detecting cardiac preload reserve ? cardiac preload reserve ? 2.1- using heart-lung interaction - SPV, PPV? NO - Inspiratory decrease in RAP? 2.2- using passive leg raising

Passive Leg Raising 45 ° Venous blood shift (Rutlen et al. 1981, Reich et al. 1989) Increase in left ventricular preload Increase in left ventricular preload ( Rocha 1987, Takagi 1989, De Hert 1999, Kyriades 1994 ) Reversible effects Reversible effects Increase in right ventricular preload Increase in right ventricular preload (Thomas et al 1965)

Base Base PLR PLR post-PLR post-PLR RAP (mmHg) PAOP (mmHg) Chest 2002; 121:

Passive Leg Raising 45 ° Venous blood shift (Rutlen et al. 1981, Reich et al. 1989) Increase in right ventricular preload Increase in right ventricular preload (Thomas et al 1965) Increase in left ventricular preload Increase in left ventricular preload ( Rocha 1987, Takagi 1989, De Hert 1999, Kyriades 1994 ) Transient effect (Gaffney 1982) PLR could be used as a test to detect volume responsiveness PLR could be used as a test to detect volume responsiveness rather than as a therapy

Hypothesis The increase in pulse pressure during PLR predicts the increase in stroke volume afler volume loading Chest 2002; 121:

PLR-induced changes in Pulse Pressure (mmHg) Fluid-induced changes in Stroke Volume (%) n = 39 r = 0.74 Chest 2002; 121:

Real-time CO monitoring is mandatory Hypothesis : a better surrogate of stroke volume than PP could do better

Hypothesis PLR-induced increase in mean aortic blood flow provides a better prediction of volume responsiveness than PLR-induced increase in pulse pressure Hypothesis : a better surrogate of stroke volume than PP could do better

Base 1 Base 2 Post VE PLR 500 mL saline

Base 1PLRBase 2Post VE Changes in aortic blood flow (%) responders nonresponders

R NR PLR-induced changes in pulse pressure * * RNR % change from Baseline 10 PLR-induced changes in aortic blood flow

specificity specificity PPV PLR-induced changes in PP PLR-induced changes in ABF sensitivity Monnet et al. Crit Care Med 2006 patients with spontaneous breathing n = 19

Hypothesis : a better surrogate of stroke volume than PP could do better Hypothesis PLR-induced increase in Pulse Contour CO PLR-induced increase in Pulse Contour CO provides a better prediction of volume responsiveness than PLR-induced increase in pulse pressure

cut-off = 12 % % increase in Pulse Contour CO during PLR Ridel ATS 2006 * Se = 70 % Sp = 92 % nonrespondersresponders

Conclusion In spontaneously breathing patients Prediction of volume responsiveness is a difficult issue markers of preload PPV, SPV, SVV Inspiratory decrease in RAP Response to passive leg raising unreliable valuable Thank you for your attention but need to be confirmed