EARLY GOAL DIRECTED THERAPY : Etat des lieux en 2017 Daniel De Backer Head Dept Intensive Care, CHIREC hospitals, Belgium Professor of Intensive Care, Université Libre de Bruxelles Past-President European Society of Intensive Care Medicine
HEMODYNAMIC ALTERATIONS IN SEPTIC SHOCK Varpula et al Factors independently associated with poor outcome ICM 31:1066;2005 Continuous hemodynamic measurements during the first 48h of shock (111 consecutive patients)
Early goal directed therapy Rationale: Prevent the development of tissue hypoperfusion by achieving specific resuscitation goals: MAP (perfusion pressure) SvO2 (adequacy of flow/o2needs) CVP (minimal preload)
2012 Recommendation for Initial Resuscitation. We recommend the protocolized, quantitative resuscitation of patients with sepsis- induced tissue hypoperfusion. During the first 6 hours of resuscitation, the goals of initial resuscitation should include all of the following as a part of a treatment protocol: a) CVP 8 12 mm Hg b) MAP 65 mm Hg c) Urine output 0.5 ml/kg/hr d) Scvo2 70%.
EARLY GOAL DIRECTED RESUSCITATION?
Severe sepsis and septic shock Rivers et al NEJM 345:1968;2001 Importance of early hemodynamic optimization (based on SvcO2 monitoring) DDB USI
Severe sepsis and septic shock Early hemodynamic optimization Rivers et al NEJM 345:1968;2001 263 patients: 2 SIRS criteria + TAsyst < 90 mmhg, after 30 ml/kg cryst in 30 min or Lactate > 4 meq/l Randomized 133: classic therapy 130: Early Goal Therapy guided on SvcO2 Emergency room DDB USI
EARLY HEMODYNAMIC OPTIMIZATION Mortality, % 60 50 40 30 20 10 0 * 28-Day * 60-Day Rivers et al NEJM 345:1968;2001 Standard Early goal * p<0.05 vs ctrl DDB USI
What are the interventions that made the diference? More fluids RBC transfusions Dobutamine and vasodilatory agents Sedation and mechanical ventilation More doctors? Rivers et al NEJM 345:1968;2001
What are the interventions that made the diference? Rivers et al NEJM 345:1968;2001
Timing of fluid resuscitation is essential! Correa T et al CCM 40:2841;2012 Delayed fluid resuscitation is associated with greater activation of inflammation Pigs CLP DDB USI
Correction of hypoxia limits the development of inflammation EGDT CTRL Rivers et al CCM 35:2016;2007 243 patients assigned to EGDT or CTRL group DDB USI
Correction of hypoxia limits the development of cellular lesions Rivers et al CCM 35:2016;2007 EGDT CTRL 243 patients assigned to EGDT or CTRL group DDB USI
Eltzschig and Carmeliet NEJM 364:656;2011
A lot of criticisms Single centre trial Potential confounders A few patients made the difference Resuscitation goals (CVP ) => Need for confirmation in a multicentric RCT
We had a dream. DDB USI
1pt/month /centre March 2014 1351 pts
EGDT in septic shock Peake S et al NEJM 2014 ARISE (1600 pts)
EGDT in septic shock Mouncey P et al NEJM 2015 PROMISE (1260 pts)
ICM 2015
Most patient already reached target ScvO2 values in the recent trials Rivers et al PROCESS ARISE PROMISE ScvO2 % 49 71 73 70 Rivers et al NEJM 2001 Angus D et al NEJM 2014 Peake S et al NEJM 2014 Mouncey P et al NEJM 2015 Inclusion: refractory hypotension and/or lactate 4 (despite fluids)
Peake S et al Most patients reached ScvO2 goal at inclusion NEJM 2014 and the proposed protocol was not able to significantly increase this proportion over time!
Major differences in mortality in control arm Mortality % (ctrl) Rivers et al PROCESS ARISE PROMISE 50 19 19 29 ScvO2 % 49 71 73 70 Rivers et al NEJM 2001 Angus D et al NEJM 2014 Peake S et al NEJM 2014 Mouncey P et al NEJM 2015 Inclusion: refractory hypotension and/or lactate 4 (despite fluids)
Major differences in mortality in control arm Reflects improvement of care! Reflects improvement of care???? De Backer D et al CCM 2016
Major differences in mortality in control arm Reflects patient selection?
Prognostic value of lacate and impact of time from diagnosis ARISE Casserly B et al CCM 43:567;20150 ProMISe ProCESS Rivers 28150 pts / 218 sites / SSC database
Inclusion rates? Angus D et al NEJM 2014 0.9 Patients/ centre / month (included) 3.9 Patients/ centre / month (screened) (ED with at least 40000 admissions/ year) 0.5 Patients/ centre / month (included) 1.6 Patients/ centre / month (screened) 7.4 Patients/ centre / month (included) 8 Patients/ centre / month (screened) (ED with??? admissions/ year) Peake S et al NEJM 2014 Rivers et al NEJM 345:1968;2001
=> Inclusion mostly during «office hours» => Is the care of the control group similar during «non- office hours»? Mouncey P et al NEJM 2015
20 % of these «septic shock» patients were not admitted to the ICU!? Angus D et al ICM 20145
And the issue of compliance in the treated arm should not be neglected
Issues with compliance Angus D et al ICM 20145
Excellent trials but. PROCESS / ARISE / PROMISE Angus D et al NEJM 2014 Peake S et al NEJM 2014 - Many patients in ctrl group were within targets - When not in target, the protocol failed to put them on target - Limited power (given the lower than expected mortality) - Limited external validity: - Low inclusion rate - Mostly office hours inclusions => This is not the end of EGDT but EGDT should not be applied in all septic patients without discrimination Mouncey P et al NEJM 2015
NEJM 2017
NEJM 2017
NEJM 2017
Conclusions from the various studies The conclusions of each of the trials are valid EGDT should not be implemented in less severe patients who already have normalized ScvO2 at baseline There was no proof for harm with EGDT, even in the less severely ill patients EGDT should probably still be implemented in the most severe patients with altered ScvO2
Rhodes et al ICM 2017 CCM 2017
Rhodes et al ICM 2017 CCM 2017
What do you recommend to use? Rhodes et al ICM 2017 CCM 2017
Variables to use to indicate further fluid administration? Heart rate / blood pressure Skin mottling CVP Lactate Veno-arterial PCO2 gradients Urine output Echo Other available hemodynamic measurements Most of these variables indicate poor tissue perfusion not that the patient will respond to fluids!
Signs of tissue hypoperfusion? no yes Expect Cardiac output low or inadapted? no Other intervention yes Hemodynamic intervention
What are the signs of tissue hypoperfusion that I use at bedside? Hypotension Skin mottling Capillary refill time Lactate Veno-arterial PCO2 gradients Urine output
Expect What do I use to evaluate whether cardiac output is low or inadapted? Signs of tissue hypoperfusion? no no Other intervention yes Cardiac output low or inadapted? yes Hemodynamic intervention
What do I use to evaluate whether cardiac output is low or inadapted? ScvO2 Cardiac output measurement by echo
Which goals? Resolution of signs of tissue hypoperfusion Which interventions? Fluids Vasopressors Inotropes Guided by? ScvO2 CVP Echo Lactate MAP/DAP
Limits in the interpretation of SvO2 A high SvO2 can be abnormal (microciculatory alterations and/or mitochondrial dysfunction)
Mortality, % ScvO2 <70% 70-90% >90% 619 pts severe sepsis (ED) Pope J et al Ann Emerg Med 15:40;2010
Resuscitation should not be guided on signs of cardiac dysfunction in isolation
A proposal to replace SvO2 by direct evaluation of preload responsiveness and ejection fraction Bouferrache K et al CCM 40:2821;2012 46 pts septic shock
Divergent information between ScvO2 and EF Should these patients be treated with inotropic agents? 46 pts septic shock Bouferrache K et al CCM 40:2821;2012 No indication for inotropic agents despite low SvO2!
Echographic evaluation of LVEF in patients with septic shock Vieillard-Baron et al AJRCCM 168:1270;2003 => Inotropic agents should not be used to correct a low EF
Central venous pressure
CVP to predict fluid responsiveness Biais M et al Crit Care 2014 556 pts
96 pts 150 fluid challenges Central venous pressure 26/81 (32%) Osman et al ICM 35:64;2007 55/81 (68%) 8 DDB USI
CVP: Never an optimal prediction but still some reasonable guidance when something better cannot be used... Eskesen et al ICM 2016 1148 pts
The increase in CVP is the price to pay, not a goal in itself!
Administration of fluids / targets need to be adapted to the situation!
Jama 2017
Jama 2017 => Patients were reassessed for tolerance, but not for indication!
Lactate guided therapy?
Vincent JL and De Backer D NEJM 369:1726; 2013
Prognostic value of lacate and impact of time from diagnosis Casserly B et al CCM 43:567;20150 28150 pts / 218 sites / SSC database
Lactate decrease and outcome Lactate decrease > 10% Lactate decrease < 10% Arnold R et al Shock 2013 110 pts with septic shock
What is the ideal decrease slope? Nichol et al Crit Care 15:R242;2011 5041 pts (ICU) DDB USI
9190 pts with sepsis Liu V et al Annals ATS 2013
Lactate guided therapy What goal? N=185 severe sepsis (ED) Puskarich M et al Chest 143:1548; 2013 Lactate at 6h
Lactate guided therapy (-20%/2h for 8h) Janssens T et al AJRCCM 2010 N=348
Our resuscitation strategies should be adapted to time Early phase Vincent JL and De Backer D NEJM 369:1726; 2013 Later stages
CONCLUSIONS Should we abandon Early Goal directed therapy? No, but it should be adapted using more physiologic variables and endpoints. Individualization of therapy is probably better than standardization to common minimal endpoints