ECMO - beyond protective ventilation
-
Upload
cosmin -
Category
Health & Medicine
-
view
31 -
download
2
Transcript of ECMO - beyond protective ventilation
Extracorporeal Life Support for
Adults
-a terse primer for the busy intensivist-
Bucuresti, Februarie 2017
Linking ECMO to ARDS
O2-CO2 diagram (ROF)
Alveolar gas equation (ROF)
Relaxation pressure-volume curves
for lung and chest wall (ROF)
WOB and contributions from elastic,
viscous and turbulent forces (Otis)
Alveolar stress distributions (J.
Mead)
Baby lung hypothesis
Initially an anatomical concept
Tidal Volume scaled to IBW
Esp is nearly normal (Esp=E×FRC)
OLV and RM
The baby lung is in fact a sponge lung
- proning changes the densities and
homogenizes the parenchyma
Body weight vs FRC/EELV scaling of
VT
Stress raisers are redefined
Transpulmonary approaches
Strain/Stress emphasis with regards
to PP, sighs, 48h-NMB, spont. (PAV,
APRV, NAVA) vs controlled MV
Inflammation(PET) may sometimes
pertai espe ially to the a y lu g dependent on VTroi/EELVroi
40’s- 0’s
0’s-2000’s
currently
The 40’s to 0’s
The 40’s to 0’s
Modified after Gattinoni
Peffective = Papplied = Pplateau-Ppl
The 40’s to 0’s
Modified after Gattinoni
Peffective = Papplied × (Vopen/Vclosed)2/3 = (Pplateau-Ppl)× (Vopen/Vclosed)2/3
The 40’s to 0’s
In uniformly expanded lungs, transpulmonary pressure is the distending
pressure
In nonuniformly expanded lungs, the effective distending pressure differs from
the transpulmonary pressure and in the appropriate sign to reduce the
nonuniformity
The principal functional risk that it entails is increase in capillary transmural
pressure in regions which become subjected to abnormally high outward-
acting stress
Mead J. et al, J. Applied Physiol,1970 May;28(5):596-608.
There might be no safe ventilation
The 40’s to 0’s
For a given alveolar ventilation optimal frequency will be lowered by
▲ the dead space, by ▲ the nonelastic resistance, or by ▼the elastic
resistance, i.e. by increasing the compliance.
It also appears that for a given dead space,compliance and resistance,
the optimal frequency will increase with increasing alveolar
ventilation.
It must be kept in mind that the above analysis assumes that
expiration is passive.
The 40’s to 0’s
Frutos-Vivar et al, Med Intensiva. 2013;37(9):605---617 .
The 40’s to 0’s
The only feasible means of expanding dependent portions of the lung is by
placement of the body in such a position that ventilation of the normally
dependent portions is facilitated. The prone position appears best suited
for this purpose and merits serious consideration.
Bryan AC, AJRCCM, Vol. 110, No. 6P2 | Dec 01, 1974.
What did the 40’s to 0’s tea h us?
Interfaces between non-uniformly expanded alveoli bring about
excessive stress
This same stress endangers capillary structure
WOB can be optimized with regards to RR and VT for the same MV
WOB can be optimized with regards to the risk of excessive stress
especially in non-uniform, diseased lungs by tweaking the MV
PP could be used to lessen stress as it might be able to educe an FRC
augmentation/ improved lung or V/Q homogeneity
There might be no safe ventilation
The 0’s to 2000’s baby lung and OLV era
Mathe ati al odeli g to syste ati ally characterize P-V curves and objectively derive physiologically and li i ally useful para eters (Venegas, 1997)
The P-V curve is in fact a recruitment curve (Hickling 1998, Albaiceta 2008)
Setting open-PEEP is an expiratory phenomenon, pertaining to the deflation limb (PMC) (Hickling 2001, Suarez-Sipmann 2007)
Open up the lung and keep it open (Lachmann, 1992) – VALI/VILI and the derived biotrauma (Slutsky, 1998) concept are undeniable and PV finds
its place (ARDSnet, 2000)
PEEP was not systematically tailored to recruitability/lung inhomogeneity (ALVEOLI 2004, EXPRESS 2008, LOVS 2008) but subgroup analysis
favours high PEEP in severe ARDS (Briel et al, 2010)
Venegas J., J Appl Physiol (1985). 1998 Jan;84(1):389-95 . Albaiceta GM, Curr Opin Crit Care. 2008 Feb;14(1):80-6.
Da of the 0’s to 2000’s strain/stress beginnings
Santos R.,Anesth Analg 2016;122:1089–100
STEPRM provides a lower mean airway pressure with a higher EELV corresponding to each step – equivalent to lower stress, lower strain
STEPRM entails a milder hemodynamic impact as it gradually improves Elung (Lim 2004, Odenstedt 2005) – equivalent to a RV protective effect,
re i i g “utter’s approa h to ards the est PEEP (Sutter 1975)
STEPRM is capable of reducing the biotrauma (Santos 2016)
Sighs provide similar long term benefits as STEPRM, recently having been shown to elicit favorable effects with regards to strain and
heterogeneity (Mauri 2015)
RMs launched the sighs and the sighs then launched the variable ventilation – delineating the fractal behavior of the lung which entails a power-
law profile as well as avalanche dynamics in opening up (Brewster 2005, Suki 1998, Suki 1994, Barabasi 1996)
Da of the 0’s to 2000’s strain/stress beginnings
Noisy ventilation is to be understood as VT variability following a random law or a fractal auto-correlated pattern
Straightforward benefits comprise an increased EELV for the same PEEP (equivalent to improved strain should the VT remain the same) and an
improved V/Q
Noisy is the e at h: oisy e tilatio , oisy perfusio / ardiopul o ary ypass (Mutch 2000)
APRV a d PAV/PP“ a d NAVA are oisier tha P“/A“B or a y IMV ode
Da of the 0’s to 2000’s strain/stress beginnings
With GE permission . With GE permission .
Karason S et al, Acta Anaesthesiol
Scand, 2000; 44: 578-585
.
FRC/EELV - side stream paramagnetic O2 analyzer with a response time of 480 ms (
AS/3 –Datex Ohmeda Helsinki-Finland, Olegard 2005 and Weissman 2007, LUFU)
Good agreement with helium dilution or CT based techniques (Chiumello 2008)
Hampered by volume-dependence of RecV instead of grams-of-tissue
Shifting the paradigm from IBW scaling to a y lu g capacity scaling
Sizing the Ba y lu g – Mattingley 2011 (Vrel/TLC at 40cmH2O close to 0.45±0.11)
Sizing the Ba y lu g – Beiltler 2016 (VRM at 40cmH2O predicts stress)
V1, V2 = recruited volume
)1(
22
2
ETNETN
V
FRCbaseline
breaths
N
First modern automated approximation of strain analysis (Gonzalez-Lopez 2012),
although in this study a different approach to strain was used (VT/EELV)
EELV derived RecV predo i ately measures a slow fraction of inflation of already
aerated lung tissue and not recruitment of collapsed al eoli (Stahl & Stenqvist
2014). Thus, RecV may just represent a functional recruitment(stress relaxation).
Cdyn dynamics bears the same urse . Decremental best PEEP (Suarez-Sipmann
2007) may just as well represent functional recruitment.
Stress index (constant flow, pressure-time curve) is another functional, ursed
parameter
What did the 0’s to 2000’s tea h us?
VT needs to be tailored physiologically (to the baby lung)
Spontaneous and variable might benefit those with mild and moderate ARDS (Yoshida 2016) but asynchrony( i.e. reverse triggering) (Akoumianaki 2013) has to be
looked for sedulously
Gaining homogeneity through a 48h NMB (Papazian 2010) or/and PP (Guerin 2013) improves the strain/stress, serendipitously, where you would expect the highest
number of unstable alveoli (stress raisers) – severe ARDS
Overdistension and atelectasis coexist to a varying degree – one needs to compromise (Plataki & Hubmayr 2010)
Permissive atelectasis is a useful but risky concept (Page 2003, Fanelli 2009, Albaiaceta 2011), especially enticing when on ECMO
PEEP selection based on lung mechanics and/or absolute esophageal pressure is unrelated to recruitability and similar in mild, moderate and severe ARDS
(Chiumello 2013) (12-15cmH2O). A CT-SP based OLV PEEP was found to be 16cm H2O (Cressoni 2014) equally in recruiters and non-recruiters
Higher PEEP is a prerequisite in recruiters (Briel 2010)
There might be no safe ventilation
Where Lately…i ARD“
ARDS may very well be preventable by means of a primary, secondary and tetiary strategy. Prevention is supported by chaos theory (i.e.
butterfly effect) (Yadav & Gajic 2016)
ARDS mimickers have been described as CRF – with much too often DAD – and have been ascribed a dismal prognosis, sometimes
benefiting from CS (Gibelin & Mekonto Dessap 2016)
Berlin based ARDS comprises only 56.4 % DAD which is a marker of higher hospital mortality (Kuo-Chin Kao 2015)
Latent Class Analysis (LCA) was used to differentiate and characterize two ARDS subphenotypes responding differently to fluid and PEEP
– Phenotype 2 has higher levels of inflammatory biomarkers ,a higher prevalence of shock, lower serum bicarbonate, and a higher
prevalence of sepsis, compared with Phenotype 1 (Calfee 2014-2016)
Toward Smarter Lumping and Smarter Splitting: Rethinking Strategies for Sepsis and Acute Respiratory Distress Syndrome Clinical Trial
Design (Prescott & Calfee 2016)
Functional instead of anatomical recruitment will suffice and Cdyn based strategies seem adequate (Santos 2015)
Inhomogeneity indexes using rapid response O2 analyzer such as LUFU (Weissman 2007) will be able in the future to guide MV at the
bedside similarly to ALPE, MIGET (Bikker & Gommers 2014)
Driving Pressure has been shown to be a rather useful bedside touchstone for the a y lu g (Amato 2015)
Transpulmonary approaches could be able to yield a transpulmonary DP (Chiumello 2016, Loring & Hubmayr 2016)
ACP scores have been described and MV is more hemodynamic as never before (Mekonto-Dessap & Vieillard Baron 2016)
Transpulmonary approaches
Chiumello, Intensive Care Med (2014) 40:1670–1678.
Talmor,Crit Care Med 2013;41:0–0.
(Δ PPL/ΔV) / (Δ PAO/ΔV) = ECW / E RS (static)
PPL = PAO × ECW/ERS
PL,EXP =Total PEEP - PPL at end expiration
PL,PLAT = PAO,PLAT - PPL at end inspiration
PL=PAO×EL/ERS
Directly measured end-expiratory
transpulmonary pressure: Airway pressure at PEEP – esophageal pressure at PEEP
Release-derived end-inspiratory
transpulmonary pressure:
(Airway pressure at endinspiration-
atmospheric pressure)-(esophageal
pressure at end-inspiration-esophageal
pressure at atmospheric pressure)
Release-derived end-expiratory
transpulmonary pressure: PEEP-(esophageal pressure at PEEP-
esophageal pressure at atmospheric
pressure)
Strain/Stress concepts
ΔPL=(Pawplateau-Pesplateau)-
(Patm-Pes at atm pressure)
Strain= ΔV/FRC=(ΔVT+ΔPEEP)/FRC
ΔPL=ΔV×EL=ΔV/FRC×FRC×EL=ΔV/
FRC×ELsp=Strain×ELsp
ELsp=13.5
Max Strain = 2
Max Stress=27cmH2O
ΔPL=ΔPaw×EL/(EL+ECW)
The strain/stress concept is only a GLOBAL refinement
Strain/Stress caveats
In the presence of intratidal recruitment, strain will be lower
Chiumello, AJRCCM, Vol 178. pp 346–355, 2008.
Strain/Stress implications
Scaling VT to IBW is inadequate
Chiumello, AJRCCM, Vol 178. pp 346–355, 2008.
Strain/Stress implications
Plateau pressure based PV is inadequate
Chiumello, AJRCCM, Vol 178. pp 346–355, 2008.
Statics and dynamics in strain analysis
Alveolar stability may be the key as Albaiceta had stated in 2011.
Maximizing recruitment and low VTs are the equivalent to Protti’s study.
This is in fact equivalent to a low DP. Protti, Crit Care Med 2013; 41:1046–1055
Strain and stress at the bedside
EL=ΔPEEP/ΔEELV the chest wall and abdomen gradually can accommodate
changes in lung volume Stenqvist,Acta Anaesthesiol Scand 2012
Strain and stress at the bedside
• EL= ΔPEEP/ΔEELV
• ELSP = EL × FRC
• Stress = Strain × ELSP
• Strain global = (VT+VPEEP)/FRC
Acta Anaesthesiologica Scandinavica 60 (2016) 69–78
Caveat: there is VT and/or VPEEP dependent recruitment which will alter the strain as they are not taken into
account. The actual strain will be lower. Using EELV instead of FRC will underestimate the strain. Gonzalez-
Lopez et al. found 0.27 using EELV in 2012.
Let us keep it simple - DP
For a lung stress of 24 and 26 cmH2O, the optimal cutoff value for the airway driving pressure were 15.0
cmH2O (ROC AUC 0.85, 95 % CI = 0.782–0.922); and 16.7 (ROC AUC 0.84, 95 % CI = 0.742–0.936).
Conclusions: Airway driving pressure can detect lung overstress with an acceptable accuracy. However, further
studies are needed to establish if these limits could be used for ventilator settings.
Chiumello et al. Critical Care (2016) 20:276
It’s all a out the e ergy i put
Gattinoni et al, Intensive Care Med (2016) 42:1567–1575
Dissipated Undissipated
Bei g glo al does ’t help lung inhomogeneities
Cressoni et al, AJRCCM, Vol 189, Iss 2, pp 149–158
There might be no safe ventilation
HEALTHY MILD
MODERATE
SEVERE
PEEP
?
Bei g glo al does ’t help lung inhomogeneities
High regional lung strains may be present even when global lung strain is within acceptable limits.
Such localized metabolic activation is prevented by reducing and homogenizing regional tidal strain
with high PEEP and low VT.
Wellman et al,Crit Care Med 2014; 42:e491–e500
There might be no safe ventilation
Dynamic [18F]fluoro-2-deoxyd-glucose
scans to quantify metabolic activation,
indicating local neutrophilic
inflammation.
Bei g glo al does ’t help lung inhomogeneities
Borges et al, Acta Anaesthesiologica Scandinavica (2016)
There might be no safe ventilation
Synchrotron imaging was used
to measure lung aeration and
regional-specific ventilation (sV ̇ ).
LUNG PROTECTIVE VENTILATION
RV PROTECTIVE VENTILATION
LESS MECHANICAL POWER PRESERVE/INDUCE HOMOGENEITY
“OMETIME“ IT’“ NOT ENOUGH
THERE MIGHT BE NO SAFE VENTILATION
DISSOCIATE GAS EXCHANGE FROM MECHANICS
ECLS – ECMO/ECCO2R
LUNG PROTECTIVE VENTILATION
RV PROTECTIVE VENTILATION
LESS MECHANICAL POWER PRESERVE/INDUCE HOMOGENEITY
“OMETIME“ IT’“ NOT ENOUGH
THERE MIGHT BE NO SAFE VENTILATION
DISSOCIATE GAS EXCHANGE FROM MECHANICS
ECCO2R – SUPERNOVA TRIAL = OLV (EX) + UPV
Friday Night Ventilation
300 200 100
MILD MODERATE SEVERE
DP = Paw_pl-PEEPt = �
<20cmH20 <15cmH20 <12cmH20
INHOMOGENEITIES
� 6ml/kg IBW <6ml/kg IBW <4ml/kg IBW
� � + ≈ H >15cmH20
�
ECMO
ECCO2R
NMB 48h
Prone Position
Modified after Gattinoni L.
ECMO didactics
Introduction to ECMO History
Current Status
Risks and benefits
Membrane gas exchange physics and
physiology
Oxygen content, delivery and consumption
Shunt physiology
Types of ECMO
Future applications
Research
Physiology of the diseases treated with
ECMO Neonatal RF
Pneumonia
ARDS
Pulmonary embolism
Sepsis
Postoperative congenital heart disease
Heart transplantation
Cardiomyopathy and myocarditis
PreECMO procedures Notification of the ECMO Team
Cannulation procedures
Initiation of bypass
Responsibility of team members
Criteria for ECMO Patient selection
Selection criteria
PreECMO evaluation
Contraindications
Selection of ECLS support(VA,VV, VA-V)
Blood products and coagulation Blood products and interactions
Blood product management of the bleeding
patient
Coagulation cascade
Blood surface interactions
Heparin pharmacology
Activated clotting times
Anticoagulant monitoring studies
Protamine, Amicar and other drugs
Recombinant clotting factors
Disseminated intravascular coagulation
Mechanical emergencies and
complications on ECMO Circuit disruption
Raceway rupture
Cavitation
System failure
Air embolus
Inadvertent decannulation
Clots
Management of complex ECMO cases Surgery on ECMO
Transport on ECMO
Weaning from ECMO Technique s and complications
Pump and gas flow weaning techniques
ACT during weaning
Ventilator changes during weaning
Trial off
Decannulation from low flow
Decannulation Personnel needed
Medications required
Potential complications
Vessel ligation
Vessel reconstruction
Percutaneous approach
Post ECMO complications Platelet and electrolyte alterations
Short and long term development
outcome Institutional follow-up protocol
Literature review
Ethical and social issues Consent process
Parental and family support
Withdrawal of ECMO
Medical emergencies and complications
during ECMO Intracranial and other haemorrhages
Pneumothorax and pneumopericardium
Cardiac arrest
Arrhythmias
Hypotension and hypovolemia
Hypertension
Severe coagulopathy
Seizures
Hemothorax and hemopericardium
Uncontrolled bleeding
Renal failure
Daily circuit management on ECMO Aseptic techniques
Pump and gas flow
Pressure monitoring
Blood product infusion techniques
Circuit infusions
Management of anticoagulation
Circuit checks
Hemofiltration setup
ECMO equipment
Physiology of VA and VV ECMO Indications
Vessel cannulation
Physiology
Advantages and disadvantages
Cannulation and initiation of ECMO support
Daily patient management on ECMO Bedside care
Fluid, electrolytes and nutrition
Infection control
Respiratory support
Sedation and pain control
Hematology
Cardiac support
Pharmacological issues
Psychosocial
Abbreviation Definition Details
ECLS Extracorporeal life
support
All EC technologies and life support components including
oxygenation, CO2 removal, HD support; renal and liver
support may also be incorporated
ECMO
Extracorporeal
membrane
oxygenation
Older traditional term for EC life support that omits
reference to inherent additional life supports such as
haemodynamic support and CO2removal
VV ECLS
Veno-venous
extracorporeal life
support
Deoxygenated blood is drained from one or more major
vein and oxygenated blood returned to the RA; supports
respiratory function only and requires native heart
function to deliver oxygenated blood to the tissues
VA ECLS
Veno-arterial
extracorporeal life
support
Deoxygenated blood is drained from one or more major
vein and oxygenated blood pumped back into a major
artery, thus providing tissue perfusion in the absence of
adequate native heart function
ECPR
Extracorporeal
cardiopulmonary
resuscitation
Extracorporeal life support instituted during, and as an
adjunct to conventional CPR
VV-ECCO2R
Extracorporeal
membrane carbon
dioxide removal
Selective CO2 removal
AV-ECCO2R Arterio-venous
extracorporeal life
support
Pumpless ECLS, driving pressure is patie t’s arterio-
venous ΔP, supports CO2 removal
Gaffney et al,BMJ 2010;341:c5317)
Gaffney et al,BMJ 2010;341:c5317)
ECLS – most common conditions
ECLS – beginnings
Laffey & Kavanagh,Am J Respir Crit Care Med. 2017 Feb 1
1667 - Jean Baptiste Denis, physician to King Louis XIV experimented on cross-transfusing the blood of a human with that of a lamb, patient survived
1930 - John Gibbon MD and Mary Gibbon MD created a roller pump device for EC support as a response to the death of a patient suffering with PE
1953 - first use of the device by JG on 18 year old Cecilia Bavolek with ASD
1954 - cardiac surgeon C. Walton Lillehei operated via cross circulation with a bubble oxygenator he and Richard DeWall invented
1956 - Clowes invented membrane oxygenator
1957 - Kammermeyer invented the silicone rubber which was strong enough to withstand hydrostatic pressure and yet permeable to gas transfer
1972 - JD Hill first successful adult use of ECLS outside of the OR (posttraumatic ARDS, male, 24year old, aortic rupture)
1972 - Bartlett first to use ECLS on neonates (two year old boy, correction of TGV, cardiac failure)
1975 - Bartlett first to use ECLS for newborn with RF after meconium aspiration-Esperanza
1985 - first RCT o eo atal RF, Bartlett; play the i er ra do izatio ’
1989 - O’Rourke, ECL“ for eo ates ith PPH
1996 - UK Collaborative ECMO trial group – These preliminary results demonstrate the clinical effectiveness of a well-staffed and organised neonatal ECMO service. ECMO support should be actively considered for neonates with severe but potentially reversible respiratory failure
1979 - Zapol et al, RCT in VA ECMO for adult RF, 10% SR for both groups
1994 - Morris et al, RCT in VA ECCO2R for adult RF
2009 - Peek at al, CESAR trial, regionalized VV ECMO approach for adult RF
.
We recommend transferring of adult patients with severe but potentially reversible respiratory failure, whose
Murray score exceeds 3·0 or who have a pH of less than 7·20 on optimum conventional management, to a centre
with an ECMO-based management protocol to significantly improve survival without severe disability. This
strategy is also likely to be cost effective in settings with similar services to those in the UK.
ECLS – beginnings
ELSO registry January 2016
Let’s get i
ECMO is coming to your adult ICU
Barbaro et al, 2015
Let’s get read
ECMO is coming to your adult ICU
Let’s get read
ECMO is coming to your Adult ICU
ELSO Registry data
Let’s get read
ECMO configurations
ML is parallel to NL
ML is in series with
NL
Euler Liljestrand ▼
Lung alkalosis▲
RECIRCULATION
-
+
Resorbtion
atelectasis if
FiO2ML > FiO2NL
ECMO membrane
Cross current flow between gas and blood
Fi k’s La of Diffsion
Effective Diffusivity
=porosity
= constrictivity
τ= tortuosity
ECMO membrane
≈ hours weeks
Oxygenator overview
ECMO circuit
P1 = negative suction
pressure, see for
chattering/cavitation(
▼preload?)
P2 = positive ejection
pressure
P3 = pressure drop of
the oxygenator
ECMO gas transfers
O2content/dl = (Hb×SatO2×1.39)+(pO2×0.0031)
VO2ML = BF×(CoutO2×CinO2)
VO2NL = CO×(CaO2-CvmixO2)
VO2tot = VO2ML+VO2NL
VCO2NL = alvPCO2×RR×(VT-VD)=NLETCO2×MV
VCO2ML = MLETCO2×GF
DO2 depends on Hb and BF/CO/R
VCO2 depends on GF and is relatively
independent of BF
BF/CO > 60% => adequate oxygenation dis
soci
ati
on
Sch
mid
t et a
l
Inte
nsiv
e C
are
Me
d (2
01
3) 3
9:8
38–
84
6
VV ECMO recirculation
VO2ML = BF×(CoutO2×CinO2)
VV ECMO recirculation
VV ECMO recirculation
Gillon et al, Crit Care Med 2016; 44:e583–e586
VV ECMO – which shunt to begin with
VV ECMO – PaO2 as a function of R/BF -different BF-
Increasing R/BF will forestall oxygenation especially at low CO/BF
At high CO/BF, it does ’t atter a y ay
VV ECMO – SvmixO2 as a function of BF/CO -different VO2s-
SvmixO2 chiefly depends on SvO2(tissues), BF/CO and R when Hb is constant
At increasing VO2, for steady SvmixO2 one needs to increase the BF
Increasing Hb makes sense especially when not on ECMO
Full ECMO support should help the clinician resist unnecessary transfusion
VV ECMO – SvmixO2 as a function of Hb -different BF-
VV ECMO – PaO2 as a function of BF -different CO-
I reasi g the BF ill saturate the shu t a d e e tually the PaO2 ill soar, quite similar to the O2 dissociation curve
At the same BF, increasing the CO will bring about an increase in DO2 and
consequently PvO2 will rise. The same holds true if CO is stable and BF will
vary
VV ECMO – PvO2 as a function of BF -different CO-
VV ECMO – PaO2 as a function of BF -different shunts-
Increasing shunts(severity of lung injury) will mandate higher BF(BF/CO) to
achieve the same PaO2
VV ECMO – PaO2 as a function of BF -different FiO2NL-
I reasi g the BF ill e e tually saturate the urre t shu t a d the PaO2 will then soar, similar to the ODC. The higher the FiO2NL, the quicker it will
soar.
Increasing FiO2NL creates the chance of O2 toxicity.
VV ECMO – PaO2 as a function of BF -different FiO2ML-
I reasi g the BF ill e e tually saturate the urre t shu t a d the PaO2 will then soar, similar to the ODC. The higher the FiO2ML, the quicker it
will soar.
I reasi g FiO2ML ay ot e ise as it’s ee sho to eli it resorbtion
atelectasis when FiO2ML > FiO2NL .
Cannulae - definitions
Access cannulae – drain blood from the venous system into the ECMO circuit
Single-stage cannulae
Multi – stage cannulae
Return cannulae – deliver blood back to the patient
Single stage
Distal perfusion cannula – deliver blood antegradely into the FA
Double lumen cannulae
Cannula length
Long cannulae (55cm) – e ous
Short cannulae (15-25cm) – arterial . Return blood in both VA ECMO and VV ECMO
(fem-jug) and to access blood in high flow VV or in VAV. They have side ports to connect
to distal perfusion cannulae.
Multi-stage access cannula
Return cannula
Distal perfusion cannula
Double lumen cannula
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
Doufle, Crit Care. 2015; 19: 326
ECMO MODES
VV ECMO VA ECMO
Femoro-femoral
High-flow
Femoro-jugular
Dual lumen(Avalon)
Standard fem-fem
Emergency fem-fem
High-flow
Central
VPA ECMO
Femoro-femoral VV
Cavo-atrial flow direction
Access cannula(multistage) within hepatic IVC, 21-25 F
Return cannula(single stage) within RA, 21-25 F
Δd=10-15cm
Limited maximum flow rates, bed bound
High-flow VV
Starts with bi-femoral approach
An additional short access (arterial) cannula is
inserted in the RJV with the tip in the SVC, at a
safe distance from the RA cannula in order to
prevent visible recirculation
Bi-cavo-atrial flow direction
Allows high flows, easy to start off CRRT but
increases rate of complications
Fem-jug VV
Access cannula tip at the inferior cavo-atrial
junction, 21-25 F multistage
Return cannula is short (arterial), tip in the very low
SVC, 19-23 F single-stage
Cavo-atrial flow direction
Allows high flows, easy to start off CRRT
Bed bound
Single cannula VV
Single cannula, two lumens, RJV
Two access stages (SVC & IVC), return port towards
TV
Bi-Cavo-atrial flow direction
Allows ambulation but technical difficulties at
insertion, large cannula (27-31 F), awkward to
implement CRRT
Fem-fem VA
Cavo-atrial flow direction
Access cannula(multistage) within hepatic IVC, close to
RA, 21-25 F
Return cannula(single stage) within the common FA, 17-
21 F, lying in the common IA/lower aorta
Backflow cannula is mandatory, inserted in the common
FA and directed into the superficial FA
May need conversion to V-A-V if lung injury coxists with
preserved cardiac function
Emergency Fem-fem VA
Cavo-atrial flow direction
Access cannula(multistage) within hepatic IVC, close to
RA, 19-21 F
Return cannula(single stage) within the common FA, 15
F, lying in the common IA/lower aorta
Ba kflo a ula is a dator , BUT it’s i serted o l after ECMO is established
Same disadvantage as the previous plus possible
inability to support CRRT
Much easier to initiate due to smaller cannulae
High-flow VA
Access cannula tip at the inferior cavo-atrial
junction, 21-25 F multistage
Return cannula is short (arterial), tip in the very low
SVC, 19-23 F single-stage
Cavo-atrial flow direction
Allows high flows, easy to start off CRRT
Bed bound
ECMO START-UP TOOLKIT
Patient Selection -VV ECMO-
1. In hypoxic respiratory failure due to any cause (primary or secondary) ECLS should be considered when the risk
of mortality is 50% or greater, and is indicated when the risk of mortality is 80% or greater.
a. 50% mortality risk is associated with a PaO2/FiO2 < 150 on FiO2 > 90% and/or Murray score 2-3.
b. 80% mortality risk is associated with a PaO2/FiO2 < 100 on FiO2> 90% and/or Murray score 3-4 despite
optimal care for 6 hours or more.
2. CO2 retention on mechanical ventilation despite high Pplat (>30 cm H2O)
3. Severe air leak syndromes ELSO 2013
Ca 't make my own decisions
Or make any with pre isio
Ca 't make my own decisions
Or make any with pre isio
Absolute contraindications to all forms of ECMO 1. Progressive and non-recoverable heart disease (and not
suitable for transplant)
2. Progressive and non-recoverable respiratory disease
(irrespective of transplant status)
3. Chronic severe pulmonary hypertension
4. Advanced malignancy
5. Graft versus host disease
6. Unwitnessed cardiac arrest
7. Cachexia due to an underlying progressive chronic
disease
Specific absolute contraindications to veno-venous
ECMO 1. Severe (medically unsupportable) heart
failure/cardiogenic shock
2. Severe chronic pulmonary hypertension and right
ventricular failure (mean pulmonary artery
pressure approaching systemic blood pressure)
3. Cardiac arrest (ongoing)
Relative contraindications to all forms of ECMO 1. Age>70
2. Duration of conventional mechanical ventilation >7
days, with high inspiratory pressures
(Pplat>30cmH20), high FiO2 (FiO2 >0.8)
3. Trauma with multiple bleeding sites
4. Severe immunosuppression (transplant recipients >30
days, advanced HIV, recent diagnosis of
haematological malignancy, bone marrow transplant
recipients)
5. Intracranial bleeding
6. CPR duration >30 min without documented
neurological recovery
7. Severe multiple organ failure
8. CNS injury
9. BMI <18 or >40
Adapted after
Guy’s a d “t Tho as’ NH“ Fou datio Trust
Hemodynamics – mercurial behaviour?
In our cohort, an increase in pH was the only ECMO-related effect that
correlated with the reduction in pressor requirements we observed.
Improved PVR/RV function
(ACP score▼)
Improved ino-vasopressor
responsiveness
COMPREHENSIVE TTE/TOE before assignment
to rule out severe RV/LV failure
The ACP score -predicting VV ECMO dependent hemodynamic benefit-
Pneumonia as cause of ARDS
Driving pressure > 18 cmH2O
PaO2/FiO2ratio < 150 mmHg
PaCO2 > 48 mmHg
1
1
1
1
> 2
Comprehensive echo exam is
mandatory
Mekontso Dessap et al
Intensive Care Med (2016) 42:862–870
ACP defined as a dilated Rv in ME4CV
[end-diastolic RV/left ventricle (LV) area
ratio [0.6] + septal dyskinesia in the
transgastric short-axis view of the heart
Having a plan might be of help
We recommend transferring of adult patients with severe but potentially reversible respiratory failure, whose
Murray score exceeds 3·0 or who have a pH of less than 7·20 on optimum conventional management, to a centre
with an ECMO-based management protocol to significantly improve survival without severe disability. This
strategy is also likely to be cost effective in settings with similar services to those in the UK. CESAR TRIAL
Optimise ventilator settings
Slow recruitment manoeuvre, Cdyn-
PEEP/EELV-PEEP(Engstrom)/FiO2-
table based/Express-trial/Gattinoni
VT<6ml/kg/IBW, Pplat<30, DP<15
I:E ratio 1:1
RR 20-30
Optimise IAP (<20)
Consider PP, NMB
Optimise hemodynamics
ITBVI 850-1000ml/m2
EVLWI<15
PMSA approach (Parkin)
TTE/TOE screening
Flow monitoring (lactate, ScVO2,
diuresis, CRT, mottled skin score)
Keep alb > 30g/L
Hb>7-8g/dl
+
Negative fluid balance –
deresucitation
Assess FR, then fill
IF PaO2/FiO2<80-100 OR pH<7.2(RA) OR not able to achieve PV AND <7days of MV
Assess Murray score, if >3 OR ph<7.2
CT chest + RM
Significant LR
APRV
Phigh 24-28
Plow 0
I:E 10:1
HFOV
RM, bias flow 40
Initial f 6Hz
CDP<30 cmH2o
Significant ΔD-V
PP
Minimal LR OR
recruitable BUT barotrauma present
ECMO
FiO2 > 0.5
PEEP > 10
ECCO2R
pH<7.2
AND
FiO2<0.5
At 24-48h
APRV
PaO2/FiO2<80 OR
pH < 7.2 (RA) OR
Phigh>30
HFOV
PaO2/FiO2<80 OR ΔPaO2/FiO2<38% OR
pH < 7.2 (RA) OR
CDP >30
Prone position
PaO2/FiO2<80 OR
pH < 7.2 (RA) OR
Phigh>30
ECMO/ECCO2R
Patient Selection -VA ECMO-
Potentially reversible, severe, refractory cardiogenic shock in patients who have
failed alternative therapy.
DECATECHOLAMINIZATION Singer & Matthay, Crit Care. 2011;15:225
Selective V1A(ie
Selepressin) (He X et al) Clonidine
&Dexmedetomidine
Beta-blockers (Morelli
et al) Ivabradine (MODIFY)
Levosimendan
(LEOPARD)
Omecamtiv Mecarbil ?
Specific physiological indications (after 1-12
hours of commencement of inotropic support)
1. persisting lactate >3mmol/L
2. persisting cardiac index <2L/min/m2
3. evidence of end organ dysfunction
4. trans-thoracic echocardiography with left
ventricular ejection fraction <30% and aortic
velocity time integral (Ao VTI) <8cm.
Causes of cardiogenic shock
1. Viral cardiomyopathy
2. Cardiac toxic drug overdose
3. Septic cardiomyopathy
4. Peripartum cardiomyopathy
5. Massive pulmonary embolism
To be considered patients should:
1. Have no significant chronic medical co-morbidities
2. Be aged 60 years old or less
3. Be deemed to have a potentially reversible
aetiology
4. Be within 24 hours of the onset of cardiogenic shock
5. Not to be excluded from cardiac transplantation
Contraindications to VA-ECMO
1. Symptomatic chronic cardiac failure (NYHA 3 or 4)
2. Progressive chronic respiratory failure
3. Advanced malignancy
4. End-stage renal failure requiring dialysis
5. Advanced HIV
6. Significant immunocompromise from any cause
7. Child-Pugh B or C chronic hepatic failure
8. Any other significant medical co-morbidity
Contraindications to VA-ECMO
9. Age >60 years old
10. Severe peripheral arterial disease, severe aortic
valve regurgitation or aortic dissection precluding
cannulation
11. Any condition precluding heparinisation, including
active haemorrhage, intracerebral haemorrhage or
allergy
12. CPR >45 minutes (estimated time to cannulation)
13. Out of hospital cardiac arrest
Patient Selection -MODS precludes ECMO success-
Respiratory failure with septic shock & >3 of:
1. Lactate > 10
2. Norepinephrine > 1.5mcg/kg/min
3. Severe myocardial dysfunction
4. Advanced microcirculatory failure
Cardiogenic shock & > 3 of:
1. Lactate > 15
2. Advanced microcirculatory failure
3. AST/ALT > 2000 or/and INR > 4.5
4. Anuria > 4h
ECMO Guideline, Alfred Hospital, Melbourne
Mo za’s flo hart, “a galli et al, ECMO-
Extracorporeal life support in adults, 123-124
B. Riou, Réanimation 28 (2009) 182–186
Ca 't make my own decisions
Or make any with pre isio
Ca 't make my own decisions
Or make any with pre isio
Ca 't make my own decisions
Or make any with pre isio
Ca 't make my own decisions
Or make any with pre isio
To su arize…
Eddy Fan et al,Intensive Care Med (2016)
42:712–724
Basic tips and tricks
1. VV ECMO – bifemoral vein approach is preferred especially in
retrieval/emergency situations
2. Cannulation SHOULD be US-guided
3. Fluoroscopy SHOULD be used to document position except for peri-
arrest/arrest situations
4. US may be used instead of fluoroscopy
5. Once wires are in place, give 50UI/Kg of heparin systemically
6. Before starting ECMO, check ACT > 200sec
7. Establish baseline anticoagulation 10UI/Kg/H, APTT 1.5-2
8. Single dose Teicoplanin 400mg and Gentamicin 5mg/kg at line
insertion. Atb are not otherwise specifically prescribed.
9. VV ECMO – target BF/CO > 60%
10. VA ECMO – target 70-80% of COpred, aim for PP>10-20mmHg, SR
11. VAV – aim 1L venous flow using gate clamp
12. VA ECMO - Flow meters on the combined arterial return and the
aortic return to assess backflow cannula flow
13. VA ECMO – monitor using NIRS/pulse oximeter bilaterally
14. Assess for Harlequin syndrome
15. Never stop gas flow when on VA
Circuit flow 50-80 mL/kg/min
Sweep gas flow 50-80 mL/kg/min
Fractional inspired oxygen (sweep gas) 100%
► reduce accordingly when on VA
► reduce so that ML FiO2 ≈ NL FiO2
Inlet pressure (centrifugal pump) > -100 mmHg
Oxygen saturation (return cannula) 100%
Oxygen saturation (drainage cannula) > 65%
Arterial oxygen saturation VA > 95%; VV: 85%-92%
Mixed venous oxygen saturation > 65%
Arterial carbon dioxide tension 35-45 mmHg
pH 7.35-7.45
Mean arterial pressure 65-95 mmHg
Hematocrit 30%-40%
Activated clotting time 1.5-2.0 times normal
Activated partial thromboplastin time 1.5-2.0 times normal
Platelet count > 100,000/mm3 or >100,000 if bleeding occurs
Basic tips and tricks
Initial Settings and Goals After the
Institution of ECMO
Adapted after Sidebotham et al, Journal of
Cardiothoracic and Vascular Anesthesia, Vol
24, No 1 (February), 2010:
Arterial blood gases 3-4 hourly
Pre - and post - oxygenator blood gases daily
Activated coagulation time 1-2 hourly
Complete blood count 6 hourly
Coagulation tests 6 hourly
D dimer & AT3 daily
Thromboelastograph 12 hourly
Blood chemistry, renal function, and liver
function 12 hourly
Plasma free hemoglobin (<0.1) 12 hourly
Blood cultures from the circuit daily
Triglycerides daily
CK, Troponin, NT ProBNP/BNP (BNP t1/2 is
shorter) if VA ECMO daily
Basic tips and tricks
Schedule of Initial Point of Care Testing
Adapted after Sidebotham et al, Journal of
Cardiothoracic and Vascular Anesthesia, Vol
24, No 1 (February), 2010:
Comprehensive cardiac
echo TTE/TOE daily
pre/post ECMO
Comprehensive lung US
daily
Douflé et al. Critical Care (2015)
Basic tips and tricks
100% FiO2 test (Cilley) – VV daily
Cdyn trends – VV daily
SRM to challenge the baby lung – VV ?
Transmembrane pressure gradients/R – VV/VA daily
Pre and post oxygenator blood gas – VV/VA daily
Sedation management – no fentanyl/midazolam
(Shekar K, Crit Care 2012), choose propofol and
alfentanil
Inotropic support to preserve PP 10-20mmHg – VA
Daily echo + AoVTI – VA
Cerebral oximetry ± lower/upper limbs bilaterally – VA
Right radial/brachial ± anywhere but the cannulated
femoral artery – VA
VV ECMO – refractory hO2
Strategies to i rease the lood’s o ge o te t:
i.increase of ECMO flow;
ii.increase of blood oxygen-carrying capacity
Strategies to reduce recirculation
Strategies to reduce oxygen consumption:
i.sedation and neuromuscular blockade
ii.therapeutic hypothermia
Manipulation of CO and intrapulmonary shunt:
i.β-blockers infusion
ii.prone positioning
Switch to VA ECMO or a hybrid configuration(VAV)
Adapted after Montisci, ASAIO Journal
2015; 61:227–236:
VV ECMO – refractory hO2
Levy B,Intensive Care Med (2015) 41:508–510:
VV/VA ECMO – clot formation
Pump head (centrifugal)
►Rising plasma free hemoglobin
►Change in sound of pump
Oxygenator
►Increasing pressure gradient across the oxygenator
►Fall in postoxygenator PO2
►Increase in sweep gas needed to maintain PaCO2
Tubing/bridge
►Visible fibrin strands or thrombus
Nonspecific markers of blood clot formation
►Increasing D-dimers or fibrin degradation products
►Increased bleeding
Adapted after Sidebotham et al, Journal of
Cardiothoracic and Vascular Anesthesia, Vol
24, No 1 (February), 2010:
VV ECMO – recirculation
At higher circuit flows, recirculation
decreases effective circuit blood flow
Abrams et al, ASAIO
Journal 2015; 61:115–121
VV ECMO – recirculation
ELSO Guidelines, 2015
Increasing the distance between the cannulae
Use a bicaval, dual lumen cannula
Use a high flow configuration – it will cause less negative pressure => less recirculation
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
Constant improvement over the days
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
One patient lived
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
Patie t a gai s VO through his o NL hilst patie t’s NL de ours O o i g from the ML.
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
One patient lived
Sangalli et al, ECMO-Extracorporeal life
support in adults, 123-124
Trends from a progressive reduction in sweep flow
VA ECMO fem-fem – specifics
VA ECMO fem-fem – specifics (1st)
Optimise ventilation
Increase ECMO flow to push the mixing cloud
Switch to V-A-V configuration or even VV if cardiac function has improved
Switch to central VA (right subclavian or true central)
Christopher Lotz et al. Circulation.
2014;130:1095-1104
HARLEQUIN
Brain ischaemia
Cardiac ischaemia
VA ECMO fem-fem – specifics (2nd)
Increase ECMO BF to reduce pulmonary BF in the setting of a good RH but you may risk
an even more increased LVEDP
Get an IABP at once if not already initiated (A. Combes 2016, Reanimation)
Do an atrial septostomy
Switch to central VA (right subclavian or true central)
Christopher Lotz et al. Circulation.
2014;130:1095-1104
Start Levosimendan
Not so decatecholaminizing after
all, is t’t it?
After septostomy
VA ECMO fem-fem – specifics (3rd)
Backflow cannula is mandatory if not in periarrest context
Post decannulation period is also critical as it can elicit a reperfusion compartment
syndrome
Ventilation strategies
Permissive atelectasis is tricky
As the lung was conceived to contain air
I many patients the lung may proceed to total
consolidation before recovery occurs, but this might be
avoided by maintaining some inflation pressure as high
pressures are decreased, and by supplying nitrogen to
prevent adsorption atele tasis ELSO 2016
Ventilation strategies
Decreasing ΔP/Pplateau/Pinsp/RR decreased meanPaw gravitational atelectasis
FiO2ML > FiO2NL � � ��� � (VV)
VCO2ML ↓RQNL PAO2 ↓ if PiO2 and PaCO2 are the same
Sighs are part of our normal physiology
When on VA, beware of pulmonary tissue alkalosis; PV still aplies
a Assuming airway resistance of
10cmH2O/l/s
Gattinoni L, Curr Opin Crit Care 2017 &i
Critical Care (2016) 20:130
Ventilation strategies
Weaning strategies -VV ECMO-
If CXR/ LUNG ultrasound improves, Cdyn/ TV/ Cilley test improve, consider weaning
Decrease flow in steps to 1L/min at sweep FiO2 100% OR decrease flow to 2L/min then
decrease sweep FiO2 to maintain SaO2 > 95%
When SaO2 stable on these settings, on VV, trial off by clamping sweep on vent rest
settings PSV or CPAP 20 cm H2O. If SaO2 >95and PaCO2 <50 x 60 mins, come off OR
further observe on minimum PS
If PaCO2 >50 stay on at low flow, go to selective CO2 clearance mode
Venous duplex 24-48 hours following the removal of cannulae to assess for DVT
after ELSO Guidelines 2016
Weaning strategies -VA ECMO-
Daily comprehensive echo (LUNG + CARDIAC) exam is mandatory
Never stop the sweep gas flow!
As the BF is reduced in steps (0.5l) to 25%, perform repeated echos – observe AoVTI
(>12cm), RV-LV interaction, LV/ RV distension
Minimum low flow is 1l/min for 30 min max OR clamp circuit and allow recirculation
for trial period of 30 minutes to 4 hours AND flush cannulae with heparinized saline
continuously
Anticoagulation is paramount during weaning sessions (consider APTT 2-2.5)
Consider paired central venous and arterial blood gases to assess tissue perfusion
Consider Levosimendan if not already administered (Affronti, ASAIO Journal 2013;
59:554–557)
Venous/arterial duplex 24-48 hours following the removal of cannulae to assess for
DVT
after ELSO Guidelines 2016
Weaning strategies -VA ECMO-
LVOT_VTI > 10cm
LVEF > 20-25%
TDSa > 6cm/s
ECMO flow 1-1.5l/min
LVOT_VTI > 12cm
LVEF > 20-25%
TDSa > 6cm/s
Aissaoui ,Réanimation 2015
EC
MO
flo
w 1
-1.5
l/m
in
Weaning strategies -VA ECMO-
ΔRVEDV > 0
AND
ΔLVEDV < 0
As ECMO flow is weaned