Pharmaceutical Considerations

during ECMO
John McGuinness – Cardiothoracic
Critical Care Pharmacist
 Objectives
• To understand the complex
interaction of critical illness, ECMO
and organ dysfunction on the
pharmacokinetics of drugs.
• The need for good research in drug
handling in adults.
• General principles of dosing and the
importance of therapeutic drug
monitoring.
 ECMO
• The purpose of ECMO is to allow time for intrinsic
recovery of the lungs and heart.

• VV ECMO involves a double lumen into the vena-

cava, blood is drained and oxygenated and is
returned and angled to direct blood across the
Tricuspid valve. Pulmonary blood flow is
maintained. – used in reversible severe
respiratory failure.

• VA ECMO involves blood drained from the Right

Atrium and oxygenated blood returned via the
ascending aorta thereby bypassing the pulmonary
circulation, it decreases PAP and provides
cardiac/lung support. – used pre/post transplant
• Resent studies in adults report a survival rate to
discharge of 78% in cases of severe respiratory
failure. Survival for VA ECMO in adults with cardiac
failure is up to 53%.

• A typical circuit consist of PVC tubing, a hollow fibre

oxygenator and a heat exchanger. For lipophilic drugs
circuit size and type of oxygenator seem to affect
aborption.

• Alterations in pharmokinetics (Pk) are infrequently

addressed despite the data from 35 neonatal studies.

• Physiologic processes that influence absorption,

distribution, metabolism and excretion are immature
in neonates and cannot be extrapolated to adults.
• Pharmacokinetic changes during ECMO

• A recent review of PK studies found only 5 case

reports in adults.

• In critically ill patients PK changes can be significant –

systemic inflammatory responses, organ dysfunction
and organ support and drug interactions are all known
to affect PK.

• ECMO can alter PK by increasing the volume of

distribution Vd, decreasing drug elimination and
sequestering drugs in the circuit.

• Effective dosing of analgesia, sedation and antibiotics

are a significant challenge.
• Sequestration in ECMO circuits is influenced
by the drugs molecular size, degree of
ionisation, lipophilicity and plasma protein
binding.

• Given the large surface area of the tubing and

membranes a significant amount of drug can
be sequestered over time – increasing the Vd.

• The circuit may then continue to release the

drug once the infusion has been stopped –
prolonging the effect
• The type and age of the circuit also appear to
be a factor in the level of sequestration.

• In a single dose experiment using a membrane

oxygenator and roller pump circuit up to 50%
of morphine was sequestered at 24hrs with
older circuits recording higher losses.

• This may partly explain the need to escalate

sedative doses overtime.

• Most of the drug losses occurred in the

uncoated PVC tubing.
• A comparison of crystalloid and blood primed
circuits showed a loss of 33% and 53% of
heparin and 87% and 100% of Fentanyl
respectively.

• It could be that priming and adjusting pH

could affect the protein binding of the drug.

• Hemodilution from priming, on going blood

product transfusions and fluids to maintain
circuit flows may all increase Vd particularly
for hydrophilic drugs but may enhance the
effects of highly protein bound drugs by
increasing the unbound fraction.
• Renal Dysfunction and ECMO.

• The incidence of renal dysfunction in adults

has been reported as 50% and 41% for VV
ECMO and VA ECMO respectively.

• Patients often have hypoxia &, hypoperfusion

pre ECMO with a decrease in glomerular
filtration & decreases in clearance.

• Decreased clearance in neonates is a result of

immature glomerular and tubuler function
(renal) and enzymatic pathways (hepatic).
• Decreased pulmonary blood flow during VA
ECMO may affect metabolism of many
sedative and analgesics by the lungs.

• Altered Pk in patients on RRT while on ECMO

is complex.
• Variability in the techniques used for RRT and
ECMO is a significant limitation to Pk studies.
• Clinical approaches to dosing remain
theoretical rather than evidence based and
highlights the importance of therapeutic drug
monitoring.
 So what do we know?
• Optimum sedation is not clearly defined.

• Evidence based practices of lighter

sedation with daily interruptions and
minimising paralytic agents may not be
practical, despite the awareness of the
morbidity of excessive sedation.
• Sedation needs to minimise the risk of
catheter malposition or dislodgement so
deep sedation and paralysis may be
required especially in VV ECMO to
optimise circuit flow and ventilation.
• In practice patients receiving ECMO
often have substantially higher sedative
dosing requirements.
• Neonatal studies demonstrate a need
to escalate fentanyl doses with time (9-
71mcg/kg/hr
• Midazolam dose requirement increased
after 5-7 days despite accumulation of
the active metabolite. The Vd increased
from 4.2 to 14L/Kg on the
commencement of ECMO.
• Propofol which is lipophilic and
highly protein bound is significantly
sequestered in the circuit, only 25%
of predicted concentrations are
recoverable after 2hrs.
• Although there are no evidence
based guidelines for sedation it is
prudent to slowly taper opioid and
benzodiazepine dosages after
decannulation to prevent
withdrawal.
 Antibiotics
• Often the success of ECMO may rely on the
success of antibiotic therapy.
• Antibiotic Pk data in adults on ECMO is
sparse.
• By default clinicians prescribe standard doses
risking therapeutic failure or toxicity.
• Vancomycin and gentamycin studies in
neonates are less relevant due to changes in
ECMO technology over the last 10 years as
well as the immaturity of there organ
function.
• Studies in neonates, older children and adults
confirm that the clearance of vancomycin
was decreased and its Vd increased during
ECMO.

• Loading doses 20-30mg/kg are suggested for

the larger Vd and therapeutic drug
monitoring and extending the dosing interval
should guide maintenance.

• Caspafungin peak and trough levels are

maintained during ECMO. However
Voriconazole which is more lipophilic is
significantly sequestered with a loss of 71% in
the circuit necessitating higher doses. TDM is
suggested to guide dosing.
• Oseltamivir Pk was not significantly
influenced by ECMO (150mg 12hrly in pts
over 40kg).

• A resent report on Meropenim 1g every 8hrs

showed an increased Cl of 20L/hr (13L/kg in
normal renal function, 6L/hr during CVVH)
and a high Vd. Meropenim conc was only
maintained above 4x MIC for only 30% of the
dosing interval

• A second patient on ECMO and RRT received

a high dose continuous infusion 6.5g every
24hrs. Again Cl remained high 20.8L/hr
however plasma levels were maintained
above the target 4x MIC
• There is significant variability in
antibiotic trough concentrations in
critically ill patients receiving continuous
RRT. Empirical dosing failed to achieve
the target trough concentration during
25% of the dosing interval.
• The variation in antibiotic Pk in ECMO
are concerning. Toxicity and therapeutic
failure add to the morbidity and
mortality of individual patients but the
emergence of resistant microorganisms
may have wider implications.
 Other Drugs
• Significant loss of Frusemide in the circuit
has been reported. Given that the drug is
dosed to clinical end points clinicians need
to be aware larger than usual doses may
be required.
• More than 50% of Heparin is eliminated
during ECMO which may have significant
implications and makes anticoagulation on
ECMO a challenge and a high risk therapy.
• Standard activated partial thromboplastin
time monitoring is essential and more
frequent monitoring may be appropriate.
 Conclusions
• Current data is insufficient to make
any meaningful recommendations
for adjusting drug dosing.
• Lipophilic drugs are significantly
sequestered in the circuit. Higher
Loading doses may be required.
• Hydrophilic drugs are significantly
affected by hemodilution and the
increased Vd associated with sepsis.
• Meropenim sequestration is
problematic as it depends on a time-
dependant bacterial kill. It may require
larger doses or more frequent
administration. Physical instability of
meropenim may affect its delivery by
continuous infusion.
• Maximising clinical outcomes and
minimising antibiotic resistance using
individualised doses may be best
achieved with therapeutic drug
monitoring.
 Future Directions
• Present data indicate substantial
variability and a lack of predictability in
drug behaviour in the presence of
ECMO.
• There is an urgent need to fully
elucidate the behaviour of drugs during
ECMO in adults.
• Standardisation of ECMO circuitry will
make Pk data more widely applicable.
• PK modelling of routinely used drugs
may help in drug dosing strategies to
optimise outcome and develop circuitry
that has the least influence on Pk.
• Currently a multi centre study to
optimise drug therapy in ECMO is being
undertaken in Australia and New
Zealand – ASAP ECMO.
• The study will look at 19 drugs
commonly used in ECMO with a target
of 10-12 patients for each study drug. (3
years)
 References
• Kiran Shekar et al. Journal of Critical
Care 2012 27,741.e9- 741.e18
• Kiran Shekar et al. BMC Anaesthesiol.
2012;12(29)
• E.D.Wildschut et al.Intensive Care Med
(2010) 36:2109-2116
• D.M.Roberts et al.Critical Care Med
2012 40 (5) 1523-28
• Jason A. Roberts et al. Critical Care Med
2009 37 (3) 840-51