9.3 Extracorporeal Membrane Oxygenation (ECMO) & Critical Care PK

9.3 Extracorporeal Membrane Oxygenation (ECMO) & Critical Care PK

Extracorporeal Membrane Oxygenation (ECMO) provides temporary, life-sustaining cardiopulmonary support for neonates, infants, and children with refractory respiratory or cardiac failure failing maximal conventional medical management. While ECMO maintains end-organ perfusion and gas exchange, introducing an extracorporeal circuit composed of large synthetic surface areas, an artificial membrane oxygenator, and a non-biological priming volume creates profound pharmacokinetic (PK) disruptions. Clinicians must adjust dosing regimens based on drug physicochemical properties to prevent therapeutic failures or severe toxicities.

---

Circuit Physiology: VA vs. VV ECMO Configurations

ECMO Circuit Architectures:

[VV ECMO - Isolated Pulmonary Support]       [VA ECMO - Cardiopulmonary Support]
   Venous Drainage (IVC / Femoral V)           Venous Drainage (Right Atrium / SVC)
                 │                                           │
                 ▼                                           ▼
       [Centrifugal Pump]                          [Centrifugal Pump]
                 │                                           │
                 ▼                                           ▼
   [PMP Membrane Oxygenator]                   [PMP Membrane Oxygenator]
  (CO2 removed, O2 added)                     (CO2 removed, O2 added)
                 │                                           │
                 ▼                                           ▼
    Venous Return (Right Atrium)                Arterial Return (Aorta / Carotid A)
  (Relies on native cardiac output)           (Provides systemic perfusion & MAP)

Circuit Hardware Components

• Blood Tubing: Medical-grade polyvinyl chloride (PVC) or heparin-coated polyurethane.
• Blood Pump: Centrifugal pumps utilizing magnetic levitation impellers are standard; they minimize shear stress, turbulence, and red blood cell hemolysis compared to older roller pumps.
• Membrane Oxygenator: Hollow-fiber polymethylpentene (PMP) diffusion membranes. PMP prevents plasma leakage across hollow fibers, provides high gas-exchange efficiency, and resists protein deposition better than historic silicone membranes.
• Heat Exchanger: Integrated within the oxygenator housing to regulate core temperature.

---

Critical Pharmacokinetic Alterations on ECMO

Pharmacokinetic changes on ECMO are driven by two distinct mechanisms:

1. Expansion of Apparent Volume of Distribution ($V_d$)
2. Drug Sequestration & Adsorption into Circuit Components

1. Volume of Distribution ($V_d$) Expansion (Hydrophilic Drugs)

In neonates and infants, circulating blood volume is approximately 80 to 90 mL/kg (e.g., a 3 kg term neonate has a total circulating blood volume of ~250 mL). An assembled neonatal/pediatric ECMO circuit requires a priming volume of 300 to 600 mL consisting of balanced crystalloids, 5% albumin, packed red blood cells (PRBCs), and fresh frozen plasma (FFP).

Circulating Blood Volume vs. Circuit Priming Volume:

Neonate (3 kg):   [ 250 mL Native Blood ]
ECMO Circuit:     [ ======= 350 to 500 mL Prime ======= ]
Total Volume:     [ ================== 600 to 750 mL ================== ]
                  ▲ 150% to 200% EXPANSION OF CIRCULATING FLUID COMPARTMENT

This immediate 100% to 300% expansion of the intravascular fluid compartment dramatically dilutes circulating drug concentrations. Hydrophilic medications ($V_d \le 0.6 \text{ L/kg}$, low lipophilicity $\log P < 0$) distribute extensively throughout this expanded aqueous priming volume:

$C_{\text{peak}} = \frac{\text{Dose}}{V_d}$

If a standard weight-based loading dose is administered, peak serum concentrations ($C_{\max}$) will be severely subtherapeutic. Therefore, hydrophilic antimicrobials require significantly larger loading doses to achieve target therapeutic peaks:

• Vancomycin: Loading dose increased to 25 to 30 mg/kg IV (target peak 25–35 mcg/mL; monitor via AUC/MIC or TDM troughs).
• Aminoglycosides (Gentamicin/Tobramycin): Loading dose increased to 4.5 to 5 mg/kg IV in neonates (target peak 8–12 mcg/mL).
• Beta-Lactams (Meropenem, Cefepime): Require maximum-tier weight-based dosing; consider extended or continuous infusions to maintain free time above MIC ($fT_{>\text{MIC}} \ge 70\text{--}100%$).

2. Drug Sequestration & Circuit Adsorption (Lipophilic Drugs)

Sequestration into the ECMO circuit is primarily governed by two physicochemical parameters: lipophilicity (octanol-water partition coefficient, $\log P$) and plasma protein binding ($f_b$).

• Mechanism: PVC tubing contains lipophilic plasticizers (e.g., DEHP), and PMP hollow fibers present a vast hydrophobic surface area. Medications with high lipophilicity ($\log P > 2.0$) and high protein binding ($>80%$) partition out of the aqueous plasma phase and adsorb directly into the synthetic polymers.
• Consequences: Marked reduction in circulating active drug, rapid loss of clinical efficacy, and severe underdosing if conventional doses are used.

Drug Partitioning in the ECMO Circuit:

[Hydrophilic Drugs: log P < 1.0]        [Lipophilic Drugs: log P > 2.0]
• Aminoglycosides (log P -3.1)          • Fentanyl (log P +4.05)
• Beta-Lactams    (log P -1.5)          • Midazolam (log P +3.90)
• Vancomycin      (log P -3.1)          • Voriconazole (log P +2.56)
• Morphine        (log P +0.89)         • Propofol (log P +3.80)
        │                                       │
        ▼                                       ▼
[Minimal Circuit Binding]              [EXTENSIVE CIRCUIT ADSORPTION]
- Diluted by priming fluid             - Up to 80-95% loss into PVC / PMP
- Requires HIGHER LOADING DOSE         - Requires HIGHER MAINTENANCE DOSE
- Normal clearance once loaded         - Saturates over time; lost on circuit change

The "Circuit Saturation & Changeout" Phenomenon

Over the initial 48 to 72 hours of an ECMO run, binding sites on the PVC tubing and PMP membrane become progressively saturated with lipophilic drugs, and drug requirements may stabilize. However, if acute oxygenator thrombosis, membrane failure, or mechanical rupture necessitates an emergent circuit changeout, all drug binding sites are newly empty.

[!IMPORTANT] Acute Circuit Changeout Pearl: Changing out the ECMO circuit instantly precipitates acute drug loss as the new, unsaturated tubing rapidly strips lipophilic drugs (fentanyl, midazolam) from the circulation. This induces acute sedative withdrawal and rapid drops in antimicrobial levels. Clinicians must re-bolus lipophilic sedatives and hydrophilic antibiotics immediately following circuit changeout.

Anticoagulation on ECMO: Heparin vs. Bivalirudin

Continuous blood exposure to foreign non-endothelial surfaces triggers the intrinsic coagulation cascade (Factor XII contact activation), platelet consumption, and systemic inflammation. Therapeutic anticoagulation is mandatory to prevent circuit thrombosis while avoiding catastrophic intracranial hemorrhage.

1. Unfractionated Heparin (UFH) Protocol

• Mechanism: Binds to circulating Antithrombin III (ATIII) via a specific pentasaccharide sequence, inducing a conformational change that accelerates ATIII-mediated inactivation of Factor IIa (Thrombin) and Factor Xa by ~1,000-fold.
• Dosing: Continuous IV infusion initiated at 20 to 50 units/kg/h in neonates/infants (15 to 30 units/kg/h in older children), titrated without bolus in non-bleeding patients.
• Monitoring Modalities:
  • Anti-Factor Xa Assay (Preferred Standard): Target 0.3 to 0.7 units/mL. Anti-Xa specifically measures heparin-dependent Xa inhibition and is not confounded by hemodilution, hypofibrinogenemia, or baseline coagulopathy.
  • Activated Clotting Time (ACT): Target 180 to 220 seconds (bedside point-of-care test, but heavily confounded by thrombocytopenia, hemodilution, and hypothermia).
  • Activated Partial Thromboplastin Time (aPTT): Target 60 to 80 seconds (or 1.5 to 2.5 times patient baseline).

Heparin Resistance & Antithrombin III Management

• Physiology of Heparin Resistance: Unfractionated heparin has no intrinsic anticoagulant activity—it is entirely dependent on adequate functional levels of antithrombin III. Neonatal baseline ATIII activity is only 40% to 60% of adult levels due to hepatic immaturity, and critically ill children rapidly consume ATIII through systemic inflammation and circuit contact.
• Clinical Recognition: Suspect heparin resistance when the heparin infusion rate escalates beyond $>40 \text{ to } 60 \text{ units/kg/h}$ without achieving therapeutic anti-Factor Xa ($>0.3 \text{ units/mL}$) or target ACT levels.
• Target ATIII Activity: Maintain functional ATIII activity $>80% \text{ to } 100%$.
• Antithrombin Concentrate Dosing: Administer human plasma-derived antithrombin concentrate (Thrombate III) using the formula:

$\text{Antithrombin III Dose (Units)} = \frac{[\text{Target Activity (\%)} - \text{Baseline Activity (\%)}] \times \text{Weight (kg)}}{1.4}$

(In neonates and infants with expanded extracellular volume, some institutional protocols utilize a divisor of 1.0 to 1.2 instead of 1.4). Fresh frozen plasma (FFP) contains physiologic ATIII (1 unit/mL) but delivers significant volume and is less effective than purified ATIII concentrate.

2. Bivalirudin: Direct Thrombin Inhibitor (DTI)

Bivalirudin is increasingly adopted as the primary anticoagulant in pediatric ECMO centers and represents the gold standard alternative for heparin resistance or Heparin-Induced Thrombocytopenia (HIT).

Mechanism of Bivalirudin vs. Unfractionated Heparin:

[UNFRACTIONATED HEPARIN]                     [BIVALIRUDIN]
Requires Antithrombin III (Cofactor)          Direct Thrombin Inhibitor (Independent of ATIII)
           │                                                │
           ▼                                                ▼
[Heparin + ATIII Complex]                    [Binds Free & Clot-Bound Thrombin (IIa)]
           │                                                │
           ▼                                                ▼
Inactivates Thrombin & Factor Xa             Reversible inhibition of active site & exosite 1
(Ineffective if ATIII deficient)             (Fully effective regardless of ATIII levels)

• Mechanism of Action: Synthetic 20-amino-acid bivalent peptide that binds directly, specifically, and reversibly to both the catalytic active site and exosite 1 of free and clot-bound thrombin (Factor IIa). Does not require antithrombin III.
• Key Advantages over Heparin:
  • Complete independence from antithrombin III concentrations.
  • Zero cross-reactivity with anti-platelet factor 4 (PF4)/heparin antibodies; drug of choice for Heparin-Induced Thrombocytopenia (HIT).
  • Inhibits clot-bound thrombin within existing microvascular thrombi.
  • Superior time in therapeutic range and significantly reduced circuit change frequency in pediatric clinical trials.
• Dosing & Titration: Continuous infusion initiated at 0.1 to 0.3 mg/kg/h (no loading bolus in non-arrest ECMO to avoid bleeding).
• Monitoring: Titrate to target aPTT of 60 to 80 seconds (or 1.5 to 2.5 times baseline), diluted thrombin time (dTT), or chromogenic anti-IIa assay.
• Clearance Mechanics: 80% cleared via intravascular non-organ enzymatic proteolysis (cleaved by thrombin itself); 20% cleared via renal excretion. In severe renal dysfunction or anuria, reduce starting infusion rate to 0.05 to 0.1 mg/kg/h.