Authors: Beukers A et al.
Anesthesia & Analgesia, Volume 142, Issue 1, pages 5–14, January 2026
Summary:
This narrative review synthesizes how cardiopulmonary bypass (CPB) alters the pharmacokinetics (PK) and pharmacodynamics (PD) of commonly used continuously administered anesthetic and analgesic drugs during on-pump cardiac surgery, and why those shifts can translate into clinically meaningful under- or overdosing. From 197 screened publications (January 2000–January 2022, updated through January 8, 2025), 22 studies were included, with the largest evidence base for propofol (9 studies), followed by sevoflurane (4), remifentanil (3), isoflurane (2), fentanyl (2), sufentanil (2), and alfentanil (1).
Across drug classes, CPB initiation typically introduces hemodilution and hypothermia. The common pattern described is a decrease in total plasma concentration at CPB onset (dilution and altered distribution), paired with an increase in the unbound (free) fraction for highly protein-bound drugs due to reduced plasma proteins (especially albumin). Because the free fraction drives effect, the “total concentration falls but effect may not” paradox is a recurring theme—and a major reason why conventional dosing habits can mislead clinicians during CPB.
For hypnotics, propofol shows pronounced CPB-related PK/PD disruption. Total concentrations tend to drop at CPB initiation, but the unbound fraction can rise substantially (reported as a multi-fold increase in some studies), which may amplify hemodynamic depression and deepen cortical suppression despite “lower totals.” Several included studies associated higher propofol infusion rates during CPB with more burst suppression, supporting the practical implication that continuing pre-CPB propofol rates into hypothermic CPB can produce excessive effective drug levels. The authors also highlight that commonly used target-controlled infusion (TCI) models for propofol (Schnider, Marsh, PGIMER) are unreliable in CPB—some underpredict and others overpredict plasma concentrations—so model-driven dosing during CPB risks systematic bias.
Volatile anesthetics (sevoflurane, isoflurane) also show CPB-related shifts. Plasma concentrations may drop at CPB initiation and then drift upward during CPB, influenced by temperature, altered solubility, circuit/oxygenator interactions, and changes in perfusion. Clinically, the review emphasizes that anesthetic requirements may be lower during CPB than before bypass, and that very high volatile concentrations (especially those producing burst suppression) can meaningfully affect cerebral blood flow velocity and cerebral oxygen extraction, reflecting partial uncoupling of flow-metabolism coupling—an issue potentially more pronounced with isoflurane.
For opioids, fentanyl and sufentanil commonly show an initial drop in plasma concentration at CPB initiation (averaging around a quarter in some reports), followed by redistribution-related increases during CPB and/or after separation from bypass, with sufentanil noted in some studies to rise substantially post-CPB. This pattern supports a “front-load then back off” risk: a pre-CPB bolus might blunt the initial dip (helping avoid inadequate analgesia), but ongoing maintenance often needs reduction to avoid prolonged sedation and delayed emergence as redistribution and hypothermia-related clearance changes accumulate. Remifentanil behaves differently because it is rapidly metabolized and strongly temperature-sensitive: CPB initiation can decrease concentrations via increased volume of distribution, but hypothermia reduces clearance so the net effect depends heavily on patient temperature and timing. The review summarizes a practical temperature-based adjustment concept (reducing infusion as temperature falls) while noting that TCI using the Minto model is not consistently accurate across pre-, intra-, and post-CPB phases, with particular problems in hypothermia. Alfentanil is presented as the most stable of the reviewed opioids in CPB: total levels can fall with hemodilution while free fraction remains relatively unchanged, suggesting minimal adjustment may be needed when using a constant infusion strategy.
The review closes by emphasizing that translating CPB-related PK/PD changes into precise, individualized dosing remains difficult because studies are small, heterogeneous, often use different models and monitoring methods, and rarely link drug-level changes to hard clinical outcomes. Nonetheless, the authors argue that CPB-related dosing inaccuracy plausibly contributes to competing risks (awareness from underdosing versus hypotension, burst suppression, delayed awakening, and delirium from overdosing), and that clinicians should incorporate patient-specific factors, objective monitoring, temperature effects, and the limitations of TCI models into CPB anesthetic management.
What You Should Know:
CPB can lower total plasma concentrations while increasing free (active) drug fractions; effect may increase even when “levels” appear lower.
Propofol and volatile anesthetics may require lower dosing during CPB (especially with hypothermia and hemodilution), with higher risk of burst suppression and hemodynamic depression if pre-CPB dosing is continued unchanged.
Fentanyl and sufentanil often drop at CPB initiation and then rise during/after CPB due to redistribution and changing clearance; pre-CPB bolusing can help but increases the need to reduce maintenance dosing to avoid prolonged sedation.
Remifentanil dosing is highly temperature-dependent; hypothermia reduces clearance and can lead to higher-than-expected concentrations if infusions are not reduced.
Standard propofol (Schnider/Marsh/PGIMER) and remifentanil (Minto) TCI models are not dependable during CPB; relying on them may cause systematic under- or overdosing.
Key Points:
CPB changes volume of distribution, protein binding, and clearance (especially with hypothermia), producing predictable but drug-specific PK/PD distortions.
Total concentration trends during CPB can be misleading; free fraction and clinical effect can diverge from total levels.
Propofol and volatile agents commonly need downward titration during CPB; opioid strategies often require anticipating an initial dilutional drop and later rebound/accumulation.
Temperature is a dominant driver for remifentanil behavior during CPB; dosing should be temperature-aware.
Existing TCI models frequently misestimate concentrations during CPB, limiting their safety and usefulness without additional clinical context and monitoring.
References:
Thank you for allowing us to highlight this article from Anesthesia & Analgesia.