A007: Dose Optimization of Cefazolin in Adult Cardiac Surgery With Cardiopulmonary Bypass After Albumin Supplementation

Statement of Purpose, Innovation or Hypothesis: Surgical site infections postcardiac surgery (occurring in 0.5-7.8% of the patients) causes extended hospital stays, higher costs, and increased morbidity/mortality. Cefazolin (CFZ) prevents such infections. Its efficacy is driven by the time during which the unbound drug concentration (fT) is above the minimum inhibitory concentration (MIC) for the pathogen (fT >MIC). Cardiac surgeries with CPB may require albumin supplementation to counter hypoalbuminemia, associated with higher mortality. This supplementation may alter cefazolin's protein binding, impacting its pharmacokinetics. This study evaluates how albumin affects CFZ PK in cardiac surgery patients, necessitating potential dose adjustments for optimal efficacy.

Description of Methods & Materials: Total and unbound concentrations of CFZ in plasma and tissues were collected from patients (N = 17) receiving 4g before the skin incision and 2g after the CPB (30-min infusions). The intervention group received human albumin with the second dose. The unbound concentrations in plasma and tissue were collected through the microdialysis (MD) technique. A nonlinear mixed-effect modeling approach has been used in NONMEM (version 7.4.4) to develop a PK model. The model and PKPD index (fT >MIC) were used to predict the probability of target attainment (PTA) against Staphylococcus aureus, being the target the time that CFZ concentration is higher 4xMIC values. For the PTA analysis, different dosing regimens were tested to predict a best-case scenario.

Data & Results: The unbound and total plasma concentrations were well described by a 1-compartment model (Vd, 35.9 L), with first-order elimination (CL, 10.0 L/h), associated with a compartment accounting for the nonlinear binding kinetic to albumin where the maximum protein binding capacity (Bmax, 210 mg/L) and the dissociation constant (K_D 53.7 mg/L) were fixed based on literature. An extra compartment accounting for the CPB (V_CPB, 1.6 L; Q_CPB, 2.4L/min/m2 BSA) is also attached when the bypass is activated. An integrated method was applied to predict average unbound concentrations over the time ranges of the MD collection. MD probe recovery has the estimative value of 35.7%. Gender and weight were applied as covariates on both CL and Vd, while albumin drives the changes of Bmax. The intervention group's Bmax increased by 32.3% compared to the control group, leading to a higher binding capacity and lower unbound concentrations. An exposure corresponding to 4 times the MIC could not be maintained (for at least 95% of the patients) for MIC equal or greater than 4 mg/L for the control group and 2mg/L for the intervention group, when the actual dosing regimen was evaluated. On the other hand, when a continuous infusion of 15 mg/min after 2g IV bolus loading dose was evaluated, the PTA is higher than 90% for at least 95% of patients until the MIC reaches 8mg/L in both groups.

Interpretation, Conclusion or Significance: For both groups, the dosing regimen used during this study does not cover the full MIC distribution of the pathogen of interest. The patients receiving fresh albumin supplementation have a lower unbound concentration in plasma and a lower PTA compared to the control group. A continuous infusion at a lower rate associated with a loading bolus dose is a better alternative.

Disclosures: Ana Carolina Conchon Costa is funded through UF/GSK Post-Doctoral Fellowship; All authors have no conflict of interest to declare.

Citations/References:

Additional Information/Authors: Gebhart A, Costa ACC, Edlinger-Stanger M, Poschner S, Jäger W, Zeitlinger M, Hutschala D, Azeredo FJ, Schmidt S.