Abstract
Vancomycin is widely used as a prophylactic antibiotic for ventricular assist device (VAD) implantation to prevent infections, the most common complication. As vancomycin is renally eliminated, an accurate renal function estimation is essential. However, it has been reported that creatinine-based models inaccurately estimate renal function in VAD recipients, and cystatin C could alleviate the limitation. This study analyzed the association between renal function estimation methods and vancomycin trough concentrations in VAD recipients. Clinical data of VAD recipients who received prophylactic vancomycin at Seoul National University Hospital between 2014 and 2023 were retrospectively analyzed. Recipients were categorized into high trough (> 15 mg/dL) or non-high trough (≤ 15 mg/dL) groups based on the 1st vancomycin trough concentration after surgery. The estimated glomerular filtration rates (eGFRs) based on creatinine alone (eGFR(Cr)), combined with cystatin C (eGFR(Cr-Cys-C)), and their differences (eGFR(diff)) were compared between the 2 groups. The association between the eGFRs and vancomycin trough concentrations was evaluated using Pearson's method. Among the 20 recipients, 13 were the high trough group and 7 were the non-high trough group. The high trough group had a significantly higher eGFR(diff) than non-high trough group (8.9 vs. -5.1 mL/min/1.73 m(2), p = 0.0265), while other eGFRs were comparable. Among the three eGFR estimates, eGFR(diff) showed the strongest correlation (r = 0.41) with the first measured vancomycin trough levels. In conclusion, creatinine-based eGFR might not fully capture vancomycin pharmacokinetics in VAD recipients. The difference between eGFR(Cr) and eGFR(Cr-Cys-C) is associated with vancomycin trough concentration in VAD recipients.