Cystatin C‐based <scp>eGFR</scp> better predicts renal vancomycin clearance than creatinine‐based <scp>eGFR</scp> in patients with allogeneic stem cell transplantation

Knowledge of the glomerular filtration rate (GFR) is mandatory when dosing renally eliminated drugs such as vancomycin. In clinical practice, different biomarkers and various equations are used to estimate GFR (eGFR), resulting in varying estimates. These variations may be explained by nonrenal factors, such as muscle status or glucocorticoid administration. This study aimed to evaluate the performance of different eGFR equations in terms of accuracy and precision compared to renal vancomycin clearance, including subgroup analyses for nonrenal confounders. We retrospectively analyzed data from 121 adult allogeneic hematopoietic stem cell transplant (allo-HSCT) patients. All patients received vancomycin treatment including trough concentration therapeutic drug monitoring. The eGFR was calculated using eight equations and compared to the renal vancomycin clearance that was calculated using a pharmacokinetic model and served as the reference. Individual muscle status was determined by computed tomography scans. Median renal vancomycin clearance was 49 mL/minute/1.73 m² (range 24-96). All eight eGFR equations overestimated renal vancomycin clearance. The six (partially) creatinine-based equations were significantly less accurate (bias: 24.0-62.8 mL/minute/1.73 m²) than both cystatin C-based equations (bias: 6.3-9.5 mL/minute/1.73 m²). This decreased accuracy for creatinine-based eGFR was more pronounced in patients with reduced muscle status or glucocorticoid medication. All CKD-EPI equations and the Hoek equation were more precise with an IQR of the difference to renal vancomycin clearance ≤22.5 mL/minute/1.73 m² compared to ≥35.5 mL/minute/1.73 m² (Cockcroft-Gault, MDRD). In conclusion, cystatin C-based eGFR equations are preferable to creatinine-based approaches for vancomycin dosing in allo-HSCT patients.