Background Clinicians frequently require an estimate of renal function when determining the dosage of drugs with a narrow therapeutic index, in particular, cytotoxic chemotherapeutic agents. In clinical practice indirect methods, such as formula-based equations, are used to approximate GFR. The MDRD and Cockcroft and Gault (CG) formulae are the most frequently used. A new equation, the CKD-EPI formula, has recently been published.

Purpose To assess the accuracy and impact of formula-based estimates of renal function on dosage selection of renally cleared chemotherapy agents in adult oncology patients.

Materials and methods GFR was determined using technetium-99m diethyl triamine penta-acetic acid (Tc99mDTPA) clearance, serum creatinine (Jaffe method) was measured and renal function estimates calculated using MDRD, Cockcroft and Gault (CG), Wright, Martin, and CKD-EPI formulae (MDRD and CKD-EPI adjusted for patients BSA). The Cancer Institute NSW, ‘Cancer treatments online’ was used to identify chemotherapy agents requiring dosage adjustment in renal impairment. The accuracy of the formulae were compared to measured GFR, concordance for dosage adjustment were compared and also stratified for gender, age and body mass index (BMI).

Results 311 patients were included (64% male, mean age 63 yrs). The mean measured GFR was 84 mL/min (SD 31, range 16-205 mL/min). Overall the least biased estimates of renal function were the MDRD, Wright and Martin formulae (<+/-5%); the bias of the CKD-EPI and CG formulae were -6.5% and -10.4% respectively. The overall concordance for chemotherapy dosing, based on ‘break points’ of <30 ml/min and 30-50 ml/min, were 92% for MDRD, Wright, Martin, and CKD-EPI, and 88% for CG (p=0.06).

Conclusions All renal function estimates provide similarly accurate estimates for chemotherapy dosage adjustment in renally impaired patients. However, clinicians should be aware of the limitations of any bedside approximation.