Adjustment of creatinine clearance improves accuracy of Calvert's formula for carboplatin dosing

Carboplatin clearance depends on the glomerular filtration rate (GFR), and Calvert's formula is frequently used to achieve a target area under the time vs concentration curve (mg ml^-1 min). Creatinine clearance is a substitute for GFR when creatinine values are determined by the Jaffe method, which is being replaced by the enzymatic method. When the enzymatic method is used, the corresponding creatinine clearance theoretically exceeds GFR, and the use of creatinine clearance as GFR in Calvert's formula results, accordingly, in overdosing of carboplatin. In this study, we have established a model for adjusting the creatinine clearance to offset this bias based on a relationship between creatinine values measured by the Jaffe method and by the enzymatic method: adjusted creatinine clearance (ml min^-1) = creatinine clearance (ml min^-1) x [serum creatinine (mg dl^-1)]/[serum creatinine (mg dl^-1) + 0.2]. Subsequently, we validated this model using the data from 35 lung cancer patients. Estimated clearances of carboplatin with the original equation [creatinine clearance + 25] were systematically higher than observed clearances [mean prediction error (MPE) ± standard error (s.e.) = 26 ± 5%]. This positive bias was corrected by the adjustment (MPE ± s.e. = 5 ± 4%). When the enzymatic method is used, the adjusted creatinine clearance should be used in Calvert's formula.