Prediction of hematologic toxicity of carboplatin by creatinine clearance rate.

Twenty-six patients with histologically proven lung cancer, treated with carboplatin at the National Cancer Center Hospital between October 1985 and October 1986, were retrospectively analyzed to determine the hematologic toxicity of carboplatin (CBDCA) and to develop guidelines for dose modification. A total of 34 courses of CBDCA were administered, of which 21 were adequate for assessment of the myelosuppression in relation to the renal function. One of three doses of CBDCA was administered by iv drip infusion over one hour (450 mg/m2, 19 courses; 400 12; 300, 3). Myelosuppression was dose-limiting, with thrombocytopenia being more sensitive than leukopenia, neutropenia or anemia. No significant correlation of the absolute count of platelets, white blood cells, polymorphoneutrophils, or hemoglobin with patient's creatinine clearance (Ccr) and dose of CBDCA administered was found. However, the percent reduction in platelets, white blood cells, polymorphoneutrophils, or hemoglobin correlated well with the relative dose of CBDCA [RD = total dose of carboplatin administered (mg/m2)/pretreatment Ccr/m2]. As thrombocytopenia was dose-limiting, we have developed an equation for modification of the dose of CBDCA from the relationship between the relative dose and percent reduction at platelet count nadir: Dosage (mg/m2) = (Ccr/m2/5.34) x [(1-desired platelet count nadir/pretreatment platelet count) x 100-12.9]. After consideration of the range of thrombocytopenia, we have further developed a simple equation to use CBDCA easily and safely: Dosage (mg/m2) = (1/10) x Ccr/m2 x desired % reduction in platelet count nadir = 10 x Ccr/m2 x (1-desired platelet count nadir/pretreatment platelet count. The clinical validity of these two equations is now being evaluated in prospective studies.