Background and purpose: Cisplatin-based chemotherapy is the main therapy for patients with advanced stage non-small cell lung cancer (NSCLC). The dose of cisplatin is adjusted according to the patient's renal function. Calculation of creatinine clearance (CCr) by 24-hour urine collection is the most common method for estimating the glomerular filtration rate but is time-consuming and inconvenient. Estimation of CCr using the Cockcroft-Gault formula has been suggested to be accurate, reproducible, and less costly. This study compared CCr values obtained by measured and estimated methods during cisplatin-based chemotherapy in NSCLC patients in Taiwan.
Methods: A total of 92 patients (58 men, 34 women) with advanced NSCLC who completed 6 cycles of chemotherapy participated in the study. The dose of cisplatin per cycle was 80 mg/m(2) every 28 days, reduced to 50 mg/m(2) if CCr was 30 to 60 mL/min by the measured method. When urine collection was finished, serum and urine creatinine levels were measured simultaneously. Estimated values were calculated before each cycle of chemotherapy.
Results: The mean measured CCr was 85.2 mL/min, 25.7 mL/min higher than the mean estimated value. CCr values obtained by both methods were significantly reduced during the 6 cycles of chemotherapy. There was no significant difference in CCr values between patients aged < 65 years or >/= 65 years (-19.9 vs -15.1 mL/min, p = 0.15). Using a cut-off of measured CCr >/= 60 or < 60 mL/min, agreement on the dosage for both methods was 51% for all patients, 77.7% for patients < 65 years, and 26.7% for patients >/= 65 years.
Conclusions: The Cockcroft-Gault formula underestimated measured CCr by about 25 mL/min in this study. Cisplatin-based chemotherapy reduced CCr, with no significant difference between older and younger patients. Use of the estimated method would result in significant under-dosing, especially for patients >/= 65 years old.