Influence of co-morbidity on renal function assessment by Cockcroft–Gault calculation in lung cancer and mesothelioma patients receiving platinum-based chemotherapy

Abstract

BackgroundCreatinine clearance (CrCl) estimation by Cockcroft–Gault calculation (CG) often replaces measurement of glomerular filtration rate (GFR) by [51Cr]–ethylenediaminetetraacetic acid clearance (EDTA). Co-morbidity, age, and renal impairment influence the accuracy of CG, whilst the relationship between CG and EDTA has been poorly assessed in lung cancer patients, a population significantly affected by these covariates. MethodsRetrospective analysis of co-morbidity, nephrotoxic drug use, chemotherapy toxicity, and correlation between paired CG and EDTA, in 388 lung cancer and mesothelioma patients receiving platinum-based chemotherapy. ResultsPotentially nephrotoxic co-morbidity or medication use occurred in 47% of patients, and was twice as likely in those aged >70 years (OR=2.07; 95%CI: 1.25–3.44, p=0.003). Patients with co-morbidity or nephrotoxic medication use had a lower EDTA compared to those without these baseline factors (p=0.02), but were not significantly more likely to experience chemotherapy toxicity. CG and EDTA correlation was high (r2=0.68), but reduced in patients with ETDA<50ml/min (r2=0.26, p=0.02) or >120ml/min (r2=0.32, p=0.09), and in those with CG>120ml/min (r2=0.20, p=0.01). The correlation between CG and EDTA was not significantly altered in patients with co-morbidity or nephrotoxic medication use. CG bias (mean percentage error) and precision (mean absolute percentage error, MAPE) were 7% and 26%, respectively, and precision was impaired in patients with abnormally raised serum creatinine (MAPE 65%, p<0.0001). ConclusionCG estimation of CrCl is accurate and safe in lung cancer patients with potentially nephrotoxic co-morbidity or concomitant medication, but should not be used when values are outside the range 50–120ml/min, or with abnormally elevated serum creatinine.