Background The glomerular filtration rate (GFR) is used to calculate a dose for carboplatin using the Calvert equation. Our software (Asclepios) uses the Cockcroft-Gault (CG) equation to estimate GFR. According to the recommendations, the Modified Diet in Renal Disease (MDRD) equation appears to be a more accurate estimate of GFR in patients aged over 65 years old.

Purpose To determine whether there is a significant difference between the dose administered (using the CG equation to estimate GFR) and the calculated dose (using the MDRD equation to estimate GFR) for patients over 65 years old.

Materials and methods Retrospective research (1 October 2012 to 31 March 2013) was conducted on carboplatin prescription in monotherapy or combination therapy for patients over 65 years old. For each patient, the absolute difference between the dose of carboplatin administered (x) and the dose calculated (y) was determined and compared (di = xi -yi, paired t-test). From this comparison, the patients were divided into a divergent group: group 1 (difference <-10%), group 2 (difference between -10% and -5%), group 3 (difference between 5% and 10%), group 4 (difference >10%), or into a non-divergent group (difference between -5% and +5%).

Results A total of 148 prescriptions were evaluated. The median age of men and women was 70 and 75 years old. The median serum creatinine clearance for men and women was 83 and 62 ml/min. 74 prescriptions corresponded to female patients. 54 (73%) were assigned to group 1, 3 (4%) to group 2, 2 (3%) to group 3, and 15 (20%) to the non-divergent group. The mean target AUC values for divergent and non-divergent groups were 5 and 4.85 mg/ml/min. 77% of female patients would have received a lower dose than the calculated dose. A significant decrease was found between the difference (di) of the two doses (mean difference (d) = -96.03, p < 0.0001).

74 prescriptions corresponded to male patients. 42 (57%) were assigned to group 1, 17 (23%) to group 2, 4 (5%) to group 4, and 11 (15%) to the non-divergent group. The mean AUC target values for divergent and non-divergent group were 5 and 5.01 mg/ml/min. 80% of male patients would have received a lower dose than the calculated dose. A significant decrease was found between the difference (di) of the two doses (=-88.93, p < 0.0001).

Conclusions The dose of carboplatin administered to patients aged over 65 years old is mostly under-dosed. The frequency and severity of neutropenia and thrombocytopenia in both groups leads us to suggest using the MDRD equation to calculate the carboplatin dose.

No conflict of interest.