Skip to content | Change text size

Michael Dooley's Project

Factors influencing dose selection and administration of cytotoxic chemotherapy agents in adult cancer patients.

Cytotoxic chemotherapy is one of the key treatment strategies for the management of malignancy. Despite the availability of the majority of these drugs for many years the determination of the appropriate dose is based on many historical practices. These practices including dosing by body surface area, dose capping and dose rounding. In addition, dosing modifications based on factors such as obesity, renal function, liver function and drug presentation are not consistent and vary between clinicians and institutions. The purported relationship between renal function and body surface area (BSA) has been one of the reasons that chemotherapy has been dose based on BSA. The rationale for the use of BSA as the criterion on which to calculate doses of chemotherapy was proposed in 1958. The original derivation for this postulated relationship is actually from 1928, from work that involved the study of urea excretion and the influence of body size on urea output in a very small number of patients. An accurate measurement of glomerular filtration rate (GFR) is now possible by measuring the clearance of various radioisotopes. For the proposed PhD, work was undertaken to determine the relationship between BSA and GFR measured by Tc99mDTPA clearance in adult patients with cancer.1,2 This study demonstrated a poor correlation between GFR, determined by Tc99mDTPA clearance, and BSA, measured by the Du Bois and Du Bois linear method. The study highlighted that the convention of utilising BSA as the criterion for dosing of cytotoxic chemotherapy agents has many limitations. A suggested improvement on BSA dosing has included the application of population-based formulae. The only example used in current practice is that of carboplatin with doses calculated from either the Chatelut formula or Calvert formula to achieve a targeted area under the plasma concentration versus time curve (AUC). However, there is debate as to the most appropriate and practical formula to use. The Chatelut formula was derived from population pharmacokinetic analysis. Carboplatin clearance is estimated from serum creatinine and other variables including age, height, weight and gender. The use of this formula has been supported, although it has not been formally evaluated in the randomised clinical trial setting. Local experience has suggested that the gender factor in the formula may result in biased dosing. There are no published studies that validate the Chatelut formula with an analysis by gender. The Calvert formula was developed using GFR measured by the clearance Cr51EDTA. The use of radioisotopes in the assessment of renal function is not universally available. Consequently, simpler methods of GFR estimation, using single plasma creatinine measurement to calculate creatinine clearance, have been used and substituted into the formula. Work was undertaken to compare the doses of carboplatin calculated using the Chatelut formula with the Calvert formula, and secondly, to also compare carboplatin doses calculated using the Calvert formula when modified with non-isotopic estimations of GFR.

The results of the work performed questioned the validity of the Chatelut formula in accurately estimating the dose of carboplatin when compared to the Calvert formula. This was the first time that the potential inaccuracies relating to gender differences have been demonstrated. In light of this work, the routine application of the Chatelut formula cannot be supported. It was concluded that the most appropriate method of dosing carboplatin is the Calvert formula using either Cr51EDTA or Tc99mDTPA clearance. When an isotopic measurement of GFR is not possible, incorporation of Cockcroft and Gault formula or Jelliffe formula estimation of creatinine clearance can be applied. However the formula should be modified to account for the underestimation of renal clearance of carboplatin.

As noted before, direct measures of renal function are ideal but in most situations this is not possible. Consequently various equations and nomograms have been developed to estimate creatinine clearance from serum creatinine concentration and are used in daily practice in oncology In an attempt to improve on the estimates derived from available bedside methods, two groups have recently developed formulas to predict GFR in cancer patients, using the population pharmacokinetic approach. A comparison of these formulae to a direct measurement of GFR has not yet been published by an external group. Work was undertaken to compare GFR measured by Tc99mDTPA clearance, with estimates of GFR derived by population pharmacokinetic methods of Wright et al, Martin et al and CrCl estimates of Cockcroft and Gault as well as Jelliffe.4 This study demonstrated that the renal clearance estimates of Wright, Martin, Cockcroft and Gault and Jelliffe provide a biased and imprecise estimate of GFR.

Another source of variability in the determination of dosing of chemotherapy agents is arbitrary modification of doses for patients with BSA>2m2, that is, dose capping. Surprisingly, dose adjustment using ideal body weight (IBW) is advocated by some for obese patients. Body mass index (BMI) is a more robust methodology to determine and classify obesity than IBW but is not used in practice. Work was undertaken to assess the distribution of BMI and BSA in adult oncology patients treated at the Peter MacCallum Cancer Centre and to determine the actual dose intensity administered and degree of toxicity experienced in patients with BSA>2m2 and/or BMI>30kg/m2. This work demonstrated discordance between BMI and BSA and implies that BSA should not be applied in isolation when determining dosages and that consideration should be given to dose modification based on obesity (BMI) irrespective of BSA.

In addition to the work outlined above other issues of oncological therapeutic practice have also been explored and include:

  • the variability in the methods of administration of 5-fu, a very common cytotoxic agents.
  • the practice of rounding of low serum creatinine levels and consequent impact on accuracy of bedside estimates of renal function.
  • the implication of dosing patients to the nearest commercial preparation of common cytotoxic chemotherapy agents.
  • limitations of dosage recommendations for cytotoxic chemotherapy in patients with hepatic or renal impairment.