Developing a New Framework for Dose Calculation

Abstract

Drugs are to the medical oncologist what the scalpel is to the surgical oncologist. These are our basic tools, and using them correctly and deftly should be the most fundamental action in our day-to-day practice. Correct use of drugs with an extremely narrow therapeutic index is a key element that sets us apart from most other physicians. Critical decisions include the appropriateness and timing of treatment, which drugs to use, and at what dose. Appropriately, a lot of attention is given to the first two elements. However, as far as dose selection is concerned, often a dose is picked from a protocol, then set and forgotten, with the dose only changed if excessive toxicity occurs. Because of the nature of the drugs with which we deal, oncologists should understand drug disposition and dose selection better than any other medical specialists. We have known for some time that body surface area (BSA) -based dosing is inaccurate, on the whole favoring underdosing of individuals. The accumulated knowledge over the last decade indicates that we should move away from dose calculation based solely on BSA. Unfortunately, no practical alternative has been suggested, and we continue to use an imperfect system, each oncologist patching together a method that works to a degree for his or her patients. In the absence of a workable structure, how do we teach our trainees to calculate dose effectively? Dose calculation of chemotherapy could be described as more artistic than scientific, and therefore, a structure that is difficult to replicate. We must develop a new framework for dose calculation that is logical and that can be used in the clinic to form a foundation that can be further improved by clinical research. In this issue of the Journal of Clinical Oncology, Loos et al propose such a framework using cisplatin as a model. This group of investigators had previously shown that BSA does not reduce the interpatient variation of cisplatin disposition and had recommended fixed dosing of this drug for patients of all body sizes. However, studies using fixed dosing of paclitaxel, irinotecan, cisplatin, epirubicin, or vinorelbine still show a wide variation in drug exposure and/or drug-induced neutropenia. In the study reported in this issue, the investigators refined their proposition regarding cisplatin by studying individuals at the extremes of BSA (males 1.75 or 2.05 m; females 1.60 or 1.90 m), treating patients with BSA-based or fixed doses for the first two cycles (or vice versa according to random assignment) and then comparing pharmacokinetic parameters of unbound platinum. Comparison of the first and second cycles in each patient showed an effect of BSA on area-under-the-curve (AUC) of unbound platinum. In high-BSA patients, BSA-based treatment led to a higher AUC than fixed doses. Conversely, in low-BSA patients, BSA-based treatment led to a lower AUC than fixed dosing. Comparison of clearance of unbound platinum between patient groups on the first cycle also showed a significant correlation with BSA in the study population. However, interpatient variability was extensive, and correction using BSA minimally reduced this variation (coefficient of variance, 20.8% v 17.1%). Additionally, a retrospective analysis of a larger group across all BSA values showed an association between body size and platinum clearance when examined in BSA clusters ( 1.65 m; 1.66-2.04 m; 2.05 m). Although variability of platinum clearance within clusters remained high, mean clearance in the low-BSA group was 16% slower, and in the high-BSA group, 18% higher when compared with averageBSA patients. It is important to recognize that this study does not support the strict use of BSA for dose calculation of cisplatin. As with their previous study, BSA-based dose was associated with considerable variability in platinum clearance across all patients. However, the results suggested that fixed dosing for all patients might lead to excessive toxicity in small patients and underdosing in large patients. The authors maintained that fixed dose of cisplatin was a reasonable proposition, but in the presence of extremes of body size, BSA should be taken into account. This resulted in fixed-dose recommendations for BSA clusters. For example, for a protocol calling for a cisplatin dose of 70 mg/m, all patients with BSA 1.65 m should be given 110 mg; those with BSA more than 1.66 to less than 2.04 m, 130 mg; and those 2.05 m, 150 mg. This fixed-dose schema per BSA cluster is a sensible one and gives as much importance to BSA as is warranted. A follow-up study is important to examine drug effects such as toxicity to ensure that the variability is not unbalanced between or within clusters using this dosing schema. It is also important that similar BSA-cluster recommendations be developed for other commonly used drugs. Although the link between drug exposure and body size is a tenuous one, it seems that extremes of body size do have an impact on the interaction between the drug and the recipient. A retrospective analysis of the International Breast Cancer Study Group showed that obese women with estrogen receptor–negative breast cancer treated with less than 85% of the BSA dose of adjuvant CMF (cyclophosphamide, methotrexate, fluorouracil) had a significantly worse survival compared with those receiving the protocol dose. Another retrospective analysis of 9,672 women with breast cancer treated with adjuvant doxorubicin and cyclophosphamide showed that overweight, obese, JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 24 NUMBER 10 APRIL 1 2006