|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2003 American Society for Clinical Oncology
Horseshoes, Hand Grenades, and Body-Surface Area–Based Dosing: Aiming for a TargetUniversity of Pittsburgh School of Medicine, University of Pittsburgh Cancer Institute, Pittsburgh, PA ADMINISTRATION OF a drug to a patient carries with it the implicit assumption that the drug will do something to that patient. There are two obvious clinical outcomes related to drug delivery. The first of these is therapeutic and desirable, whereas the other is toxic and not desirable. The obvious goal in treating a patient with a drug is to maximize the likelihood of producing a therapeutic response while minimizing the likelihood of producing unacceptable toxicity. The dosing variables available to achieve this goal include the amount of drug delivered and the interval or frequency at which drug can be given. Under ideal circumstances, these variables would be based on sound knowledge of the relationship between the dose, or concentration, of the agent; the likelihood of therapeutic and toxic consequences resulting from its delivery; and the knowledge of how long drug effects persist. Unfortunately, such precise information is lacking for most antineoplastic chemotherapeutic agents. Furthermore, although it may be possible to generalize dose-toxicity-response relationships in patients with different types of tumors, it is obvious that relationships between dose and therapeutic response cannot be extrapolated from one tumor type to another, because drugs known to be active in one tumor type may be relatively inactive against others. Moreover, it is obvious that not all patients are the same. As a result of genotypic and phenotypic differences, the same dose of drug will produce a range of concentration-versus-time profiles in any given group of patients, with the resulting range of therapeutic and toxic responses corresponding to that pharmacokinetic variability. For drugs that produce therapeutic effects at doses far less than those that cause toxicity, the incentive for precise dosing is far less than for drugs with a narrower therapeutic index. The narrow therapeutic index of most antineoplastic agents has provided great impetus to deliver doses as precisely as possible. One of the practices imbedded in dosing of antitumor drugs is dosing by body-surface area (BSA), most commonly milligrams per square meter (mg/m2). It is important to understand that this practice is a legacy of species scale-up1 and the attempt to initiate clinical studies with rational doses on the basis of animal toxicity data. The foundation for this practice is described in a seminal article by Freireich et al,2 in which quantitative comparisons were made in the toxicity of anticancer agents in mice, rats, monkeys, and man. Therefore, when starting doses were defined for early phase I clinical studies of potential new antitumor agents, they were expressed on an mg/m2 basis, reflecting the scale-up from preclinical toxicology studies. What may have been lost in this nearly 40-year-old practice is that the difference in size between mouse and man, or dog and man, is far greater than the variability in size among patients.3–6 As with many practices ingrained in the practice of medicine, mg/m2 dosing of antitumor drugs has become accepted without questioning the soundness or the validity of those underlying assumptions, despite the appearance and availability of newer technologies and theories that would make testing of such assumptions possible. As analytic chemical instrumentation, pharmacokinetic modeling, and increasingly sophisticated means of assessing molecular and clinical outcomes of drug therapy have been developed, there has been an increasing call to test the validity of the assumptions behind mg/m2 dosing of antitumor drugs in humans. The advantages of using the same dose of drug in all patients are obvious. Drug companies could be more efficient and decrease costs by manufacturing, storing, and shipping fewer unit sizes of drug doses. Pharmacies and hospitals could be more efficient and decrease costs by simply formulating a unit dose of drug rather than measuring out individual doses of varying size for different patients. Physicians could be more efficient by being able to prescribe a known, specific dose of drug without having to resort to nomograms of questionable validity to relate height and weight to BSA. For example, such nomograms do not even begin to consider the question of whether a kilogram of fat should be given the same weight in the calculation as a kilogram of muscle or kilogram of lower-extremity edema. Finally, delivery of a standard, fixed dose of drug would eliminate the potential for giving a patient the wrong dose of drug because that patient’s BSA was incorrectly calculated or an error was made in preparing the patient-specific, correct dose of drug. Unfortunately, a shift in standard of practice away from mg/m2-based dosing and toward fixed dosing of antitumor drugs will first require generation of data supporting such a change and then the daunting task of changing habits ingrained in generations of medical oncologists, drug company personnel, and government officials involved in development of cancer chemotherapeutic agents. The article by Smorenburg et al7 in this issue of the Journal of Clinical Oncology is an attempt to provide data that would either support or question the validity of mg/m2-based dosing of paclitaxel. In a carefully designed randomized, cross-over study, these investigators prospectively evaluated whether mg/m2-based dosing of paclitaxel reduced the pharmacokinetic variability in a population of patients treated with that drug. The statistical portion of their article is a model for other investigators to follow when designing other studies of this type. A hypothesis was clearly stated and a degree of difference clearly defined so that an appropriate sample size could be calculated and studied. The analytic chemical methodology and pharmacokinetic methods used were carefully controlled and of a quality resulting in confidence in the data generated. As a result, the authors can state that BSA-based dosing of paclitaxel was associated with less variability in the area under the plasma paclitaxel versus time curve (AUC) than was delivery of a fixed dose of 300 mg of paclitaxel. However, careful consideration of this study raises more questions than it answered. It is unfortunate that the authors focused on AUC as the pharmacokinetic parameter of paclitaxel to study. Although AUC is a standard pharmacokinetic parameter and is intimately related to clearance in drugs with linear pharmacokinetics, its relevance to the pharmacodynamic consequences of paclitaxel therapy has not been clearly demonstrated. It is a shame that the authors did not also study the time that plasma paclitaxel concentrations remained greater than 0.05 µmol/L, because a number of investigators have demonstrated that this pharmacokinetic parameter is related to the degree of neutropenia produced by paclitaxel. Unfortunately, no study has shown a relationship between an individual patient’s paclitaxel AUC and the time that plasma paclitaxel concentrations remained greater than 0.05 µmol/L in that patient. Furthermore, to date, no convincing pharmacokinetic/pharmacodynamic relationship has been established between paclitaxel AUC and the likelihood of response to treatment with that drug. Moreover, because the dose-limiting toxicity of 3-hour infusions of paclitaxel is peripheral neuropathy, which has not been related to AUC and may well be related to peak plasma concentration of paclitaxel or a component of the paclitaxel formulation, the importance of decreasing variability in AUC may be clinically unimportant. Another potential concern about the current study is related to the generalizability of its results. Paclitaxel pharmacokinetics are known to be nonlinear, with higher doses producing a greater-than-proportional AUC than predicted from lower doses. This concern is relevant to shifts in clinical practice. Many clinicians use a 1-hour infusion to deliver doses of paclitaxel similar to those used in the current study, thereby producing paclitaxel concentrations greater than those encountered in the current study. Many other clinicians are routinely using 1-hour infusions to deliver lower doses of paclitaxel (80 to 100 mg/m2) on a weekly basis rather than the 175 mg/m2 every-3-weeks schedule studied in the current article.7 Moreover, assumption of an average BSA of 1.73 m2 is not supported by data from the National Cancer Institute/Cancer Therapy Evaluation Program database, which indicates that the average BSA of patients entered onto its studies is closer to 1.8 m2. What may be the most critical assessment of the relevance of the pharmacokinetic results presented by the authors is in the final paragraph of Results, wherein they state that no patient developed nonhematologic toxicity greater than grade 2, and there were no episodes of neutropenic fever or treatment-related deaths. Moreover, there was no difference between the mean neutrophil nadir of the BSA-based dosing group and that of the flat-fixed dosing group. The fact that the coefficient of variation in neutrophil nadir in the BSA-based dosing group (70.2%) was smaller than that of the flat-fixed dosing group (94.6%) is less impressive when considered in light of how much variability existed within both groups and the fact that neutrophil nadir is in part dependent on pretreatment neutrophil counts, data that are not included in the current article. This failure to demonstrate any meaningful clinical difference between the two groups greatly reduces the effect of the documented decreased variability in AUC related to BSA-based dosing and might well argue strongly for studies using flat-fixed dosing of paclitaxel rather than those carrying the obligatory logistic encumbrances related to calculating BSA-based doses. REFERENCES 1. Dedrick RL: Animal scale-up. J Pharmacokinet Biopharm 1:435–462, 1973[CrossRef][Medline] 2. Freireich EJ, Gehan EA, Rall DP, et al: Quantitative comparison of toxicity of anticancer agents in mouse, rat, dog, monkey, and man. Cancer Chemother Rep 50:219–244, 1966[Medline]
3. Grochow LB, Baraldi C, Noe D: Is dose normalization to weight or body surface area useful in adults? J Natl Cancer Inst 82:323–325, 1990 4. Reilly JJ, Workman P: Normalization of anti-cancer drug dosage using body weight and surface area: Is it worthwhile? Cancer Chemother Pharmacol 32:411–418, 1993[CrossRef][Medline] 5. Gurney H: Dose calculation of anticancer drugs: A review of the current practice and introduction of an alternative. J Clin Oncol 14:2590–2611, 1996[Abstract] 6. Sawyer M, Ratain MJ: Body surface area as a determinant of pharmacokinetics and drug dosing. Invest New Drugs 19:171–177, 2001[CrossRef][Medline]
7. Smorenburg C, Sparreboom A, Bontenbal M et al: Randomized-Crossover Evaluation of Body-Surface Area-Based Dosing Versus Flat-Fixed Dosing of Paclitaxel. J Clin Oncol 21:197–202, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2003 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
