Maximum tolerated exposure: A more rational approach to drug dosing

The standard of care in dosing chemotherapeutic agents is based on body surface area — or BSA. The BSA-based dose for any given drug is recommended by the manufacturer according to the maximum tolerated dose established during phase 1 studies. MTD, in turn, is generally determined using the Fibonacci dose escalation “3 + 3 schema” in which three patients are enrolled at each proposed dose level and if one patient experiences a drug limiting toxicity, three additional patients are then enrolled at that same dose level. If two or more additional patients at a particular dose level present a drug limiting toxicity, enrollment of patients to that dose level is discontinued and the immediate preceding dose level is considered the MTD.

The most commonly used formula for BSA dosing was derived in 1916 by DuBois and DuBois. This formula was used in studies in the 1950s in an attempt to achieve uniformity in drug administration. However, there is actually no rigorous scientific basis for the use of BSA for anticancer agents, and numerous studies have shown that this approach to dosing is in fact not valid.

BSA-based dosing is associated with drug plasma level variability as high as 30-fold. Strong evidence exists that such inter- and intra-patient pharmacokinetic variability of chemotherapy drugs is a major contributor to toxicity and treatment failure. Given that BSA-based dosing is associated with high drug plasma level variability, how valid is selection of a MTD based on using this approach in as few as six patients?

Joseph R. Bertino

Indeed, many studies have shown that BSA is not a good indicator of systemic exposure. Baker et al demonstrated that of 33 new investigational agents tested in phase 1 trials between 1991 and 2001, only five agents (15%) showed a significant association between BSA-based dosing and reduced inter-patient clearance. In fact, for the five agents where dosing by BSA was statistically significant, the relative reduction in variability in clearance was between 15% and 35%, which indicates that only up to one-third of the total variability can be explained by differences in BSA.

In a more recent phase 3 study of colorectal cancer patients treated with leucovorin/5-FU, Gamelin et al demonstrated that by measuring 5-FU levels in plasma and adjusting doses in subsequent cycles to achieve target concentrations, severe toxicities were significantly reduced (P=.003) and response rates nearly doubled (18.3% vs. 33.7%, P=.004). Medium OS was also improved from 16 months to 22 months (P=.08). Interestingly, 68% of the patients in this study required increased dose intensity and 17% of the patients required lower doses to achieve target levels. Gamelin et al reported similar findings in a phase 2 FOLFOX4 dose adjustment study in which 50% of the patients required increased doses and 30% required dose reduction to achieve target levels. A 70% response rate, 16-month PFS and 28-month OS were observed. Severe toxicities were also lower than reported in other FOLFOX4 studies without dose adjustment.

As oncologists, we often use the word “MTD” to describe treatment with a particular regimen, but given the lack of correlation of BSA dosing with systemic exposure, are we just using this as a feel-good statement? MTD suggests that the patient is getting an optimal dose of the drug when, in fact, there is little evidence to support this. In cases when a patient shows little toxicity, are they tolerating the drug well? Or, alternatively, are they being given suboptimal exposure? There are very few oncologists today who increase the dose in the absence of toxicity for fear of creating undue toxic side effects.

This is illustrated in a 2003 study published by Lyman in which 20,799 randomly selected breast cancer patients at 1,243 sites in the United States were evaluated for the chemotherapy dosing they received. Fewer than 5% of patients received an increase in dose intensity relative to the standard reference dose. On the other hand, 36% of the total patients were given greater than a 15% reduction in dose intensity and 41% of them received an initial lower than standard dose, obviously without any indication of toxicity.

Maximum tolerated exposure

A more rational approach to dosing would be to define in the early stages of drug development an exposure level of a drug that produces a biological effect that is associated with efficacy and/or avoidance of toxicity. For many drugs, the exposure level can be determined by measuring the active form or relevant surrogate of the drug in the plasma and correlating these levels with outcomes and/or toxicity.

By defining the plasma level at which toxicity begins, individual patient dosing can be appropriately adjusted during treatment so that plasma levels just below the level that causes dose limiting toxicity can be achieved. By defining the exposure level where dose limiting toxicity is presented, a maximum tolerated exposure (MTE) can be identified.

Salvatore J. Salamone

MTE, unlike MTD, provides a true measure of the drug intensity in the individual and is an effective tool to delivering treatment that BSA dosing is supposed to provide but falls short of in practice. Although this approach has been utilized in oncology only to a limited extent so far, it has been shown to produce impressive results.

In fact, measurement of drug exposure is the phenotypic expression of all the factors, physiological, genetic and external, that can affect systemic drug levels and is an effective step to personalizing drug treatment. The use of MTE in practice will require a shift in the way we develop drugs and treat patients. It will require regulatory agencies, drug companies, oncologists and testing laboratories to recognize that all patients are not the same and that optimal treatment of individual patients requires individualized dosing decisions.

Routine tests to determine drug exposure can and should be developed early in the drug development process so that the relationship of the drug and the exposure can be elucidated for routing clinical practice. Additionally, the use of such monitoring tests may increase the success rate of new investigational drugs by lowering dose limiting toxicity and improving efficacy.

MTD based on BSA is a bad habit we have fallen into and if we are truly motivated to have more effective and personalized treatment, we need to move away from it. MTE to determine proper dosing of the drug represents a more rational path and has the potential to make drug treatment more effective and less toxic.

Joseph R. Bertino, MD, is university professor of medicine and pharmacology at the University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School and associate medical editor for HemOnc Today.

Salvatore J. Salamone, PhD, is CEO and chief scientific officer of Saladax Biomedical.

For more information:

• Baker SD. J Nat Cancer Inst. 2002;94:1883-1888.
• de Jonge ME. Clin Pharmacokinet. 2005; 44:147-173.
• Du Bois D. Arch Intern Med. 1916;17:863-871.
• Felici A. Eur J Cancer. 2002;38:1677-1684.
• Gamelin E. J Clin Oncol. 2008;26:2099-2105.
• Hon YY. Clinical Chemistry. 1998;44:388-400.
• Lyman GH. J Clin Oncol. 2003;21:4524-4531.
• Pinkel D. Cancer Res. 1958;18:853-856.
• Saif MW. J Natl Cancer Inst. 2009;101:1543- 1552.
• Takimoto CH. Targ Oncol. 2009;4:143-147.
• Undevia SD. Nat Rev Cancer. 2005;5:447-458.