Maximum-Tolerated Dose, Optimum Biologic Dose, or Optimum Clinical Value: Dosing Determination of Cancer Therapies

One of the key goals of early-phase cancer clinical trials is to determine the best dose of a drug to take forward into subsequent, outcomes-oriented, clinical trials. Typically, once a dose and schedule are recommended for phase II and III trials, we rarely, if ever, go back to explore alternatives. We then assume a one-size-fits-all dosing of our drugs, modifying only using our traditional mg/m or mg/kg dosing approaches, neither of which have solid pharmacologic rationale themselves, and ignore any personalized medicine approach. Instead of pausing after phase I to refine our understanding of how to choose the right dose and schedule for a given patient, we push forward as quickly as possible, urgently moving toward the critical phase III trial—the straightest line between drug discovery and approval. We do this because in today’s world, finding a small benefit among a large patient population is often enough to obtain regulatory approval and to influence equally important practice guidelines. The tools and methods we traditionally employ to make a dose recommendation are based primarily on convention, not true scientific rigor. Our most traditional phase I dose-finding design seeks the maximum-tolerated dose (MTD). Establishing the MTD is based in the traditional oncologic belief that more is better, and that if only we could achieve a higher dose intensity, we will perhaps be able to overcome resistance and achieve significantly higher levels of clinical activity. Almost all of our current cancer therapies are dosed according to this principle, and yet there are little data to support the fundamental notion of dose intensity in solid tumor therapeutics. The era of targeted therapy has added a second dose-finding strategy—determination of the optimum biologic dose (OBD). Thought about broadly, the recommended OBD would not be based on toxicity but would instead be based on a more rational, scientifically derived set of end points. Examples include escalating doses to reach a predefined pharmacologic parameter, escalating doses until a target becomes saturated with the drug, or escalating until a targetmediated biologic pathway is optimally altered. While we all consider the OBD an ideal target for dose finding, we have not been highly successful with this strategy in the clinic. To incorporate the OBD strategy in early trials, a few critical pieces of knowledge are required: (1) the drug hits the target, (2) the target is altered by the drug, (3) the tumor is altered by hitting the target, and (4) giving a higher dose fails to improve outcomes further. If each of these elements is met at a certain dose, then we have defined the OBD. Neither our preclinical or clinical research infrastructure is adequately equipped to routinely perform such trials in this ideal way, primarily due to our lack of complete understanding of the biology of a given drug/target interaction. In the article that accompanies this editorial, Van Cutsem et al present the Evaluation of Various Erbitux Regimens by Means of Skin and Tumor Biopsies (EVEREST) study, a trial designed to optimize and personalize the dosing of cetuximab in the treatment of colon cancer using the now established finding that skin rash, while mechanistically poorly understood, serves as a biomarker of clinical response. To understand the rationale for this trial, we must briefly revisit the initial phase I trial that recommended the eventual approved dose, one of the earliest trials to incorporate pharmacologically driven OBD as the primary end point. Baselga et al designed the trial with a pharmacologic end point which was based on preclinical data and the belief that clearance of the drug would depend on epidermal growth factor receptor (EGFR) receptor binding. No traditional MTD was defined, but the OBD was defined to be doses between 200 and 400 mg/m. With remarkable insight, Baselga et al were fascinated by the novel skin reactions observed, considering them to be a potential dose-determining biomarker: “This skin toxicity could be an indication of receptor targeting by C225 [now known as cetuximab], since the basal layer of the epidermis has high levels of EGF receptor expression and the receptor plays a critical role in the regulation of epidermal biology.” The authors went on to say, “It is possible that the observed skin toxicity could be a result of increased antibody binding to the receptor and hence an indirect measure of receptor saturation at higher dose levels.” A second study, combining cetuximab with radiation likewise did not find an MTD as no dose-limiting toxicities were observed at higher dose levels and Robert et al recommended a loading dose of 400 to 500 mg/m and a maintenance weekly dose of 250 mg/m. Using this recommended dose and schedule, cetuximab gained US Food and Drug Administration approval in patients with EGFRexpressing colon cancer based on a phase II randomized trial with response rate (11% as a single agent, 23% with irinotecan) as the primary end point. Looking back, we were naive. Tumor EGFR expression, still a feature on the US Food and Drug Administration JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 30 NUMBER 23 AUGUST 1