Research Paradigm (2006): Reinventing drug development at NCI: PK/PD

DTP is reexamining its discovery and development paradigm and is considering moving from using only target-based discovery tools and efficacy models for phase I toxicity-based clinical trials to using pharmacokinetic and pharmacodynamic (PK/PD)-guided clinical trials. Such studies are used to study how drugs react once administered to animals and humans. By studying PK/PD responses, researchers will be better able to administer the appropriate dose to achieve the desired therapeutic response with a minimum risk of toxic effects.

Various tools are available for pharmacology and toxicology studies leading to and supporting clinical trials, including targeted assays, standard and genetically engineered mouse models, bioluminescent techniques, and various "omics." The problem, however, is that resources—both monetary and human—are finite. Setting priorities is a difficult task.

For example, a combination of mouse and dog in vivo toxicology studies predicted 91% of human maximum tolerated doses (MTD). By testing other animal species, DTP could conceivably increase its success rate, but the cost and time required to determine if the drug should be advanced to human clinical trials would also increase.

A solution would be to use less costly and less time-consuming in vitro toxicity assays in both human and animal tissues. The targets of these studies would include brain, gastrointestinal tract, heart, kidney, liver, and lung tissues. The results of assays that compare drug toxicity to drug efficacy would yield an in vitro human therapeutic index for each test agent. This index would give DTP the data necessary to adopt the "fail fast and fail early" paradigm for new anticancer agents and advance novel drug candidates into clinical trials with an improved probability of success.

A preclinical PK/PD program is being proposed by the NCI to facilitate the entry of new molecular entities (NMEs) into clinical trials. This program could be used to test traditional cytotoxic agents as well as molecular-target-based, small-molecule NMEs.

The PK/PD paradigm embodies the ideal drug development scenario required to conduct a limited, single-dose PK/PD dose-escalation study in humans, which the NCI refers to as a phase 0 clinical trial. To make this program successful, researchers would perform real-time PK/PD studies to guide dose escalation instead of escalation to MTD as now the norm in phase I trials. This approach is essential for patient safety in early human clinical trials.

Methodology:

Develop and validate methodology to determine pharmacokinetics in various species and evaluate plasma-protein binding to various components of human plasma proteins, including albumin and alpha-1 acid glycoprotein. It may also be necessary to evaluate biodistribution.
Determine if metabolism is important and identify metabolic pathway. Evaluate metabolism in various species, including human in vitro studies using an appropriate metabolic system such as liver slices.
Evaluate pharmacokinetics in various species, beginning with the animal tumor model. Determine whether peak plasma levels, AUC, or time above a threshold is required for efficacy. Compare data from in vitro time versus concentration efficacy studies to design appropriate PK and toxicology studies.
Select appropriate biomarkers to assess target modulation and develop and validate methodology to determine impact of drug treatment on target(s) in tumors and on normal tissues. Evaluate pharmacodynamics in various species, starting with the animal tumor model to correlate doses' and/or plasma drug levels' impact on target in tumor and in normal tissues. Determine whether peak plasma levels, AUC, or time above a threshold is required for target modulation. Develop appropriate surrogate for tumor (e.g., peripheral blood mononuclear cells; blood, serum, plasma, skin biopsy, saliva, buccal mucosa cells).
Conduct in vitro toxicity assays in relevant target organ systems.
Conduct single-dose safety/toxicity studies in rodents to determine if efficacious drug concentrations (i.e., peak plasma concentration, AUC, or threshold) can be safely attained and maintained for an appropriate time period. Determine the impact of these concentrations on selected biomarkers (i.e., genomics, proteomics), safety, and toxicity. Evaluate reversibility of toxicity or whether toxicity is delayed. Evaluate histopathology at top dose only.
Take samples for possible genomics and proteomics studies to be conducted in conjunction with pharmacokinetic and toxicity studies.

Each NME development plan may require modification; however, individual programs should not stray too far from this methodology. For example, a highly potent compound may not be appropriate for the development of a sensitive assay for plasma drug analysis because of the challenges involved with developing the assay. In this situation, a pharmacodynamic biomarker assay would be required. In another example, the limited toxicity studies that the NCI proposes should be conducted in at least two species so that a predictive safe starting dose can be established. Researchers should adhere to the requirement of the development plan unless a sound scientific rationale is presented for the study to use one species.

¹ U.S. Food and Drug Administration. Innovation/Stagnation: Challenge and Opportunity on the Critical Path to New Medical Products. March 2004.