The role of model-based methods in the development of single scan techniques

Abstract

Single scan techniques are highly desirable for clinical trials involving radiotracers because they increase logistical feasibility, improve patient compliance, and decrease the cost associated with the study. However, the information derived from single scans usually are biased by factors unrelated to the process of interest. Therefore, identification of these factors and evaluation of their impact on the proposed outcome measure is important. In this paper, the impact of confounding factors on single scan measurements is illustrated by discussing the effect of between-subject or between-condition differences in radiotracer plasma clearance on normalized activity ratios (specific to nonspecific ratios) in the tissue of interest. Computer simulation based on kinetic analyses are presented to demonstrate this effect. It is proposed that the presence of this and other confounding factors should not necessarily preclude clinical trials based on single scan techniques. First, knowledge of the distribution of plasma clearance values in a sample of the investigated population allows researchers to assign limits to this potential bias. This information can be integrated in the power analysis. Second, the impact of this problem will vary according to the characteristic of the radiotracer, and this information can be used in the development and selection of the radiotracer. Third, simple modification of the experimental design (such as administration of the radiotracer as a bolus, followed by constant infusion, rather than as a single bolus) might remove this potential confounding factor and allow appropriate quantification within the limits of a single scanning session. In conclusion, model-based kinetic characterization of radiotracer distribution and uptake is critical to the design and interpretation of clinical trials based on single scan techniques.