Radiolabeled monoclonal antibodies: radiochemical pharmacokinetic and clinical challenges.

JL our basic science articles and two special contributions included in this issue of the Journal address some of the persistent challenges that continue to moderate the progress of radioimmunoimaging and therapy (1-6). The reports are important because of the new data they provide about current methodologies for radiolabeling antibodies for imaging and therapy. This editorial will review, and expand upon, several points brought out in these experiments. One focus will be on the chemistry issues now before us, thereby supplementing the excellent review article on monoclonal antibodies presented by Keenan (5), as well as some discussion about clinical considerations. The importance of radiolabeling antibodies which have been chemically activated by the addition of a chelating agent should be emphasized. The prototype method utilized a bidentate chelate to couple heavy metals to proteins (7). Subsequently diethylenetriaminepentaacetic acid (DTPA) was identified as a more desirable alternative (8) since synthesis of intermediates is not required and the chemistry is less complex. Though the basic methodology of using chelate-coupled antibody has changed little, many variations now exist among investigators. For example, Goodwin (/) employs bromoacetamidobenzyl ethylenediaminetetraacetic acid, while Paik et al. (2) and Fawwaz et al. (4) employ DTPA anhydride. Reported radiolabeling efficiencies among these investigators range from 35-95%, figures which depend upon two variables for proper interpretation: (a) whether postlabeling purification methods were used; and (b) the statistical variability involved in these procedures. The explanation for the first is obvious. If postlabeling unbound radioactivity is removed, reported labeling efficiencies will be higher. An example of the second variable is the situation wherein an average of three chelators are attached to each of a large population of antibody molecules. The use of Poisson statistics (°) reveals that, though we may wish to target three chelate molecules per antibody, 5% of the resulting antibody population will fail to couple any chelate, 23% will possess three chelates, and 13% will possess five. This variability is very difficult to control but may be a factor in suboptimal images, as will be seen below. Establishing reaction conditions which skew the degree of coupling in the desired direction would be a useful approach to minimizing this source of variability. Regardless of the variability, it is clear that optimizing the integrity of the Final product depends upon carefully controlling the molar ratio of DTPA to antibody prior to performing the labeling step. The importance of a well-documented inverse relationship between the average number of chelate molecules per antibody and the subsequent immunoreactivity of the Final labeled product also deserves emphasis. Just as important are efforts now underway to site-direct DTPA coupling so that antibody binding sites will not be masked.