AJR:191, July 2008 ing that contrast media bind almost entirely to albumin and that the presence or absence of prosthetic groups at both the 3 and 5 positions determines the relative weakness or strength of binding with albumin. Lasser et al. [1] then compared these results with a synthesis of the LD50 toxicity data that had been previously published and concluded that there is a consistent correspondence between binding and toxicity. The better-bound materials are also the most toxic. This also explained the known higher toxicity of the cholangiographic agent meglumine iodipamide (Cholegrafin, E. R. Squibb and Sons [now Bracco Diagnostics]). Lasser et al. [1] then correlated characteristics of contrast media with alteration in blood and blood vessels. They observed that those agents with the higher albumin binding capacity produced a greater degree of RBC morphologic changes (crenation) than those agents that had poor protein binding. Evidence from their investigation and from the literature suggested that local and systemic toxicity from contrast materials is mediated via derangements of the blood elements and blood vessels. It was hypothesized that this is related to the proven presence of contrast-binding albumin sites in or on these structures. Next, Lasser et al. [1] turned their attention to the relationship of contrast material protein binding and pathways of contrast material excretion. They observed that the more highly protein-bound contrast media appear to be preferentially excreted in the bile and the less highly bound media in the urine. This allows characterization of the types of contrast media best suited for hepatic versus renal excretion and imaging. All of the available chole cystographic contrast media in use at that time had the chemical configuration that confers good albumin binding. In contrast, all of the urographic contrast media were relatively poor albumin binders. “The Significance of Protein Binding of Contrast Media in Roentgen Diagnosis”— A Commentary