Mapping Drug Distribution Using CT Imaging Following Direct Tissue Injection in Ex Vivo Liver: Informing Clinical Implementation

Direct intratumoral injection of therapeutic drugs can minimize total dose and adverse effects compared to systemic administration. Leakage from the injection site may cause variable drug distribution and efficacy, and off-target toxicity. This study aimed to evaluate the co-injection of an iodinated contrast agent with an imageable surrogate drug (fluorescent albumin) to estimate spatial drug distribution. Fluorescent albumin and iodixanol were injected (1, 2, or 4 mL at 1 mL/min) into ex vivo bovine liver. Distribution of iodine on CT was compared to fluorescent albumin on fluorescence microscopy, including comparison of distribution volume in liver as a function of injected volume. Physical properties (hydrodynamic diameter, zeta potential) of both iodixanol and fluorescent albumin were measured. In comparison to fluorescent albumin, iodixanol is smaller (2.7 ± 0.4 nm vs. 17.0 ± 1.7 nm) and more neutrally charged (2.3 ± 0.5 mV vs. − 17.7 ± 1.4 mV). The distribution volume of iodixanol in tissue is approximately 7 times greater than that of fluorescent albumin. However, the correlation of distribution volumes of both agents is R2 = 0.89. The variance in distribution volume in tissue of iodixanol and fluorescent albumin increased with injection volume. The distributions of contrast and surrogate drug were correlated; however, differing physicochemical properties caused differences in their distribution. Although drug and contrast may not colocalize, contrast may serve as an imageable surrogate to inform preclinical development and clinical applications. This study suggests that dividing an interstitial injection into multifocal small-volume injections results in better localization of the injected drug at the target. NCT04840615.