Distribution of Lipiodol and evidence for tumor necrosis in hepatocellular carcinoma.

To study the distribution and thromboembolic effect of Ultrafluid Lipiodol, 15 surgically removed hepatocellular carcinomas with selective intraarterial Lipiodol injection 7 to 10 days before surgery and 15 noninjected controls were studied radiologically and histologically. Tissue blocks were processed with an en bloc silver impregnation technique for Lipiodol localization in histologic sections. Lipiodol was distributed evenly in tumors measuring less than 5 cm in diameter and peripherally in tumors measuring 10 cm or more. Lipiodol droplets were mainly extracellular. There was no difference in tumor architecture or in hemorrhage and necrosis scores between Lipiodol-injected cases and negative controls (1.18 versus 0.92). Similarly, in injected cases, no differences were observed between Lipiodol-positive and Lipiodol-negative areas (scores of x-ray Lipiodol-positive versus Lipiodol-negative areas: 1.17 versus 1.36; scores of microscopic Lipiodol-positive versus Lipiodol-negative areas: 1.18 versus 1.14). Lipiodol-negative but hypodense areas examined by x-ray proved to be necrosis or fibrosis with or without viable tumor islands. Lipiodol has no thromboembolic effects. The uneven Lipiodol distribution may account for its failure as a carrier for chemotherapeutic agents in large tumors.