Methods have recently been developed (17,19) for detailed biochemical analysis of tissues exposed in vivo to 5-fluorouracil (FUra). These include determination of 5-fluorodeoxyuridylate (FdUMP), the metabolite that inhibits thymidylate synthetase (TS); deoxyuridylate (dUMP), which limits this inhibition; free, non-FdUMP-bound enzyme (TSf) and total enzyme (TStot); and assay of levels of non-radiolabeled FUra incorporated into RNA (F-RNA) and DNA (F-DNA). Preliminary application of the TS methods in patients with adenocarcinomas given bolus FUra has suggested that the degree of TS inhibition may correlate with clinical response, but this may reflect histologic differences (18). In this proposal, comprehensive biochemical and pathologic analyses will be done on samples of hepatic tumor nodules and bone marrow leukocytes 24 h following infusional or bolus intravenous FUra 500 mg/m sq. These samples will be obtained at laporatomy of patients with colorectal adenocarcinoma at the University of Goteborg, Sweden, who are treated for hepatic metastases by hepatic artery ligation and portal vein infusion with FUra. Determinations of FdUMP, dUMP, TSf, TStot, (3H)FdUMP-binding capacity, F-RNA, and F-DNA will be done on fresh frozen tumor samples and bone marrow leukocytes isolated by density gradient centrifugation. Pathologic evaluations of paired tissue samples will include semi-quantitative histologic scoring; quantitative cytomorphometric determinations of cellularity, mitotic index, nuclear indices, and mucin content; and flow cytometry analysis of relative DNA content. Statistical analyses will be done to determine if the biochemical mechanism(s) of action of FUra varies between adenocarcinomas and bone marrow leukocytes or varies with the schedule of FUra administration, and to determine fundamental correlations among biochemical, pathologic, and clinical data. These studies ultimately may enable identification of mechanisms of resistance to FUra, that could be overcome by rational biochemical approaches.