Bioavailability and pharmacodynamics of curcumin in humans with resectable colorectal primaries and secondary hepatic metastases

Curcumin, a constituent of the spice turmeric, is a potent antioxidant with chemopreventive properties in several rodent models of carcinogenesis. In animal models of inherited colorectal cancer dietary curcumin reduced the levels of malondialdehyde-DNA (MiG) adduct, which reflects oxidative DNA changes, and reduced adenoma burden. We tested the hypothesis that pharmacologically active levels of curcumin can be achieved in humans by oral dosing, as measured by the biomarkers MiG (immunoslotblot assay), COX-2 (Western blotting) and P-catenin (immunohistochemistry). Furthermore, we attempted to determine the metabolism and distribution of curcumin in humans in vivo following oral administration. The concentrations of curcumin in normal and malignant colorectal tissue were 12.7 5.7 and 7.7 1.8 nmol/g, respectively, after the 3.6 g dose, 19.6 14.8 and 6.7 1.8 nmol/g, respectively, after 1.8 g and 0.9 0.4 nmol/g and zero respectively following 0.45g (mean SD). Trace levels of curcumin glucuronide and sulphate were detected in the colorectal mucosa and portal blood. All results were validated by co-elution chromatography with authentic reference compounds and by mass spectrometry. In line with most translational research, the study suffered from inherent errors due to difficulties in standardising warm ishcaemic time and in standardising sampling techniques in the recruited patients. In spite of these caveats, administration of curcumin (3,600 mg) appeared to decrease MjG levels in malignant colorectal tissue significantly to 2.0 1.8 adducts per 107 nucleotides (p<0.05), but did not affect MjG levels in normal mucosa. MiG levels in normal and malignant liver tissue were not affected by curcumin treatment. COX-2 expression in primary colorectal tumours was not inhibited by curcumin treatment. Curcumin administration had no effect of expression of p-catenin in either colorectal or hepatic malignant tissue. Concentrations of curcumin achieved in human colorectal mucosa were equivalent to those shown to elicit biochemical changes in vitro consistent with chemopreventive activity. In addition, short-term administration of 3.6g oral curcumin inhibited oxidative DNA damage in human primary colorectal tumours. These results suggest that Phase II clinical trials of curcumin should use at least 3.6g as a dosing regimen.