Dose de-escalation chemoprevention trial of alpha-difluoromethylornithine in patients with colon polyps.

BACKGROUND: alpha-Difluoromethylornithine (DFMO) is a potent inhibitor of carcinogenesis in experimental animal models. In these animal models, DFMO has been especially active in preventing carcinogen-induced epithelial cancers, including those of the skin, colon, breast, and urinary bladder. Although DFMO is known to exert its diverse biological effects by suppressing intracellular pools of the polyamines putrescine and spermidine, the precise mechanism by which polyamine depletion, induced by DFMO, suppresses carcinogenesis is unknown. PURPOSE: The specific aim of our study was to determine the lowest dose of DFMO that would deplete target tissue (colorectal mucosa) levels of these polyamines in humans who had undergone prior removal of colon polyps while producing minimal toxic effects. METHODS: A dose de-escalation chemoprevention trial of DFMO was conducted in 111 patients (36 female and 75 male) who were in generally good health, aged 39-79, and who had undergone colonoscopy for surgical removal of an adenomatous colon polyp greater than 3 mm within 5 years prior to entering the study. Groups of patients (12-20 patients per group) were orally treated with single, daily doses of DFMO ranging from 3.0 to 0.1 g/m2 for 4 weeks (28 days). Prior to initiation of DFMO treatment and at the end of treatment, six colorectal biopsy specimens were collected from each patient, along with serum samples. All biopsies were performed between 9 AM and noon to avoid possible effects of diurnal variations in laboratory end points. Samples for analysis of plasma DFMO levels were also collected during this time period on the day after the last day of drug administration. RESULTS: DFMO caused a decrease in both putrescine content and the ratio of spermidine to spermine for all dose groups down to 0.25 g/m2. Both putrescine content and the ratio of spermidine to spermine and changes in these parameters as a function of DFMO treatment decreased as a function of donor age. None of the 30 patients receiving either 0.25 or 0.5 g/m2 experienced any clinical ototoxicity in this trial. CONCLUSIONS: DFMO is both safe and effective in reducing colorectal mucosal polyamine contents when it is administered orally to patients at doses as low as 0.25 g/m2 for 28 days. No ototoxicity was observed at doses up to twice this amount. IMPLICATIONS: If DFMO is also found to be effective in suppressing polyamine contents in other target tissues, it may be useful in preventing a wide range of human epithelial cancers, including those of the prostate and breast.

Sources of variability in estimating ornithine decarboxylase activity and polyamine contents in human colorectal mucosa.

The activity of ornithine decarboxylase (ODC), the first enzyme in polyamine synthesis, is elevated during epithelial carcinogenesis. Since this enzyme is a target for colon and other cancer chemoprevention strategies, we sought to identify sources of variability affecting the measurement of tissue ODC activities and polyamine contents. Multiple colorectal biopsies were obtained from 39 patients undergoing colonoscopy. Biopsy size affected polyamine but not ODC values. Spermidine (spd):spermine (spm) ratios varied less than the contents of the individual amines. Bowel preparation methods did not affect any of the measurements. ODC activities and spd:spm ratios did not vary with bowel location. Lab assay methods contributed to sources of error. Variability was greatest for polyamine content measurements but was reduced when polyamine contents were analyzed as spd:spm ratios. Intrapatient variability of these parameters was as great or greater than interpatient variability. When measured in apparently unaffected colorectal mucosa, none of these parameters were significantly correlated with prior polyp history, number of prevalent polyps found at current colonoscopy, or polyp size. Thus, neither ODC activity nor polyamine contents of normal mucosa appear to be discriminatory markers of colorectal carcinogenesis. However, spd:spm ratios, which show the least variability among measures of polyamine contents, should be a good marker of the consequence of polyamine synthesis inhibition in chemoprevention trials.

Expression of the human spermidine/spermine N1-acetyltransferase in spermidine acetylation-deficient Escherichia coli.

A cDNA encoding the human spermidine/spermine N1-acetyltransferase (N1SSAT) was conditionally expressed in a strain of Escherichia coli deficient in spermidine-acetylating activity. Conditional expression of this cDNA was performed under the control of the lac promoter, by addition of the non-hydrolysable lactose analogue isopropyl beta-D-thiogalactoside. Expression of the N1SSAT cDNA oriented in the sense direction resulted in the acetylation of spermidine at the N1 but not the N8 position and a decrease in endogenous spermidine contents and growth rates in these bacteria. When this cDNA was expressed in the antisense orientation, spermidine acetylation was not detected and endogenous spermidine contents and growth rates were unaffected. Increasing the endogenous N1-acetylspermidine concentration by addition of this amine to the culture medium did not suppress growth, and increasing endogenous spermidine pools by exogenous addition was not sufficient to restore optimal growth in cells expressing the human N1SSAT. Exogenous spermidine, but neither N1- nor N8-acetylspermidine, stimulated cell growth in strains unable to synthesize spermidine. These results suggest that one physiological consequence of spermidine acetylation in E. coli is growth inhibition. The mechanism of this inhibition seems to involve the formation of acetylspermidine, and is not simply due to a decrease in the intracellular concentration of non-acetylated spermidine.

Polyamine-dependent expression of the matrix metalloproteinase matrilysin in a human colon cancer-derived cell line.

Matrilysin, which is a member of the matrix metalloproteinase family and is implicated in colon cancer invasion, is expressed in human colon adenocarcinoma-derived SW1116 cells. We investigated the effect of alpha-difluoromethylornithine (DFMO) on matrilysin expression in this cell line because others have shown that DFMO can inhibit invasion and carcinogenesis in epithelial tissues, including the colon, in experimental models. DFMO reduced extracellular levels of matrilysin protein after 4 d of treatment. Intracellular levels of matrilysin protein were minimally affected by DFMO treatment. The decrease in extracellular matrilysin protein levels caused by DFMO was not a consequence of lowered steady-state levels of matrilysin mRNA. After 4 d of exposure, the amount of this transcript was higher in DFMO-treated cells than in untreated cultures, whereas the mRNA stabilities were similar. These data show that polyamine depletion by DFMO can suppress the expression of matrilysin, a gene product thought to be involved in tumor invasion. The decrease in extracellular matrilysin protein caused by DFMO treatment appears to be due to a posttranscriptional mechanism, although transcription of this gene also seems to be affected by polyamines in SW1116 cells.