Fenretinide activity in retinoid-resistant oral leukoplakia.

PURPOSE To test the hypothesis that the retinamide N-(4-hydroxyphenyl)retinamide (fenretinide) would be clinically active potentially via receptor-independent apoptosis and receptor-dependent effects in natural retinoid-resistant oral leukoplakia patients--the first test of this hypothesis in any in vivo setting. EXPERIMENTAL DESIGN A phase II trial of fenretinide (200 mg/d for 3 months) in oral leukoplakia patients who had not responded (de novo resistance) or who had responded and then relapsed (acquired resistance) to previous treatment with natural retinoids. We analyzed apoptosis via the terminal deoxynucleotidyl transferase-mediated nick end labeling in situ DNA fragmentation assay. RESULTS We accrued 35 evaluable patients with retinoid-resistant oral leukoplakia, 12 (34.3%) had partial responses to fenretinide (95% confidence interval, 19.2-52.4%), and response was associated with acquired resistance to natural retinoids (P = 0.015, Fisher's exact test). Nine responders progressed within 9 months of stopping fenretinide. Toxicity was minimal and compliance was excellent. Mean apoptosis values (SE) increased from 0.35% (0.25%) at baseline to 1.18% (0.64%) at 3 months (P = 0.001, sign test); this increase did not correlate with clinical response. The increases in 3-month mean serum concentrations of fenretinide (0.23 micromol/L) and N-(4-methoxyphenyl)retinamide (0.57 micromol/L) correlated with decreased retinol concentrations [Spearman correlation coefficient of -0.57 (P = 0.001) and -0.43 (P = 0.01), respectively]. CONCLUSIONS Low-dose fenretinide was clinically active and produced a small increase in apoptosis in retinoid-resistant oral leukoplakia.