Low-Fat, High Fruit and Vegetable Diets and Weight Loss Do Not Affect Biomarkers of Cellular Proliferation in Barrett Esophagus

Abstract

Risk factors for esophageal adenocarcinoma include obesity, high fat intake, and low consumption of fruits and vegetables. This trial tested whether an intervention to reduce these risk factors in patients with Barrett esophagus, a preneoplastic condition for esophageal adenocarcinoma, could reduce biomarkers of cellular proliferation and, by inference, the risk of neoplastic progression. Eighty-seven men and women with Barrett esophagus were randomized to an intensive dietary intervention or control group. At baseline, 18 and 36 months after intervention, biopsies were obtained at 2-cm intervals throughout the length of the Barrett segment. Ki67/DNA content flow cytometry was used to assess (a) % Ki67-positive proliferating diploid G₁ cells, (b) % total Ki67-positive proliferating cells, (c) presence of aneuploidy, and (d) presence of >6% of cells in the 4N (G₂/tetraploid) fraction of the cell cycle. We also assessed re-epithelialization and length of the Barrett segment, reflux symptoms, and medication use. The intervention effects for energy, fat, fruits and vegetables, and weight were, respectively, −314 kcal, −12.2% energy, 1.8 servings/d, and −4.0 kg at 18 months (all P < 0.005) and were smaller but remained significant at 36 months. There were no significant effects of the intervention on any biomarker of cellular proliferation. The intervention effects ± SE for mean %G₁ Ki67+ cells were 0.98 ± 1.58 at 18 months and 1.79 ± 1.31 at 36 months; the relative risks (95% confidence interval) for developing >6% of cells in 4N were 0.5 (0.1-2.6) at 18 months and 0.75 (0.2-3.1) at 36 months. A single control participant developed aneuploidy. There were no significant effects on re-epithelialization, segment length, or reflux medication use. We conclude that substantial dietary change has no short-term effects on biomarkers of cellular proliferation in Barrett esophagus or on clinical observations of the Barrrett segment.