Cell proliferation, cell cycle abnormalities, and cancer outcome in patients with Barrett's esophagus: a long-term prospective study

Abstract

Purpose: Elevated cellular proliferation and cell cycle abnormalities, which have been associated with premalignant lesions, may be caused by inactivation of tumor suppressor genes. We measured proliferative and cell cycle fractions of biopsies from a cohort of patients with Barrett's esophagus to better understand the role of proliferation in early neoplastic progression and the association between cell cycle dysregulation and tumor suppressor gene inactivation.

Experimental design: Cell proliferative fractions (determined by Ki67/DNA multiparameter flow cytometry) and cell cycle fractions (DNA content flow cytometry) were measured in 853 diploid biopsies from 362 patients with Barrett's esophagus. The inactivation status of CDKN2A and TP53 was assessed in a subset of these biopsies in a cross-sectional study. A prospective study followed 276 of the patients without detectable aneuploidy for an average of 6.3 years with esophageal adenocarcinoma as an end point.

Results: Diploid S and 4N (G(2)/tetraploid) fractions were significantly higher in biopsies with TP53 mutation and loss of heterozygosity. CDKN2A inactivation was not associated with higher Ki67-positive, diploid S, G(1), or 4N fractions. High Ki67-positive and G(1)-phase fractions were not associated with the future development of esophageal adenocarcinoma (P=0.13 and P=0.15, respectively), whereas high diploid S-phase and 4N fractions were (P=0.03 and P<0.0001, respectively).

Conclusions: High Ki67-positive proliferative fractions were not associated with inactivation of CDKN2A and TP53 or future development of cancer in our cohort of patients with Barrett's esophagus. Biallelic inactivation of TP53 was associated with elevated 4N fractions, which have been associated with the future development of esophageal adenocarcinoma.

Figure 1

Ki67-Phycoerthyrin (PE) and DNA content flow cytometry. Panel A: Bivariate histogram displaying DAPI fluorescence on the x-axis (linear) and Ki67-PE fluorescence on the y-axis (logarithmic; 4 decades) of diploid cells in a Barrett’s esophagus biopsy. Panel B: Ki67-positive proliferating diploid cells, as determined using the bivariate subtraction algorithm (Multicycle). The total Ki67-positive fraction of this sample was calculated to be 35.5%. A representation of the diploid G₁ sorting gate is shown. Panel C: A single parameter histogram of DNA content, with S and 4N (G₂/M) cells indicated. The S phase fraction is 6.5% and the 4N fraction is 9.9%. There was no evidence of a tetraploid cell cycle.

Figure 2

Ki67-positive, G₁ phase, S phase, and 4N fractions of biopsies by TP53 inactivation status. Measurements from biopsies no TP53 inactivation (TP53 wt/17p het), with either TP53 mutation or 17p LOH (TP53 mutation or 17p LOH), and both mutation and LOH (TP53 mutation and 17p LOH) are summarized. The boxes indicate median and middle two quartiles, the whiskers the 95th percentiles, and the outliers are plotted as circles. The number of samples in each group is printed below each box. The biopsies with no TP53 inactivation (TP53+/+) were compared with the other two groups (TP53+/− and TP53−/−) using the Wilcoxon rank sum test, and the p values are shown on the plot.

Figure 3

Receiver operating characteristic (ROC) curves for detecting TP53 inactivation by mutation or LOH in diploid samples using S or 4N fractions. The upper plots are ROC curves for discriminating between biopsies with no detectable TP53 inactivation and those with any inactivation. 42/344 biopsies in which TP53 inactivation and S phase fractions were ascertained had detectable TP53 mutation or LOH, and 41/342 biopsies in which TP53 inactivation and 4N fractions were ascertained had detectable TP53 mutation or LOH. The associated areas under the curves (AUC) are indicated. The lower plots are ROC curves for discriminating between biopsies with both forms of TP53 inactivation (mutation and LOH) and all other biopsies. 16/344 biopsies had detectable TP53 mutation and LOH in the S phase plot, and 15/342 biopsies had detectable TP53 mutation and LOH for the 4N plot.