BACKGROUND: Ulcerative colitis (UC) is associated with an increased risk of colorectal cancer (CRC). Aberrant crypt foci (ACF) are important biomarkers of sporadic CRC risk. Their correlation with the risk of intraepithelial neoplasia (IN) in UC remains unclear. AIMS: To assess whether ACF are a risk factor for IN in long-standing UC and to investigate any correlation between the clinico-epidemiological characteristics and prevalence/number of ACF in these patients. METHODS: Seventy-six patients with long-standing UC were prospectively screened by colonoscopy with chromoendoscopy-guided endomicroscopy. ACF were sought in the lower rectum. RESULTS: Eight INs were detected in seven (9.2%) patients. The ACF prevalence and mean number were 60.5% and 2.4 ± 2.8, respectively. The number of ACF was independently associated with the risk of having IN (odds ratio = 1.338; 95% confidence interval 1.030-1.738). ACF number revealed a good calibration (area under the receiver operating characteristic curve = 0.829) and discriminative ability (p = 0.205, Hosmer-Lemeshow test) for the prediction of synchronous IN. Patients with ≥3 ACF have a significantly higher prevalence of IN than patients with <3 ACF (22.6% vs. 0%, p = 0.001). Using this cut-off value, the performance of ACF in predicting the presence of IN was as follows: sensitivity = 100%, specificity = 65.2%, positive predictive value = 22.6%, and negative predictive value = 100%. Age >40 years, family history of CRC, and increased body mass index (BMI) were associated with a significantly higher number of ACF. CONCLUSION: Long-standing UC patients with ≥3 ACF have a significantly higher likelihood of having IN. Age >40 years, family history of CRC, and increased BMI have significant positive associations with the number of ACF.
BACKGROUND:Ulcerative colitis (UC) is associated with an increased risk of colorectal cancer (CRC). Aberrant crypt foci (ACF) are important biomarkers of sporadic CRC risk. Their correlation with the risk of intraepithelial neoplasia (IN) in UC remains unclear. AIMS: To assess whether ACF are a risk factor for IN in long-standing UC and to investigate any correlation between the clinico-epidemiological characteristics and prevalence/number of ACF in these patients. METHODS: Seventy-six patients with long-standing UC were prospectively screened by colonoscopy with chromoendoscopy-guided endomicroscopy. ACF were sought in the lower rectum. RESULTS: Eight INs were detected in seven (9.2%) patients. The ACF prevalence and mean number were 60.5% and 2.4 ± 2.8, respectively. The number of ACF was independently associated with the risk of having IN (odds ratio = 1.338; 95% confidence interval 1.030-1.738). ACF number revealed a good calibration (area under the receiver operating characteristic curve = 0.829) and discriminative ability (p = 0.205, Hosmer-Lemeshow test) for the prediction of synchronous IN. Patients with ≥3 ACF have a significantly higher prevalence of IN than patients with <3 ACF (22.6% vs. 0%, p = 0.001). Using this cut-off value, the performance of ACF in predicting the presence of IN was as follows: sensitivity = 100%, specificity = 65.2%, positive predictive value = 22.6%, and negative predictive value = 100%. Age >40 years, family history of CRC, and increased body mass index (BMI) were associated with a significantly higher number of ACF. CONCLUSION: Long-standing UC patients with ≥3 ACF have a significantly higher likelihood of having IN. Age >40 years, family history of CRC, and increased BMI have significant positive associations with the number of ACF.