Association Between Concentrations of Hemoglobin Determined by Fecal Immunochemical Tests and Long-term Development of Advanced Colorectal Neoplasia
Association Between Concentrations of Hemoglobin Determined by Fecal Immunochemical Tests and Long-term Development of Advanced Colorectal Neoplasia
复制标题
DOI：
10.1053/j.gastro.2017.07.034
复制
发表时间：
2017-11-01
期刊：
影响因子：
29.4
通讯作者：
Kuipers, Ernst J.
中科院分区：
文献类型：
--
作者：
Grobbee, Esmee J.;Schreuders, Eline H.;Kuipers, Ernst J.
背景与目的：使用定量粪便免疫化学试验（FITs）筛查结直肠癌（CRC）在全球范围内迅速普及。fit总是以二分法使用预先指定的截止值。为了优化基于FIT的筛查方案，我们研究了低于FIT临界值的粪便血红蛋白（(f)Hb）浓度与结直肠癌晚期肿瘤（AN）的后期发展之间的关系。方法：我们分析了一项基于人群的研究收集的数据，该研究收集了9561名荷兰平均风险受试者（50-74岁）的数据，这些受试者从2006年11月至2014年12月接受了4轮FIT CRC筛查。我们分析了7663名至少筛查一次的参与者的数据，发现在基线时FIT结果为阴性（低于10 μ g Hb/g粪便的临界值）。参与者的随访时间中位数为4.7年（四分位数间距为2.0-6.1年）；在筛查项目之外诊断出的crc是从荷兰综合癌症中心的数据库中确定的。采用Cox比例风险回归分析确定AN的风险比。使用逻辑回归技术计算连续fHb浓度低于临界值后AN的风险。结果：经过8年的随访，基线浓度为8-10 μ g fHb/g的参与者AN的累积发生率（33%）高于基线浓度为0 μ g fHb/g的参与者（5%）（P < 0.001）。多因素风险比从0-2 μ g fHb/g受试者的1.2增加到8-10 μ g fHb/g受试者的8.2 （P < 0.001）。与连续2次fHb浓度为0 μ g Hb/g的参与者相比，连续2次fHb浓度为8 μ g Hb/g的参与者AN的风险增加了14倍（P < 0.001）。结论：在一项基于人群的研究中，FIT结果低于临界值的平均风险个体，我们将基线浓度为8-10 μ g fHb/g与基线浓度为0 μ g fHb/g的an风险增加联系起来。基线和连续fHb浓度是AN事件的独立预测因子。这些信息可用于设计基于人群的CRC筛查的个性化策略，并减少不必要的重复检查。
BACKGROUND & AIMS: Colorectal cancer (CRC) screening using quantitative fecal immunochemical tests (FITs) is rapidly gaining ground worldwide. FITs are invariably used in a dichotomous manner using pre-specified cut-off values. To optimize FIT-based screening programs, we investigated the association between fecal hemoglobin ((f)Hb) concentrations below the FIT cut-off value and later development of colorectal advanced neoplasia (AN). METHODS: We analyzed data collected from a population-based study of 9561 average-risk subjects (50-74 years old) in the Netherlands who were offered 4 rounds of FIT screening for CRC from November 2006 through December 2014. We analyzed data from 7663 participants screened at least once and found to have a negative FIT result at baseline (below the cut-off value of 10 mu g Hb/g feces). Participants were followed for a median of 4.7 years (interquartile range, 2.0-6.1 years); CRCs diagnosed outside the screening program were identified from the Dutch Comprehensive Cancer Centre database. Hazard ratios for AN were determined using Cox proportional hazard regression analyses. Logistic regression techniques were used to calculate risks of AN after consecutive fHb concentrations below the cut-off value. RESULTS: After 8 years of follow-up, participants with baseline concentrations of 8-10 mu g fHb/g had a higher cumulative incidence of AN (33%) than participants with 0 mu g fHb/g (5%) (P < .001). Multi-variate hazard ratios increased from 1.2 for subjects with concentrations of 0-2 mu g fHb/g to 8.2 for subjects with concentrations of 8-10 mu g fHb/g (P < .001). Participants with 2 consecutive fHb concentrations of 8 mu g Hb/g had a 14-fold increase in risk of AN compared with participants with 2 consecutive fHb concentrations of 0 mu g Hb/g (P < .001). CONCLUSIONS: In a population-based study of average-risk individuals with a FIT result below the cut-off value, we associated baseline concentrations of 8-10 mu g fHb/g with an increased risk of AN compared with baseline concentrations of 0 mu g fHb/g. Baseline and consecutive fHb concentrations are independent predictors for incident AN. This information might be used in designing personalized strategies for population-based CRC screening and reduce unnecessary repeat tests.