Gies Anton, Cuk Katarina, Schrotz-King Petra, Brenner Hermann
Division of Preventive Oncology, German Cancer Research Center (DKFZ) and National Center for Tumor Diseases (NCT), 69120 Heidelberg, Germany.
Medical Faculty Heidelberg, University of Heidelberg, 69120 Heidelberg, Germany.
Cancers (Basel). 2019 Jan 20;11(1):120. doi: 10.3390/cancers11010120.
A variety of fecal immunochemical tests (FITs) are used for colorectal cancer screening. FIT performance could be improved further. It is unclear, whether the combination of different FITs with different analytical characteristics (such as, different antibodies for the detection of fecal hemoglobin) can yield a better diagnostic performance. Fecal samples were obtained from 2042 participants of screening colonoscopy. All participants with advanced neoplasm (AN, colorectal cancer ( = 16) or advanced adenoma ( = 200)) and 300 randomly selected participants without AN were included. Nine quantitative FITs were evaluated simultaneously. Sensitivity and specificity was calculated for single tests ( = 9) and for their pairwise test combinations ( = 36) (requiring either both FITs (P++) or at least one FIT (P+) to be positive for defining a positive test result). Mean age of the participants ( = 516) was 63 (range: 50⁻79) years and 56% were men. At cutoffs yielding a specificity of 96.7% for single FITs, the median gain in specificity by P++ combination was +1.0%, whereas the median loss in sensitivity for AN was -4.2%. For P+ combination the median gain in sensitivity for AN was +2.8%, at a prize of median loss of -1.0% of specificity. Combinations of different FITs do not yield any relevant gain in diagnostic performance.
多种粪便免疫化学检测（FIT）用于结直肠癌筛查。FIT的性能可进一步提高。目前尚不清楚，具有不同分析特征的不同FIT（例如，用于检测粪便血红蛋白的不同抗体）组合是否能产生更好的诊断性能。从2042名接受结肠镜筛查的参与者中获取粪便样本。纳入所有患有晚期肿瘤（AN，结直肠癌（n = 16）或晚期腺瘤（n = 200））的参与者以及300名随机选择的无AN的参与者。同时评估了九种定量FIT。计算了单项检测（n = 9）及其两两检测组合（n = 36）的敏感性和特异性（定义阳性检测结果要求两个FIT（P++）或至少一个FIT（P+）为阳性）。参与者（n = 516）的平均年龄为63岁（范围：50 - 79岁），56%为男性。对于单项FIT，在特异性为96.7%的临界值下，P++组合的特异性中位数增加为+1.0%，而AN的敏感性中位数损失为 - 4.2%。对于P+组合，AN的敏感性中位数增加为+2.8%，代价是特异性中位数损失 - 1.0%。不同FIT的组合在诊断性能上没有产生任何相关的提升。
