研究背景：

    结直肠癌（Colorectal Cancer，CRC）筛查与早诊早治是降低其发病率与死亡率的有效手段。粪便免疫化学检测（Fecal Immunochemical Test，FIT）作为国际结直肠癌筛查指南推荐的技术之一，已在多个国家和地区的人群筛查项目中得到广泛应用。FIT筛查的有效性依赖于定期重复检测，但目前各国对筛查间隔的推荐尚存在差异。不同筛查间隔策略下的人群依从性、结直肠癌和进展期腺瘤（Advanced Adenoma，AA）的检出率等关键指标仍缺乏系统性评估。我国指南推荐采用每年一次的FIT筛查策略，但关于高质量人群重复筛查的实证研究仍相对有限，FIT在中国人群中应用的长期筛查效果及其对结肠镜资源的影响尚不明确。

研究目的：

本研究旨在基于既往文献证据，系统评价多轮次FIT筛查的人群依从性及其对结直肠癌和进展期腺瘤检出情况的影响；同时结合真实世界筛查数据，评估4轮年度FIT筛查在人群中的短期效果；进一步，借助模型预测分析不同筛查间隔策略下重复FIT筛查的中长期效果与卫生资源消耗，为优化我国基于FIT的结直肠癌筛查策略提供科学依据。

研究方法：

在PubMed、Embase和Cochrane数据库中检索基于FIT的重复结直肠癌筛查研究，检索时间范围为2002年1 月1日至 2024 年4月16日，通过阅读文献摘要和全文确定符合纳入标准的研究，并提取每项研究的发表时间、作者、筛查对象的年龄范围、FIT性质和阳性阈值、筛查间隔、每一轮FIT筛查的邀请人数、参与筛查人数、结直肠癌和进展期腺瘤检出数等数据，对不同筛查间隔下每轮筛查的人群参与率和进展期结直肠肿瘤（Advanced Neoplasia，AN）检出率进行Meta分析，探索参与率和检出率随筛查轮次的变化趋势。

基于中国人群新型结直肠癌筛查方案比较的多中心随机对照试验（TARGET-C），评价分配至FIT组中的7,793名50~74岁受试者4轮FIT筛查的效果。受试者接受每年一次的FIT筛查，通过流行病学调查问卷收集受试者的个人基本信息、生活方式和疾病史等信息。通过结肠镜和病理检查获取受试者的肠道病变情况。主要结局为结直肠癌和进展期腺瘤的检出率，次要结局包括4轮筛查的参与率、FIT阳性率和结肠镜资源使用情况。计算4轮筛查指标的变化情况。

基于TARGET-C中的筛查参数和公开数据，建立基于FIT的结直肠癌筛查数理模型，预测每年一次、每两年一次、每三年一次筛查间隔下筛查约10年的累积结直肠癌和进展期腺瘤检出率和每筛检1,000例肿瘤所需结肠镜数（Number of Colonoscopies Needed to be Performed to Detect 1,000 Advanced Neoplasia，NNS）。考虑到在真实筛查场景下，10年以上的筛查较难实施，且10年为评估结直肠癌长期效果的一个重要时间窗口。因此，本研究探讨了在12年内，基于三种不同间隔的FIT筛查策略所形成的14种筛查策略的筛查效果和结肠镜资源需求，以探索较为优化的筛查间隔策略。

研究结果：

系统综述共纳入了19项符合要求的研究，包含2,296,071名参与者。随着筛查轮次的增加，参与率呈现出逐步稳定上升的趋势，每年一次筛查和每两年一次筛查的参与率分别达到78.45%和74.97%。在每年一次筛查策略下，报告累积进展期结直肠肿瘤检出率的最高筛查轮次为三轮，每一轮的累积检出率分别为1.38%（95% CI：1.18%–1.63%）、1.95%（95% CI：1.72%–2.21%）和2.50%（95% CI：2.29%–2.72%）。在两年一次筛查策略下，筛查最多进行四轮，前四轮的累积检出率分别为2.22%（95% CI：1.22%–3.22%）、3.44%（95% CI：2.06%–4.82%）、4.26%（95% CI：2.70%–5.83%）和5.10%（95% CI：3.28%–7.29%）。尽管筛查的累积检出率不断增加，但每轮筛查的单轮肿瘤检出率随着筛查的推进逐渐下降。

TARGET-C的FIT组在基线纳入的7,793例受试者中，平均年龄为60.5±6.5岁，42.5%为男性。该筛查队列的人群依从性较高，4轮筛查的参与率均超过80%，共有7,775人至少参加了一轮筛查，累积参与率达到99.77%。4轮进展期结直肠肿瘤单轮检出率出现下降的趋势，分别为1.23%（95% CI：0.99%–1.52%）、0.66%（95% CI：0.48–0.91%）、0.64%（95% CI：0.46–0.88%）、0.16%（95% CI：0.05–0.26%），但累积检出率逐年上升，截至第4轮，达到2.43%（95% CI：2.09%–2.81%）。经过4轮的筛查，累积NNS为7,860。

   3. 基于FIT的结直肠癌筛查模型的预测结果显示，随着筛查轮次的增加，每年一次、每两年一次、每三年一次筛查的单轮进展期结直肠肿瘤检出率出现下降趋势，但累积检出率逐轮上升，筛查至第12年时，三种策略的累积检出率分别达到6.64%（95% CI：6.18%–7.13%）、4.97%（95% CI：4.54%–5.43%）和4.03%（95% CI：3.70%–4.40%）。各轮的NNS和累积NNS均出现上升趋势，筛查进行至12年时，每年一次筛查的NNS最高，其次是每两年一次筛查和每三年一次，分别为15,821、10,731、 9,059。在12年筛查期内，除每年一次筛查12轮外的13种策略中，在保持累积检出率与每年一次筛查无统计学差异的前提下，每年一次筛查6轮后延长筛查间隔至每两年一次的策略，累积NNS最少，为11,084。

研究结论：

基于FIT的重复结直肠癌筛查在人群中具有较高的依从性与筛查效果，支持其作为常规筛查策略持续开展。尽管每年一次的FIT筛查可实现较高的病变检出率，但其伴随的筛查成本和结肠镜资源消耗亦不可忽视。鉴于我国当前结直肠癌的发病率水平，建议在连续多轮FIT结果为阴性的前提下，可以延长筛查间隔至每两年一次，以在维持良好筛查效益的同时，提高筛查资源的使用效率，减轻医疗系统负担。

Background:

Screening and early detection of colorectal cancer (CRC) are effective strategies for reducing its incidence and mortality. As one of the recommended methods in international colorectal cancer screening guidelines, the fecal immunochemical test (FIT) has been widely implemented in population-based screening programs across various countries and regions. The effectiveness of FIT-based screening relies on regular and repeated testing. However, recommendations on screening intervals vary among countries. Key indicators such as population adherence and detection rates of colorectal cancer and advanced adenoma (AA) under different interval strategies have yet to be systematically evaluated. The current Chinese guidelines recommend an annual FIT screening strategy, but high-quality empirical research on repeated screening in target populations remains limited. The long-term effectiveness of FIT-based screening and its impact on colonoscopy resource utilization in the Chinese population are not yet well understood.

Objective:

This study aimed to systematically evaluate, based on existing literature, the adherence to repeated rounds of FIT screening and its impact on the detection of colorectal neoplasms. In addition, using real-world screening data, we assessed the short-term effectiveness of four rounds of annual FIT screening in the population. Furthermore, predictive modeling was employed to estimate the medium- to long-term outcomes and healthcare resource utilization associated with different FIT screening interval strategies, with the goal of providing scientific evidence to optimize FIT-based colorectal cancer screening strategies in China.

Methods:

A systematic search was conducted in PubMed, Embase, and the Cochrane Library for studies on repeated FIT-based colorectal cancer screening published between January 1, 2002, and April 16, 2024. Titles, abstracts, and full texts were screened to identify studies that met the predefined inclusion criteria. For each eligible study, data were extracted on publication year, authors, target population age range, FIT characteristics and positivity thresholds, screening interval, the number of individuals invited and those who participated in each screening round, and the number of detected colorectal cancers and advanced adenomas. A meta-analysis was conducted to evaluate the per-round participation rate and detection rate of advanced colorectal neoplasia (AN) under different screening intervals, aiming to explore the trends in participation and detection rates across successive screening rounds.

Based on the multicenter randomized controlled trial (Comparative Evaluation of Novel Screening Strategies for Colorectal Cancer Screening in China, TARGET-C), the effectiveness of four rounds of FIT screening was evaluated among 7,793 individuals aged 50–74 who were assigned to the FIT group. Participants underwent annual FIT screening. Demographic characteristics, lifestyle factors, and medical history were collected through epidemiological questionnaires. Colorectal lesions were identified through colonoscopy and pathological examinations. The primary outcomes were the detection rates of colorectal cancer and advanced adenomas. Secondary outcomes included participation rates across four screening rounds, FIT positivity rates, and colonoscopy resource utilization. Changes in screening indicators over the four rounds were calculated.

Based on screening parameters from the TARGET-C study and publicly available data, a mathematical model of FIT-based colorectal cancer screening was developed to predict the cumulative detection rate of advanced colorectal neoplasms and the number of colonoscopies needed to detect 1,000 advanced neoplasms (NNS) under different screening intervals: annual, biennial, and triennial. Considering that it is challenging to implement screening programs lasting over 10 years in real-world settings, and that 10 years is an important time frame for evaluating the long-term effects of colorectal cancer screening, this study examined the screening effectiveness and colonoscopy resource demands of 14 FIT-based strategies within a 12-year period, based on three different screening intervals, in order to explore optimized interval strategies.

Results:

A total of 19 eligible studies involving 2,296,071 participants were included in

the systematic review. Participation rates showed a steadily increasing trend across screening rounds, reaching 78.45% for annual screening and 74.97% for biennial screening. Under the annual screening strategy, studies reported data for up to three screening rounds, with cumulative detection rates of advanced colorectal neoplasms being 1.38% (95% CI: 1.18%–1.63%), 1.95% (95% CI: 1.72%–2.21%), and 2.50% (95% CI: 2.29%–2.72%), respectively. For biennial screening, up to four rounds were reported, with cumulative detection rates of 2.22% (95% CI: 1.22%–3.22%), 3.44% (95% CI: 2.06%–4.82%), 4.26% (95% CI: 2.70%–5.83%), and 5.10% (95% CI: 3.28%–7.29%). Although cumulative detection rates increased over time, the per-round detection rate of advanced colorectal neoplasms declined progressively with successive rounds.

   2. In the TARGET-C trial, 7,793 participants were enrolled in the FIT group, with a mean age of 60.5±6.5 years and 42.5% being male. High adherence was observed, with participation rates exceeding 80% in each of the four screening rounds. A total of 7,775 individuals completed at least one round, yielding a cumulative participation rate of 99.77%. The detection rate of colorectal neoplasms showed a decreasing trend across the four rounds—1.23% (95% CI: 0.99%–1.52%), 0.66% (95% CI: 0.48–0.91%), 0.64% (95% CI: 0.46–0.88%), 0.16% (95% CI: 0.05–0.26%), but the cumulative detection rate increased annually, reaching 2.43% (95% CI: 2.09%–2.81%) after the fourth round. The NNS over four rounds was 7,860.

   3. Model-based predictions of FIT screening showed that while the per-round detection rate of advanced colorectal neoplasms decreased with more screening rounds, cumulative detection rates rose. By the 12th year of screening, the cumulative detection rates for annual, biennial, and triennial strategies reached 6.64% (95% CI: 6.18%–7.13%), 4.97% (95% CI: 4.54%–5.43%), and 4.03% (95% CI: 3.70%–4.40%), respectively. Both round-specific and cumulative NNS increased with time. By year 12, the NNS for annual, biennial, and triennial screening were 15,821, 10,731, and 9,059, respectively. Within the 12-year screening period, among the 13 strategies other than annual screening for 12 rounds, the strategy of conducting annual screening for six rounds followed by biennial screening, while maintaining no statistically significant difference in cumulative detection rate compared to annual screening, resulted in the lowest cumulative NNS, at 11,084.

Conclusions:

Repeated colorectal cancer screening using FIT demonstrates high compliance and effectiveness in the target population, supporting its continued use as a routine screening strategy. Although annual FIT screening is associated with a relatively high detection rate of advanced neoplasms, the accompanying screening costs and colonoscopy burden are considerable. Given the current incidence of colorectal cancer in China, it may be appropriate to extend the screening interval to biennial screening after several consecutive negative FIT results. This approach may help maintain the effectiveness of screening while enhancing the efficiency of resource utilization and reducing the burden on the healthcare system.