A Systematic Review and Meta-Analysis of Interval Cancer of Stool Testing for Colorectal Cancer Screening

Speaker(s)

Hamashima C1, Abe K2, Terasawa T3, Katayama T4, Hoshi K5, Hosono S6, Sasaki S7, Tadano T8
1Teikyo University, Tokyo, Japan, 2Teikyo University, Itabashi, Tokyo, Japan, 3Fujita Health University, Toyoake, Aichi, Japan, 4Hyogo Prefecture University, Akashi, Hyogo, Japan, 5National Institute of Public Health, Wako, Saitama, Japan, 6National Cancer Center, Chuo, Tokyo, Japan, 7St. Luke’s International Hospital, Tokyo, Japan, 8Miyagi Cancer Association, Sendai, Miyagi, Japan

Presentation Documents

OBJECTIVES: Colorectal cancer (CRC) is a heavy burden worldwide, and CRC screening has been implemented nationally in developed countries. The guaiac fecal occult blood test (gFOBT) has been the primary screening modality for CRC screening, but it has recently been replaced with fecal immunochemical testing (FIT). Interval cancer is diagnosed between the last negative screening and the next scheduled screening. Interval cancer rate is a critical factor in evaluating the effectiveness of cancer screening programs. The interval CRC rate between gFOBT and FIT was compared based on a systematic review and meta-analysis.

METHODS: A literature search was conducted in the Ovid-MEDLINE, Embase, and Ichushi-Web databases for citations related to CRC screening based on stool tests, covering primarily the period from inception to April 2024. Population-based screening for asymptomatic individuals aged 40 years and above was also included. Screenings were divided into three types: prevalence screening, incidence screening, and combined screening. Interval CRC rates per 100,000 person-years following negative results were calculated for each round of CRC screening and compared between gFOBT and FIT.

RESULTS: Of the 7,971 articles, 35 studies remained and were included in the meta-analysis. There were 11 studies on gFOBT and 27 studies on FIT. Twenty-nine studies were reported from Western countries, and 6 were from Asia. The incidence rates of interval CRC following gFOBT and FIT were 70 (95% credible interval: 63-78) and 36 (95% credible interval: 30-40) per 100,000 person-years respectively. Interval CRC rates were nearly equal between prevalence and incidence screening in FIT screening.

CONCLUSIONS: Although these studies could not be directly compared due to their varying contexts, including differences in devices and cut-off values for stool testing, interval CRC rates were lower in FIT than in gFOBT. This result also supports the superiority of test accuracy in FIT for CRC screening.

Code

EPH71

Topic

Clinical Outcomes, Epidemiology & Public Health, Study Approaches

Topic Subcategory

Literature Review & Synthesis, Meta-Analysis & Indirect Comparisons, Performance-based Outcomes, Public Health

Disease

Gastrointestinal Disorders, Oncology