Screening for colorectal cancer

Colorectal cancer is the fourth most common cancer in Austria [1] overall (third most common for men and women separately) [2, p. 55] and the second most common cause of overall cancer deaths (third most common for men and women separately). More men (57.1% of all new cases) than women (42.9%) are affected by colorectal cancer. Between 2017 and 2019, approximately 60% of all colorectal cancer diagnoses in Austria were made after the tumor had already penetrated the organ boundaries (localized stage 40.1%; regionalized stage 43.7%; disseminated stage 16.2%). In Austria, the colorectal cancer incidence rates decreased from 1983 to 2016, with similar decreases from 2003 to 2016 for all 5‑year age groups 45 years or older. While colorectal cancer is exceedingly rare among younger individuals, an increase in incidence was observed in the Austrian population younger than 35 years. Furthermore, the colon cancer incidence rates appear to increase in men and women aged 45–49 years (just outside the currently recommended screening age range), but not in 50–54-year-olds. Although based on small numbers, this increase in incidence rates among younger adults is consistent with that observed in other European countries. Among the risk factors for colorectal cancer that have been attributed to the recent rise of colorectal cancer in the younger age groups are obesity, an unhealthy (processed meat-based) diet, smoking, excess of alcohol consumption and low levels of physical exercise [3]. Smoking also increases the risk of colorectal cancer in individuals with cancer predisposition syndromes such as Lynch syndrome [3].

To date, colorectal cancer screening in Austria is opportunistic including the use of colonoscopy or stool-based blood tests.

Currently used stool-based blood tests include guaiac-based fecal occult blood tests (gFOBT), which have been available since the early 1980s or the more recently available fecal immunochemical tests (FIT). If a stool-based blood test is positive, colonoscopy is always indicated as a diagnostic follow-up examination, regardless of the usual screening intervals. The main imaging procedure used for colorectal cancer screening in Austria is colonoscopy. Public health authorities have offered colonoscopy in asymptomatic populations since 2005. The most recent clinical practice guidelines on colonoscopy screening were published by the Austrian Society of Gastroenterology and Hepatology (ÖGGH) in 2016 [4]. For asymptomatic persons, the guidelines recommend colonoscopy screening every 10 years starting at age 50 years up to age 75 years. Other recommendations are based on family or personal history and screening intervals or starting ages vary depending on these factors. Stool-based blood tests are not addressed in the screening recommendations.

In January 2021, the Austrian Ministry of Social Affairs, Health, Care and Consumer Protection established the Austrian National Committee for Cancer Screening (ANCCS) as an advisory committee to the Minister of Health. The tasks of the ANCCS are to develop independent evidence-based recommendations on cancer screening and to support the implementation of organized cancer screening programs in Austria. Its members, represented by the co-authors of this publication, are experts in public health, epidemiology, health decision science, oncology, pathology, health economics, law, and ethics, and patients’ or citizens’ representatives. Consultation by the committee members is honorary and occurs free of charge. All members must declare their potential conflicts of interest.

In an iterative process involving the Ministry of Health and stakeholders, members of the ANCCS developed research questions and refined a population, intervention, comparator and outcomes (PICO) framework for inclusion criteria for studies of interest. A modified Delphi approach was used to select screening interventions of interest for the Austrian context and to determine which outcomes are relevant for informed decision-making about colorectal cancer screening. Figure 1 presents an analytic framework for colorectal cancer screening. Table 1 presents inclusion and exclusion criteria to define the scope of the recommendations. The evidence review was guided by two research questions:

The Austrian National Public Health Institute for Health Promotion, Quality, Planning and Research (Gesundheit Österreich GmbH) conducted an evidence review based on the research questions and the inclusion criteria outlined above [7].

A net benefit exists when the benefits of a preventive intervention outweigh the harms. An important aspect in determining a net benefit is the assessment of the certainty of the available evidence. The certainty of evidence reflects the extent to which the confidence in an estimate of the effect is adequate to support a particular recommendation [8]. To determine the certainty of the evidence, the Austrian National Public Health Institute for Health Promotion, Quality, Planning and Research used the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach [9, 10]. When assessing the net benefit of a screening intervention, the ANCCS primarily considered patient-relevant health outcomes, i.e., outcomes that individuals can feel or experience (e.g., cancer-specific mortality, incidence of cancer, major bleeding, perforation, potential psychological harms).

In a structured process, members of the ANCCS considered 1) the magnitude of the net benefit, 2) the certainty of evidence, and 3) the level of intervention acceptability in the target population. The ANCCS did not consider the financial costs or cost-effectiveness of the screening interventions. The ANCCS used the US Preventive Services Task Force recommendation grades, which are summarized in Table 2.

The Austrian National Public Health Institute for Health Promotion, Quality, Planning and Research organized a meeting with Austrian stakeholders to present and discuss draft recommendations. Comments from the review process were taken into consideration by the ANCCS.

A systematic review by Lin et al. identified various studies that confirmed the general efficacy of stool-based tests to reduce colorectal cancer-specific mortality [11]. Based on five randomized controlled trials (RCTs; N = 404,396), Lin et al.’s pooled results reported that biennial screening with gFOBT (2–9 rounds of screening) was associated with a reduction in colorectal cancer-specific mortality compared with no screening after 11–30 years (risk ratio, RR 0.78; 95% confidence interval, CI 0.65–0.93 at 30 years) [11].

Likewise, a prospective cohort study (N = 5,417,699) evaluating a Taiwanese screening program reported that 1–3 rounds of screening with a biennial FIT led to a lower colorectal cancer mortality than no screening (adjusted RR 0.90; 95% CI 0.84–0.95) [12].

Evidence on the comparative effectiveness of screening with gFOBT and FIT was limited to two RCTs [13, 14] and a prospective cohort study [15]. These studies assessed colorectal cancer detection rates but did not assess cancer-specific mortality. Nevertheless, the results showed consistently higher detection rates for FIT than for gFOBT. The RRs for colorectal cancer detection ranged from 1.84 (95% CI 0.71–4.79) to 2.46 (95% CI 1.82–3.33), favoring FIT over gFOBT [11]. The pooled sensitivity for FIT was 0.74 (95% CI 0.64–0.83) [11], for gFOBT sensitivities ranged from 0.50–0.75 (95% CI 0.09–1.0) in 2 studies [11]. The specificities were similar between FIT and gFOBT [11].

A decision-analytic modeling study for the Austrian context performed by Jahn et al. on behalf of the Austrian Colorectal Cancer Screening Model Group reported that FIT led to more life-years gained with fewer (false) positive tests than gFOBT, regardless of the screening intervals and the start or end screening age [16]. Biennial screening with FIT (from 45–75 years of age) would lead to 436 life-years gained compared with 407 when using gFOBT. At the same time, FIT would lead to 1514 additional colonoscopies compared to 1743 with gFOBT [16].

Apart from the risk of missing cancers or advanced adenomas (false negative results), no serious adverse events occur with noninvasive stool-based tests. Serious adverse events, however, may result from follow-up colonoscopy for abnormal stool-based tests. Therefore, the rate of false positive results, which can lead to unnecessary follow-up colonoscopy, is an important parameter to consider. Because of the similar specificities of gFOBT and FIT, however, false positive findings are likely to occur at similar rates.

Based on the available evidence, the ANCCS recommends the use of FIT as a stool-based test for colorectal cancer screening (A recommendation, moderate evidence).

Based on the review by Lin et al., two prospective observational studies reported a lower incidence of colorectal cancer and colorectal cancer-specific mortality for people who underwent colonoscopy compared with no screening [11]. In one study (N = 88,902) the colorectal cancer-specific mortality after 24 years was lower in people who self-reported at least one screening colonoscopy compared with those who had never had a screening colonoscopy (adjusted hazard ratio, HR 0.32; 95% CI 0.24–0.45) [17]. The other study (N = 348,025) reported a lower risk for developing colorectal cancer in people aged 70–74 years after 8 years of follow-up [18] (absolute risk reduction −0.42 percentage points; 95% CI −0.24 to −0.63). In people aged 75–79 years, the benefit was smaller and no longer statistically significant. The Nordic-European Initiative on Colorectal Cancer (NordICC) [19], a pragmatic RCT (N = 84,585) from Norway, Poland, and Sweden, was published after the ANCCS recommendations meeting. The NordICC reported smaller screening effects than the abovementioned observational studies after 10 years of follow-up but, overall, confirmed the benefits of screening with colonoscopy.

Regarding the harms of colonoscopies, the systematic review by Lin et al. included data of 21 observational studies (N = 903,872) to assess serious adverse events. Major bleedings occurred in 17.5 per 10,000 procedures (95% CI 7.6–27.5) and perforations in 5.7 per 10,000 procedures (95% CI 2.8–8.7) [11].

The systematic review by Lin et al. did not detect any prospective studies that directly compared the screening strategies of stool-based tests and colonoscopy [11]. The best available evidence was an RCT that compared a single colonoscopy or flexible sigmoidoscopy with four rounds of FIT [20]. The colorectal cancer detection rates were similar between participants receiving colonoscopy and those receiving FIT (0.63% vs. 0.77%). Other studies compared the colorectal cancer detection rates of a single colonoscopy with one-time testing with FIT [21‐23]. No eligible evidence was available on the screening intervals and start and end age to enable colonoscopy and FIT screening comparisons. Due to the lack of direct evidence regarding the screening intervals and start and end screening age, the ANCCS took modeling studies into consideration.

The UMIT TIROL was tasked with updating a former decision-analytic modeling study for the Austrian context to assess different screening intervals as well as different start and end ages for colorectal cancer screening [16]. Based on the Austrian experience with its current opportunistic colorectal cancer screening, an attendance rate of 29% for colonoscopy and 39% for FIT was used in the base-case analysis of the modeling study. All screening strategies in the model were more beneficial than no screening with respect to life-years gained (range 366–488 life-years gained for 1000 participants from age 40 years to death, depending on the strategy). By lowering the starting age of biennial FIT screening from 50 years to 45 years, 2 additional deaths per 1000 persons screened were prevented (41 years of life gained); however, this change also resulted in 242 additional colonoscopies (6 additional colonoscopies per life-year gained). Likewise, by commencing colonoscopy screening at the age of 45 instead of 50 years (with 3 screening rounds from 45–75 years), 2 additional deaths per 1000 persons were prevented (41 years of life gained). This change, however, resulted in 798 additional colonoscopies (19 additional colonoscopies per life-year gained).

A modeling study by Knudsen et al. simulated 163 different screening strategies in a hypothetical cohort of US adults aged 40 years with no prior cancer diagnosis [24]. Similar to the results from the Austrian study by Jahn et al., the findings showed that commencing screening at age 45 years results in more life-years gained and fewer colorectal cancer cases and deaths than similar strategies with screening initiation at ages 50 or 55 years, albeit with a higher lifetime burden of colonoscopy and non-colonoscopy examinations and a slightly higher lifetime risk of complications [24].

Based on the available evidence, primarily from the decision-analytic modeling study by Jahn et al. reflecting the Austrian healthcare context, the ANCCS recommends a quality-assured colorectal cancer screening program with colonoscopy or FIT for individuals aged 45–75 years (individualized from age 65 years). Both screening strategies are considered equivalent, and citizens should be able to make an informed decision about their preferred method (A recommendation, moderate evidence). The ANCCS recommends colonoscopy and FIT as equivalent screening strategies because the acceptance of colonoscopy in the Austrian population is low (currently about 29%). Offering FIT as a second equivalent strategy can help reach persons who are unwilling to undergo colonoscopy screening.