Background: Colorectal Carcinoma (CRC) is the third most frequent cancer type, with an incidence and prevalence of approximately 4,800 and 40,000 cases respectively each year in Austria. Multi-target stool DNA (MT-sDNA) testing is a novel non-invasive screening test that may be able to supplement or replace established stool tests by analysing (stool-based) tumour DNA. In this policy brief, the latest available evidence with regard to the clinical benefits and harms of MT-sDNA tests is summarised.


Methods: The quality of a recent EUnetHTA report published in 2019 was critically appraised using the AMSTAR-II checklist. Further, an update systematic rapid review was conducted: a systematic literature search was conducted in three databases (8/2018-5/2021) using the search strategy deployed in the EUnetHTA report. A qualitative synthesis of the evidence was conducted with a focus on effectiveness/safety and diagnostic test accuracy.

Results: The current available evidence consists of five diagnostic test accuracy observational studies (three studies in the EUnetHTA report and two newly identified studies). Hence, there is currently no direct evidence evaluating the clinical effectiveness and safety of the screening device under investigation. Indirect evidence from the EUnetHTA report (benefit-harm analysis considering a lifetime time horizon) showed the following results based on a cohort of 1,000 individuals (comparison: no screening; screening interval: 50 to 74 year olds; based on test accuracy data from two studies and assumptions regarding test frequency tailored to the Austrian system):

• Life years gained: 394 with colonoscopy, 385 with Cologuard®, 365 with FIT, and 358 with ColoAlert®.

• Colorectal cancer-related deaths prevented: 31 with colonoscopy, 30 with Cologuard®, 28 with FIT, and 27 with ColoAlert®.

• Number of Colonoscopies: 904 with FIT, 1,053 with ColoAlert®, 1,292 with Cologuard® and 2,777 with colonoscopy.

Diagnostic test accuracy data from the five currently available studies consistently indicates that the sensitivity of MT-sDNA tests in an average screening population is higher than sensitivities of FIT, yet with a lower specificity. For patients aged 45 to 49 years only, one study found considerably high specificity, although no direct comparison was conducted within this study.

Conclusion: The current evidence is insufficient to demonstrate superiority of MT-sDNA testing compared with other conventional screening methods. Based on diagnostic test accuracy data, modelling analysis shows that MT-sDNA with Cologuard® could be an option among other screening strategies such as FIT or colonoscopy, although 10-yearly colonoscopy is still the most effective strategy. There is still high uncertainty with regard to the use of MT-sDNA with ColoAlert®.
More high quality evidence derived from randomised controlled trials is needed to clearly show whether a screening strategy with MT-sDNA yields clinically relevant benefits in terms of reduced CRC mortality. Suffice to say that adherence, patient preferences and costs and cost-effectiveness further need to be considered in decision making. The screening strategy as such (opportunistic vs. organised) plays a further pivotal role in reducing the burden of CRC that should lastly be prioritised. Next to clinical trials, further decision-analytic evaluations would be necessary to demonstrate an acceptable benefit-harm ratio and cost-effectiveness of screening strategies with varying test time intervals (start and end ages) and a combination of tests.