The current study was conducted in the Northwest of Iran as a feasibility study of a screening program of colorectal cancer and setting the new Mt-sDNA test in the region. Also, we tried to evaluate the diagnostic accuracy of Mt-sDNA test panel, the FIT, and gFOBT tests compared with colonoscopy, as the first step to provide the best CRC screening modality in our population. All individuals were first assessed by our established CRC risk assessment tool, then eligible volunteers from average and high-risk groups and relative ages entered the study. Colonoscopy was performed for all participants, also stool-based tests including traditional guaiac, high-sensitivity guaiac-based, the FIT and Mt-sDNA panel tests were performed.
According to the results of the current study, Mt-sDNA test panel had a sensitivity of 77.8% for the detection of colorectal cancer with a specificity of 91.2%. The FIT test alone had a lower sensitivity (66.7%) and almost the same specificity of 93.9% for cancer detection. Also, ROC analytic results showed that both Mt-sDNA panel and the FIT test had acceptable cut-off points for cancer detection, however, Mt-sDNA test had better diagnostic accuracy than the FIT (AUC=0.85 vs 0.80), and is a more useful screening test. The positive likelihood ratio for Mt-sDNA test was in a large and conclusive range (LR+ 8.79), which indicated the absolute probability of cancer detection by Mt-sDNA test. Negative likelihood ratio test which shows the probability of negative cases for cancer was 0.24 for Mt-sDNA, which was relatively in the high probability range. The LR+ was higher for the FIT test alone with an LR+ of 10.9, but the FIT test alone had a moderate probability for the negative outcome (LR- 0.36). Positive and negative predictive values for cancer detection for both Mt-sDNA and the FIT tests had almost the same results, however Mt-sDNA test had better NPV results than the FIT test alone for cancer, AA, and combined CRC and AA.
Comparing these results with the last published NCCN guideline for colorectal cancer screening revealed interesting results (Table 3). The first colonoscopy still has the highest sensitivity for CRC (95%) and AA (89-98%), with a specificity of 90%. The last provided results of diagnostic accuracy of other stool-based tests are comparable with our results, while in this guideline Mt-sDNA test had a sensitivity of 92% for CRC (vs our results: 91.2%), 42% for AA (vs our results: 33.3%), with a specificity of 87% (vs our results: 92.8%). Although the FIT was still more specific for both CRC and AA in both results which indicates that the FIT had fewer false-positive results, we are aware that sensitivity is the most important characteristic of any screening modality [17].
Table 3
Comparing Diagnostic Accuracy Results with the last published NCCN guideline for colorectal cancer screening
|
Sensitivity
|
Specificity
|
|
Colorectal Cancer
|
Advanced Adenoma
|
|
NCCN Guideline
|
Our Results
|
NCCN Guideline
|
Our Results
|
NCCN Guideline
|
Our Results
|
high-sensitivity guaiac-based test
|
62-79%
|
66.7%
|
7%
|
22.2%
|
87-96%
|
96.4%
|
FIT
|
76-95%
|
66.7%
|
27-47%
|
22.2%
|
89-96%
|
94.9%
|
Mt-sDNA panel
|
92%
|
91.2%
|
42%
|
33.3%
|
87%
|
92.8%
|
Despite wide use of colonoscopy in many countries as the golden standard for CRC screening and early diagnosis, it has some main disadvantages which play the role of the predominant barriers and limitations for implementation on a large scale in different populations [21]. Missed diagnosis of AA, side-effecting detection biases, fear and discomfort related to colonoscopy, bowel preparation difficulties, and possible harms have been presented as the predominant barriers [22-24]. Both Mt-sDNA and the FIT tests are noninvasive and patient-friendly, have acceptable diagnostic accuracy do not need any food and drug restrictions, and have higher adherence in the younger age group and average-risk population [16, 11, 25]. Despite the advantages and harmless of stool-based tests, colonoscopy is still the most accepted and available screening modality in numerous countries, with the advantage of higher sensitivity, ability to remove pre-cancerous lesions and with a longer rescreening interval [16, 19, 2, 24]. Although current guidelines provided using non-invasive screening modalities, it has been strongly suggested that all cases with positive stool-based tests should undergo a visual colonoscopy to detect and remove any precancerous lesions [26, 27].
Familial colorectal cancer accounts for 35% of CRC cases, however, 85% of sporadic CRCs have molecular changes like chromosomal instability (CIN), microsatellite instabilities (MSI). However, colorectal cancer has diverse pathogenesis, with different molecular changes and affected pathways [28, 3, 29]. Therefore, biomarkers recently received important attention for using “in screening and early detection programs” of colorectal cancer [30]. Since the approval of Mt-sDNA test as screening modality in CRC screening from US Food and Drug Administration (FDA) in 2014, there was an increase in the use of this test by providers and patients [16, 11]. Most recently, the development of next-generation mt-sDNA testing with higher specificity, has led to detecting at least 3 novel DNA methylated markers [31, 16, 11].
Most recently Imperial et al., conducted a prospective cross-sectional study on 983 participants to determine the specificity of the multitarget stool DNA test in younger volunteers (45-49 YO) and the average-risk population. They revealed that Mt-sDNA panel had higher specificity (95.2%) in this age group, even better than colonoscopy and the FIT [2]. Recently, US Preventive Services Task Force endorsed Mt-sDNA test as a first-line CRC screening modality, by focusing on use in patients 45-49 years old [11]. The lower prevalence of any precancerous lesion and CRC in younger and average age groups leads to higher specificity, which means lower false-positive cases [16, 18, 11]. While according to last released screening guidelines from the American Cancer Society (ACS) which recommend commencing CRC screening from age 45, these pieces of evidence should be taken in mind in terms of advantages of noninvasive, higher specificity and sensitivity and better acceptance in young age groups [2].
The superior sensitivity of Mt-sDNA tests is still under fire because of its high cost and poor availability, and it is still equivocal how test results could be corroborated independently by other tests. However, recently cost-effective analysis revealed that colorectal cancer screening with triennial Mt-sDNA test reduced significantly CRC incidence and mortality compared with no screening; however, was the most costly and less available strategy compared to colonoscopy and/or the FIT tests[32].There is still a need for larger studies and screening trials with evidence-based analyses to prove the benefits of this test, despite its cost and availability.
Recently CRC screening recommended commencing at age 45, according to the American Cancer Society, and this may be implemented in Iran as well [11, 33]. Overall, developing a comprehensive strategy for providing the best screening method and the best age for screening for colorectal cancer in Iran is strongly recommended. However, the design and implementation of screening programs in each community depend on identifying average and high-risk people, and launching a primary risk assessment tool.
Despite available evidence-based guidelines for colorectal cancer screening, for establishing an effective screening program in each population several joint studies and pieces of evidence should be considered. , therefore, we are trying to introduce a suitable screening program to be used in the country Consequently, several factors are considered to determine whether the benefits of screening modality outweigh the risks and costs of the tests and according to WHO recommendations, these factors will include:
• Possible harms from the screening test
• The likelihood of the test correctly identifying cancer
• The likelihood of colorectal cancer
• Possible harms from follow-up procedures
• Whether suitable treatment is available and appropriate
• Whether early detection improves treatment outcomes
• Whether the cancer will ever need treatment
• Whether the test is acceptable to the patients
• Cost of the test
• The extent to which a cancer is treatable