Limitations regarding the FIT studies were also noted. First, a lack of diverse populations proved to be a limit for some studies. Chang et al included only participants of Chinese descent,11 and de Wijkerslooth et al noticed the majority of their participants were Caucasian.7 An inability to obtain information from 3 testing programs to assess FITs limited the generalizability of 1 study’s results.8 The inclusion of younger participants in the systematic review could lead to a bias.12 Not all study samples were limited to persons of average CRC risk, and inclusion of participants with high risk may limit the applicability of findings to the general population.6,7,9,11 One study demonstrated a high rate of compliance for the FIT screening test (60%-62%), but a low compliance rate with colonoscopy (22%), which made it difficult to compare and verify results.7 Another limitation exists when including voluntary participants who were not screening-naive. Chang et al included participants who volunteered for health check-ups and were concerned this may have led to selection bias with higher-socioeconomic-status participants.11 Although Grobbee et al included 4 rounds of FIT testing, a different method was used in the final round, so participants were less familiar with it.9 Concerns regarding the accuracy of the specificity and sensitivity results represent another limitation: Having a limited number of positive results (8 with CRC, 24 with isolated proximal AN, and 83 with isolated distal AN from 1256 total participants) may have affected the precision of estimates regarding FIT sensitivity in CRC screening and AN detection.7 Two studies included only 1 round of FIT screening to assess accuracy.7,10 This may underestimate the sensitivity of the FIT compared with studies including data from several screenings. The systematic review noted 7 studies with at least a 2-year longitudinal follow-up, which could have led to an overestimation of sensitivity and an underestimation of specificity.12 The same review also noted the inability to determine sensitivity and specificity of the FIT for proximal vs distal sites because most of the studies in their review did not include this information.
Two of the 3 head-to-head studies with the FIT and MT-sDNA tests had limitations worth noting. These studies had small sample sizes (456 and 661, respectively) which interfered with being able to analyze their results.16,17 One sample only included the Native Alaskan population.16
Considerations and implications for practice
The goal of this literature review is to provide information on both the FIT and MT-sDNA test to assist providers in advising patients on other options for colorectal cancer screening in situations when patients refuse colonoscopy. To further understand the implications for the MT-sDNA test and FIT, it is important to evaluate the performance, cost, and ease of use by patients for each of the tests. Both tests are approved by the US FDA, US Preventive Services Task Force, and several other organizations for use as screening tests.
With the MT-sDNA test, there is a high degree of detection for CRC as well as precancerous lesions. Lidgard et al reported the MT-sDNA’s high sensitivity and specificity in identifying CRC (98% and 90%, respectively).14 In addition, the MT-sDNA detected lesions at both proximal and distal sites, which supports its reliability. Some authors even suggest that the use of a MT-sDNA test before a colonoscopy may aid in the detection of proximal lesions that are often missed by conventional colonoscopy.14,17 SSAs and AAs can be detected by sDNA methylation markers. Neither colonoscopy nor FIT alone is sensitive enough to detect SSAs, which makes the sDNA methylation markers useful in CRC screening. Furthermore, the FIT was unable to detect sessile serrated polyps, whereas the sDNA test detected sessile serrated polyps ≥ 1 cm.17 However, there are still questions that need answered to truly validate the MT-sDNA test. Storage and collection should be optimized and consistent. Further studies need to explore how inconsistent collection procedures affect the results of the test. Additional research efforts are also needed regarding testing intervals, costs, and patient acceptance and adherence. The manufacturer for the MT-sDNA test suggests its use every 1 to 3 years for CRC screening.2 A 3-year interval seems to provide reasonable performance at acceptable costs with lower patient, clinician, and administrative burdens compared with annual screening.18 The MT-sDNA test (Cologuard) had a negative predictive value of a single test event of 99.94% for CRC and 95% for advanced adenoma when using a multiyear testing interval.15 With these data, the presence of missed significant lesions before the next screening interval would be rare.
There is a valid concern with the MT-sDNA test regarding its lower specificity that may lead to a higher number of false-positive results. Patients need to understand that a positive result warrants a follow-up colonoscopy. This could lead to additional cost and risk for the patient, as well as concern about screening compliance. Another consideration that is important to note is the lack of evidence regarding appropriate longitudinal follow-up of abnormal results after a negative diagnostic colonoscopy.2 With a positive MT-sDNA test and negative follow-up colonoscopy, unnecessary close surveillance could occur because of both patient and provider concerns over the genetic component of the test.