Advances in Faecal DNA Testing for Colorectal Cancer Screening: A Literature Review for Primary Care Providers

Abstract

Recent epidemiological data gathered by the Centers for Disease Control and Prevention (CDC) suggest that colorectal cancer (CRC) is the second leading cause of cancer-related deaths in the USA. Clinicians currently use five types of test to screen for CRC. Two of these five types, the DNA stool test and the faecal occult blood test, are non-invasive. The DNA stool test successfully detects both advanced neoplasias and non-advanced adenomas with more sensitivity than the faecal occult blood test. However, data suggest that it also generates more false-positive results. There is only one DNA stool test that has been approved by the U.S. Food and Drug administration (FDA): the approved Cologuard® (Exact Sciences, Madison, Wisconsin, USA) test. This test screens for nine different DNA biomarkers, one haemoglobin biomarker, and one β-actin. This article is a literature review of research on faecal DNA biomarkers conducted in the past 5 years from four large databases. Key findings include the ability to reach a sensitivity as high as 98% to detect abnormalities in the colon using a multi-target stool DNA-based assay. In comparison, the Cologuard offers 92% sensitivity and 87% specificity for all stages of CRC. Testing DNA biomarkers can serve as an adequate screen for cancer and adenomas in average-risk adults. Areas for further research include implementing studies to compare long-term health consequences for patients who receive colonoscopies versus DNA stool tests, finding ways to improve both the sensitivity and specificity of screening tests, and finding ways to improve the detection of those biomarkers most associated with CRC, including microRNA detection in the marking panel.

Full text

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
84 85
ADVANCES IN FAECAL DNA TESTING FOR
COLORECTAL CANCER SCREENING: A LITERATURE
REVIEW FOR PRIMARY CARE PROVIDERS
*Louise Babikow, Adelle Grant McAuley, Jenny Calhoun
University of Pennsylvania, Philadelphia, Pennsylvania, USA
*Correspondence to louisebabikow@gmail.com
Disclosure: The authors have declared no conflicts of interest.
Received: 19.06.17 Accepted: 13.11.17
Citation: EMJ. 2018;3[1]:84-89.
ABSTRACT
Recent epidemiological data gathered by the Centers for Disease Control and Prevention (CDC)
suggest that colorectal cancer (CRC) is the second leading cause of cancer-related deaths in the USA.
Clinicians currently use five types of test to screen for CRC. Two of these five types, the DNA stool
test and the faecal occult blood test, are non-invasive. The DNA stool test successfully detects both
advanced neoplasias and non-advanced adenomas with more sensitivity than the faecal occult blood test.
However, data suggest that it also generates more false-positive results. There is only one DNA stool
test that has been approved by the U.S. Food and Drug Administration (FDA): the approved Cologuard®
(Exact Sciences, Madison, Wisconsin, USA) test. This test screens for nine dierent DNA biomarkers,
one haemoglobin biomarker, and one β-actin. This article is a literature review of research on faecal DNA
biomarkers conducted in the past 5 years from four large databases. Key findings include the ability to
reach a sensitivity as high as 98% to detect abnormalities in the colon using a multi-target stool DNA-based
assay. In comparison, the Cologuard oers 92% sensitivity and 87% specificity for all stages of CRC.
Testing DNA biomarkers can serve as an adequate screen for cancer and adenomas in average-risk adults.
Areas for further research include implementing studies to compare long-term health consequences for
patients who receive colonoscopies versus DNA stool tests, finding ways to improve both the sensitivity
and specificity of screening tests, and finding ways to improve the detection of those biomarkers most
associated with CRC, including microRNA detection in the marking panel.
Keywords: Colorectal cancer (CRC), early markers, faecal immunochemical test, screening test, stool
DNA (sDNA).
INTRODUCTION
According to the Centers for Disease Control
and Prevention (CDC), colorectal cancer (CRC)
is currently the second leading cause of cancer-
related deaths in the USA.1 In 2013, approximately
71,000 men and 65,000 women were diagnosed
with CRC.1 Primary care providers use five
types of test to screen for CRC: two faecal occult
blood tests (the guaiac faecal occult blood test
and the faecal immunochemical test [FIT]),
a DNA stool test (also referred to as the
FIT-DNA), sigmoidoscopy, colonoscopy, and virtual
colonoscopy.2 Unfortunately, many symptoms of
CRC, such as a change in bowel habits, blood
in the stool, narrow stools, weight loss, fatigue,
and vomiting, only manifest once the cancer
has approached late stages. Almost half of CRC
incidences are detected at late stages.3 Therefore,
early screening, accurate detection, and early
intervention are all key to improved outcomes and
decreased mortality.
Early detection of CRC is crucial to treatment and
survival. The most reliable way to detect early
CRC is through colonoscopy.4 Colonoscopy oers
high accuracy and allows for longer time between
screening but also has numerous disadvantages,
including an extensive bowel preparation process,
risks due to general anaesthesia, risk of bowel
perforation, and length of the procedure.4
The U.S. Preventive Services Task Force

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
84 85
recommends that all adults have a colonoscopy
for screening purposes, beginning at 50 years old,
and that they undergo a repeat colonoscopy every
10 years if the results are normal, until the age
of 75 years.4
In contrast, faecal occult blood tests and DNA stool
tests oer a non-invasive screening alternative that
involves no preparation and can be completed at
home.4 The development of CRC is a multistep
process that begins with the normal mucosa of
the large intestine mutating into abnormal lesions.5
This causes a continuous shedding of mutated
cells that are then excreted into the faeces.5
Since human DNA is stable in faeces, it can be
separated out and analysed for tumour-associated
alterations.5 Many tests have been conducted to
identify the mutations most associated with CRC;
however, there is currently just one DNA stool test
approved by the U.S. Food and Drug Administration
(FDA): the Cologuard® (Exact Sciences, Madison,
Wisconsin, USA). The National Cancer Institute
determined that Cologuard can discern microscopic
blood as well as nine DNA biomarkers that code for
three dierent genes (NDRG4, BMP3, and KRAS).
These genes have been linked to both CRC and
adenomas, which are precancerous but advanced
growths found in the gut. As cells pass through a
patient’s colon and rectum, DNA from these cells
shed and bind together in the stool. A specialised
computer program can successfully analyse these
cells and categorise them into positive or negative
findings. Patients who are given a positive result are
recommended to proceed with a colonoscopy.5
In a substantial study, it was demonstrated that
in patients at average risk of developing CRC
with no cancer-suggestive symptoms at the time,
Cologuard could detect both more adenomas and
more cancerous lesions than the FIT test.5 In other
words, Cologuard was more sensitive than the FIT
test. It should be noted, however, that Cologuard
produced more false-positive results than the FIT
test.5 Cologuard tests for seven DNA mutation
biomarkers (KRAS) and two DNA methylation
biomarkers (NDRG4 and BMP3), as well as one
haemoglobin biomarker and one β-actin marker.6
Cologuard is currently covered under Traditional
Medicare and Medicare Advantage plans for
patients between the ages of 50 and 85 years who
show no signs or symptoms of CRC and are at an
average risk of developing CRC.7 Medicaid coverage
varies by state, and private insurance policies vary
greatly. The out-of-pocket cost for one kit is $649.6
While the future for DNA stool tests is promising,
there is a strong need for further research into
DNA biomarkers beyond the nine currently
used in Cologuard, as well as for research that
validates the sensitivity and specificity of these
biomarkers, particularly in detecting adenomas.
In the current article, the authors investigated
the following population, intervention, control,
and outcomes (PICO) clinical question: Is faecal
DNA testing accurate in screening for CRC?
The following four online databases were used:
PubMed, Scopus, Web of Science, and Cumulative
Index to Nursing and Allied Health Literature.
A review was conducted of the literature spanning
the past 5 years from 2012 to January 2017.
METHODS
Search Strategies and Key Term Definitions
PubMed was the first database searched.
The search began with ‘colorectal cancer screening’
and results were further narrowed by adding the
additional keywords ‘fecal DNA biomarkers’ and
‘noninvasive’. This yielded a total of 17 suitable
articles from the PubMed database. Secondly, those
same search terms were used in Scopus, which
resulted in 11 articles. Thirdly, a search was conducted
in Web of Science using the same key terms, which
yielded 10 articles. Finally, a search search of the
Cumulative Index to Nursing and Allied Health
Literature resulted in no further articles after all
the key terms were entered. This review of the
literature was conducted on 11th February 2017.
The authors maintained the following inclusion
criteria: articles published in English, published
within the past 5 years, articles must pertain to
faecal DNA biomarker testing rather than serum
DNA, and articles must pertain to DNA testing
rather than RNA. The terms ‘faecal DNA biomarkers’
and ‘non-invasive’ can be defined as follows:
a faecal DNA biomarker is a biological DNA
molecule found in faecal matter that is a sign of a
normal or abnormal process, or of a condition or
disease. A biomarker may be used to see how well
the body responds to a treatment for a disease or
condition. Biomarkers are also called molecular
markers and signature molecules.8 Non-invasive
is defined as: “A procedure that does not require
inserting an instrument through the skin or into a
body opening.”9

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
86 87
Organising the Evidence and Assessing
Evidence Quality
To organise these articles, the Preferred Reporting
Items for Systematic Reviews and Meta-Analyses
flow diagram was used (Figure 1). This tool
illustrates the stages of research conducted in a
systematic review.10 At the end of the searches
from all four databases, 38 articles were identified.
Nine articles were removed because they were
duplicates (Figure 1). One article was not accessible
through electronic means and was therefore
excluded. Nineteen articles were removed due
to irrelevance; many articles in this group were
investigating other types of cancers, blood
screening tests, and RNA testing. The nine
articles that remained focussed specifically on the
eectiveness of faecal DNA testing in the screening
of CRC. These nine articles were evaluated using
the Grading of Recommendations Assessment,
Development, and Evaluation method. This is
a system used for rating the quality of research
evidence and considers the study design and
magnitude of eect, along with the risk and
presence of bias, imprecision, inconsistency, and
indirectness. It assigns one of four levels of quality:
high, medium, low, and very low (Table 1).11
RESULTS AND COMMON THEMES
Summary of Studies found
in the Literature Review
Given the vast range of possibilities in the field of
genetics and epigenetics, the studies analysed
each focussed on dierent DNA markers (Table 1).
Many focussed on determining the specificity of a
single DNA mutation in detecting CRC.12-14 Lu et al.12
and Babaei et al.13 both studied the SFRP2.
Carmona et al.15 looked at three genes: AGTR1,
WNT2, and SLIT2, while Baxter et al.16 conducted
a study that isolated the 16S ribosomal RNA gene.
Other studies focussed on epigenetics, specifically
the methylation of promoter genes, which causes
the potential for the hyperexpression or silencing
of genes.17,18 Several hypermethylated genes have
been linked with CRC; Wu et al.17 specifically looked
at microRNA promoters in this context. Lastly, two
articles tested stools for both genetic and epigenetic
mutations, which is how Cologuard screens for
CRC;19,20 Lidgard et al.19 and Imperiale et al.20 both
Figure 1: Preferred reporting items for systematic reviews and meta-analyses.
IdentificationScreeningEligibilityIncluded
Records identified through
database searching
(n=38)
Additional records identified
through other sources
(n=0)
Records after duplicates removed
(n=29)
Full-text articles assessed
for eligibility
(n=28)
Records excluded
(n=1)
• Not accessible via the University of
Pennsylvania’s database (n=1)
Full-text articles excluded
(n=19)
• Irrelevant (n=8)
- Studies not focussed on
colorectal cancer
- Studies on RNA rather
than DNA
• Systematic reviews of research
conducted in 2011–2015 (n=11)
Studies included in
table of evidence
(n=9)
Records screened
(n=29)

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
86 87
used a multi-target stool DNA (sDNA) assay to
measure sDNA markers as well as methylation.
All studies diered regarding sample size, DNA
markers, control groups, and findings. However,
analysing the sensitivity and specificity of dierent
biomarkers is beneficial in drawing conclusions
about the usefulness of these tests relative to
one another.
AGTR1: angiotensin II receptor type 1; FBN1: fibrillin 1; FIT: faecal immunochemical test; GATA5: GATA
binding protein 5; GRADE: Grading of Recommendations Assessment, Development, and Evaluation;
miRNA: microRNA; sDNA: stool DNA; SFRP2: secreted frizzled-related protein 2 gene; SLIT2: Slit homolog
2 protein; WNT2: WNT family member 2.
Author Type of study Method Sample or setting Biomarker
or detection
method
used
Sensitivity and
specificity
GRADE11
Lu et al.,12
2014
Case-control Quantitative N=96
Outpatient
university research
institute
SFRP2,
GATA 5
SFRP2:
Sensitivity: 57.0%
Specificity: 70.0%
GATA 5 :
Sensitivity: 83.9%
Specificity: 82.5%
Very low
Babaei et al.,13
2016
Cross-
sectional
Quantitative N=50
Outpatient
university research
institute
SFRP2 Sensitivity: 60.0%
Specificity: 92.0%
Very low
Guo et al.,14
2013
Case-control Quantitative N=105
Outpatient
university research
institute
FBN1 Sensitivity: 72.0%
Specificity: 93.3%
Low
Carmona et al.,15
2013
Case-control Quantitative N=151
Outpatient
hospital oce,
public data sets
AGTR1,
WNT2,
SLIT2
Specificity:
AGTR1: 21.0%
WNT2: 40.0%
SLIT2: 52.0%
Low
Baxter et al.,16
2016
Cross-
sectional
Quantitative N=404
Outpatient
medical oces (4)
Detection of
microbiota
in FIT test
-
Low
Wu et al.,17
2014
Case-control Quantitative N=122
Outpatient
university research
institute
miR-34a and
miR-34b/c
miRNA
methylation
miR-34a:
Sensitivity: 76.8%
Specificity: 93.6%
miR-34b/c:
Sensitivity: 95.0%
Specificity: 100.0%
Low
Ghanbari et al.,18
2016
Observational Quantitative N=77
Outpatient
university research
institute
Let-7a-5p
and
Let-7f-5p
miRNA
“Significant to
discriminate
between CRC
subjects and
healthy subjects.”
Very low
Lidgard et al.,19
2013
Case-control Quantitative N=1,003
Outpatient
medical oces
(21)
Multi-target
sDNA assay
Sensitivity: 98.0%
Specificity: 90.0%
Low
Imperiale et al.,20
2014
Cross-
sectional
Quantitative N=9,989
Outpatient
university research
institute, medical
oces (90)
Multi-target
sDNA
(Fit-DNA /
Cologuard)
Sensitivity: 92.3%
Specificity: 86.6%
Moderate
Table 1: Grading of recommendations assessment, development, and evaluation system applied to
articles pertaining to faecal DNA testing in colorectal cancer screening.

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
88 89
Sensitivity and Specificity of Biomarkers
Of the nine studies found through the literature
search, seven studies researched the sensitivity
and specificity of stool biomarkers. Sensitivity is
the extent to which a test can correctly identify a
positive result (true-positive) and specificity is
the extent to which a test can correctly identify a
negative result (true-negative). A test with low
sensitivity would produce more false-negatives
and a test with low specificity would produce more
false-positives. The sensitivities of the tests in the
studies identified ranged from 57–98% (Table 1).
Lidgard et al.19 found the highest sensitivity (98%)
using a multi-target sDNA assay to measure
β-actin (a marker of total human DNA), mutant
KRAS, aberrantly methylated BMP3 and NDRG4,
and faecal haemoglobin. It is possible that the
high sensitivity was due to their test being an
assay of multiple biomarkers for CRC. The study
with the lowest sensitivity was Lu et al.,12 which
tested the biomarker SFRP2 (57%). Babaei et al.13
also studied SFRP2 in stool samples and found
a similar specificity of 60%. For specificity,
the tests ranged from 21–100% (Table 1). Wu et al.17
studied the methylation status of miR-34a and
miR-34b/c promoter in CRC patients’ stool samples
and found the specificity of miR-34b/c was 100%.
Carmona et al.15 studied three genes in relation to
detecting CRC in patients’ stool: AGTR1, which had
a specificity of 21%; WNT2, which had a specificity
of 40%; and SLIT2, with a specificity of 52%.
CONCLUSION
This research is extremely relevant to many
healthcare providers since it pertains to common
screening tools. While providers currently just
have one FDA-approved option (Cologuard),
genetic detection of cancer risk continues to be an
area of research and development. The Cologuard
test is indicated in the screening of all adults ≥50
years old who are at an average risk for developing
CRC. It is contraindicated in patients with a history
of any form of cancer, if they had a previous
positive result from any CRC screening method, have
a family history of CRC, or have an inflammatory
bowel disease.6 Although this is an approved
screening method that oers convenience and
no direct risks to the colon, not all providers
have embraced this new recommendation and
the CDC continues to advise adults to have
colonoscopies.2 Hopefully, with continued research
and improvements in the testing methods, this
screening option will become more available and
accurate in detecting CRC in adults.
Need for Further Research
There is still a great need for research in this area.
To date, there have been no randomised controlled
trials that compare dierent screening tests,
such as colonoscopy and DNA stool-based tests,
to one another; only studies that test the reliability,
sensitivity, and specificity of specific screening
tests have been performed. Because of this, it is
dicult to know which tests are the most eective.
Analysing long-term outcomes, including morbidity
and mortality, is necessary to assess the role of
multi-target sDNA testing. Another area for future
research is on issues related to testing intervals,
patient acceptance, compliance of screening
guidelines, and barriers that prevent individuals
from being screened for CRC. It is unclear if patient
preference for non-invasive screening tools along
with more eective methods of non-invasive
screening will one day diminish the importance of
the use of colonoscopies in the average-risk adult.
Additionally, mRNA and microRNA are areas of
current research that oer potential for highly
specific and sensitive CRC screening. While more
research is needed in several areas, the role of
multi-target sDNA testing is continuously
advancing and oers an aordable, convenient,
and safe option for adults who are at an average
risk for developing CRC.
1. Centers for Disease Control and
Prevention. Colorectal Cancer Statistics.
2017. Available at: https://www.cdc.gov/
cancer/colorectal/statistics/index.htm.
Last accessed: 29 November 2017.
2. Centers for Disease Control and
Prevention. Colorectal cancer screening
tests. 2017. Available at: https://www.
cdc.gov/cancer/colorectal/basic_info/
screening/guidelines.htm. Last accessed:
15 November 2017.
3. Centers for Disease Control and
Prevention. Cancers diagnosed at late-
stages despite available screening tests.
2010. Available at: https://www.cdc.gov/
media/pressrel/2010/r101124.html. Last
accessed: 15 November 2017.
4. US Preventive Services Task Force;
Bibbins-Domingo K et al. Screening for
colorectal cancer: US preventive services
task force recommendation statement.
JAMA. 2016;315(23)2564-75.
5. National Cancer Institute. Tests to
detect colorectal cancer and polyps.
2016. Available at: https://www.cancer.
gov/types/colorectal/screening-fact-
sheet. Last accessed: 15 November 2017.
REFERENCES

EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL EUROPEAN MEDICAL JOURNAL • March 2018 • Creative Commons Attribution-Non Commercial 4.0 EMJ EUROPEAN MEDICAL JOURNAL
88 89
6. Exact Science. The science behind
the test: Stool DNA technology. 2017.
Available at: http://www.cologuardtest.
com/hcp/about-cologuard/the-
science-behind-the-test. Last accessed:
15 November 2017.
7. Medicare. Your Medicare coverage.
2017. Available at: https://www.medicare.
gov/coverage/colorectal-cancer-
screenings.html. Last accessed:
15 November 2017.
8. National Cancer Institute. National
Cancer Institute dictionary of cancer
terms. Biomarker. Available at: https://
www.cancer.gov/publications/
dictionaries/cancer-terms?cdrid=45618.
Last accessed: 29 November 2017.
9. National Cancer Institute. National
Cancer Institute dictionary of cancer
terms. Noninvasive. Available at:
https://www.cancer.gov/publications/
dictionaries/cancer-terms?cdrid=455364.
Last accessed: 29 November 2017.
10. Moher D et al. Preferred reporting
items for systematic reviews and
meta-analyses: The PRISMA statement.
PLOS Med. 2009;62(10):1006-12.
11. Balshem H et al. GRADE guidelines:
3. Rating the quality of evidence. J Clin
Epidemiol. 2011;64(4):401-6.
12. Lu H et al. DNA methylation analysis
of SFRP2, GATA4/5, NDRG4 and VIM
for the detection of colorectal cancer in
fecal DNA. Oncol Lett. 2014;8(4):1751-6.
13. Babaei H et al. DNA methylation
analysis of secreted frizzled-related
protein 2 gene for the early detection of
colorectal cancer in fecal DNA. Niger Med
J. 2016;57(4):242-5.
14. Guo Q et al. Detection of
hypermethylated fibrillin-1 in the stool
samples of colorectal cancer patients.
Med Oncol. 2013;30(4):695.
15. Carmona FJ et al. DNA methylation
biomarkers for noninvasive diagnosis of
colorectal cancer. Cancer Prev Res (Phila).
2013;6(7):656-65.
16. Baxter NT et al. DNA from fecal
immunochemical test can replace stool
for detection of colonic lesions using a
microbiota-based model. Microbiome.
2016;4(1):59.
17. Wu XD et al. Detection of miR-34a and
miR-34b/c in stool sample as potential
screening biomarkers for non-invasive
diagnosis of colorectal cancer. Med
Oncol. 2014;31(4):894.
18. Ghanbari R et al. Simultaneous
underexpression of let-7a-5p and
let-7f-5p microRNAs in plasma and stool
samples from early stage colorectal
carcinoma. Biomark Cancer. 2016;
7(Suppl 1):39-48.
19. Lidgard GP et al. Clinical performance
of an automated stool DNA assay for
detection of colorectal neoplasia. Clin
Gastroenterol Hepatol. 2013;11(10):1313-8.
20. Imperiale TF et al. Multitarget stool
DNA testing for colorectal-cancer
screening. N Engl J Med. 2014;371(2):
187-8.
If you would like reprints of any article, contact: +44 (0) 1245 334450.