Earlier this year the FDA’s
Molecular and Clinical Genetics Panel of the Medical Devices Advisory Committee
determined that Exact Science Corp’s new stool based DNA test Cologaurd, a non-invasive
colorectal cancer screening test, had demonstrated effectiveness, safety and
low risk. Why is this important? Cologuard screens for colorectal cancer, a
cancer that continues to be a major cause of death to men and women in the United
States. It kills up to 600,000 people per year. As the current screening
strategies for colorectal cancer have been found to be costly scientists
working for Exact Science Corp. have been researching cost-effective tests with
greater sensitivity. This research led
to a simple, non-invasive test with high sensitivity for both colorectal cancer
and advanced precancerous lesions. The clinical trial for Cologuard was quite
successful compared to the standard fecal immunochemical test (FIT), which is
the current standard colorectal screening test.
A total of 12,776 applicants
participated in this trial and 78.2% of them had results that could be
evaluated. Out of the 78.2% of the results that could be evaluated, 65 patients
were found to have colorectal cancer on colonoscopy and757 participants had
advanced precancerous lesions. Precancerous lesions are found in the inner
lining of the colon and can grow into the lumen. They are dangerous because
cancer can arise from some of these tissues. The stool DNA testing, used in Cologuard,
was able to identify 60 out of the 65 participants with cancer, a sensitivity
of 92.3%. Among the 757 participants with advanced precancerous lesions, the
DNA testing correctly identified 321 of these participants with a sensitivity
of 42.4%. In terms of specificity among the 9167 participants who had no traces
of colorectal cancer or advanced precancerous lesions the specificity was
86.6%. The fecal immunochemical test(FIT) the compared test to Cologuard. FIT’s
detection rate was 48 out of 65 colorectal cancers, a sensitivity of 73.8%, and
180 out of 757 advanced precancerous lesions, a sensitivity of 23.8%. These
findings from the FIT test were all inferior compared to the results from the
DNA test. Among the 9167 participants with findings other than colorectal
cancer or advanced precancerous lesions, the specificity of the FIT was 94.9%.
This was the only instance where the FIT was superior to the DNA test. The
colonoscopy is slightly more efficient than the multi-target DNA test, but the
colonoscopy is very invasive while the multi-target DNA test is non-invasive
and more cost-effective.
Previous stool
based DNA screening tests, such as the FIT, only had medium sensitivity for the
determination of colorectal cancer, and very low sensitivity for the detection
of advanced adenomas. The new advancements that lead to higher sensitivity
include stabilizing buffers, and more discriminating markers. The table above
supports this conclusion as we can see that the multi-target DNA test is much
more efficient at detecting the disease. If we look at the data for Stage 1 to
stage 3 colorectal cancer we see an increase of 20 percent in sensitivity in
the multi-target DNA test. Stabilizing buffers are used to stop degradations of
human long DNA. Human long DNA in stool is often used as a colorectal cancer
marker. While the samples of stool are in storage waiting to be screened for if
a stabilizing buffer such as EDTA is missing, the essential DNA degrades. This
was found to be caused by bacterial DNAases, which would degrade human DNA
during fecal storage. The EDTA buffer preserves the DNA integrity during
storage by acting as a DNAase inhibitor. The graph below shows the difference in the amount of human long
DNA left in stool samples ,which have been in storage for 8 days. The two
graphs differ in whether there is any stabilizing buffer in the sample. It is
quite evident that in the sample that does contain EDTA the DNA was viable even
after 8days, whereas in the sample with no EDTA the amount of DNA began to
decrease rapidly from day 0 and by day 8 there is very little DNA left. The
more DNA is available the more efficient the DNA test will be for screening for
disease. In terms of more discriminating markers in these new DNA tests
only two markers are utilizes; these being hypermethylated vimentin gene, and a
two site DNA integrity assay. Improved stool DNA tests while incorporating only
two markers which have high sensitivity, lead to reduced costs, ease of use,
and highly reproducible results.
Figure 5.
Long DNA levels of CRC stools stored at room temperature for 1,
3, and 8 days after defecation fell median percentages of 75%, 81%, and 89%,
respectively. B.Buffers with 100 mmol/L EDTA and
16 mmol/L EDTA showed different effects on preserving human DNA added in stools
stored at room temperature.
Two points that can explain the DNA
testing’s low specificity. Firstly, the analysis of the measure of specificity,
86.6%, includes the participants with non-advanced adenomas. This can cause for
the test to show positive when the colonoscopy turns up a true negative result.
For example, among those with only negative results on colonoscopy, the
specificity of the DNA test was almost 90%. Secondly, specificity and age are
inversely related. Among participants between the ages of 50 to 64 with a
negative result on colonoscopy, the specificity of the DNA test was 94%. The
relationship between specificity and age can be due to the presence of lesions
that were missed on colonoscopy or age-related DNA methylation.
If we apply this topic to what we
just learned we can see that it is probably a phase 3 clinical trial because it
is testing thousands of people, and also the test is being compared to the
standard test, FIT. The main issue I have with the test as well as other
colorectal cancer tests is that they have very low sensitivity for detecting
non-advanced adenomas. Non-advanced adenomas are dangerous due to the fact that
they are quite prevalent in the colon. This is dangerous becomes these adenomas
often become malignant. In terms of the multi-target DNA test the sensitivity
for non-advanced adenomas was 17.2% while the sensitivity using the FIT was
7.6%. However, in terms of specificity for adenomas using the DNA test the value
was 86.6%, and for the FIT we get a value of 94.9%.
This paper is an excellent piece of
research in showing the importance and use of sensitivity and specificity. When
we first learned about screening tests for cancers, I understood how to calculate
for sensitivity and specificity but I did not realize the significance of these
two pieces of data. In the case of the stool DNA test we see that it has a high
sensitivity and a relatively low specificity. This means that if one takes the
test and receives a negative result they can conclude that it is a true
negative because the test is so good at detecting cancer. For this reason a high
specificity test is good at ruling out a disease, because it rarely
misdiagnoses a disease We should be
excited by these results because this is a non-invasive test with a high
sensitivity that screens for one of the most preventable yet least prevented
cancers. The main reason why colorectal cancer does not get screened for is
because many of the screening tests are quite invasive such as a colonoscopy.
With the help of tools that give us a higher sensitivity such as stabilizing
buffers, Cologuard has the potential to be the most effective and easiest
screens for colorectal cancer.
Sources:
Ahlquist DA, Zou H, Domanico M, et al. Next-generation stool DNA test
accurately detects colorectal cancer and large adenomas. Gastroenterology 2012;142:248-256
Ahlquist DA, Taylor WR, Yab TC, et al. Aberrantly methylated gene marker
levels in stool: effects of demographic, exposure, body mass, and other patient
characteristics. J Mol Biomark Diagn 2012;3:e1000133-e1000133
Itzkowitz S, Brand R, Jandorf L, et al. A simplified, noninvasive stool DNA
test for colorectal cancer detection. Am J Gastroenterol 2008;103:2862-2870
Zou H, Harrington
JJ, Klatt KK, Ahlquist DA. A sensitive method to quantify human long
DNA in stool: relevance to colorectal cancer screening. Cancer Epidemiol Biomarkers Prev2006;15:1115-1119
