In a previous post, I exhorted people to heed the American Cancer Society recommendation to have a colonoscopy when you turn 50. Indeed an increase in colonoscopy screening has led to a dramatic decline (>30% over last decade) in deaths caused by colorectal cancer (CRC). The discomfort of colonoscopies has led to avoidance; a new stool diagnostic (fecal immunochemical test or FIT) that tests for blood in stool is effective (~70% sensitivity) but no substitute for a colonoscopy, which is the gold standard.
Well there is new technology in town (stool DNA test) and the results (reported in the New England Journal of Medicine last week) are impressive. Like FIT, it is a noninvasive stool test, but instead of blood, it measures changes in the DNA of cells sloughed off in your stool. Colon cancer has been well-characterized, and it is known that certain genes are deleted, while others are amplified in the tumor genome. In addition, some genes may be modified (e.g. hypermethylated) differently from normal intestinal cells. PCR and DNA sequencing can detect these changes in the DNA from a small sample of cells.
In the study, the patient pool consisted of 9989 asymptomatic patients between the ages of 50 and 84 years. All participants submitted a stool specimen and then underwent screening colonoscopy. Both DNA and FIT tests were performed on the stool sample. The verdict of the colonoscopy was treated as the correct diagnosis (ground Truth).
The sensitivity of the DNA test was quite impressive for patients who had CRC; stool DNA testing identified 60 of 65 participants with cancer (92.3%). This result exceeded the FIT test (73.8%) by a significant margin. 0.7% of all the patients had colorectal cancer.
The sensitivity was less impressive for patients with advanced precancerous lesions (but not CRC yet, 7.6% of patients); DNA testing identified 42.4% of these individuals. However this percentage far exceeded the sensitivity of FIT (only 23.8%) on advanced precancerous lesions.
On the other hand, the specificity of the DNA test (86.6%) was worse than that of FIT (94.9%). In other words, among the participants who did not have either colorectal cancer or advanced precancerous lesions (~92% of patients), the DNA test incorrectly predicted 13.4% as being positive (false positive rate). It should be noted that a false negative (i.e patient with CRC who tests negative) is a far worse outcome than a false positive (i.e. patient without CRC who tests positive), and so the thresholds of the test should be set to maximize the number of true positives (CRC individuals who test positive).
Thus the authors concluded, "a stool test combining altered human DNA and fecal hemoglobin showed higher single-application sensitivity than a commercial FIT for both colorectal cancer and advanced precancerous lesions, although with lower specificity."
Several final comments. First, the stool test was done only once (single-application); it is possible that performing repeated tests (say once a year) would increase the true positive rate beyond 92.3% (while also increasing the false positive rate). Second, the low sensitivity detecting advanced precancerous lesions underscore the importance of colonoscopy; these advanced precancerous lesions need to be detected as soon as possible and then removed. Third, the stool DNA test is not meant to replace colonoscopies but to complement them. In particular for those who absolutely refuse a colonoscopy, an increasingly accurate stool test could be a reasonable alternative in the future:
"Offering a choice among tests may improve uptake of screening.32,36 A noninvasive test with a high single-application sensitivity for curable-stage cancer may provide an option for persons who prefer noninvasive testing."
Remember the colonosocopy is the gold standard. Even 92.3% sensitivity means that 5 out of 65 people were missed.
No comments:
Post a Comment