NEW STOOL TEST BETTER THAN HEMOCCULT

DNA Stool Tests Show Promise for Colorectal Cancer Screening

October 9, 2008 — Stool DNA testing is a new approach to screening for colorectal cancer, but it is an evolving technology. A study has found that the first-generation stool DNA test (SDT-1) is not more effective than the widely used fecal occult blood tests Hemoccult and HemoccultSensa (Beckman Coulter) for detecting screen-relevant colorectal neoplasms. In fact, it performed more poorly than HemoccultSensa in detecting cases of cancer, according to the results that appear in the October 7 issue of the Annals of Internal Medicine.

However, a second-generation stool DNA stool test (SDT-2), which uses a broader marker panel, was substantially more sensitive than fecal occult blood testing, and the higher sensitivity was particularly evident for adenomas. The sensitivities of SDT-1 and HemoccultSensa were very similar for screen-relevant neoplasms (20% vs 21%, respectively), whereas the sensitivity of SDT-2 was 40%.

"The results with SDT-2 are promising, but more research is needed," said lead author David A. Ahlquist, MD, professor of medicine at the Mayo Clinic, in Rochester, Minnesota. "We identified several correctable technical problems with DNA tests in our study, including degradation of DNA during stool mailing, inefficient recovery of DNA from stool, suboptimal analytical sensitivity of instruments used to detect DNA markers, and incomplete marker informativeness for cancer and polyps."

Improvements Introduced in Testing

Dr. Ahlquist explained that since the completion of this study, his team and other groups have introduced improvements that represent a "fix" for each of these technical problems. "The results of such improvements have been published in both manuscript and abstract form," he said. "Detection rates in these subsequent smaller studies have ranged from 80% to 100% for cancer and from 58% to 78% for advanced adenomas, and specificities with different marker panels have ranged from 82% to 96%."

The availability of a simple noninvasive test capable of detecting colorectal cancer and precancerous lesions with a reasonable sensitivity and specificity might help patients avoid the invasiveness, the unpleasant bowel preparation, and the potential complications associated with colonoscopy. But although stool DNA analysis might be more effective in detecting colorectal neoplasia than fecal occult blood testing, the tests are still not comparable in efficacy to colonoscopy, as previously reported by Medscape Oncology.

"Stool DNA testing is certain to get better but will likely never match the exquisite point-sensitivity of colonoscopy," Dr. Ahlquist told Medscape Oncology. "However, many patients refuse invasive and costly procedures like colonoscopy. If compliance to a noninvasive stool DNA test is higher than to colonoscopy, and if stool testing is performed more frequently than colonoscopy, then programmatic sensitivity with an accurate stool DNA test could exceed that of colonoscopy."

DNA Tests Compared With Occult Blood Testing

Dr. Ahlquist and colleagues compared stool DNA and fecal blood testing for the detection of screen-relevant neoplasia in 3764 healthy adults between the ages of 50 and 80 years who had undergone screening colonoscopy. SDT-1 is a precommercial 23-marker assay, and SDT-2 targets 3 broadly informative markers.

Colonoscopy screening revealed screen-relevant neoplasms in 290 (7.7%) patients. Of this group, 39 had nonmetastatic cancer or high-grade dysplasia and 251 had adenomas that were 1 cm or larger.

The researchers found that the precommercial SDT-1 test did not demonstrate any real improvement over the widely used fecal occult blood tests. Because of these poor results, they decided to also evaluate SDT-2, which potentially targets a more informative marker panel. Although SDT-2 was examined in a smaller subgroup, it was significantly better at detecting neoplasms than either the fecal occult blood tests or SDT-1.

Summary of Test Performance

Index Test	Screen Relevant Neoplasia (n)	Positive Test Result (n)	Sensitivity (96% CI)	Specificity (96% CI)
Hemoccult (n = 2497)	157	17	11	98
HemoccultSensa (n = 2497	157	33	21	97
SDT-1 (n = 2497)	157	31	20	96
SDT-2 (n = 217)	142	66	40	—

For Hemoccult, HemoccultSensa, and SDT-1 test numbers are based on the first 2497 evaluable patients.

The SDT-2 detected 46% of screen-relevant neoplasms, compared with Hemoccult, which detected 16%, and HemoccultSensa, which detected 24%. SDT-2 also detected 46% of adenomas 1 cm or larger, compared with Hemoccult, which detected 10%, and HemoccultSensa, which detected 17%. With SDT-2, test positivity was significantly higher in individuals with normal colonoscopy results (13%) than it was with Hemoccult (4%) or HemoccultSensa (5%).

"This is an important study because it highlights the promise of stool DNA technology for colorectal screening," commented Steven Itzkowitz, MD, FACP, professor of medicine and associate director of the division of gastroenterology at Mount Sinai School of Medicine, in New York, New York. Dr. Itzkowitz was not involved in the study.

"Unfortunately, like many studies that take several years to complete, the technology you start off with undergoes changes, so the final results often have to be taken with a grain of salt," he told Medscape Oncology. "Thus, while this multicenter study began with the original prototype version of the stool DNA test [SDT-1] . . ., this rather cumbersome assay that had to analyze 23 different markers was later replaced with newer technology [SDT-2] that only needed to analyze for k-ras, APC mutation cluster region, and methylated vimentin markers."

This resulted in the SDT-2 having significantly greater sensitivity for detecting screen-relevant neoplasia than the guaiac-based tests, including HemoccultSensa. "This occurred despite the fact that the guaiac-based tests were interpreted by a centralized expert laboratory, and not the usual . . . local labs or doctors offices," he said.

Main Points of Study

Dr. Itzkowitz pointed out that there are 3 things worth emphasizing about this study:

• The stools were collected without using a DNA-stabilizing buffer and were shipped on ice. Studies performed since the onset of this study have shown that stool DNA degrades rapidly during transit to the lab, despite the use of ice packs and express shipping, resulting in a significant underperformance of some of the stool markers. When a stabilizing buffer is added to the stool by patients, the DNA is preserved for several days, even at room temperature.
• SDT-2 detected considerably more adenomas than either of the guaiac-based fecal occult blood tests, and did so regardless of location in the colon. This is an important advance over guaiac-based tests because people with large adenomas and those that have high-grade dysplasia are important surrogate end points in colorectal-cancer-prevention studies.
• The marker panel in the SDT-2 assay was used only in this study and is not commercially available. However, there is now a newer commercially available stool DNA test (ColoSure, LabCorp) that tests for hypermethylated vimentin (1 of the markers in the SDT-2 panel) using buffer-stabilized stool DNA. Recent studies of hypermethylated vimentin plus DNA integrity assay have demonstrated high sensitivities for colon cancer, in the 83% to 88% range.

The SDT-1 assay is no longer offered commercially, explained Dr. Ahlquist and, assuming that next-generation commercial stool DNA tests show the same or better performance, as was noted for SDT-2, there would be a number of benefits to using stool DNA testing. These include ease of use, no diet or medication restriction, and superior detection of premalignant polyps, all of which would lead to more effective colorectal cancer prevention. "In the future, stool DNA tests may be able to screen for cancers throughout the [gastrointestinal] tract," he said.

Although Dr. Itzkowitz agreed that colonoscopy is the preferred screening test for all people older than 50 years and does not view any of the noninvasive stool-based tests as a replacement for colonoscopy, the DNA tests and fecal immunochemical tests appear to be more effective than the guaiac-based tests, even the more sensitive ones.

The cost of the stool DNA tests was an initial barrier to their use, but Dr. Itzkowitz explained that that is changing. "The cost of the original stool DNA test was rather high, approximately $800, because it had to analyze more than 20 markers. However, the latest version of the test that includes only methylated vimentin is in the $200 range."

Although still more expensive than fecal occult blood tests, it is becoming a very competitive option. "Cost-effective analyses indicate that if a stool DNA test had a sensitivity of 65% for colorectal cancer and 40% for large adenomas, 95% specificity, a test cost of $195, and a screening interval of 2 years, it would be comparable to the cost-effectiveness of colonoscopy," Dr. Itzkowitz said. "We are quite close to this goal."

The study was funded by the National Institutes of Health. Several of the authors have served as consultants or own stock in EXACT Sciences, the company that performed the genetic assays on tissue and stool for this study. The Mayo Clinic is a minor equity investor in EXACT Sciences.

Ann Intern Med. 2008;149:441-450.