|
By Clark Rundell
PhD, DABCC, FACB Maine Molecular Quality Controls, Inc.
03/17/08
Recently, there have been discussions in the popular and scientific literature
about the effect of putative Her-2 testing problems.1,2 Her-2 was chosen for
this discussion because of the current interest, however, there are other equally
interesting reports where medical laboratory results did not always meet expectations
A contributing factor in the real or perceived performance problems is in the definition and evolution of the goals of our laboratory tests, medical treatments, and even our institutions. For the Her-2 test, there is a specific therapy that is indicated if expansion of the Her-2 gene is present in a tumor. To the patient the implications were serious and indicated definitive action. A positive Her-2 test meant the likelihood of an aggressive cancer and one that might be effectively treated with Herceptin®. A laboratory testing scheme was designed to detect specific expansions in the Her-2 gene and, since it was important to detect the lesion even if present in small amounts, the tests were optimized for sensitivity. The test was also investigated for specificity to ensure that it would not give a high percentage of false positive values for patients that did not have the Her-2 expansion. Note that a laboratory would have difficulty determining the exact specificity and sensitivity for a new test due to a lack of an accurate "gold standard" test and a lack of suitable reference materials. In any case, there were many publications stating the value of the Her-2 test in triaging patients for beneficial outcomes and laboratories were encouraged to set up the test.
What happened?
First, laboratories instituted the test using state of the art technologies. Second, proficiency surveys with Her-2 positive and negative samples later became available and laboratories validated their work by passing the proficiency tests with an 80 person score. Third, better technologies became available, comparison testing was done, and results from long term monitoring of patient outcomes were published. Fourth, the results of number three suggested that for some laboratories, as many as 20 percent of samples that were formerly thought to be positive, are negative for Her-2 expansion by today's best test schemes. It is uncertain whether therapy based on the earlier Her-2 test schemes was always appropriate. Although the current situation is problematic, consider that this is after all a natural progression for any new test because the best algorithms will not be developed until extensive testing and data collection has been done.
Regarding improved efficiency in achieving a goal of satisfactory testing, the following questions occur. First, was there ever, and is there now, a defined allowable error within which the test would still be useful for medical decision making? And, second, is today's understanding of the allowable error different from that understood when the testing was started. For example, in 2007 the American Society of Clinical Oncology and College of American Pathologists put forward the following goal, "It is recommended that to perform Her-2 testing, laboratories show 95 percent concordance with another validated test for positive and negative assay values."1 Some tests were no doubt instituted without sufficient concern for their performance. However, in other cases, tests that were once satisfactory as far as could be determined, are now judged to be unsatisfactory in the light of new information. It may also be that the tests have performed well and the problem lies in interpretation of the results.
The conclusion, Dr. Paik of the National Surgical Adjuvant Breast and Bowel Project is quoted as saying "To me, the take-home message is that we don't have a perfect test, unfortunately."3 No test is perfect. However, to attain and maintain satisfactory performance of laboratory tests, particularly new tests such as in the molecular field, it behooves laboratories to determine, very early, the medical goals to be achieved (total allowable false positives and false negatives) and ensure that the test results are sufficiently accurate to support those goals.4 Test maintenance requires frequent reassessment of test performance visa vie the medical goals as new resources, new technologies, and new data become available.
- Wolff, Antonio C.et. al. American Society of Clinical Oncology/College of American Pathologists Guideline Recommendations for Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer. J Clin Oncol. 2007 Jan 1; 25(1):118-145.
- Rabiya S. Tuma. Inconsistency of HER2 Test Raises Questions. J Natl Cancer Inst. 2007;99(14):1064-1065
- Pollack, A. Cancer Drug May Elude Many Women Who Need It. New York Times, June 12, 2007. Downloaded 1-30-2008 from http://www.nytimes.com/2007/06/12/health/12canc.html.
- Westgard JO. Assuring the Right Quality Right. Madison WI: Westgard QC, Inc.,2007.
|
