Forgot your password?
typodupeerror
Medicine

UK Team Claims Breakthrough In Universal Cancer Test 63

Posted by Unknown Lamer
from the coming-to-a-patent-office-near-you dept.
An anonymous reader writes UK researchers say they've devised a simple blood test that can be used to diagnose whether people have cancer or not. The Lymphocyte Genome Sensitivity (LGS) test looks at white blood cells and measures the damage caused to their DNA when subjected to different intensities of ultraviolet light (UVA), which is known to damage DNA. The results of the empirical study show a distinction between the damage to the white blood cells from patients with cancer, with pre-cancerous conditions and from healthy patients. "Whilst the numbers of people we tested are, in epidemiological terms, quite small (208), in molecular epidemiological terms, the results are powerful," said the team's lead researcher. "We've identified significant differences between the healthy volunteers, suspected cancer patients and confirmed cancer patients of mixed ages at a statistically significant level .... This means that the possibility of these results happening by chance is 1 in 1000." The research is published online in the FASEB Journal, the U.S. Journal of the Federation of American Societies for Experimental Biology.
This discussion has been archived. No new comments can be posted.

UK Team Claims Breakthrough In Universal Cancer Test

Comments Filter:
  • Link to abstract (Score:5, Insightful)

    by Michael Woodhams (112247) on Monday July 28, 2014 @11:46PM (#47555199) Journal

    Here is the abstract [fasebj.org]. The actual paper is behind a paywall.

    "ROC analysis of [the test statistic], for cancers plus precancerous/suspect conditions vs. controls, cancer vs. precancerous/suspect conditions plus controls, and cancer vs. controls, gave areas under the curve of 0.87, 0.89, and 0.93, respectively (P<0.001). Optimization allowed test sensitivity or specificity to approach 100% with acceptable complementary measures."

    The ROC curve [wikipedia.org] has area under it of 1 for a perfect classifier and 0.5 for wild guessing. This is a more useful measurement than the p-value. (E.g. if I look at height vs sex for humans, it won't take too big a sample to get a great p-value for there being a difference, yet classifying people as male/female depending on whether they exceed some height threshold is a very poor diagnostic system.) I don't have much of a feel for how good ROC area of about 0.9 is for a medical test. I'd guess it is good enough to be useful, but you'd not want to rely on that test alone.

  • Re: So... (Score:3, Insightful)

    by pla (258480) on Tuesday July 29, 2014 @06:34AM (#47556089) Journal
    Don't do it! Everyone will be diagnosed.

    Bizarre trolling aside, You have the right idea - Virtually everyone over the age of 50 has dozens of "cancerous" cell clusters scattered around their bodies, all more-or-less harmless unless juuust the right combination of environmental conditions triggers a few to start growing (and spreading) uncontrollably.

    I find it easy to believe that a universal test for "cancer" would have a near-perfect success rate, because nearly everyone has it, to some degree. I find the negative side much harder to believe, because it means differentiating between cancer-but-harmless and cancer-gonna-kill-you.

    Or looked at another way, consider recent changes in attitude regarding breast and prostate cancers. 20 years ago, detecting either meant immediately scheduling a radical mastectomy/prostatectomy. Today, unless you have a family history of aggressive cancers, your oncologist will likely suggest watching and waiting for at least a few months to see if it actually does anything more that sit there harmlessly. Yet, even if it does - still cancer. Much like we don't universally vaccinate people against TB because it makes TB antibody tests diagnostically useless, I see this test as having the same issue, accurate but useless.

If the facts don't fit the theory, change the facts. -- Albert Einstein

Working...