Please Read if You Are Considering Genetic Testing

Options
Bestbird
Bestbird Member Posts: 2,818
edited February 2017 in Stage IV/Metastatic Breast Cancer ONLY

Two of the most widely used genetic sequencing tests — FoundationOne and Guardant360 — showed "widely discordant" results in the same patients, according to a small online study published in JAMA Oncology.

The FoundationOne test sequences clinical tumor samples to characterize the exons of 315 cancer-associated genes and introns from 28 genes involved in rearrangements. The Guardant360 test uses cell-free circulating DNA from blood to sequence 70 genes.

The study evaluated nine patientswith breast cancer (n = 5), pancreatic cancer, thymic carcinoma, lung cancer and salivary gland cancer (n = 1 for each).

Results:

One patient had no identified genetic alteration using either sequencing test. The remaining eight patients harbored 45 alterations, only 10 (22%) of which were concordant (in agreement) between FoundationOne and Guardant360 platforms.

In two (25%) of the other eight patients, there was absolutely no concordance among the described alterations. And although total of 36 drugs were recommended for the eight patients, only nine (24%) of these drugs were recommended for the same patients by both platforms. Notably, in five patients (more than 50%), there was no overlap whatsoever between the drugs recommended by the FoundationOne test and those recommended by the Guardant360 test.

The researchers concluded that the output from genetic testing can differ markedly depending on which genetic test is applied. Furthermore, they stated that these findings are clinically relevant because both the FoundationOne and the Guardant360 tests are performed on thousands of cancer patients each year.

From: http://www.healio.com/hematology-oncology/breast-cancer/news/in-the-journals/{7ad8fdf6-0661-4d67-a0b2-773251903520}/genetic-sequencing-tests-show-widely-discordant-results

· Share on FacebookShare on Twitter

Comments

  • MTwoman
    MTwoman Member Posts: 2,704
    edited February 2017

    Hi Bestbird,

    While that information is certainly interesting and concerning, n=9 is not nearly anywhere close to a number that is generalizable (with only 1 person from 4 different types of cancer and 5 from breast cancer - notoriously heterogeneous). It isn't powered well enough to draw any conclusions whatsoever, only speculations. Hopefully, the recommendations from this study was that a much larger study is warranted.

    · Share on FacebookShare on Twitter
  • Bestbird
    Bestbird Member Posts: 2,818
    edited February 2017

    I agree that the study is small and that wider testing should be recommended. That said, the fact that there was such a degree of disparity in a small population might lead one to hypothesize that lack of concordance would be even more pronounced . Furthermore, one might wonder whether the fact that breast cancer is considerably heterogeneous might lead to even more diverse results. Again, it'll be interesting to hear what comes of this relative to further studies, and I'd also like to see study results among two different companies that use the same genetic testing technique.

    Thank you for sharing your thoughts!

    · Share on FacebookShare on Twitter
  • MTwoman
    MTwoman Member Posts: 2,704
    edited February 2017

    What would also be interesting is what is the overlap in the genes that are sequenced given "FoundationOne test sequences clinical tumor samples to characterize the exons of 315 cancer-associated genes and introns from 28 genes involved in rearrangements. The Guardant360 test uses cell-free circulating DNA from blood to sequence 70 genes". That may account for some of the vastly different reported findings. I wonder what would be the concordance if you limit the results to only those genes sequenced by both tests. (I didn't have a chance to read the whole study yet, so maybe they did report this, sorry!)

    · Share on FacebookShare on Twitter
  • Bestbird
    Bestbird Member Posts: 2,818
    edited February 2017

    You raise an excellent point! Yet since cancer patients and their medical teams may be making medical decisions as the result of these tests, the fact that there is discord (for potentially a myriad of reasons) remains important. It'll be interesting to read what other researchers may have to say.

    · Share on FacebookShare on Twitter
  • MTwoman
    MTwoman Member Posts: 2,704
    edited February 2017

    I agree that it is concerning that these tests are used for important decision making, given that there doesn't seem to be good concordance (or reproducibility?). I was reading an article that raised the current issue of data sharing. One of the companies that first started sequencing genes (which has the largest single database) refuses to share their information regarding on which genes they have clear evidence - either positive or negative links to cancers. They insist that when people are buying their test (at a premium) that they are paying for that database, while newer companies are pooling their findings to create a massive database to benefit patients. But as some patients are paying for genetic testing out of their pockets, a lower cost test is more attractive to some patients. There is now a site for sharing genome information that is being created by Washington University in St. Louis ( https://genome.github.io/civic-api-docs/ ). Patients are actually allowed to enter their own data (that needs to be validated by one other professional) to help grow the database. This is clearly an issue that needs further study and cooperation by researchers and testing providers.

    · Share on FacebookShare on Twitter
  • Bestbird
    Bestbird Member Posts: 2,818
    edited February 2017

    Indeed, there are several initiatives that encourage patients to enter their own data. To my mind, these would be of even more value if they downloaded the information to a centralized database that could be interrogated by authorized researchers. The issue of researchers conducting their analyses in silos is one of the main reasons why cancer treatment is not as far along as it should be, given the billions of dollars spent over the past 5 decades.

    · Share on FacebookShare on Twitter
  • MTwoman
    MTwoman Member Posts: 2,704
    edited February 2017

    The genome site is one that is trying to include other (smaller or newer) testing companies' as well as other researchers' data. They are trying to compile a large enough data warehouse where anyone can go if they get genomic information without understanding it's importance. It will be a slow process, but worth it. Some of the individual patients (of the large company refusing to share it's data) are actually inputting their data, which allows for that proprietary data to be shared, albeit in smaller quantities.

    · Share on FacebookShare on Twitter

Categories