Friday, 20 November 2015

Which mutations to test for?

Hello guys!

I have surgery of my low grade glioma scheduled in less than 2 weeks, but the neurosurgeon told me they don't test for many mutations/markers unless I specifically ask them to.

At the moment I know of:

-MGMT methylation

Is there any other mutation/marker I should ask them to test?

Your help is greatly appreciated,


  1. 1. Standard practice is to test for IDH1 R132H mutation by immunohistochemistry (IHC). If that test is negative, then DNA sequencing of IDH1 and IDH2 would be a next step.

    2. If positive for IDH1 or IDH2 mutation (IDH1 is by far more common), the next step would be distinguish between astrocytoma and oligodendroglioma. Strong positivity for p53 or absence of ATRX (both by IHC testing) would suggest astrocytoma. 1p/19q codeletion would confirm oligodendroglioma.

    3. Most IDH-mutant tumors will have MGMT promoter methylation, but good to confirm this with testing.

    4. EGFRvIII is unlikely. That is more of a GBM marker. I would put the chance of having any EGFR mutation at ~5% or less, and only if the tumor is IDH non-mutant. There are exceptions to this rule but they are very rare.

    The first 3 steps are higher priority: IDH1 R132H, p53, ATRX (all by immunohistochemistry), and MGMT promoter methylation status.

    If it turns out to be an IDH1-mutant astrocytoma, which is most likely, with strong p53 staining and loss of ATRX, then you could consider having the TP53 gene sequenced to determine the specific mutation, (but this would be more expensive and would not change the treatment plan.)

    1. Stephen you are amazing.

      I am trying to see what of the above can be useful for my friend with recurrent GBM. We know already he is EGFR vIII positive and MGMT unmethylated. But these are the only two tests he ran.

      Any other mutations we should be looking for? And if so, could you let us know what targeted treatments we should be evaluating?

      Many thanks again !

    2. Guy,
      EGFRvIII is typically associated with "classical" type GBM. EGFRvIII is pretty much always found in conjunction with EGFR gene amplification (many extra copies of the gene, rather than the two copies found in normal cells). Loss of one copy of chromosome 10 (which contains the tumor suppressor gene PTEN) is also very likely. Also, frequently the second copy of PTEN is taken out by mutation or deletion.

      In addition to EGFRvIII and EGFR amplification, there may be other EGFR mutations.

      TP53 mutations are rare in the "classical" subtype.

      ALDH1A1 is not a standard protein to test for in GBM, but is a stem cell marker, often found in EGFR overexpressing cells, and can be targeted by the drug disulfiram.

      To summarize, EGFRvIII virtually guarantees EGFR amplification. There are a number of EGFR inhibitors approved for other cancers, the newer generation being represented by afatinib. Metronomic TMZ schedules might be more effective than the standard schedule for EGFR+, MGMT unmethylated GBM. Chloroquine may also be preferentially beneficial for EGFRvIII positive GBM.

      PTEN de-activation through deletion or mutation leads to increased Akt and mTOR activity. Rapamycin inhibits mTOR, but has not been too successful in GBM trials so far. There are a number of dual PI3K/mTOR inhibitors in trial, but none approved yet. Honokiol is a commercially available supplement that inhibits PI3K/mTOR pathway in GBM cells in the lab, and prolongs mouse survival in vivo.

      So, as far as cocktail ingredients I would emphasize chloroquine, disulfiram, honokiol, perhaps an EGFR inhibitor like afatinib if an oncologist would prescribe it (and it would likely need some help crossing the blood-brain barrier), and perhaps metronomic temozolomide, depending on his prior exposure and history with TMZ.

  2. Stephen, Wouldn't all this information be found in the Foundation Report? And is the Foundation Report not part of standard protocol when sending the biopsy to the lab?

    1. Everything but the MGMT methylation status and 1p/19q codeletion would be in the Foundation Report. However, Foundation testing is not part of the standard protocol, but a separate commercial service, that is quite expensive, and as I was told, only can be covered by American insurance providers. So my (Canadian) friends were out of luck.

      At most centers, you're lucky to get testing for MGMT status, EGFR amplification and EGFRvIII, chromosome 10 deletion, p53 staining, and IDH1 status. In contrast, FoundationOne does DNA sequencing of over 300 cancer-related genes, and also detects amplifications or double deletions of those same genes. This is why it is so expensive - it is far beyond the testing you would get in most hospital pathology departments.

      But it doesn't do everything. It tests only for gene mutations/amplifications/deletions. It does not test for gene methylation status (for example MGMT methylation).

  3. Good luck Matjaz, we'll be thinking of you!

    1. Welcome to the blog my friend!

    2. Thank you both! And welcome Jessica :)

  4. I think if I (or anyone else in similar situation) will want to try with immunotherapy and checkpoint inhibitor, they should also test for PDL1.

    Dr. Van Gool wrote: "I know that PDL1 is expressed in high grade glioma. But I am not immediately aware for low grade glioma. In a first search, I find nothing on Pubmed. Eventually one can ask the pathologist to stain on the tissue obtained."