Tuesday, 2 January 2018

P53 loss questions

Hi all,

I'm in the middle of determining what my next steps will be following my AA3 (previously thought to be an Oligo 2) diagnoses.

The tumour is listed as having P53 loss (or truncated mutation), which I know is more unusual in this type of tumour. While doing research on a ketogenic diet I repeatedly came across brief descriptions of p53 playing a role in metobolism...so I decided to look into it more. Below are a few links to different papers I've read. It's seems like you can exploit this loss in a number of different ways.

Firstly, there is evidence that it has some role in fatty acid oxidation and glutathione oxidation (I do not have a mutation to the MYC gene - which also plays a role in glutathione oxidation). According to these papers, these tumours (with p53 loss) are highly vulnerable to glucose restriction, they allow the PPP to become unchecked - which is a pathway that metabolises glucose. (I have read that p53 loss may prevent gene mutations (I think I was understanding that right), which maybe why I do not have an MYC mutation (the mutations I do have are listed in my post called Pathology Report from a few weeks ago).

It also regulates ROS homeostatsis, either through pro or anti-oxidant means (would loss of p53 mean that ROS was unchecked and how would this impact IDH1 mutant gliomas where ROS is already significantly altered, right?) In general, I'm questioning how IDH1 mutation works with loss of p53? IDH mutation metabolises glutamate at high rates, correct? So if p53 is lost (therefore the tumour would not generate GLS2 (at least through P53) and I do not have an MYC mutation (which allows the tumour to uptake large amounts of GLS1 - both GLS1 and 2 metabolise glutathione, then what others ways what the cells use glutathione?

I'm also questioning whether a ketogenic diet would be useful if I am able to use Proton therapy. I believe all radiation therapy uses ROS as one mechanism, proton therapy using more. A ketogenic diet has been said to limit ROS (sort of like antioxidants) but if it will limited it in standard therapy will more be better in Proton therapy? And then, the P53 in relationship to ROS question also comes in here.

I have read that keto may not be good for an IDH mutant tumour because of NAD+. Can someone elaborate more on that?

Just some more info - I had a gross total resection (surgeon said 99.9%). I'm considering Proton Therapy (do people have a hard time getting that covered by insurance), CBD/THC is try to block glutathione uptake during radiation, ketogenic diet to block glucose with attention to glutamate and methionine intake and DHC fatty acids, if needed boswellia and celebrex to prevent need for steroids, Stephen has recommended disulfiram, which I will look into and I'm going to look into metformin (seems to have a special interest with P53 loss), curcumin ( I was taking it before surgery) and I'm considering not taking any chemo at this time.

I apologize that I'm not as organized as I could be in asking these questions. I am experiencing aphasia following surgery.

Here are the links:


Thank you all.



  1. Please check out my article called "Exploring Strategies for TP53 Mutated Gliomas"

    Here I summarize some experimental evidence on the usefulness of glucose restriction in TP53 mutant tumors.

    1. Hi Stephen,

      Thank you. I have read that page as well. Reviewing it now, it doesn’t go into why these tumours may be smaller. The studies above might allude to that.

      I believe what the studies are saying is that p53 mutant tumours can eventually override glucose restriction because (via p53) they can metabolize fatty acids and gluthiathione. Loss p53 would not have those abilities, at least as far as p53 goes.

      P53 can also regulate the cells response to ROS.

      I’m not sure what wild type p53 is. Is it mutated, normal, both? Could it become mutated? It’s a bit confusing to read these studies because they use mutant p53 to describe the protein a lot but loss p53 is also considered mutated.

      Thank you for your help.


    2. PS - to make things a little more confusing on the gene mutation pathology report TP53 is listed as no change detected for me. I’m not sure what that means. On the IHM (Immunohistochemistry) report p53 is listed as being lost.

    3. "Wild-type" is kind of a strange term that just means the gene is not mutated. In other words "normal".

      Loss of detectable p53 protein may occur when one copy of the TP53 gene is deleted and the other copy has a frameshift mutation, often leading to a truncated protein that is typically non-functional and often not detectable by standard immunohistochemistry methods, especially if the mutation occurs early in the genetic sequence for that gene.

      Missense or point mutations in TP53 are more common, and lead to an accumulation of mutated p53 proteins, detectable by standard IHC methods.

    4. ..."the gene mutation pathology report TP53 is listed as no change detected for me."

      Are you saying they sequenced the TP53 gene? Who did this testing?

    5. Hi Stephen,

      Thank you.

      Brigham and Women’s/Dana Farber did the testing. They sequenced 48 or 49 items (nf1, nf2, myc (and other much things), braf, Mdm2, mdm4, PTEN, a few things that start with C....a bunch of things. I’m in the car right now, so I don’t have the list. I can type it for you late today.)

      I’ve been looking through other genes that may create a p53 truncated mutation. I don’t have any of them (all are listed as no change detected). There is one on chromosome 7 (from my memory), I think it’s 7p21. Oki is on 7p26. That one is listed as having something like 1.2 mb gain but no gene disruption. I am wondering if P53 loss could be an initiating factor with this tumour. I’ve only read one mouse study about that and I’m not sure what that would mean.

      I’ve also read that P53 loss is often radio-resistant and perhaps I read chemo-resistant. But I wonder if that not when the tumour has returned and turned of p53 through a number of these gene mutations.

      I did request a second opinion on the pathology report.


    6. Other much is supposed to say other myc things.

    7. I haven’t decided whether to get the second opinion at Johns Hopkins, Mayo Clinic or Duke (maybe there are other suggestions).

    8. I'd be interested in seeing the Dana Farber testing if you have a digital copy you could email. It would help me understand what's going on, because virtually all IDH1 mutant astrocytoma do also have TP53 mutations. It would be odd if yours didn't.

    9. Hi Stephen,

      We left a few sheets in RI (where my family is) but I do have the sheets that list all of the copy changes (or not) they tested. I can scan them in and email them. Where should I email them?

      On one of the other pages it did say something like - these changes are not specific for any brain tumour but are compatible with an AA3. I’m not sure if that is something every report says.

      I requested digital copies today.


    10. I just talked to Dr. Wen and he says copy number increases do not matter. That there is probably a mutation in TP53.

      He also says that the tumour in unmethlyated.

      I’m quite confused about all of this.


    11. Ah okay, I understand now. Sounds like they didn't do DNA sequencing of TP53 or other genes, the analysis was just a copy number analysis (which wouldn't detect mutations within genes, just gains or deletions of larger regions of genetic material). So yes, I would agree with Dr. Wen that there is probably a mutation in TP53.

    12. Should I request a DNA sequencing?

    13. What about unmethlyated status? I thought that virtually all IDH mutants where methylated?

      What does the copy sequencing mean? I’m not sure what I should take from it if it seems to mean nothing.

    14. Hi Maria. Not all IDH mutant tumors are methylated, even though the vast majority are I think. In my case, tumor was also IDH1 mutated but not MGMT methylated. So it may be rare, but the combination still exists.

    15. Hi John. Thanks. I do know that I guess I’m just shocked at this pathology. We’ve been watching the tumour for a year, because I was pregnant when it was diagnosed, and every doctor I spoke to thought it was an Oligo 2.

      It’s diagnosed as a AA3, unmethlyated with ATRX loss and P53 loss. MIB is 7.7% and mitosis is 3 of 10 HFP - I might have that terminology wrong but you’ll get the idea.

      Do you mind telling me about your pathology report?


    16. I'm very sorry to hear that - I'm sure it was a big shock for you. When I was diagnosed, I was also led to some wrong hope as my neurosurgeon initially thought the MRI looked benign.

      My pathology report is GBM IV, IDH1 (R132H) mutated, MGMT unmethylated, ATRX loss, P53 positive in 20% of cells, Mib-1 15%. To be honest, I never really looked too deeply into the meaning of these so far, but ATRX loss seems to be a good prognostic factor based on NOA-04 assessment:


    17. I’m sorry that you were shocked too. I guess that may be the way these types of stories go.

      My pathology is negative p53. I’ve read this can have positive and negative effects, so I’m not sure which to believe. I feel as though it’s probably less positive, since the research isn’t really done on lower grade gliomas with p53 loss, since it is also rare (like unmethlyated).

      I wish you well, John.


  2. "Should I request a DNA sequencing?"

    For the TP53 gene? Probably not worth it if you have to pay out-of-pocket for it. We already know p53 protein expression is lost (most likely due to a mutation leading to a truncated protein). Knowing the exact mutation causing this loss doesn't give us any more actionable information.

  3. "What about unmethlyated status? I thought that virtually all IDH mutants where methylated?"

    It's important to remember that MGMT methylation results aren't set in stone. See this recent abstract from the SNO 2017 conference showing *discordant* results for MGMT methylation status between local and central laboratories in 39% of cases.


    A different study a few years ago showed that in some cases an "unmethylated" result by methylation-specific PCR became a "methylated" result with pyrosequencing.

    Some tumors can have heterogenous MGMT status, with high methylation of the MGMT promoter in some parts of the tumor, and relatively low levels in other parts.

    In short, MGMT methylation status is not a simple binary, and can be more complicated that people realize, and the methylation status might not be in agreement between different labs or using different techniques.

    1. I am really glad to hear that MGMT methylation status is not set in stone – thanks for pointing that out and for the studies. My partner recently had 2 debunking surgeries (awake component and non- awake a few weeks apart) for a grade 2 astrocytoma and we had a Caris Molecular Test performed and the test showed (via pyrosequencing) MGNT as unmethylated. This was concerning to us and also surprising since given that all low grade idh1 mutated gliomas are invariably methylated! So there is still some hope at least that it may have just tested one of the unmethylated regions of the tumour.

      Another strange thing is the discordance between the 2 pathology reports from the 2 surgeries (conducted by different doctors) that show ATRX as mutated, whereas the Caris test shows no ATRX mutation. To add to the confusion, FoundationOne testing on a sample from a 2015 surgery also showed no ATRX mutation – so pathology reports detected ATRX as mutated/lost (via staining), vs 2 epigenetic tests showing ATRX as not mutated/lost.. confusing stuff. Any idea which one would be more likely to be correct?

      Both pathology and epigenetic testing agree on IDH1 mutated, TP53 mutated, proliferation index (ki-67) of ~3%, and low mutational burden and conclude low grade astrocytoma.

      Thanks for any insights.

    2. Low grade IDH1 mutant gliomas aren't invariably methylated for MGMT. Most are, but some aren't, even when tested with pyrosequencing. But yes, the result could potentially be different if a different part of the tumor were to be tested.

      ATRX mutation occur later in the course of disease than IDH1 mutations. It's possible that most of the tumor carries no ATRX mutation (as in the original report and the later Caris reports) but that a smaller subclone exists within the tumor that has picked up an ATRX mutation. That is not necessarily the correct explanation, but is within the realm of possibility.

      Discordant test results doesn't always mean one is correct and the other isn't. Instead it could reflect intratumoral heterogeneity (both in space and time).

    3. Thank you very much for your insights Stephen, they are much appreciated. When I mentioned that the IDH1 mutant gliomas are invariably methylated, I was quoting the article that you posted - https://www.ncbi.nlm.nih.gov/pubmed/22020830 - but thanks for clarifying that it is not always the case.

      Your explaination for a possible reason for the discordant results for the ATRX testing also makes sense, thanks.

      It does make it difficult when deciding on whether or not to go for chemo (with TMZ and CCNU) and risk hyper-mutation, which is obviously worth the risk if the tumour is definitely methylated and there would be a definite advantageous effect. My partner is lucky as it seems that she will be admitted to the Bayer idh1 inhibitor trial and will be screening for enrolment next week – so hopefully she will satisfy the blood tests etc. This would give us time before needing to decide on chemo.

      We are hoping that the Bayer inhibitor will at least match the great results (which you recently posted from the SNO) achieved by the AG120 inhibitor!

      I just had another question, if you don’t mind.. the http://astrocytomaoptions.com/idh1-mutation/ page near the bottom, you mention a cautionary note that was raised in a 2014 publication where it is mentioned that, “Patients should be monitored carefully in near-future clinical trials, because IDH1/2 mutant-specific inhibitors may counteract the putative survival-prolonging effects of IDH1/2 mutations and result in a decrease, rather than an increase in survival”, do you think that this is somewhat less of a concern given the great results todate of the AG120 trial?

  4. Thank you.

    Rich G. has recommended that I have the methylation retested.

    So, if the copy number analysis I have is unreliable how do must people here know the mutations of their tumour? Does everyone pay to get the DNA sequencing done?


    1. The copy number analysis isn't unreliable, just different from DNA sequencing. It only looks at copy number changes (gains or losses) of genetic material, but can't detect changes in the DNA code.

      Most people don't know the mutations of their tumor unless they are fortunate enough to be at an instituation that does that kind of testing, or have insurance that can cover the expense of testing such as FoundationOne or similar services.

      For your type of tumor (IDH1 mutant astrocytoma) at first diagnosis, there are only 3 genes that are commonly mutated: IDH1, TP53, and ATRX. We already know your tumor has the IDH1 mutation, p53 expression is lost (I would guess due to a frameshift mutation in TP53), and the loss of ATRX expression is almost certianly due to an ATRX mutation.

      In your case further genetic testing won't likely discover any additional actionable mutations. In your case I would save that money for other uses, unless your insurance would cover further testing.

    2. Except I would agree with Rich that having MGMT methylation status re-tested from a different part of the tumor would be interesting.

    3. Ok. I guess I’m having trouble understanding what CNV is in aid of then. Would it help them to see if there maybe a change in the DNA code but some changes would go unnoticed? Like TP53 in this case?

      Would a copy number change or loss indicate there maybe a mutation on that chromosome?

      Thanks for helping me understand, Stephen. I appreciate it.


    4. It's still not clear to me what testing was actually done, as I haven't seen the report. It's possible there was some DNA sequencing done as well in addition to the gains of parts of chromosomes you listed in your original post.

      504 kb gain of SOX2
      1.9 mb gain of OKI (no gene disruption)
      21.2 Mb broad gain of 7q, an
      8.7 Mb gain of 12q

      I still can't figure out what OKI refers to but if I saw the report it might make sense.

      Broad copy number changes such as those listed in your original post don't often contribute to therapeutic decisions for brain tumors, although focal gene amplifications (such as EGFR) sometimes lead to use of a targeted drug in GBM. For your type of tumor at first diagnosis, I've never heard of treatment decisions being made on the basis of amplifications or deletions (probably because these are relatively rarely found, unlike in GBM where EGFR is often amplified, CDKN2A/B is often lost etc.).

      For a recurrent secondary glioblastoma that progressed from a lower grade glioma, looking for gene amplifications (like PDGFRA) and doing more sequencing to look for additional mutations would be more important, as these things are often driving the progression from a lower to a higher grade.

    5. In summary, I'd be satisified with the testing that has already been done, because mutant IDH1, TP53, and ATRX is the unholy trinity of lower grade astrocytomas, and other alterations are relatively rare. In the event of a recurrence, especially post-temozolomide or at a higher grade of tumor, further testing would become much more important.

      With the caveat that a re-test of MGMT methylation status would be desirable at this point.

    6. Thanks, Stephen. Dr. Wen never sent me the report digitally and I think what I have will not suffice. We will call them today because my local NO also hasn’t received them.

      All I have, basically, is the IHC report that includes IDH, ATTX and P53 like you pasted above and a copy number analysis that is the SOX2 gain, BRAF 7q34 single copy gain, OKI which is at 6q26 gain (with no gene disruption) and CCND2 12p13 gain. It then lists out everything like TP53, PTEN, 1p, 19q, CDKNA2A, CDKNA2B, CDKNA2C and EGFR and more and says no change detected. This is all Dr. Wen gave me. Although there is now the methylation status as well.

      It says that - these changes are not specific to any brain tumour but are compatible with AA3 diagnosis. I guess this is because of the chromosome 7 gain and SOX2. It doesn’t seem like the rest of the changes are compatible or common really.

      I am going to meet with my local NO and request the testing through him.


  5. We had the IHC testing evaluated at the Mayo Clinic (using the same prepared slides from Dana Farber) and they identified a small amount of cells expressing P53 (generally weak) which lead them to speculate that TP53 is not mutated but with no definitive answer because it could be mutated in some other way. Mitosis they say is 2 and MIB is between 4-6%. Although this falls within WHO guidelines as a grade 3, they felt uncomfortable about naming it a 3...and implied that it was in a grey area between 2-3.

    Proton therapy was also denied by my insurance...so we either need to appeal or go with photon. We interviewed the proton doc and she said the tumour is not in a sensitive place and it’s right guy is n the surface. No no real structured will be damaged by the conventional radiation. I’m afraid we’ve waited too long to appeal the proton decision. It’s a disappointment because they said my insurance is usually more accommodating than others.


    1. The cells showing p53 expression could have been non-cancerous cells like reactive astrocytes. If there was a very small amount of p53 expression, it's hard to prove those are cancer cells without co-staining for both IDH1 R132H and p53. It would be very rare for an IDH1-mutant astrocytoma with loss of ATRX to not have a TP53 mutation.

      Regardless of when a tumor is discovered, it's probable that all grade 3 astrocytomas have evolved from a grade 2, so it makes sense some tumors would be diagnosed during the transition to a higher grade. The difference in prognosis between a grade 2 and 3 IDH-mutant astrocytoma is not that different anyway. Much more important is IDH1 mutant versus wild-type status.

      What reasoning did your insurance use to deny coverage for proton? Because it hasn't been proven in clinical trials to be superior?

  6. I’m not sure, Stephen. We just got the message last night. So we haven’t checked on it yet. The doctor said it is usually denied for that reason.

    Regarding the grade, yes, that makes sense. I guess I’m hoping it will act like a grade two (maybe with more PFS time) than a grade 3. But I do hear often that there isn’t much difference.