Hi folks/Stephen,
Needed a bit of help from you all. My mom is a GBM patient who was diagnosed in Sep 2017. Her cancer is IDH1/2 -ve, methylated and 1p/19q codeletion negative, this is the information that we gauged from the initial gene testing that was done by our hospital post surgery. We are now considering a next generation gene sequencing test done, and I had the following questions on the same:
1. How helpful has the gene sequencing report been for you so far, if you have got one done? What questions should I look at getting answered from this test?
2. Since there is very little tumor tissue that we all have, which is the best place to get the gene sequencing done from? I currently have read and spoken to OncoDNA, StrandAdvantage, RGCC Greece and FoundationOne so far. Which one would you recommend, and why? The recommendation could be different from the mentioned five too.
I read the following doc shared by Stephen in the Brain Tumor library too, but I'm not quite able to figure out on which test is a more appropriate one for what kind of patient.
https://docs.google.com/spreadsheets/d/1653spmu9DkVCKX16G0_ZUlsTuJOx_Ln2qVysJJI-uoU/edit#gid=0
There are so many genes written in all of them, that it's hard for me to be able to compare and make an informed call :/
Want to know if the gene testing will be of help in further treatment, and the drugs/supplements that we should use for her in the near future.
Her current status
1. Her MRI three months back showed no growth, but had choline elevation at two parts where there was tumor resection. We have her next MRI in about a week from now.
2. She will do her remaining chemotherapy cycles with Temozolomide+Lomustine, her cancer being methylated.
She currently is on ketogenic diet and lot of naturopathic supplements suggested by Nutrition Solutions.
My plan going forward
After my mom is done with the chemotherapy, I plan to get her immunotherapy (Dendritic cell therapy) done. Want to know if the gene sequencing will be an add on in this plan.
Mutations commonly found in GBM are: TERT promoter (>80% of tumors), PTEN (~30%), TP53 (~30%), EGFR (~30%), PIK3R1 (~10%), PIK3CA (~10%), NF1 (~10%), and RB1 (~10%).
ReplyDeleteIf your mom's tumor is a typical GBM, next-generation sequencing would likely detect a mutation in one or more of these genes, and perhaps gene amplifications or deletions such as EGFR amplification, and CDKN2A/B deletion.
The question is, how useful would that be? EGFR is the only one of these genes directly targetable by approved drugs, while other of these mutations activate pathways that are druggable (PTEN, PIK3CA, PIK3R1 imply activation of PI3K -> AKT -> mTOR pathway; NF1 mutation could be potentially targeted with a MEK inhibitor + mTOR inhibitor. However, targeted therapy has met with limited success so far for GBM, partially because most approved targeted therapies were developed for non-central nervous system cancers and have suboptimal uptake into the CNS/brain. One of the most common overactive pathways in GBM is PI3K/mTOR, yet the mTOR inhibitor everolimus showed no benefit when added to standard therapy for unselected newly diagnosed GBM.
You might get some actionable information from sequencing of a newly diagnosed tumor, but I feel that for MGMT-methylated tumors, the sequencing becomes much more important if the tumor recurs post-temozolomide, and the possibility of a TMZ-induced hypermutation profile needs to be determined or ruled out to decide on further treatment strategies. For example, a hypermutated tumor at recurrence with mismatch repair deficiency should not be treated with further temozolomide or procarbazine, while lomustine would be a good choice. Also anti PD-1/PD-L1 treatment would be a good option in this instance.
Another instance where next-generation sequencing would be important is if pursuing a personalized multi-peptide vaccine. In my spreadsheet, I show information on the German sequencing service CeGaT. Having sequencing done with CeGaT would give the possibility of having a personalized peptide vaccine created if sufficient immunogenic mutations were found (but that can't be known beforehand). Of course if pursuing a tumor lysate-pulsed dendritic cell vaccine, this sequencing wouldn't be necessary.
I would want PD-L1 expression included in the results. This is included in FoundationOne if specially requested, and is included in the Caris Molecular Intelligence package, but the full package is much more expensive than the Next Generation sequencing only by Caris. OncoDNA includes PD-L1, as well as CD8 both of which could indicate whether it is an immunologically "hot" tumor that might respond to immune checkpoint blockade (anti PD-1 or PD-L1). The OncoDeep test (by OncoDNA) includes far fewer genes than the other companies, but it does include all the commonly mutated genes in GBM except for PIK3R1 and NF1.
ps by adding lomustine to TMZ upfront, you might be decreasing the risk for a hypermutated recurrence, as the cells that acquire mismatch repair defects are *more* sensitive to lomustine, while resistant to TMZ.
DeleteThanks for the response Stephen. A few followup questions some of your points, would be great if you can help with the answers:
DeleteYour point: You might get some actionable information from sequencing of a newly diagnosed tumor, but I feel that for MGMT-methylated tumors, the sequencing becomes much more important if the tumor recurs post-temozolomide, and the possibility of a TMZ-induced hypermutation profile needs to be determined or ruled out to decide on further treatment strategies.
My query: Do you mean the sequencing of the first diagnosed tumor will help in case of recurrence or will we have to do the sequencing of the recurred tumor for better understanding the treatment strategy, just in case we do a surgery again? I'm presuming you mean the former, but I'm not quite clear on how the sequence of the first tumor will help in case of a recurrence, given the fact that tumor cells do keep changing their mutations basis the treatments that are being thrown at them, that's what my oncologist told me.
Your point: OncoDNA includes PD-L1, as well as CD8 both of which could indicate whether it is an immunologically "hot" tumor that might respond to immune checkpoint blockade (anti PD-1 or PD-L1).
My query: I saw the list of the genes in FoundationOne, they neither have PDL-1 nor CTLA-4. I read during my research that Nivolumab is given as a checkpoint inhibitor for PDL-1 +ve and ipilimumab for CTLA-4 +ve. Should I be asking for the CTLA-4 test from the FoundationOne folks too, or does a PDL-1 -ve imply a CTLA-4 +ve? How important is the CD8 test in defining how "hot" the tumor is for immunotherapy?
Our agent from FoundationOne kept trying to dodge the question of PDL-1 test, saying that the tumor burden is good enough a marker for testing the "hotness" of the tumor for immunotherapy, how correct is that for brain tumors?
The only research that I found was this talked of tumor burden's relation with immunotherapy efficacy majorly had lung cancer patients and the data output wasn't statistically significant.
http://ascopubs.org/doi/abs/10.1200/JCO.2017.35.15_suppl.e14579
PS: For now, I'm planning to go with the FoundationOne Test with a more adamant request for the PDL-1 test :)
I meant that upon recurrence, sequencing of the recurrent tumor would be important to help prioritize treatment strategy.
DeletePD-L1 is not part of the sequencing panel in FoundationOne - the point is not to look for a mutation in that gene, but to determine protein expression levels of the normal protein by immunohistochemistry. If you download the FoundationOne Test Requisition Form from their website you'll see a check box labeled "Additional Options IHC testing: PD-L1". This is for the ordering physician to fill out.
https://assets.ctfassets.net/vhribv12lmne/28Xw5cgFrqecQa4aa2uEsO/ac6bc2a45df82c970e3f39d5606c3f29/101017_-_MKT-0088-03_-_FoundationOne_and_FoundationOneHeme_Requisition_Form.pdf
I've not seen any of these companies testing for CTLA-4 expression.
As SVG has pointed out several times, checkpoint inhibitors like anti PD-1 or PD-L1 won't work unless there is already an immune reaction within the tumor, which the tumor has suppressed by upregulating PD-L1 or other immune checkpoints. Without this pre-existing immune response, checkpoint inhibitors alone will not likely be of much value, although such an immune response can be induced by a vaccine.
High PD-L1 expression is a sign that there are infiltrating immune cells that the tumor needs to suppress. CD8 is a marker for cytotoxic T-lymphocytes, and would be direct evidence of an immune reaction within the tumor.
In theory, a tumor with high mutational burden is more likely to create an immune reaction because of a greater number of mutated proteins for the immune system to recognize as foreign. However newly diagnosed GBM usually have low mutational burden, yet even some newly diagnosed tumors with low mutational burden may be "T-cell" inflamed and respond to immune checkpoint inhibitor therapy. This might be more associated with certain GBM subtypes (mesenchymal).
Your mom's original tumor sample is likely to have low mutational burden like most newly diagnosed GBMs. Regardless, high PD-L1 expression might predict a response to anti PD-1/PD-L1 therapy. High PD-L1 and high CD8 expression combined would be even more convincing, with direct evidence of infiltrating CD8+ T-cells. Given the current level of experience with checkpoint inhibitors for GBM, it's probably only a minority that have sufficiently "hot" tumors to respond to checkpoint blockade alone, but if you're in that minority you could do quite well on them.