-
Notifications
You must be signed in to change notification settings - Fork 13
List of Relevant Publications
moskalika edited this page Dec 12, 2018
·
7 revisions
-
- Comparison of primary vs metastatic ccRCC samples shows evolutionary bottlenecking, where metastases are more homogeneous with less driver somatic alterations. Majority of the driver events accumulate in the primary tumor, with relatively low levels of driver events exclusively found in the metastasized tumors. The metastasized tumors did show increased somatic CNAs, proliferation and immune evasion (not SNV/INDEL counts). Loss of chromosomes 9p and/or 14q were found in the majority of metastasized samples.
-
- Point mutations in ccRCC patients accumulate in a linear pattern correlated to patient age.
- Mutations accumulate at a relatively similar rate between tumor subclones
- Mutations occurring prior to chromothripsis of 3p/5q will be on 2 5q chromosomes, while those occurring after will only be present on one 5q chromosome. (can correlate mutation differences between those on both chromosome and on one to calculate time point of duplication)
-
- A cohort of primary and recurrent tumor pairs that describes changes in miRNA expression based on the type of therapy pts received, and what genes are affected by this change.
- There is a nice figure showing how commonly affected genes influence treatment resistance, such as radiosensitization leading to poor radiotherapy response and what are the miRNAs that affect those genes.
- Found no difference between expression of the tested miRNA and MGMT status, but did find changes in MGMT between primary and recurrence as has been noted in previous studies.
- Samples were compared to Human Brain Reference RNA instead of pt controls
-
Wang J, et al. Clonal Evolution of Glioblastoma under Therapy. Nature Genetics. July, 2016
- On average number of somatic mutations decreased between initial and recurrent GBMs except where recurrent cases were hypermutated
- Only tumors previously treated with TMZ were hypermutated
- PTPN11, LTBP4 (10/93), DNA mismatch repair gene MutS Homolog 6, MSH6 (8/93), PRDM2 (10/93) and IGF1R (9/93) - potential new driver genes (all but PTEN11 found only in recurrents)
- LOH in TP53, NF1, PTEN, and newly discovered APC
- Hypermutated recurrent tumors are highly enriched with C>T (G>A) transition; recurrent hypermutated samples contained significantly greater numbers of silent mutations; genes containing hypermutations are expressed in larger quantities than other mutated genes; MMR genes are the usual harbors for hypermutation.
- Secondary GBMs are more stable in their genetic signatures than primary GBM
- Mutated genes can change the mutation between primary and recurrent tumors
-
- Compared 29 pairs of GBMs (later added 2 more cohorts) to see the changes in variants between primary and recurrent tumors as well as treatment vs no-treatment group.
- Showed lower number of variants at recurrence except in one hypermethylated case.
- Showed a neutral tumor evolution trajectory in primary tumors and a non-neutral in secondary tumors treated with RT+TMZ
- Treated pts showed an emergence of unique variants in the MMR pathway genes in the recurrent tumors.
-
- Used binary model for LTS (cutoff at 3 yrs) and had two groups for GBMs, primary and secondary (not in pairs)
- Clinical variables that were used for the survival prediction model originally included KPS post-op, chemotherapy, and age (gender, race,radiation and history of LGG did not show significance). When KPS and chemotherapy were controlled for age, age showed to be the only clinical variable that affected survival.
- Used germline and somatic point mutations, gene expression, DNA methylation, copy number variation (CNV) and microRNA (miRNA) expression data from TCGA GBM cohort.
- A list of molecular markers characteristic of LTS in GBM broken down by the 6 subclasses of data listed above, with CNV showing the highest correlation with LTS (0.6785) after age (0.8070)
- The combinations of the following information were more predictive than age alone were: age+mir+met, age+met+mir+exp+cnv, age+exp+met, age+met+cnv (in order of hi to lo 0.8126-0.8034)
- Genetic profiles of LTS GBMs were more similar with other non-LTS GBMs than with each other.
- While IDH1mut is a significant predictor of LTS, the majority of LTS cases in the TCGA data set are IDH1- wildtype and none of the IDH1+ genotypes was in a GBM with a prior LGG history.
- Secondary GBMs also did not show molecular profiles similar to LGG nor to each other more than primary GBMs
-
- Batch effects on DNA methylation data were not observed and the only difference in the DNA methylation dataset was observed between FFPE and fresh-frozen samples
- Deletions of 10q in recurrences were correlated w/ increased OS and sensitivity to alkylating agents.
- Transcriptional subtype was heterogeneous within multisector samples as well as between primary and recurrent pairs.
- DNA methylation can be used to infer immune cell infiltration similarly to RNA-expression profiles, extent of necrosis, shape of tumor cell nuclei, and predict OS
- Increased EZH2-binding activity was observed in mesenchymal tumors (worst survival)
- High proliferation in the recurrent tumors (but not in the primary) was associated w/ longer PFS. High proliferation was characterized by intermediate DNA methylation levels at discriminatory regions; low proliferation: more extreme DNA methylation patterns
- More temporal DNA methylation differences were observed in pts with shorter surgical interval, suggesting that more aggressive tumors may have higher epigenetic plasticity
- Promoters that gained DNA methylation were related to neural development and apoptosis while those that lost methylation were enriched in the Wnt pathway and T cell activation (the latter is associated with sig. reduced survival)
-
- Studied glycolysis level and investigated GGESS's (glycolytic gene expression signature score) clinical significance as well as its relations with GBM subtypes and IDH1 mutation status using the TCGA’s glioma dataset
- Genes used in GGESS: positive effect on glycolysis: HK2, HK3, LDHA, PKM2, GAPDH, ENO1 and GALM negative: LDHB, PKLR and ALDOB
- High GGESS predicted poor prognosis and poor response to chemotherapy in GBMs from TCGA, which was validated in GBMs from Repository for Molecular Brain Neoplasia Data (REMBRANDT) and GEO database (GSE16011).
- LGGs had lower GGESS than GBMs
- Mesenchymal GBMs exhibited high GGESS while IDH1-mutant GBMs had low GGESS.
- The promoter regions of tumor-promoting glycolytic genes composing GGESS were hypermethylated in IDH1-mutant GBMs compared with IDH1-wildtype GBMs.
- There is a figure which compares each gene’s expression between IDHwt and IDHmut GBMs
- Pts with high GGESS had worse survival when treated w/ TMZ than no TMZ group
-
- hTERT mutation was not a predictor of PFS or OS in IDHwt GBMs, it was only an independent predictor for both OS and PFS in MGMT unmethylated patients (negative effect in MGMT-U)
- Survival benefit associated with MGMT promoter methylation remained only in patients carrying the hTERT promoter mutation (need hTERT mutation to see survival benefit of MGMT) - also most pts received TMZ after surgery
-
- Found that somatic mutation count and CNV increase proportionally to grade and stage in genito-urinary cancers and inversely correlated with recurrence-free and overall survival using TCGA data
-
Jolly C, Van Loo P. Timing Somatic Events in Evolution of Cancer. Genome Biology. 2018
- Summarizes different mutational signatures and their derivation (not specific to a single tumor type).
- Use of mutation rate based on age is a close estimate to Signature 1 of predicting time point of tumorigenic events.
-
- Pairs of primary and recurrent GBM samples (58 pairs) were evaluated for the presence of ecDNA and its response to chemotherapy (showed that most ecDNA is preserved at recurrence)
- 60% of ecDNAs target an oncogene
- Three most commonly found oncogenes on ecDNA: EGFR, PDGFRA, MYC, MDM2 and CDK4 (both found on 12p)
- IDHwt samples contained relatively more ecDNAs than IDHmut, but the proportion of samples that had at least one ecDNA was the same between the two subtypes
- No significant correlations was noted between somatic mutations and the presence of ecDNA. However a significantly shorter time was observed to second surgery for patients whose primary tumor sample was predicted to carry at least one ecDNA, relative to patients with a primary tumor that contained no predicted ecDNAs.
-
Brennan CW, et al. The somatic genomic landscape of glioblastoma. Cell 155, 2013.
- Proteomic profile showed a non-linear relationship between gene alteration and effects on downstream signaling pathway, challenging therapeutic targeting of downstream components of known mutated genes.
- Chromosome 12q showed to have a high frequency of structural variants (often involving CDK4 and MDM2) and was associated w/ presence of ecDNA
- Almost one half of GBM samples in the cohort had EGFR variants, with intratumor heterogeneity
- Chromatin-modifier gene mutations were noted in 46% of cases
- Genes identifies as sig. mutated: EGFR, PTEN, TP53, PIK3CA, PIK3R1, NF1, RB1, IDH1, PDGFRA, and LZTR1 as well as other less prevalent mutations.
- Most common amplifications were on chromosomes: 4, 7, and 12
- Frequent gains of the following genes were noted: SOX2, MYCN, CCND1, and CCNE1
- QKI possibly functions as a tumor suppressor in GBM and was found to be deleted or mutated in 14 cases (6q26)
- Interesting figures
-
- Purpose: To quantify the spectrum of ECDNA in human cancer and systematically interrogate its contents.
- EcDNA was found in 90% of pt-derived brain tumor models, was seldom found in normal cells, and varied from cell to cell.
- Amplified oncogenes were found either in ecDNA or in ecDNA and corresponding chromosome, with high transcription levels and high copy-number variability compared to chromosomal loci
- EcDNA can dynamically relocate to chromosomal loci while still maintaining main structural features (can stop expression during treatment w/ EGFR inhibitors and then resume expression when therapy seizes)
-
- Overall survival based on the 2-gene model: IDHmut+MGMTmet > IDHmut+MGMTwt/IDHwt+MGMTmet > IDHwt+MGMTwt
- The pts' age, extent of resection, KFS, dosage (Gy) of radiotherapy received, mutations in IDH1, PIK3CA, PTEN and TP53, alterations of EGFR, and methylation of the MGMT promoter were found to be sig. associated w/ OS
-
- Set of 12 tumor samples from 6 pts (multisectors), all IDHwt primary GBMs
- All tumors harbored alterations in the 3 GBM core pathways: RTK/PI3K, p53, and RB regulatory pathways with aberrations of EGFR and CDKN2A/B in all (100%) patients.
-
- The most important increase of mutation rates within FAs were observed in brain tumours: LGG(9-fold) and GBM (8-fold) compared to 15 other TCGA tumor types
- In summary, showed how focal amplification by creation of ecDNA can lead to high copy number and lead to creation of driver mutations, such as seen with EGFR. EcDNA explains the ability of GBMs to quickly adapt to their environments as the ecDNAs can quickly multiple or completely vanish from a population.
-
- EGFRamp, 7+/10-, and pTERTmut is exclusively seen in IDHwt GBm. EGFRamp can be used as a single marker or in combination with either one of the other 2 markers to upgrade IDHwt astrocytoma to IDHwt GBM in the absence of histologic evidence
-
- Study aimed to compare a group of 88 primary GBMs to 22 recurrent GBMs (not paired) based on pt characteristics, gene signatures, and molecular subtypes (based on WHO 2007).
- Primary tumors exhibited gene signatures related to translation, RNA_processing, and cellular biosynthesis; while recurrent were related to DNA damage response signal transduction, DNA damage checkpoint and chromosome segregation
-
- Macrophages are more prevalent in the tumor microenvironment than T cells, and T cells that are found are, at a higher proportion, T-regulatory cells.
- TAMs originate mostly from circulatory monocytes - 85% (not microglia - 15%)
- Showed that TAMs derived from the bone marrow compartment drive gliomagenesis, whereas microglia appears to play a less significant role in tumor growth and is mostly involved in tumor cell invasion
- TAMs exhibit M1 and M2 like characteristics (unclear if a single cell can have both), attempts to polarize TAMs towards an M1 phenotype have been futile as tumor cells produce factors that revert TAMs back to an M2 phenotype, this also requires an adequate CD4+ T cells in the tumor environment
- MCP family members: CCL2, CCL7, CCL8, and CCL13, play an important role in monocyte migration and infiltration. Tumors with high levels of CCL2 were found to have sig. shorter survival than those with low levels (providing a promising avenue for future therapy strategies)
- T-cell related genes that have been implicated in GBM include: PD-1, CTLA-4, and HAVCR2 (inversely related to OS);
-
Molinero A, et al. Statistical considerations on prognostic models for glioma. Neuro-oncology. 2016
-
Cordner R, et al. Exploitation of adaptive evolution in glioma treatment. CNS Oncology. 2013
-
Kim J, et al. Spatiotemporal Evolution of the Primary Glioblastoma Genome. Cancer Cell, 2015
-
- The heterogeneity observed at the single cell level is captured in the expression profile of the bulk tumor.
- Comparison of ITGAM1 and AIF1 (markers of macrophages/microglia) levels showed a significant increase in mesenchymal subtype compared to the other two
- NF1 deactivation in glioma cells leads to recruitment of macrophages/microglia.
- Recurrent Mes GBMs showed an increase in transcriptional activity associated with non-polarized M0 macrophages and dendritic cells.
- PN recurrent GBMs showed a sig. lower level of transcriptional activity related to immune cells compared to primary tumors.
- Patients with hypermutated tumors at recurrence due to TMZ treatment show higher levels of CD8+ T cells in the tumor microenvironment and may have a more immunological reactive microenvironment that may be responsive to immune checkpoint inhibitors.
- M2 macrophages and CD4+ T cells have been found in higher amount in short term relapse vs long term in patient treated w/ radiotherapy, pointing to possible involvement of M2 in radioprotection.