See the article by Baldock et al, on pages 779–786.
The molecular genetics of glioblastoma (GBM) are complex, and efforts are underway to systematically investigate their association with tumor aggressiveness and patient outcome1–3. Mutation in the isocitrate dehydrogenase 1 (IDH1) gene produces an alternative metabolite, 2-hydroxyglutarate instead of alpha-ketoglutarate and is strongly associated with lower grade gliomas4,5. The majority of GBMs containing the IDH1 mutation are secondary, and diagnostic approaches to distinguish IDH1 wild-type and mutant tumors as well as primary and secondary GBM are under development.
Magnetic resonance imaging (MRI) is a standard diagnostic approach used to provide glioma grading based on contrast enhancement, edema, as well as necrosis6; however its diagnostic power is limited as MRI alone cannot provide an accurate measure of tumor aggressiveness based on just these parameters. Efforts have been made to increase the accuracy of MRI using additional parameters such as blood volume and metabolite measurements that enhance diagnostic sensitivity7. Despite these advances, the ability to distinguish invasive tumor cells remains a challenge. Mathematical modeling and extrapolation of MRI data is being evaluated as an alternative approach to enhance the diagnostic power of MRI, including estimating the extent of invasion, that does not involve additional imaging or probes8–12. Integrating computational modeling into predictions of GBM classification (primary versus secondary) as well as mutation status (such as wild-type or IDH1 mutant) represents the next steps for these approaches.
Using a previously described model based on serial MRI data that calculates aggressiveness based on rate constants that represent proliferation (ρ) and dispersion (D)10, Swanson and colleagues compare wild-type and IDH1 mutant gliomas. By combining these two parameters, an aggressiveness ratio can be generated (ρ/D) that predicts prognosis, extent of hypoxia, and possibly primary GBM characterized by increased proliferation and secondary GBM characterized by decreased proliferation and increased diffusion13. In a well-characterized patient cohort of 178 gliomas (158 of which were GBM) from the Cancer Genome Atlas (TCGA) and two academic medical centers, the authors found that while tumors may display identical rates of proliferation, the IDH1 mutant tumors often have increased rates of dispersion. Overall, the wild-type tumors had a significantly higher aggressiveness ratio as compared to the IDH1 mutant tumors. These differences were present both in primary and secondary GBM, with a more significant difference present in secondary GBM. Taken together, the model presented by the authors distinguished between wild-type and IDH1 mutant tumors based on differences in proliferation and dispersion (Fig. 1). The authors confirm that IDH1 mutant tumors are less aggressive and possess an elevated invasion profile based on MRI. These findings can be applied to pre-operative MRI data and allow for predictions based on IDH1 status that can be integrated into the clinical management plan.
Figure 1.
Computational models predict differences in invasion and aggressiveness between wild-type and IDH1 mutant tumors. Schematic depicting differences between wild-type (left) and IHD1 mutant (right) tumors based on proliferation (ρ) and dispersion ...
The results of the computational modeling also support the go or grow hypothesis14 that predicts that rapidly proliferating cells have a lower invasion capacity while cells with a higher propensity to invade have lower proliferation. Invasion is a hallmark of GBM and remains a challenge to the treatment of the disease. While the signaling pathways responsible for invasion are being evaluated for the development of additional anti-GBM therapies, recent observations that anti-angiogenic therapies can increase the invasiveness of GBM15 suggest that invasion can also be a therapeutic resistance mechanism. How IDH1 mutation status relates to invasion remains largely unexplored. The authors' model predicts that IDH1 mutant tumors would be more invasive, and this has been confirmed by an independent group based on histology demonstrating elevated numbers of IDH1 mutant cells in invading areas16. However, the molecular pathways responsible for this possible increase in invasion and decrease in proliferation are yet to be determined and may yield additional insight into the consequence of IDH1 mutation.
The ability to accurately predict and detect IDH1 mutation status remains a priority for the neuro-oncology community. The authors' computational modeling approach is a complementary method to antibody17 and metabolite18 based detection methods. Moving forward, accurately identifying key mutations linked to tumor grade and outcome will be essential to ensure that the most accurate diagnosis and treatment plan is provided for each patient.
[1]
R. McLendon,et al.
IDH1 and IDH2 mutations in gliomas.
,
2009,
The New England journal of medicine.
[2]
S. Gabriel,et al.
Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1.
,
2010,
Cancer cell.
[3]
H. Frieboes,et al.
An integrated computational/experimental model of tumor invasion.
,
2006,
Cancer research.
[4]
Thomas D. Wu,et al.
Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis.
,
2006,
Cancer cell.
[5]
Glyn Johnson,et al.
Glioma grading: sensitivity, specificity, and predictive values of perfusion MR imaging and proton MR spectroscopic imaging compared with conventional MR imaging.
,
2003,
AJNR. American journal of neuroradiology.
[6]
H. Frieboes,et al.
Computer simulation of glioma growth and morphology
,
2007,
NeuroImage.
[7]
D. Cheresh,et al.
VEGF inhibits tumor cell invasion and mesenchymal transition through a MET/VEGFR2 complex.
,
2012,
Cancer cell.
[8]
J. Murray,et al.
A mathematical model of glioma growth: the effect of chemotherapy on spatio‐temporal growth
,
1995,
Cell proliferation.
[9]
R. E. Steiner,et al.
Clinical NMR imaging of the brain: 140 cases
,
1982
.
[10]
Albert Lai,et al.
Prognostic significance of growth kinetics in newly diagnosed glioblastomas revealed by combining serial imaging with a novel biomathematical model.
,
2009,
Cancer research.
[11]
S. Coons,et al.
Dichotomy of astrocytoma migration and proliferation
,
1996,
International journal of cancer.
[12]
H. Zentgraf,et al.
Monoclonal antibody specific for IDH1 R132H mutation
,
2009,
Acta Neuropathologica.
[13]
J. Uhm.
An Integrated Genomic Analysis of Human Glioblastoma Multiforme
,
2009
.
[14]
J. Hamada,et al.
Characterizing invading glioma cells based on IDH1-R132H and Ki-67 immunofluorescence
,
2014,
Brain Tumor Pathology.
[15]
B. Rosen,et al.
Detection of oncogenic IDH1 mutations using magnetic resonance spectroscopy of 2-hydroxyglutarate.
,
2013,
The Journal of clinical investigation.
[16]
J. Uhm.
Comprehensive genomic characterization defines human glioblastoma genes and core pathways
,
2009
.
[17]
K. Swanson,et al.
A mathematical modelling tool for predicting survival of individual patients following resection of glioblastoma: a proof of principle
,
2007,
British Journal of Cancer.
[18]
Alexander R A Anderson,et al.
Quantifying the Role of Angiogenesis in Malignant Progression of Gliomas: in Silico Modeling Integrates Imaging and Histology Nih Public Access Author Manuscript Introduction
,
2011
.