This web page was produced as an assignment for an undergraduate course at Davidson College.

Summary: Mutational Analysis Reveals the Origin and Therapy-Driven Evolution of Recurrent Glioma (Johnson et. al., 2014)


This study examined the genomic links between initial and recurrent glioma tumors, as well as the effects of treatment. Gliomas are invasive brain tumors that often reoccur, and generally, treatments that worked on the original tumor do not have the same success on the new one. The work done by Johnson et. al. offers an explanation why: while the initial tumor has one set of mutations, which physicians may be able to treat and successfully shrink, the recurrent tumor is only distantly related to the original, often having evolved from cells early in the original tumor and picking up entirely different mutations. In fact, some initial therapies, such as TMZ (temozolomide, a common chemotherapy), catalyze further mutations in cells that later give rise to the secondary tumor.

Working with sequenced exomes from various stages of initial and recurrent gliomas, Johnson et. al. concluded that on average, about 54% of 33 somatic-cell mutations existed in both the initial and recurrent gliomas tested, while the rest of the mutations were exclusive to either the initial or the recurrent tumor cells. Across the board, the initial and recurrent tumors varied in how closely related they were, with some sharing over 75% of their mutations and others sharing less than 25%; Patient 17’s initial and recurrent tumors had only one mutation in common, in the gene for IDH1, which encodes isocitrate dehydrogenase, an important enzyme for metabolism. Interestingly, this mutation in IDH1was the only one found across every patient’s tumors, leading the researchers to conclude that it has a vital role in the development of gliomas.      

In considering the evolution of recurrent tumor cells, Johnson et. al. looked at geographically different sections of the tumor to control for intratumoral heterogeneity, where different areas of the tumor have different mutations. This intratumoral heterogeneity could cause mutations in recurrent tumors to appear exclusive to the recurrent tumor, when in fact they were present in a geographically different area from the sample of the initial tumor. However, further testing showed that only a few mutations thought to be exclusive were shared due to intratumoral heterogeneity, and so this was not the explanation for the majority of genomic differences between the initial and recurrent tumors.        

Additionally, Johnson et. al. considered the effects of TMZ, a chemotherapy drug which causes apoptosis in the glioma cells. It also has distinct characteristics of the mutagenesis it causes, which was evident in several of the recurrent tumors in patients who had been treated with TMZ; 60% of the TMZ-treated patients had hypermutated secondary tumors with characteristic TMZ mutations, over 98.7%. Two particular effects of TMZ mutagenesis in glioma cells appear to be suppression of the RB tumor-supressor pathway, and hyperactivation of the Akt-mTOR signaling pathway, which leads to reduced apoptosis in cancer cells. The researchers found no evidence of these mutations existing prior to treatment with TMZ.

Overall, I found the paper informative and fairly easy to follow. The description of and the results of testing intratumoral hetogeneity lacked clarity, but with a second reading I understood the point. Most of the figures lacked significance statistics, which I would have liked to see on the bar graphs. The figures, though daunting at first look, proved fairly simple to comprehend after reading the paper (and the descriptions). Figure 1, Panel A and Figure 2 I thought were particularly good representations of the data, understandable and visually clear.


(All figures below courtesy of Johnson et. al., 2014)

Figure 1:

  • Panel B shows Patient 27, who has more mutations private to the recurrence that are present in a higher fraction of cells than in the initial tumor; Patient 27 also has a mutation in IDH1 shared by all cells, and a few other shared that occur in all cells of the recurrence but only between 70-90% of the initial tumor.

  • Panel C is Patient 17, who has only one shared mutation (IDH1) between the initial and recurrent tumors; again, far more mutations exist in the recurrent than in the initial tumor, and these vary in frequency within the tumors themselves.

  • Panel D relates the mutations of Patient 18, who has a huge number of mutations in the recurrent tumor relative to the initial tumor. Again, IDH1 is shared by both, and Patient 18 has a few other shared mutations that are in high frequency for both recurrent and initial tumors. Panel D also includes the sequence for the BRAF V600E (a gene related to signaling for cell growth) with mutations that were sampled from geographically different locations in initial and recurrent tumors, as well as the normal sequence.

Figure 2: shows us MRI scans of two different patients at various stages of recurrence, denoting how much time passed between recurrences, and when TMZ chemotherapy occurred. It also includes a phylogenetic tree of tumors, denoting the mutations that separate the initial tumor and the recurrences.

Figure 3:


Johnson, BE et. al. Muational Analysis Reveals the Origin and Therapy-Driven Evolution of Recurrent Glioma. Science, 2014. 343:189-193.

Return to Home

Mouse Phenome Database

Total Synthesis of a Functional Designer Eukaryotic Chromosome


Genomics Page
Biology Home Page

Email Questions or Comments.

© Copyright 2014 Department of Biology, Davidson College, Davidson, NC 28035