Investigating chromosomal instability in long-term survivors with glioblastoma and grade 4 astrocytoma

Abstract

Background: Only a small group of patients with glioblastoma multiforme (GBM) survives more than 36 months, so-called long-term survivors. Recent studies have shown that chromosomal instability (CIN) plays a prognostic and predictive role among different cancer types. Here, we compared histological (chromosome missegregation) and bioinformatic metrics (CIN signatures) of CIN in tumors of GBM typical survivors (≤36 months overall survival), GBM long-term survivors and isocitrate dehydrogenase (IDH)-mutant grade 4 astrocytomas.

Methods: Tumor sections of all gliomas were examined for anaphases and chromosome missegregation. Further CIN signature activity analysis in the The Cancer Genome Atlas (TCGA)-GBM cohort was performed.

Results: Our data show that chromosome missegregation is pervasive in high grade gliomas and is not different between the 3 groups. We find only limited evidence of altered CIN levels in tumors of GBM long-term survivors relative to the other groups, since a significant depletion in CIN signature 11 relative to GBM typical survivors was the only alteration detected. In contrast, within IDH-mutant grade 4 astrocytomas we detected a significant enrichment of CIN signature 5 and 10 activities and a depletion of CIN signature 1 activity relative to tumors of GBM typical survivors.

Conclusions: Our data suggest that CIN is pervasive in high grade gliomas, however this is unlikely to be a major contributor to the phenomenon of long-term survivorship in GBM. Nevertheless, further evaluation of specific types of CIN (signatures) could have prognostic value in patients suffering from grade 4 gliomas.

Keywords: IDH; astrocytoma; chromosomal instability; genome instability; glioblastoma multiforme; high grade glioma.

Figure 1

Examples of H&E-stained samples of normal anaphases, lagging chromosomes (arrows) and chromatin bridges (arrows). Scale bar, 5 μm.

Figure 2

(A) Kaplan-Meier curve showing the difference in overall survival between GBM-TS (n = 10, grey), GBM-LTS (n =5, green) and IDH-MUT grade 4 astrocytoma patients (n = 5, blue) (log-rank test, P = 0.0005). (B) Barplots comparing the patients’ age in years. Results represent mean ± SD (one-way ANOVA with Dunnett’s multiple comparisons test).

Figure 3

(A) Number of anaphases per mm² of tumor tissue. Results represent mean ± SD (one-way ANOVA with Dunnett’s multiple comparisons test). (B) Chromosome missegregation frequency of cells in anaphase. Results represent mean ± SD (one-way ANOVA with Dunnett’s multiple comparisons test). (C) Number of cells in anaphase exhibiting chromosome missegregation per mm² of tumor tissue. Results represent mean ± SD (one-way ANOVA with Dunnett’s multiple comparisons test). (D) Kaplan-Meier curve showing the difference in overall survival between patients with a chromosome missegregation frequency below (n = 10, black) and above the median (n =10, grey) (log-rank test, P = 0.030).

Figure 4

(A) Comparison of wGII scores between GBM-TS (n = 346, grey), GBM-LTS (n =40, green) and IDH-MUT grade 4 astrocytoma patients (n = 9, blue) (Kruskal-Wallis test with Dunn’s post hoc test, corrected for multiple testing by using the Benjamini-Hochberg method, P > 0.05). (B) Relative activities of the 4 significantly altered CIN signatures (GBM-TS, n = 354; GBM-LTS, n = 41; IDH-MUT grade 4 astrocytoma, n = 10). Boxplots summarize rescaled CIN signature activities (Kruskal-Wallis test with Dunn’s post hoc test, corrected for multiple testing by using the Benjamini-Hochberg method, FDR < 0.05). Boxes represent the interquartile range with the median depicted as a bold line and the whiskers extend to 1.5 times the interquartile range from the lower or upper quartile, with datapoints outside this interval being considered as potential outliers.