Abstract 299P
Background
Radiotherapy (RT) is used in the treatment of >50% of cancer patients, but RT-resistance is common. We have recently identified RT-associated genomic small deletion signatures, marking tumors that have developed resistance. Here, we used longitudinal molecular profiles to explore non-genetic mechanisms of RT-resistance.
Methods
We have analyzed the multi-omic Glioma Longitudinal Analysis dataset, integrating DNA and RNA sequencing datasets from primary and post-treatment gliomas.
Results
We classified tumors as harboring temozolomide-associated hypermutation based on mutational burden increase (HM), and RT-associated signature based on small deletion burden increase (RTscars). By deconvoluting RNA profiles into cell state fractions, we observed an increase of a proliferating stem-like (PSL) cell state in 50% of patients with a RTscars+ signature which was significantly higher than RTscars–/HM– patients (P=8e-04, Fisher’s exact). PSL cell state increase associated with worse overall survival outcomes (P=1.5e-03, log-rank). We observed a significant correlation between the expression change of E2F cell cycle regulator genes (R=0.91, P=6.7e-04, Pearson) and EZH2 (R=0.81, P=8.3e-03) with the increase in the PSL cell state, nominating the E2F/EZH2-pathway as regulator of the PSL cell state. Furthermore, the RTscars signature was associated with an increase in frameshift neoantigens and significantly higher neoantigen burden at recurrence (P=2.4e-02, Kruskal-Wallis). This was accompanied by a significantly higher post-treatment T-cell fraction in RTscars+ tumors (P=8.2e-04, Wilcox), suggesting an increased T-cell infiltration in these patients.
Conclusions
Our analyses revealed a longitudinal increase in the proliferating stem-like cell state associated with RT-resistance and nominates the cell cycle pathway as an actionable target. The RT-associated deletion signature (RTscars) correlated with increased frameshift-neoantigens and T-cell fractions at recurrence, suggesting a potential benefit of a combinatorial immune-targeted therapy in this specific patient population.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.