Potent predictive CpG signature for temozolomide response in non-glioma-CpG island methylator phenotype glioblastomas with methylated MGMT promoter

Abstract

Aim: We aimed to identify potent CpG signatures predicting temozolomide (TMZ) response in glioblastomas (GBMs) that do not have the glioma-CpG island methylator phenotype (G-CIMP) but have a methylated promoter of MGMT (meMGMT). Materials & methods: Different datasets of non-G-CIMP meMGMT GBMs with molecular and clinical data were analyzed. Results: A panel of 77 TMZ efficacy-related CpGs and a seven-CpG risk signature were identified and validated for distinguishing differential outcomes to radiotherapy plus TMZ versus radiotherapy alone in non-G-CIMP meMGMT GBMs. An integrated classification scheme was also proposed for refining a MGMT-based TMZ-guiding approach in all G-CIMP-GBMs. Conclusion: The CpG signatures may serve as promising predictive biomarker candidates for guiding optimal TMZ usage in non-G-CIMP meMGMT GBMs.

Plain language summary

Glioblastomas that do not have the glioma-CpG island methylator phenotype (G-CIMP) but have a methylated promoter of the MGMT gene (meMGMT) show considerable variability in their response to temozolomide (TMZ). Powerful biomarkers that provide predictive information on optimal TMZ decision-making can be clinically useful. This study has identified and validated a panel of 77 TMZ efficacy-related CpGs and a seven-CpG risk signature for predicting TMZ usage in non-G-CIMP meMGMT glioblastomas. An integrated classification scheme is proposed for refining a MGMT-based TMZ-guiding approach in non-G-CIMP glioblastomas.

Keywords::

• DNA methylation
• glioblastoma
• methylated MGMT
• predictive biomarker
• temozolomide

Supplementary data

To view the supplementary data that accompany this paper please visit the journal website at: www.tandfonline.com/doi/suppl/10.2217/epi-2022-0344

Conception and design of the study: W Lin, A Yin and Y He. Provision of study material or patients: A Etcheverry, A Yin and B Song. Acquisition and assembly of data: J Wang, M Zhang, Y Liu, Y Yao and Y Ji. Project administration, Software, methodology: J Wang, M Zhang, Y Liu and K Chen. Analysis and interpretation of results: K Chen, B Song, A Yin and Y He. Manuscript writing and final approval of manuscript: all authors.

Acknowledgments

The authors gratefully acknowledge M Aubry, J Mosser and the Tumor Banks from Angers and Rennes for their work in sample collection and processing. They also gratefully thank those who were willing to share valuable scientific data (R Lai and J Barnholtz-Sloan). A Yin wishes to thank his daughter (J Yin) and his wife (Y Dong) for their great support. The results published here are in part based upon data generated by The Cancer Genome Atlas and the research teams mentioned in this study.

Financial & competing interests disclosure

This work was partially funded by grants from National Natural Science Foundation of China (no. 81402049, 81802486), Shandong Province Natural Science Foundation (no. ZR2020QH0233), and by grants from the Brittany Region (France) and the FEDER (Europe). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Ethical conduct of research

The authors state that they have obtained verbal and written informed consent from the patients for the inclusion of their medical and treatment history within this report. Informed consent was obtained for all participants from the Neurosurgery Departments of Rennes and Angers University Hospitals.

An integrated classification scheme combining the present meMGMT-specific risk signature and an unpublished Supplementary data