Meta-GWAS identifies the heritability of acute radiation-induced toxicities in head and neck cancer

Radiogenomics Consortium, Elnaz Naderi*, Line M.H. Schack, Ceilidh Welsh, Adelene Y.L. Sim, Miguel E. Aguado-Barrera, Tom Dudding, Holly Summersgil, Laura Martínez-Calvo, Enya H.W. Ong, Yasmin Odding, Ana Varela-Pazos, Roel J.H.M. Steenbakkers, Anne P.G. Crijns, Rajesh Jena, Miranda Pring, Joe Dennis, Ramón Lobato-Busto, Jan Alsner, Andy NessChristopher Nutting, David J. Thomson, Antonio Gómez-Caamaño, Jesper G. Eriksen, Steve J. Thomas, Amy M. Bates, Jens Overgaard, Luis M. Cascallar-Caneda, Fréderic Duprez, Gillian C. Barnett, Leila Dorling, Melvin L.K. Chua, Ana Vega, Catharine M.L. West, Johannes A. Langendijk, Christian Nicolaj Andreassen, Behrooz Z. Alizadeh

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

3 Downloads (Pure)

Abstract

Background and purpose: We aimed to the genetic components and susceptibility variants associated with acute radiation-induced toxicities (RITs) in patients with head and neck cancer (HNC).

Materials and methods: We performed the largest meta-GWAS of seven European cohorts (n = 4,042). Patients were scored weekly during radiotherapy for acute RITs including dysphagia, mucositis, and xerostomia. We analyzed the effect of variants on the average burden (measured as area under curve, AUC) per each RIT, and standardized total average acute toxicity (STATacute) score using a multivariate linear regression. We tested suggestive variants (p < 1.0x10-5) in discovery set (three cohorts; n = 2,640) in a replication set (four cohorts; n = 1,402). We meta-analysed all cohorts to calculate RITs specific SNP-based heritability, and effect of polygenic risk scores (PRSs), and genetic correlations among RITS.

Results: From 393 suggestive SNPs identified in discovery set; 37 were nominally significant (preplication < 0.05) in replication set, but none reached genome-wide significance (pcombined < 5 × 10-8). In-silico functional analyses identified “3′-5'-exoribonuclease activity” (FDR = 1.6e-10) for dysphagia, “inositol phosphate-mediated signalling” for mucositis (FDR = 2.20e-09), and “drug catabolic process” for STATacute (FDR = 3.57e-12) as the most enriched pathways by the RIT specific suggestive genes. The SNP-based heritability (±standard error) was 29 ± 0.08 % for dysphagia, 9 ± 0.12 % (mucositis) and 27 ± 0.09 % (STATacute). Positive genetic correlation was rg = 0.65 (p = 0.048) between dysphagia and STATacute. PRSs explained limited variation of dysphagia (3 %), mucositis (2.5 %), and STATacute (0.4 %).

Conclusion: In HNC patients, acute RITs are modestly heritable, sharing 10 % genetic susceptibility, when PRS explains < 3 % of their variance. We identified numerus suggestive SNPs, which remain to be replicated in larger studies.

Original languageEnglish
Pages (from-to)138-148
Number of pages11
JournalRadiotherapy and Oncology
Volume176
DOIs
Publication statusPublished - Nov-2022

Keywords

  • Head and neck cancer
  • Meta-GWAS
  • Polygenic risk score
  • Radiation-induced acute toxicity
  • SNP-based heritability

Cite this