Genetic determinants of fatigue up to 2 years after radiotherapy in prostate cancer patients, 2026, Heumann et al

[Nightsong]

Abstract:
Fatigue is a common symptom of cancer patients, which can impair quality of life even years after treatment. Little is known about genetic determinants of fatigue, especially in prostate cancer (PCa). This study aims to identify SNPs associated with long-term fatigue in a prospective cohort of PCa patients. A genome-wide association study was conducted in non-metastatic PCa patients treated with radiotherapy in 7 European countries and the USA. A total of 1,381 men recorded fatigue using the EORTC QLQ-C30 and 877 men additionally completed the Multidimensional Fatigue Inventory (MFI) up to two years post-radiotherapy. Clinically important fatigue is defined for the EORTC QLQ-C30 based on the proposed threshold as scores ≥39 and for the MFI as scores ≥75th percentile in the general German male population aged ≥60 years.

Regression models adjusted for demographic, disease- and treatment-specific characteristics examine SNPs associated with clinically important fatigue. Differential gene expressions are explored using expression quantitative trait analysis. rs142212041 located in the ACTR3/CBWD2 gene region is significantly associated (P = 3×10-8) with long-term physical fatigue in 643 men without physical fatigue pre-radiotherapy. Several potential risk loci (P < 5×10-6) are identified for distinct fatigue phenotypes. Gene expression differences are observed for ACTR3 and CBWD2, although not significant after correction for multiple testing. The results emphasise the multidimensionality of fatigue and suggest a plausible biological mechanism in fatigue pathophysiology, previously discussed for myalgic encephalomyelitis/chronic fatigue syndrome, which might be a potential intervention target.

Link | PDF (Nature Communications, April 2026, open access)

 

[Nightsong]

Shared genetic architecture between fatigue phenotypes in the cancer population studied, and EHR-derived malaise and fatigue and ME/CFS in white British participants of the UK Biobank was investigated using publicly available GWAS data54,55.

...

Local genetic correlations between incident physical fatigue and EHR-derived malaise and fatigue and chronic fatigue syndrome (ME/CFS) were examined. Non-associated loci on chromosome 2 were filtered using univariate analysis and a Bonferroni-corrected significance level threshold of α = 0.05/220 = 0.00023. Bivariate tests were performed for the remaining loci. For fatigue and malaise (3413 cases, 403,789 controls), bivariate tests were calculated for 158 loci on chromosome 2 with a mean rg of 0.03. 19 loci showed a Bonferroni-corrected significant bivariate rg, including the locus around the index SNP rs142212041 (rg = −0.09, P = 0.0001). For ME/CFS (590 cases, 403,791 controls), bivariate tests were calculated for 158 loci, 18 of which showed a Bonferroni-corrected significant bivariate local genetic correlation.

Across loci on chromosome 2, mean rg was 0.03. For the locus harbouring the index SNP, a highly statistically significant local genetic correlation between incident physical fatigue and ME/CFS was observed (rg = 0.18, P = 5.79 × 10−13). In the locus 200 kb around ACTR3, the P-values of the genome-wide association analysis and cis-eQTLs correlated most strongly for skeletal muscle tissue (r = 0.79, P = 0.02; Supplementary Fig. 7) and whole blood (r = 0.63, P = 0.10; Supplementary Fig. 8). No statistically significant associations were observed for pituitary gland (r = 0.47, P = 0.21; Supplementary Fig. 9) and hippocampal tissue (r = −0.17, P = 0.89; Supplementary Fig. 10). Across tissues, no evidence was found that the trait-significant variant rs142212041 was associated with ACTR3 expression levels.