Milea Timbergen

105 β-catenin and DNMT1, as reported by Song et al., was investigated 15 . Despite the use of similar immunoprecipitation conditions and an identical cell line (HCT116) DNMT1 was not pulled-down with wild-type β-catenin or any of the β-catenin mutants tested. No distinct cluster patterns were seen based on tumour size when a hierarchical clustering was performed using all 354 DMRs between S45F and T41A samples. Interestingly, when only the smallest and largest tumours were considered, DNA methylation patterns almost perfectly discriminated the two groups. Although tumour size depends on the measuring methods (radiological imaging or the dimensions of freshly resected surgical specimen) and can therefore be fairly subjective, our data suggest larger tumours display a different methylation pattern compared to smaller tumours. The phenomena that methylation patterns differ between tumour sizes has previously been described by Nishida et al., and may suggest that stepwise progression of methylation alterations may take place during the development of tumours 40 . This study has several limitations. The first limitation is the relatively small DTF sample size. DTF samples are challenging to obtain due to the rarity of these tumours and the current tendency to use an active surveillance approach instead of surgical resection 41 . Furthermore, obtaining paired control tissue such as fascia from which desmoids are believed to arise, is challenging as it would require an additional resection of adjacent fascia next to the tumour site. Due to the retrospective nature of the current study, we were not able to obtain paired control tissue samples. Furthermore, we opted not to include WT DTF samples as control as they comprise a heterogeneous group which often contains rare CTNNB1 mutations or alterations in other genes 5 . Future research should focus on the integrated genomic and molecular characterization of DTF samples and include appropriate control tissues to further delineate and understand the biological mechanisms and epigenetic changes involved in the pathogenesis of DTF. The functional significance of the observed differential gene body methylation of CCDC6 in T41A and S45F DTF should be further validated and investigated. Exploration of the dynamic changes in DNA methylation patterns and their consequences for gene expression from tumour onset to tumour progression and/or regression is of interest too and may provide an explanation for the different clinical behaviours of DTF tumours. 4