Skip to main content
Download PDF

Tumour DNA methylation markers associated with breast cancer survival: a replication study

Breast Cancer Research volume 27, Article number: 9 (2025) Cite this article

Abstract

Background

Tumour DNA methylation has been investigated as a potential marker for breast cancer survival, but findings often lack replication across studies.

Methods

This study sought to replicate previously reported associations for individual CpG sites and multi-CpG signatures using an Australian sample of 425 women with breast cancer from the Melbourne Collaborative Cohort Study (MCCS). Candidate methylation sites (N = 22) and signatures (N = 3) potentially associated with breast cancer survival were identified from five prior studies that used The Cancer Genome Atlas (TCGA) methylation dataset, which shares key characteristics with the MCCS: comparable sample size, tissue type (formalin-fixed paraffin-embedded; FFPE), technology (Illumina HumanMethylation450 array), and participant characteristics (age, ancestry, and disease subtype and severity). Cox proportional hazard regression analyses were conducted to assess associations between these markers and both breast cancer-specific survival and overall survival, adjusting for relevant participant characteristics.

Results

Our findings revealed partial replication for both individual CpG sites (9 out of 22) and multi-CpG signatures (2 out of 3). These associations were maintained after adjustment for participant characteristics and were stronger for breast cancer-specific mortality than for overall mortality. In fully-adjusted models, strong associations were observed for a CpG in PRAC2 (per standard deviation [SD], HR = 1.67, 95%CI: 1.24–2.25) and a signature based on 28 CpGs developed using elastic net (per SD, HR = 1.48, 95%CI: 1.09–2.00).

Conclusions

While further studies are needed to confirm and expand on these findings, our study suggests that DNA methylation markers hold promise for improving breast cancer prognostication.

Introduction

Breast cancer is a heterogeneous disease characterised by a wide range of alterations in the genomic, epigenomic, transcriptomic, and proteomic landscape of cancer cells [1, 2]. These alterations have the potential to meet the important need of identifying prognostic molecular markers, beyond traditional clinicopathological variables, but are seldom used in the clinic. Epigenetic changes are a ‘hallmark of cancer’ and have been widely reported in breast tumours [3]. Such changes include DNA methylation alterations occurring in tumour tissue compared with normal breast tissue as well as DNA methylation differences by key clinical variables such as hormonal receptor status, tumour stage, treatment modality or age at diagnosis [4,5,6,7]. DNA methylation is a fairly stable epigenetic modification. Modern micro-assays enable DNA methylation measurement at hundreds of thousands of sites across the genome. This makes DNA methylation a promising molecular marker for precision medicine approaches to breast cancer outcome prediction [3]. However, there is a relative lack of large genomic datasets of breast tumour tissues with DNA methylation measurements and outcome follow-up. The main publicly available resource, The Cancer Genome Atlas (TCGA) [8], has been used extensively to investigate a wide range of research questions. Among these, several studies have reported associations with breast cancer survival for individual CpG sites [9, 10], and/or developed epigenetic scores that predict outcomes, with or without replication in external small datasets [11,12,13,14,15]. To our knowledge, these findings have not received major attention and no studies have been conducted to confirm their potential clinical relevance through independent replication. Another challenge in considering the results from these studies is that they have used different methodological approaches, e.g. in terms of adjustment variables [9, 10, 13] or filtering of CpGs [11, 12, 14,15,16] tested for association with survival.

In this study, we considered several studies that have reported individual CpG sites or multi-CpG signatures in breast tumours to be associated with breast cancer survival using the TCGA DNA methylation dataset, and assessed their replication using a breast tumour dataset with similar characteristics in terms of sample size, tissue (formalin-fixed paraffin-embedded [FFPE]), technology (Illumina HumanMethylation450 array) and participant characteristics such as age, ancestry and disease subtype and severity.

Material and methods

Study sample

The Melbourne Collaborative Cohort Study (MCCS) is a cohort of 41,513 participants of white European ancestry aged 40–69 years when recruited in 1990–1994 [17]. Incident cancer cases were identified via linkage with the Victorian Cancer Registry. Breast tumours stored as FFPE (formalin-fixed paraffin-embedded) tissues were retrieved from diagnostic laboratories and reviewed by qualified pathologists [18]. Following pathologist evaluation, all FFPE tissues were sectioned at 3 µm thick. A subset of these sections underwent hematoxylin and eosin (H&E) staining for microscopic examination, while the remaining sections were left unstained. All samples were desiccated and stored at 4 °C for up to 15 years [18, 19]. Linkage with the Victorian Registry of Births, Deaths and Marriages and the National Death Index was used to ascertain deaths, including breast cancer-specific deaths [17]. Immunohistochemical staining and breast cancer subtyping were performed using the methods outlined by Blows et al. [20].

Genome‑wide DNA methylation profiling

The HumanMethylation450K (HM450K) BeadChip array was used to measure methylation in DNA extracted from FFPE tumour sections. The full protocol, detailed in Wong et al. [18], is a modification of the standard HM450K protocol for FFPE samples and incorporates an additional quality control (QC) step. Briefly, out of 474 FFPE breast tumour samples, 430 (90%) successfully passed the three mandatory QC checkpoints, which assessed DNA quality (Qubit, Life Technologies), and the presence of DNA following bisulfite conversion and restoration (qPCR with an in-house assay). The ENmix pipeline was used to pre-process and normalise the methylation data, which includes background correction, dye-bias correction, and probe-type bias correlation, and has excellent reported performance [21]. Samples and probes were excluded if the detection P-value was greater than 0.01 for over > 5% of the samples or probes respectively. The final methylation dataset consisted of 425 samples, after exclusion of 5 samples with missing data for stage and molecular subtypes define by immunohistochemistry markers (IHC-based), and 476,155 CpG sites. Beta-values, ranging from 0 to 1, were used to represent the methylation percentage at each CpG site.

Candidate studies

We searched the literature for studies that stored breast tumours as FFPE sections, measured DNA methylation using the HM450K assay, and carried out an epigenome-wide association study (EWAS) of, or developed a methylation-based risk score for either overall or breast-cancer specific survival. The following studies were retained, all of which used TCGA data:

  1. (i)

    Kim et al. [10] used 692 breast tumours from TCGA as the discovery set (cause-specific death N = 87) and 180 breast tumours (GSE72308, cause-specific death N = 29) as the replication set. From six EWASs carried out in the discovery dataset for breast cancer overall and three gene expression-based subtypes (luminal A, luminal B, basal-like), 20 associations were identified (P < 1 × 10–7, 17 unique CpGs) with little evidence of replication.

  2. (ii)

    de Almeida et al. [9] used 780 breast tumours from TCGA and pre-selected CpGs based on their correlation with gene expression, and differential methylation between tumour tissue and normal adjacent samples. The authors reported nominally significant associations (P < 0.05) for 5 CpGs using unadjusted Cox models. No validation analysis was conducted.

  3. (iii)

    Tao et al. [12] used data from 788 TCGA breast tumours; they pre-selected CpGs based on differential methylation between normal and tumour tissue and nominal associations with survival, after which they applied stepwise regression to derive a 7-CpG mortality risk score. No validation analysis was conducted.

  4. (iv)

    Du et al. [11] used a similar approach to Tao et al. [12] to pre-select CpGs (N = 1,777), from which they used penalised regression (least absolute shrinkage and selection operator, LASSO) to develop a score based on 16 CpGs. Validation was carried out in an external set of 62 women, with a reported AUC of 0.94.

  5. (v)

    Liu et al. [15] also used regularised regression (elastic net) to derive a 28-CpG signature of survival using a discovery set including 60% TCGA data. This signature was validated in a separate internal TCGA validation set (40% of the total data) and several external datasets, with AUCs ranging from 0.62 to 0.65 at various follow-up time points.

Other studies considered, but not included were: Hao et al. [16], which was similar to Du et al. [11] and Liu et al. [15], but did not provide the specific regression coefficients (weights) for each CpG site; Peng et al. [14] which focused on triple-negative tumours, hence with small number of deaths in their study and ours; Pedersen et al. [13] investigated associations with survival between two groups before and after neoadjuvant chemotherapy treatment, for which we did not have data available.

Statistical analysis

There were no overlapping CpGs across the five studies included in the analyses. As the included studies used different variables for adjustment, we carried out the analysis using four Cox proportional hazard regression models for overall and breast cancer-specific survival: Model 0: unadjusted [9, 11, 15] Model 1: adjusted for age at diagnosis and country of birth; Model 2: Model 1 + additional control for IHC-based subtypes and stage, using stratified Cox models. "Stratified" refers to the model allowing the baseline hazard rate to vary by the stratification variable while assuming a common HR for association of methylation with survival across strata [22]; Model 3: Model 2 + additional adjustment for tumour purity [10]. For overall survival models, heterogeneity between subtypes was investigated using a likelihood ratio test for the interaction between DNA methylation and subtype. To align with Kim et al. [10] for replication purposes, subgroup analyses were carried out for luminal A and luminal B subtypes. Additionally, we assessed the heterogeneity between subtypes by testing interactions between tumour subtype and DNA methylation using likelihood ratio tests. Tumour purity was estimated using the R function InfiniumPurify [23]. As the directions and effect sizes for associations in De Almeida’s study were not reported [9], we calculated them using the TCGA data. Multiple testing was considered but not explicitly accounted for. All analyses were performed using R version 4.3.0.

Results

Of 425 women included in the analysis, 168 died (follow-up time; median (interquartile range (IQR)): 15.5 (12–19.4) years, follow-up until 2019), including 66 of breast cancer (follow-up time, median (IQR): 12.1 (9.2–15.8) years, follow-up until 2015). Luminal A and B tumours represented 77% of all tumours (242 and 87, respectively), HER2-positive 7% (N = 31) and triple-negative 15% (N = 65). The estimated tumour purity of the samples was 0.63 (IQR: 0.55–0.70), similar to other studies [10, 12, 24], Table 1.

Table 1 Clinical characteristics of the study sample (Melbourne Collaborative Cohort Study, N = 425 women)
Full size table

Replication of associations at individual CpGs

For De Almeida et al. [9], 3 of 5 CpGs showed evidence of associations with breast cancer-specific survival that were consistent in direction (P ≤ 0.01, HRs per SD between 1.41 and 1.58 for HOXD9, C17orf93 and TDRD10) in Model 0 used by the authors, Table 2. For the two remaining CpGs, the HRs were consistent in direction with De Almeida et al.’s [9] study (cg04475027, HR = 1.26, P = 0.08, cg01268824, HR = 1.18, P = 0.18), Table 2. Only cg12374721 (C17orf93) retained a clear association in adjusted models (Model 3, HR = 1.67, 95%CI: 1.24–2.25, P = 0.001), Table 2. The results for overall survival were consistent with breast cancer-specific survival, with smaller HRs (Table S1).

Table 2 Associations of tumour DNA methylation with breast cancer-specific survival for 22 individual CpG sites using MCCS data (N cases = 425, N breast cancer deaths = 66)
Full size table

Most (76%) of the associations for the 17 CpGs identified by Kim et al. [10] were in the same direction in our study, with evidence of replication at P < 0.05 (Model 3, cause-specific survival) for six CpGs: cg17827670 (AHCYL2, per SD, HR = 0.66, P = 0.02), cg18703983 (KCNS3, HR = 0.77, P = 0.02), cg17735983 (MZF1, HR = 1.26, P = 0.01), cg10678486 (ELAC1, HR = 1.24, P = 0.02), cg24328142 (TSPAN15, HR = 0.77, P = 0.02), and cg09926728 (SH3PXD2A, HR = 0.80, P = 0.05), Table 2. The results for overall survival were consistent with cause-specific survival, with weaker associations, Table S1. Heterogeneity by subtype was only detected for cg17735983 (MZF1, P = 0.04). This association was also detected for luminal B tumours (HR = 1.33, P = 0.01), Table S3.

In the unadjusted model, the 16-CpG and 28-CpG scores showed a fairly strong association with cause-specific survival (HR = 1.41, P = 0.002 and HR = 1.62, P = 5.5 × 10–5), Table 3, and weaker associations with overall survival (HR = 1.26, P = 0.02 and HR = 1.34, P = 1 × 10–4), Table S2. These associations were slightly attenuated after adjustment for clinical variables with the 16-CpG (HR = 1.17, P = 0.03) and 28-CpG signatures (HR = 1.23, P = 0.007) though retained association with overall survival, Table S2, and cause-specific survival (HR = 1.40, P = 0.003 and HR = 1.16, P = 8.8 × 10–5), Table 3.

Table 3 Associations of three tumour DNA methylation-based signatures with breast cancer-specific survival using MCCS data (N cases = 425, N breast cancer deaths = 66)
Full size table

Discussion

Most studies that have investigated tumour DNA methylation markers of breast cancer survival have been limited by the small number of events and lack of independent replication. We sought to replicate findings from five previous studies that produced non-overlapping lists of prognostic CpGs. We found: (i) some evidence of replication for both individual CpGs and multi-CpG signatures; (ii) the majority of associations were stronger for cause-specific mortality, which may indicate that they were not false-positives.

The number of cause-specific deaths in TCGA is at most 101, so our study including 66 cause-specific deaths and 168 for all-cause mortality is a valuable addition. These are nevertheless small sample sizes and only allow detection of relatively large effects, in particular when considering a strict Bonferroni correction threshold for the HM450 assay of 1 × 10–7, as in Kim et al. [10]. Factors such as menopausal status and race were considered in their study [10] as these might influence the association between DNA methylation and survival, but we did not undertake these analyses as we did not have detailed data on these variables; most of our participants were post-menopausal and all women in our cohort were of White European origin. Additional subgroup analyses were not conducted, as they were anticipated to be underpowered and unlikely to yield robust conclusions. Kim et al. [10] reported minimal evidence of replication in a small external dataset, whereas we were able to corroborate their findings for at least 6 CpGs in AHCYL2, SH3PXD2A, KCNS3, MZF1, ELAC1 and TSPAN15. Most other HRs were in the same direction, which suggests additional studies with larger sample sizes might more clearly identify survival-associated CpGs.

Therefore, while further studies are required to confirm and extend these findings, it seems likely that tumour DNA methylation markers identified from epigenome-wide association studies may have a role for breast cancer prognostication.

Beyond clinical characteristics, methodological variations in pre-processing, normalization of methylation data, and subtype definition (IHC- vs gene expression-based) can further contributing to lack of replication across studies. Even when using the same dataset (TCGA), discrepancies in statistical approaches have led to different conclusions. For example, in the studies we considered, three used regularised regression but with different models: two used LASSO [11, 16], which selects a smaller, more relevant subset of CpGs by shrinking the estimates of weaker predictors to zero, while the other one used elastic net [15], which combines this selection process with a more moderate reduction in the influence of predictors that contribute less to the outcome. Another study based on TCGA used backward selection [11] of CpGs, which is considered an inferior method to derive a prediction model, and the resulting association with breast cancer survival was close to null in our study.

Using different statistical approaches and adjusting for different clinical characteristics to develop risk scores resulted in inconsistent predictive accuracy in previous studies based on the TCGA dataset. Hao et al. [16] developed mortality risk scores for various cancers, including breast cancer, and reported c-indices of 0.61–0.63 in their internal replication set. Liu et al. [15] used a similar approach to derive a methylation-based predictor of mortality, reporting AUCs ranging from 0.62 to 0.75 in the discovery set and from 0.62 to 0.65 in the validation set at various follow up times. In contrast, Du et al. [11] reported an implausibly high AUC of 0.94, raising concerns about potential overfitting when applied to the validation set. Despite the use of similar datasets, none of these studies provided the respective weights necessary for replication, limiting the ability to verify their findings. In our study, although the associations between individual CpGs and multi-CpG signatures with cause-specific survival and overall survival were somewhat attenuated after adjusting for clinical characteristics, they remained robust overall. This consistency suggests that DNA methylation plays a role in prognosis, independent of hormonal factors, tumour subtype, or stage.

The strongest association we observed was for C17orf93 also known as PRAC2 or HOXB13-AS1, which is highly expressed in prostate, rectum, colon and is a potential prostate cancer susceptibility gene [25,26,27]. To our knowledge it has not been reported in the field of breast cancer and was ranked at the bottom of the ‘OncoScore’ list used by De Almeida et al. [9]. Other genes we replicated have been reported to play a role in breast cancer progression such as HOXD9 [28, 29], TDRD10 [9, 30], SH3PXD2A [31] and TSPAN15 [32], although these studies did not involve DNA methylation data as a prognostic marker. Other genes were reported in the context of other cancer types, e.g. AHCYL2 as having a potential role in melanoma [33], colorectal [34], ovarian [35], and lung cancer [36], and KCNS3 as potentially relevant in colon and lung cancers [37]. We also found that a CpG in MZF1 had different survival associations in luminal A and luminal B subtypes, suggesting heterogeneity across subtypes. This CpG was independently associated with survival in luminal B tumours. MZF1 was implicated in the progression of several other cancers, including colorectal, cervical, liver, lung, and prostate cancer (25). Future studies investigating the functional role of DNA methylation of these genes in breast tumours would contribute to elucidate their involvement in cancer progression and establish them as promising candidates for survival stratification.

Conclusion

Our findings provide some evidence supporting the potential of DNA methylation markers to predict breast cancer outcomes. Additional larger studies are required to confirm and extend these results.

Availability of data and materials

The data that support the findings of this study can be made available upon reasonable request to the corresponding author.

References

  1. Visvader JE. Cells of origin in cancer. Nature. 2011;469(7330):314–22.

    Article  CAS  PubMed  Google Scholar 

  2. Harbeck N, Penault-Llorca F, Cortes J, Gnant M, Houssami N, Poortmans P, Ruddy K, Tsang J, Cardoso F. Breast cancer. Nat Rev Dis Primers. 2019;5(1):66.

    Article  PubMed  Google Scholar 

  3. Davalos V, Martinez-Cardus A, Esteller M. The epigenomic revolution in breast cancer: from single-gene to genome-wide next-generation approaches. Am J Pathol. 2017;187(10):2163–74.

    Article  CAS  PubMed  Google Scholar 

  4. Benevolenskaya EV, Islam ABMMK, Ahsan H, Kibriya MG, Jasmine F, Wolff B, Al-Alem U, Wiley E, Kajdacsy-Balla A, Macias V, et al. DNA methylation and hormone receptor status in breast cancer. Clin Epigenetics. 2016;8(1):17.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Klajic J, Fleischer T, Dejeux E, Edvardsen H, Warnberg F, Bukholm I, Lønning PE, Solvang H, Børresen-Dale A-L, Tost J, et al. Quantitative DNA methylation analyses reveal stage dependent DNA methylation and association to clinico-pathological factors in breast tumors. BMC Cancer. 2013;13(1):456.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Halvorsen AR, Helland A, Fleischer T, Haug KM, Grenaker Alnaes GI, Nebdal D, Syljuåsen RG, Touleimat N, Busato F, Tost J, et al. Differential DNA methylation analysis of breast cancer reveals the impact of immune signaling in radiation therapy. Int J Cancer. 2014;135(9):2085–95.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Ennour-Idrissi K, Dragic D, Durocher F, Diorio C. Epigenome-wide DNA methylation and risk of breast cancer: a systematic review. BMC Cancer. 2020;20(1):1048.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Comprehensive molecular portraits of human breast tumours. Nature 2012;490(7418):61-70.

  9. de Almeida BP, Apolónio JD, Binnie A, Castelo-Branco P. Roadmap of DNA methylation in breast cancer identifies novel prognostic biomarkers. BMC Cancer. 2019;19(1):219.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Kim DH, Binder AM, Zhou H, Jung SY. DNA methylation patterns associated with breast cancer prognosis that are specific to tumor subtype and menopausal status. Front Genet. 2023;14:1133443.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Du T, Liu B, Wang Z, Wan X, Wu Y. CpG methylation signature predicts prognosis in breast cancer. Breast Cancer Res Treat. 2019;178(3):565–72.

    Article  CAS  PubMed  Google Scholar 

  12. Tao C, Luo R, Song J, Zhang W, Ran L. A seven-DNA methylation signature as a novel prognostic biomarker in breast cancer. J Cell Biochem. 2020;121(3):2385–93.

    Article  CAS  PubMed  Google Scholar 

  13. Pedersen CA, Cao MD, Fleischer T, Rye MB, Knappskog S, Eikesdal HP, Lønning PE, Tost J, Kristensen VN, Tessem M-B, et al. DNA methylation changes in response to neoadjuvant chemotherapy are associated with breast cancer survival. Breast Cancer Res. 2022;24(1):43.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Peng Y, Shui L, Xie J, Liu S. Development and validation of a novel 15-CpG-based signature for predicting prognosis in triple-negative breast cancer. J Cell Mol Med. 2020;24(16):9378–87.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Liu XP, Hou J, Chen C, Guan L, Hu HK, Li S. A DNA methylation-based panel for the prognosis and dagnosis of patients with breast cancer and its mechanisms. Front Mol Biosci. 2020;7:118.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Hao X, Luo H, Krawczyk M, Wei W, Wang W, Wang J, Flagg K, Hou J, Zhang H, Yi S. DNA methylation markers for diagnosis and prognosis of common cancers. Proc Natl Acad Sci. 2017;114(28):7414–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Milne RL, Fletcher AS, MacInnis RJ, Hodge AM, Hopkins AH, Bassett JK, Bruinsma FJ, Lynch BM, Dugué PA, Jayasekara H, et al. Cohort profile: the melbourne collaborative cohort study (Health 2020). Int J Epidemiol. 2017;46(6):1757–1757i.

    Article  CAS  PubMed  Google Scholar 

  18. Wong EM, Joo JE, McLean CA, Baglietto L, English DR, Severi G, Hopper JL, Milne RL, FitzGerald LM, Giles GG, et al. Tools for translational epigenetic studies involving formalin-fixed paraffin-embedded human tissue: applying the Infinium HumanMethyation450 Beadchip assay to large population-based studies. BMC Res Notes. 2015;8:543.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Wong EM, Joo JE, McLean CA, Baglietto L, English DR, Severi G, Wu HC, Terry MB, Hopper JL, Milne RL, et al. Analysis of the breast cancer methylome using formalin-fixed paraffin-embedded tumour. Breast Cancer Res Treat. 2016;160(1):173–80.

    Article  CAS  PubMed  Google Scholar 

  20. Blows FM, Driver KE, Schmidt MK, Broeks A, van Leeuwen FE, Wesseling J, Cheang MC, Gelmon K, Nielsen TO, Blomqvist C, et al. Subtyping of breast cancer by immunohistochemistry to investigate a relationship between subtype and short and long term survival: a collaborative analysis of data for 10,159 cases from 12 studies. PLoS Med. 2010;7(5): e1000279.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Xu Z, Niu L, Taylor JA. The ENmix DNA methylation analysis pipeline for Illumina BeadChip and comparisons with seven other preprocessing pipelines. Clin Epigenetics. 2021;13(1):216.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Therneau T, Grambsch P. Modeling survival data: extending the cox model, vol. 48. New York: Springer; 2000.

    Book  Google Scholar 

  23. Qin Y, Feng H, Chen M, Wu H, Zheng X. InfiniumPurify: an R package for estimating and accounting for tumor purity in cancer methylation research. Genes Dis. 2018;5(1):43–5.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Jeschke J, Bizet M, Desmedt C, Calonne E, Dedeurwaerder S, Garaud S, Koch A, Larsimont D, Salgado R, Van den Eynden G, et al. DNA methylation-based immune response signature improves patient diagnosis in multiple cancers. J Clin Invest. 2017;127(8):3090–102.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Olsson P, Motegi A, Bera TK, Lee B, Pastan I. PRAC2: a new gene expressed in human prostate and prostate cancer. Prostate. 2003;56(2):123–30.

    Article  CAS  PubMed  Google Scholar 

  26. Mukund K, Syulyukina N, Ramamoorthy S, Subramaniam S. Right and left-sided colon cancers - specificity of molecular mechanisms in tumorigenesis and progression. BMC Cancer. 2020;20(1):317.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Bera TK, Egland KA, Lee B, Pastan I: Identification of novel cancer target antigens utilizing EST and genome sequence databases. The Oncogenomics Handbook 2005:31–42.

  28. Hu XC, Chu J, Zhou Y, Li CC, Zhou GJ, Jiang GQ. HOXD9 transcriptionally induced UXT facilitate breast cancer progression via epigenetic modification of RND3. Cell Signal. 2022;90: 110188.

    Article  CAS  PubMed  Google Scholar 

  29. Lin L-F, Li Y-T, Han H, Lin S-G. MicroRNA-205-5p targets the HOXD9-Snail1 axis to inhibit triple negative breast cancer cell proliferation and chemoresistance. Aging (Albany NY). 2021;13(3):3945.

    Article  CAS  PubMed  Google Scholar 

  30. Jiang Y, Liu L, Shan W, Yang Z-Q. An integrated genomic analysis of Tudor domain–containing proteins identifies PHD finger protein 20-like 1 (PHF20L1) as a candidate oncogene in breast cancer. Mol Oncol. 2016;10(2):292–302.

    Article  CAS  PubMed  Google Scholar 

  31. Wu J, Chi Y: Abstract P1–05–33: RNA in Extracellular vesicles used for the prediction of distant metastasis of breast cancer and the function of TGM2 in promoting breast cancer development. Cancer Res 2023;83(5_Supplement):P1–05.

    Article  Google Scholar 

  32. Bao Y, Wang L, Shi L, Yun F, Liu X, Chen Y, Chen C, Ren Y, Jia Y. Transcriptome profiling revealed multiple genes and ECM-receptor interaction pathways that may be associated with breast cancer. Cell Mol Biol Lett. 2019;24:1–20.

    Article  CAS  Google Scholar 

  33. Cai C, Tu J, Najarro J, Zhang R, Fan H, Zhang FQ, Li J, Xie Z, Su R, Dong L. NRAS mutant dictates AHCYL1-governed ER calcium homeostasis for melanoma tumor growth. Mol Cancer Res. 2024;22(4):386–401.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Hou P, Lin T, Meng S, Shi M, Chen F, Jiang T, Li Z, Li M, Chu S, Zheng J. Long noncoding RNA SH3PXD2A-AS1 promotes colorectal cancer progression by regulating p53-mediated gene transcription. Int J Biol Sci. 2021;17(8):1979.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Lin C-Y, Wu K-Y, Chi L-M, Tang Y-H, Huang H-J, Lai C-H, Tsai C-N, Tsai C-L. Starvation-inactivated MTOR triggers cell migration via a ULK1-SH3PXD2A/TKS5-MMP14 pathway in ovarian carcinoma. Autophagy. 2023;19(12):3151–68.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Zhou Y, Yong H, Cui W, Chu S, Li M, Li Z, Bai J, Zhang H. Long noncoding RNA SH3PXD2A-AS1 promotes NSCLC proliferation and accelerates cell cycle progression by interacting with DHX9. Cell Death Discovery. 2022;8(1):192.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Lee JH, Park JW, Byun JK, Kim HK, Ryu PD, Lee SY, Kim DY. Silencing of voltage-gated potassium channel KV9.3 inhibits proliferation in human colon and lung carcinoma cells. Oncotarget. 2015;6(10):8132–43.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Funding

MCCS cohort recruitment was funded by VicHealth and Cancer Council Victoria. The MCCS was further supported by Australian NHMRC grants 209057 and 396414 and by infrastructure provided by Cancer Council Victoria. Cases and their vital status were ascertained through the Victorian Cancer Registry (VCR) and the Australian Institute of Health and Welfare (AIHW), including the National Death Index and the Australian Cancer Database. This work was also supported by NHMRC grants GNT1011618 and GNT1074383. MCS is a NHMRC L3 Investigator Fellow (GNT2017325). P-AD is a Victoria Cancer Agency Mid-Career Fellow (MCRF22025). SL is an NHMRC EL1 Investigator Fellow (GNT2017373). EZ holds a Monash University PhD scholarship.

Author information

Authors and Affiliations

  1. Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, 246 Clayton Road, Clayton, VIC, 3168, Australia

    Elaheh Zarean, Shuai Li, Ee Ming Wong, Roger L. Milne, Graham G. Giles, Melissa C. Southey & Pierre-Antoine Dugué

  2. Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Parkville, VIC, Australia

    Shuai Li, Roger L. Milne, Graham G. Giles & Pierre-Antoine Dugué

  3. Department of Data Science and AI, Faculty of Information Technology, Monash University, Clayton, VIC, Australia

    Enes Makalic

  4. Cancer Epidemiology Division, Cancer Council Victoria, Melbourne, VIC, Australia

    Roger L. Milne, Graham G. Giles, Melissa C. Southey & Pierre-Antoine Dugué

  5. Anatomical Pathology, Alfred Health, The Alfred Hospital, Melbourne, VIC, Australia

    Catriona McLean

  6. Department of Clinical Pathology, Melbourne Medical School, The University of Melbourne, Parkville, VIC, Australia

    Melissa C. Southey

Authors
  1. Elaheh Zarean
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Shuai Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Ee Ming Wong
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Enes Makalic
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Roger L. Milne
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Graham G. Giles
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Catriona McLean
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Melissa C. Southey
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Pierre-Antoine Dugué
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

EZ: conceptualization, methodology, formal analysis, writing – original draft.SL: supervision, writing – review and editing. EMW: investigation, data curation, writing – review and editing. EM: writing – review and editing. RLM: resources, writing – review and editing. GGG: resources, funding acquisition, writing – review and editing. CM: investigation, data curation, writing – review and editing. MCS: supervision, data curation, resources, funding acquisition, writing – review and editing. PA-D: supervision, conceptualization, methodology, data curation, formal analysis, writing – original draft.

Corresponding author

Correspondence to Pierre-Antoine Dugué.

Ethics declarations

Ethics approval and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Human Research Ethics Committee of Cancer Council Victoria. Informed consent was obtained from all subjects involved in the study.

Competing Interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zarean, E., Li, S., Wong, E.M. et al. Tumour DNA methylation markers associated with breast cancer survival: a replication study. Breast Cancer Res 27, 9 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13058-024-01955-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s13058-024-01955-x

Keywords

Breast Cancer Research

ISSN: 1465-542X

Contact us