Abstract Background: Ductal carcinoma in situ (DCIS) is a non-invasive precursor lesion confined to the duct of the breast, and is diagnosed in ∼54,000 women in the USA annually. Most are treated with breast-conserving surgery (BCS) with or without breast radiotherapy (RT). Randomized control trials found that the 10-year cumulative incidence of invasive local recurrence (ILR) is estimated to be up to 14% for women treated with BCS alone and 6.5% in women who also received RT. The development of ILR necessitates significant additional treatment, since it is associated with a 7-17 fold increased subsequence risk of breast cancer mortality. There remains a critical need to identify prognostic and predictive markers for DCIS which would allow for patient-specific targeted strategies for prevention of invasive breast cancer (BC) (in aggressive cases) and the reduction of overtreatment (in indolent cases). Invasive BC has long been recognized as a heterogeneous disease consisting of at least four intrinsic subtypes (luminal A and B, basal-like, HER2-enriched) which are distinct at the molecular level, have different treatment options, and have differential patient prognoses. Several efforts have established that DCIS samples also possess these subtypes, albeit with differences in expression of some marker genes and with significant differences in the relative frequencies of the four intrinsic subtypes e.g. the HER2-enriched subtype is estimated to be twice as frequent in DCIS than invasive BC. Hypotheses/Objectives: (i) To identify the intrinsic subtypes across our large Ontario population-based cohort of pure DCIS treated with BCS and/or RT; (ii) To compare the frequency of subtypes in our DCIS cohort against previous DCIS and IDC cohorts; (iii) To evaluate the prognostic capacity of the intrinsic subtypes with respect to ILRs; and (iv) To identify associations between intrinsic DCIS subtypes and clinicopathological factors including age and grade. Methods: Using a custom Nanostring nCounter panel, we evaluated the RNA expression of the 50 genes from the PAM50 intrinsic subtyping tool across N=976 pure DCIS samples, and performed unsupervised and supervised analysis to classify intrinsic subtype and to identify molecular components that differ from the canonical patterns found in invasive BC. Cox-based regression is used to estimate hazard ratios (HRs) for other variables with respect to ILR-free survival. Results: Our cohort’s increased size (in comparison to 8 previous studies) and the fact that it is close to a true population-based study provide for more robust estimates of intrinsic subtype frequencies. Within luminal samples, we see an increase in luminal B (18%) and a decrease in luminal A (40%) compared to previous findings. Surprisingly, we observe a significant reduction in the HER2-enriched subtype (16.6%) below previous reports (∼25%), although this remains elevated from their observed frequency in invasive BC (∼11.5%). We concur with previous findings that the basal-like subtype is molecularly more heterogeneous than their invasive counterparts. ILR was significantly elevated only in the HER2-enriched subtype (HR=1.69 p0.036; baseline luminal A). The prognostic capacity of some clinicopathological factors were strongly dependent on subtype e.g. reduced HR for age only in the basal-like. Adjuvant RT reduced ILR across the entire cohort (HR=0.70; p0.055), however it was most beneficial in HER2-enriched DCIS (HR=0.44; p0.035). Significance: Since the molecular differences between subtype are so large, they must be considered when designing studies. Our results provide more robust, closer to population-based estimates for several important variables including intrinsic subtype and clinicopathological factors. Citation Format: M. T. Hallett, S. Nofech-Mozes, E. J. Mucaki, S. Trebinjac, Y. Amemiya, A. Seth, E. Hahn, L. Paszat, E. Rakovitch. The intrinsic subtypes in a large cohort of pure DCIS - frequencies and associations with local recurrence abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-08-13.
Building similarity graph...
Analyzing shared references across papers
Loading...
M. T. Hallett
Sharon Nofech‐Mozes
E. J. Mucaki
Clinical Cancer Research
Western University
Sunnybrook Health Science Centre
Princess Margaret Cancer Centre
Building similarity graph...
Analyzing shared references across papers
Loading...
Hallett et al. (Tue,) studied this question.
www.synapsesocial.com/papers/6996a8efecb39a600b3f0396 — DOI: https://doi.org/10.1158/1557-3265.sabcs25-ps4-08-13
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: