Exploring the role of tumour microenvironment in the behaviour of breast ductal carcinoma in situ

Background: Breast ductal carcinoma in situ (DCIS) accounts for more than 20% of screen-detected breast cancer. To date, predictors of DCIS behaviour are not well defined. Tumour microenvironment (TME) plays a role in tumour progression; therefore, the aim of the current study was to characterise a large cohort of DCIS with long term follow-up data and to decipher the role of TME in DCIS outcome. Patients and methods: This study includes 1,488 cases comprises pure DCIS (n=1,249) and DCIS associated with synchronous invasive carcinoma (n=239). Clinicopathological and outcome data was collected. All cases were histologically reviewed, and representative tumour blocks were retrieved. Tissue microarrays (TMAs) were constructed followed by molecular characterisation for oestrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2) and Ki67. Tumour infiltrating lymphocytes (TILs) were assessed using novel method and their association with local recurrence was investigated. Mining for biomarkers that regulate remodelling and degradation of the surrounding stroma and have potential prognostic values in DCIS was performed. This includes prolyl-4-hydroxylase Α subunit 2 (P4HA2), cathepsin A (CTSA), cathepsin V (CTSV), legumain (LGMN) and Collagen (XI) alpha-1 chain (COL11A1). The expression of these biomarkers was evaluated in the TMA using immunohistochemistry and their prognostic value was analysed. In addition, image analysis assessment of the geometric characterisation of collagen surrounding DCIS was performed and collagen prognostic index (CPI) was developed. Moreover, the role of hypoxia and centrosome amplification (CA) has been investigated. Results: There was a significant increase in the rate of breast conserving surgery (BCS) and the use of radiotherapy (RT) in DCIS management with time, along with a decline in the rate of re- excisions. Large size, high grade, and omission of adjuvant RT were independent predictors of the development of ipsilateral local recurrence. The highest proportion (60%) of cases were of Luminal A subtype, and the lowest attribution (11%) was for HER2 positive tumours. HER2 positive and Triple Negative subtypes were of larger tumour size, higher grade, and more frequently treated with mastectomy. Dense TILs were associated with adverse clinicopathological parameters and high recurrence rate. DCIS associated with synchronous invasive carcinoma showed higher TILs density than pure DCIS. The assessment of the selected biomarker panel revealed a significant difference in their protein expression between pure DCIS and those associated with invasion. An independent association between high expression of P4HA2, CTSA, CTSV, LGMN and COL11A1 with the development of local recurrence was observed. DCIS with high CPI and harbouring higher structural and/or numerical centrosomal aberrations were associated poor outcome. Conclusions: This study shows that TME plays a key role in DCIS behaviour and outcome. In addition to the conventional clinicopathological variables, dense TILs, high expression of P4HA2, CTSA, CTSV, LGMN, COL11A1, high CPI and centrosomal aberrations were independent predictors of DCIS outcome. These variables could potentially be used in combination to refine the prognostic stratification of DCIS patients in routine practice