# Influence of Cancerization of Lobules in Ductal Carcinoma In Situ of the Breast on the Pathological Outcomes in Mastectomy Specimens

## TL;DR

This study explores how cancerization of lobules in breast cancer relates to disease severity but does not predict invasive cancer.

## Contribution

The study identifies cancerization of lobules as a marker of higher disease burden in DCIS but not a predictor of invasion.

## Key findings

- Cancerization of lobules is associated with higher DCIS grade, central necrosis, and extensive intraductal component.
- Cancerization of lobules is not linked to an increased likelihood of invasive carcinoma.
- Most patients with cancerization of lobules did not experience local recurrence or metastasis.

## Abstract

Cancerization of lobules (COL) occurs when the breast lobular acini are involved by ductal carcinoma in situ (DCIS), but its role in predicting invasive carcinoma remains unclear. Our study evaluated whether COL is linked to worse pathological outcomes in mastectomy specimens. We analyzed 171 cases of DCIS and found that COL is associated with factors indicating a higher disease burden, such as higher DCIS grade, central (“comedo”) necrosis, and extensive intraductal component (EIC). However, the presence of COL was not associated with an increased likelihood of detecting an invasive component. These findings suggest that COL may reflect a greater extent of disease but is not a direct predictor of invasion.

Cancerization of lobules (COL) is defined as the involvement of lobular acini by ductal carcinoma in situ (DCIS). Whether it represents a morphological variation in DCIS or a secondary extension of DCIS into lobules is debatable. The relation between COL and the probability of invasion is conflicting among different studies. We assessed if COL is a predictor of adverse pathological outcomes in mastectomy specimens. We reviewed the clinicopathological data of patients who underwent partial or total mastectomy for DCIS during a 3-year period (January 2015 until December 2017). Pathological parameters and follow-up data were collected. Whole-tissue hematoxylin and eosin (H&E) slides were reviewed and re-evaluated for COL. Cases with COL were stained immunohistochemically for E-cadherin and p120 to confirm the ductal phenotype of the neoplasms. In total, 171 mastectomies were identified including 65 specimens with pure DCIS and 106 specimens with DCIS with invasive carcinoma. COL was identified in 73 specimens. COL was significantly associated with adverse pathological outcomes including higher DCIS nuclear grade (p-value = 0.006), central (expansive “comedo”) necrosis (p-value = 0.008), presence of DCIS within or less than 2 mm from the surgical resection margin(s) (p-value = 0.004), higher percentage of blocks/slides with DCIS (p-value < 0.001), and extensive intraductal component (EIC) (applicable in cases with invasion) (p-value < 0.001). Invasion was seen in approximately two-thirds of the cases regardless of the presence of COL, with no statistical significance. Ninety-eight patients achieved 60 months of follow-up, of which only one patient developed local DCIS recurrence and had COL and EIC. Four other patients developed metastatic disease related to the invasive component. While other studies have previously hypothesized that COL may be associated with a worse pathological outcome at mastectomy, our results show that it may indeed be a measure of a higher disease burden representing EIC; however, it is not associated with an increased risk of detecting invasive carcinoma.