Abstract
Tumor stroma properties distinguish ductal carcinoma in situ (DCIS) lesions of patients who relapse.
Major Finding: Tumor stroma properties distinguish ductal carcinoma in situ (DCIS) lesions of patients who relapse.
Concept: The transition from DCIS to invasive breast cancer involves coordinated changes in stromal structure.
Impact: This study suggests a protective role for myoepithelial disruption against invasive breast cancer.
Accounting for 20% of new breast cancer diagnoses, ductal carcinoma in situ (DCIS) is a preinvasive breast malignancy localized within the breast duct and separated from the stroma by a layer of myoepithelial cells. Although many DCIS lesions may remain indolent, up to half of patients with DCIS develop invasive breast cancer (IBC) within 10 years if untreated. To understand factors that underlie the progression from DCIS to IBC, as well as features distinguishing DCIS lesions from patients who do and do not progress, Risom, Glass, Averbukh, and colleagues performed a high-dimensional analysis of tissue from a longitudinal cohort of patients with DCIS, consisting of normal breast tissue (n = 9), primary DCIS (n = 58), and IBC from subsequent relapse (n = 12). Multiplex ion beam imaging by time of flight (MIBI-TOF) enabled antibody-mediated visualization of 37 markers to distinguish epithelial, stromal, and immune cells and to demarcate epithelial, myoepithelial, and stromal compartments within the DCIS tumor microenvironment (TME). Assessment of spatial distribution and compartment morphology highlighted three clusters of TME parameters that uniquely enriched for normal, DCIS, or IBC samples, in addition to a fourth cluster specifically depleted in DCIS. Comparison of parameters within normal-, DCIS-, and IBC-enriched clusters suggested a coordinated shift in tumor, myoepithelial, and stromal cell function across disease stages. DCIS tumors displayed enhanced myoepithelial proliferation, as well as a distinct increase in stromal desmoplasia, a property not seen in IBC. Training a random forest classifier to predict DCIS lesions that will progress to IBC revealed myoepithelial phenotype and spatial distribution of immune cells as primary features that distinguished progressors from nonprogressors. Notably, whereas the myoepithelium of progressor DCIS resembled that of normal, nonprogressor DCIS exhibited myoepithelial breakdown, as evidenced by a thinner, less continuous structure. In summary, this comprehensive spatial imaging atlas of DCIS progression highlights coordinated changes in the TME and implicates a protective role of myoepithelial disruption in preventing invasive disease. ■
Note:Research Watch is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details. For more Research Watch, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/CDNews.