Age and Grade as a Function of Decision Score in Women Diagnosed with DCIS

Abstract

Proper selection of patients with ductal carcinoma in situ (DCIS) for adjuvant radiation therapy (RT) after breast conserving surgery (BCS) remains a challenge. Clinico-pathologic factors are routinely used to grossly stratify women into low, intermediate and high-risk groups for treatment. However, EBCTCG data has shown that clinico-pathologic provides limited utility for risk stratification. A new predictive and prognostic DCIS biosignature (PreludeDx, Laguna Hills, CA) validated in multiple cohorts, assessed discordance of Decision Score (DS) with a variety of clinical factors used to make treatment decisions. De-identified datasets from DCIS studies performed in Sweden (n = 1,046) and the US (n = 455 and 296) were combined into a single dataset, totaling 1,553 women. Women with positive surgical margins (n = 204) were excluded. Women were grouped by age, tumor size, nuclear grade and RTOG 9804-like criteria (screen detected, extent < 2.5 cm, grade < 3 and no ink on tumor, since distance to margin status was not available for some of the cases). Differences in contingency tables were assessed using Chi-Square analysis with an alpha level of 0.05. Using 9804-like criteria (low and intermediate grade, size <2.5 cm, negative margins), 44% of women presenting as 9804 low risk were reclassified to elevated risk by DS (P < .0001). In women <50, 42% were at low risk by DS, while 58% of women were at elevated 10-yr total IBE risk. In women over 70, 25% were low risk by DS and 75% were at elevated risk (P < .0001). In women with low to intermediate grade, 53% were at low risk by DS, while 47% of women were at elevated 10-yr total IBE risk. In women with high grade, 34% were low risk by DS and 66% were at elevated risk (P < .0001). In this analysis, DS for women <50 demonstrated a bi-modal distribution to the very low and very high biological risk of DCIS recurrence or development of invasive breast cancer over 10 years. This differs from the opinion that the majority of women <50 are at high risk. Additionally, there is a significant subset of women >70 years old treated with BCS who remain at elevated risk of recurrence or development of invasive disease and may require the addition of RT to effectively manage risk. Lastly, DS reclassified half of women with low risk by 9804-like criteria to an elevated risk profile, potentially identifying a group of women who would be under-treated.Abstract 2034; TableGroupCriteriaLow Risk%Elevated Risk%TotalAge (P<.0001)Under 5015342%21458%36750-7038740%58160%968Over 705525%16375%218Tumor Size (P<.0001)< = 20mm43940%65160%1090>20mm5326%15474%207Nuclear Grade (P<.0001)Low-int grade40953%36347%772High grade16434%32166%485RTOG9804-like (P<.0001)Good risk27456%21644%490Not good risk23932%50168%740 Open table in a new tab