Abstract P1-11-02: A comparison of models (physician, the Van Nuys prognostic index, the Memorial-Sloan-Kettering Cancer Center DCIS nomogram)to predict ipsilateral breast events in patients with ductal carcinoma in situ (DCIS) of the breast after breast-conserving surgery failed to replicate results of the oncotype DCIS recurrence score

Abstract

Purpose: To compare ipsilateral breast event (IBE) risks in patients with DCIS of the breast post-lumpectomy, as estimated by breast radiation oncologists, the Van Nuys Prognostic Index (VNPI) studies, the Memorial Sloan Kettering Cancer Center (MSK) DCIS nomogram, and the 12-gene DCIS Recurrence Score assay. Materials/Methods: 12-gene DCIS Recurrence Score of 91 DCIS lumpectomy cases were identified from our practice. Clinicopathologic factors (excluding DCIS score) were summarized for blinded review. 3 radiation oncologists independently estimated the 10-year IBE risk and rated the impact of age/menopausal status, tumor morphology, tumor span, and margin width. Corresponding VNPI and MSK nomogram estimates were generated. Differences and correlations between the IBE estimates and clinicopathologic factors were evaluated with univariate and multivariate analysis. Results: Median age was 60, 25% margins < 3 mm, 48% DCIS span > 10 mm, 48% would have been ineligible for E5194. All comparisons were initially examined for similarity (non-significant differences) and susequently tested for a significant correlation. In three possible physician comparisons only 1 (AA-CL) had similar IBE estimates and strong correlation (r=0.792). Comparison of physicians to either VNPI nomogram or MSK DCIS calculator had similar estimates with strong correlation in only 2 of 6 possible comparisons in the all cases and the E5194 eligible patient cohorts respectively; and in 1 of 6 comparisons in the E5194 ineligible cohort. The DCIS Recurrence Score showed no strong correlation in regard to similar IBE estimates for any comparisons including either physicians, VNPI nomogram or MSK DCIS calculator. Tumor size was highly correlated with all physician IBE estimates (r = 0.547 to 0.799), while margin size showed moderate to high inverse correlations (r = -0.324 to -0.619). All 3 physicians rated margin width as having the most impact on their IBE risk estimates, then tumor morphology and tumor span. Source of EstimateAll Cases (n = 91)E5194 Eligible (n = 47)E5194 Ineligible (n = 44)DCIS Score15.1%, 5.3 (14.0-16.2)14.2%, 4.8 (12.8-15.6)16.1%, 5.7 (14.3-17.8)MSK DCIS Nomogram (Rudloff et al., 2010)18.8%, 7.7 (17.2-20.4)23.4%, 18.6 (19.6-27.3)16.1%, 4.1 (14.9-17.3)21.7%, 9.4 (18.8-24.6)VNPI studies23.4%, 18.6 (19.6-27.3)15.6%, 15.5 (11.0-20.1)31.9%, 18.1 (26.4-37.4)AA18.9%, 14.1 (16.0-21.9)11.7%, 8.8 (9.1-14.3)26.6%, 14.7 (22.2-31.1)CL19.9%, 11.8 (17.5-22.4)14.5%, 6.6 (12.5-16.4)25.8%, 13.3 (21.7-29.8)DC26.8%, 13.9 (23.9-29.7)19.8%, 5.9 (18.1-21.6)34.2%, 16.1 (29.3-39.1) Conclusion: IBE risk estimates for this cohort of DCIS cases vary significantly among commonly available clinical predictive tools and individual physician estimates. Surgical margins and tumor size continue to factor prominently in physician decision algorithms. In our observations, neither the VNPI nomogram, MSK DCIS calculator nor physician IBE risk estimates have similar values with strong enough correlation to reliably replicate the DCIS Recurrence Score. Citation Format: Leonard C, Lei R, Antell A, Nowels M, Fryman S, Howell K, Dennis C. A comparison of models (physician, the Van Nuys prognostic index, the Memorial-Sloan-Kettering Cancer Center DCIS nomogram)to predict ipsilateral breast events in patients with ductal carcinoma in situ (DCIS) of the breast after breast-conserving surgery failed to replicate results of the oncotype DCIS recurrence score [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P1-11-02.