Abstract P3-07-17: Analysis of the National cancer data base 2010-2012 oncotype DX breast cancer assay: Lessons learned

Abstract

Abstract Oncotype DX (OD) 21-gene breast cancer (BC) assay is used for testing of estrogen receptor positive (ER+) early stage BC and provides a low, intermediate or high 10-year risk recurrence score (LRS, IRS, HRS) for BC. Scores are used as guidelines to predict the probability of successfully adding adjuvant chemotherapy (AC) to endocrine therapy (ET) to reduce the risk of BC recurrence. In retrospective analysis of 2 clinical trials, patients (pts) with HRS benefited from addition of AC to ET, no benefits were shown for LRS, and benefits of AC for IRS were not clear. The OD assay has been utilized since 2004, but data on the impact of using assay results in clinical practice across the US are lacking. The TAILORx and RxPONDER prospective clinical trials which are using OD scores are ongoing (results will be available in 2017 and 2022, respectively) and may help clinicians better understand the predictive capabilities of the IRS OD score. We analyzed the current impact of IRS OD score in a retrospective observational study of the National Cancer Data Base (NCDB) (which represents 75% of the US population) from 2010-2012. This time period encompasses the beginning of required recording of molecular assays and the latest data released by NCDB in 4/2015. Demographic and clinical variables of all pts with IRS results were analyzed using frequency statistics, chi-square and logistic regression analysis. Data from 24260/27995 pts with IRS and documented AC information was analyzed. 11520=47.4% pts received AC. Age ranged from 20-90 (mean 58.4, median 59 years); 99.2% were ER+ and females, 2.7% were HER2+; 6143=25.3% had T1a, T1b or T3 tumors; 19791=81.5% were N0, 3684=15.1% N1; 17950=73.6% had ≥G2 tumors and 3389=14% had lymphvascular invasion (LVI). Selected demographic and clinical characteristics of patients (pts) with -/+ chemotherapy (AC) ACAC NoYes# Pts12740=52.6%11520=47.4%**Mean age61.255.3*Race White11328=46.9%9973=41.1%Black868=3.5%966=3.9%*Other431=1.7%491=2%Unknown113=0.4%90=0.3%Insurance (I) Private7181=29.6%8227=33.9%Medicaid550=2.2%646=2.6%Medicare4605=18.9%2201=9%*No I182=0.7%197=0.8%Tumor size mm 1-5487=2%238=1%6-103162=13%1890=7.8%11-508874=36.5%9060=37.3%*>50132=0.5%234=1%*Grade G13450=14.2%1760=7.2%G26887=28.3%6210=25.5%G3 & undifferentiated1775=7.2%3078=12.6%*LVI No9911=40.8%8045=33.1%Yes1288=5.3%2101=8.7%*Stage I9586=39.5%7028=28.9%II2970=12.2%4172=17.1%*III-IV99=0.4%243=1%*Lymph node+ (LN+) 010984=45.2%8807=36.3%1-31346=5.5%2338=9.6%*>328=0.1%154=0.6%**p<.001 **positive predictive value=PPV IRS OD result had poor PPV for the administration of AC (47.4%). AC administration was significantly associated with larger tumors, LN+, LVI, high tumor grade, higher TNM stage, younger age and black race (p<.001), but not with facility type (community/comprehensive/academic-research program), income, or education (p>0.05). Medicare pts were less likely to receive AC than ones with private insurance (p<.001). Our data analysis reveals that additional guidelines for selection of pts for OD testing and new algorithms for AC administration in the IRS subsets are needed. The authors of this abstract solicit a call to action from interested parties on behalf of the patients affected by this conundrum. Citation Format: Orucevic A, Heidel RE, Bell JL. Analysis of the National cancer data base 2010-2012 oncotype DX breast cancer assay: Lessons learned. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P3-07-17.