An estimation model for Oncotype Dx score using routine clinical and pathological parameters.

BACKGROUND: Oncotype DX (ODX) is a validated recurrence score (RS) used to predict the risk of recurrence and benefit of chemotherapy in ER positive, node negative breast cancer patients. Prior to ODX, treatment recommendations regarding adjuvant chemotherapy and mortality approximation have taken into account clinical and pathologic risk factors. A discordance rate of 7-19% between risk allocating pathologic factors and ODX RS has been previously reported with progesterone receptor (PR) negativity noted as a defining clinical feature in numerous cases. The association between other clinicopathologic features and discordance is less certain. METHODS: ODX data and clinicopathologic features were retrospectively reviewed for 724 breast cancer tumors belonging to 704 patients between 2006 and 2016. ODX discordance was defined as either 1-step discordance or 2-step discordance between ODX risk group (low, intermediate, high) and tumor grade (TG) (well differentiated, moderately differentiated, poorly differentiated). Tumors with 1-step discordance received a discordance score (DS) of 1 while those with 2-step discordance received a DS of 2. The database was subsequently analyzed using Paik's RS cutoffs as well as those outlined in the TAILORx trial. An odds ratio (OR) of >1 was consistent with discordance. RESULTS: Among 724 tumor samples, ODX ER score (p=0.000), ODX PR score (p=0.000), ODX HER2 score (p=0.000), TG (p=0.000), mitotic count (MC) (p=0.0012), DCIS grade (p=0.0046), DCIS type (comedo necrosis vs. non-comedo necrosis) (p=0.0335) and micropapillary features (p=0.0044) were significantly associated with RS. Median age of cohort was 59 years and median tumor size was 1.2 cm. Of 724 tumors, 619 from 604 subjects were eligible for assessment of discordance. Median RS was 16. Using Paik's RS cutoffs, 64.3% discordance was observed: 52.5% 1-step discordance (DS 1) and 11.8% 2-step discordance (DS 2). The TAILORx categorization yielded a discordance rate of 44.3%: 40.1% 1-step discordance and 4.2% 2-step discordance. On univariate analysis and using Paik's RS cutoffs, young age (p= 0.0240), high MC (p=0.0006), large tumor size (>20 mm) (p=0.0209), the presence of DCIS (p=0.0480), high DCIS grade (p= 0.0033), and high ODX PR and ER scores (p= 0.0000) were significant clinicopathologic features predictive of discordance. On multivariate analysis, high MC (p= 0.0000), high ODX PR and ER scores (p=0.0000) remained significant as well as premenopausal status (p=0.026). Per TAILORx cutoffs, univariate analysis revealed younger age (p= 0.0060), high MC (p= 0.0270), premenopausal status (p= 0.0124), and high ODX PR and ER scores (p= 0.0000) as significant for discordance. On multivariate analysis, high ODX PR and ER scores (p= 0.0000) remained significant. CONCLUSION: In this retrospective ODX database, premenopausal status, high MC, high ODX PR and ER scores as per Paik's RS cutoffs were significant predictors for ODX discordance while high ODX PR and ER scores were significant predictors per the RS's outlined in the TAILORx trial. RS cutoffs per the TAILORX trial appear to create less discordance between RS and TG than the original cutoffs outlined by Paik and colleagues. Citation Format: Cascetta KP, Zimmerman BS, Eggert L, Molot MC, Ru M, Nayak A, Bleiweiss I, Tiersten A. Retrospective analysis of clinicopathologic features predictive of oncotype DX discordance in estrogen receptor positive, node negative breast cancer patients [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-28.

Introduction: Oncotype Dx (ODX) is the most widely used genomic test for adjuvant treatment decision-making in patients with hormone receptor (HR)+ early breast cancer (EBC). Risk stratification in this clinical setting is crucial and often challenging, thus additional molecular information could significantly enhance patients’ management. We hypothesized that Prosigna, a prognostic assay estimating distant relapse-free survival in postmenopausal women with HR+ EBC, could serve as a complementary test to ODX for risk stratification. This study aims to evaluate the concordance between ODX and Prosigna in a real-world scenario. Methods: A total of 30 postmenopausal HR+ EBC patients, who were previously tested with ODX and classified as low (n=12), intermediate (n=5), and high (n=13) genomic risk, were included. The three risk categories were assigned considering the recurrence score (RS) and lymph node status: low (RS 0-10), intermediate (RS 11-25), and high risk (RS 26-100) for node-negative (pN0), as reported in TAILORx trial; low (RS ≤ 25) and high risk (RS>25) for 1-3 node-positive (N1-3), as reported in RxPONDER trial. For each case, RNA was extracted from the same formalin-fixed paraffin-embedded (FFPE) tumor block that was used for ODX, and subjected to Prosigna testing on a NanoString nCounter® DX Analysis. The Prosigna risk of recurrence (ROR) score also considers lymph node status: for pN0, 0-40 indicates low, 41-60 intermediate, and 61-100 high risk; for N1-3, 0-15 indicates low, 16-40 intermediate, and 41-100 high risk. Descriptive analyses of clinicopathological features were performed. Cohen's Kappa and Spearman (rs) correlation analyses were calculated between RS, ROR, and clinicopathological factors. Results: The overall agreement between the two platforms was the same as in the TransATAC study (κ=0.32; p<0.001). Considering ODX as the gold standard, a low-high disagreement was found in n=6 (20.0%) cases, while a low-intermediate or intermediate-high disagreement was observed in n=7 (23.3%) cases. The lowest concordance was seen in the ODX low (1/12) and intermediate (3/5) risk groups, with Prosigna tending to assign a higher risk. This tendency was confirmed by the 100% concordance (13/13) in the high-risk group. Prosigna assigned luminal A, luminal B, and HER2-enriched molecular subtypes to n=7, 23.3%; n=22, 73.3%; and n=1, 3.3% cases, respectively. Most luminal B (n=20; 90.9%) cases had high ROR scores, whereas n=12 (54.5%) cases were classified as high RS. A significant proportion of luminal A (6/7; 85.7%) cases were classified as intermediate ROR, while ODX classified n=1 (14.3%) of them as intermediate and n=6 (85.7%) as low RS. Tumor samples with high ROR were more likely G3 (p=0.002), pT2 (p=0.021), and Ki67>20% (p=0.01) in comparison with the intermediate risk ones. ROR showed a strong correlation with Ki67 expression (rs=0.72, p<0.001), whereas RS demonstrated a moderate correlation (rs=0.57, p<0.001). According to available follow-up (F/U) data (median 6 months; range 1-36), recurrence was detected in only one case (G2; Ki67=40%; F/U=14 months), which both assays classified as intermediate risk. Conclusions: Our study confirms that ODX and Prosigna provide different types of clinical information for postmenopausal HR+ EBC patients, as they analyze distinct genes and pathways. In our real-world cohort, Prosigna generally assigned a higher risk category compared to ODX. Therefore, integrating multiple molecular assays could potentially refine risk assessment and enhance personalized treatment strategies in this patient population. However, it remains to be defined which specific subpopulation might benefit from this integrated approach for molecular testing. Further investigations with larger cohorts and long-term F/U data are needed to validate these preliminary findings. Citation Format: Nicola Fusco, Giulia Cursano, Konstantinos Venetis, Elisabetta Munzone, Eltjona Mane, Mariia Ivanova, Chiara Frascarelli, Elisa De Camilli, Oriana Pala, Giovanni Mazzarol, Silvia Dellapasqua, Antonio Marra, Carmen Criscitiello, Giuseppe Viale, Giuseppe Curigliano, Elena Guerini-Rocco. Enhancing Risk Stratification in HR+ Early Breast Cancer: A Real-World Evaluation of Oncotype Dx and Prosigna [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P1-11-15.

Background: Oncotype DX(ODX) is a clinically validated, commercially available multi-gene assay that provides prognostic and predictive information in estrogen receptor(ER)-positive, Her2-negative breast cancer. Because health care system does not cover this expensive assay in many countries, physicians and researchers try to use routine histopathological variables predict ODX Recurrence Score (RS). Hyunseok Kim et.al presented at ASCO that Estimated Recurrence Score (ERS) model could identify those patients most likely to benefit from including ODX RS results in therapeutic decision-making (J Clin Oncol 32:5s, 2014 suppl; abstr 559). Methods: We retrospectively reviewed histopathological slides between July 2007 and April 2014 from Toranomon Hospital patients (n=149) with early stage ER-positive, Her2-negative breast cancer, for whom ODX was ordered. We developed an original linear regression model using routine histopathological markers (Allred Score for progesterone receptor (PR), nuclear grade(NG), and Ki67) to calculate an Estimated Recurrence Score (ERS), and correlated it with the observed ODX RS assay result. ERS=20.46+0.298xKi67–1.48xPR+1.41xNG predicts the ODX RS with an R2 of 0.48. In order to internally validate this model in the Toranomon Hospital cohort in Japan, we adapted Hyunseok’s algorithm. Results: 132 patients had an observed RS ≤ 25 (89%) and 17 patients had an observed RS above 25 (11%). When the ERS was < 21 (n=124), we accurately classified 97% of them (120) who were found to have a low risk (observed RS ≤ 25). Similarly, 83% of those (5/6) with an ERS > 30 fell into a high risk category with observed RS > 25. Table1 (TAILORx risk categories)Toranomon cohortODX ≤25ODX RS>25TotalERS < 211204124ERS(21-30)11819ERS(> 30)156Total13217149 We also presented the comparison of ERS and observed RS based on original risk categories. Table2 (Original risk categories )Toranomon cohortODX RS<18ODX RS 18-30ODX RS>30TotalERS(<18)6633099ERS(18-30)1725244ERS(>30)0156Total83597149 Conclusions: We developed an ERS model to find those patients most likely to benefit from ODX RS results in clinical practice setting. Although further external and prospective validation is mandatory, preliminary result supports the usefulness of this ERS model. Citation Format: Hidetaka Kawabata, Keiichi Kinowaki, Nobuko Tamura, Yoko Kobayashi, Masami Kadowaki, Daishu Miura, Toshimi Takano, Akihiko Shimomura, Tsuguo Iwatani, Yuko Kitawaki, Takeshi Fujii. Routine histopathological variables predict oncotype DX risk categories [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-11-32.

INTRODUCTION: The recent TAILORx results suggest that adjuvant endocrine therapy and chemoendocrine therapy had similar efficacy in women with hormone-positive, HER2-negative, node-negative invasive breast carcinomas with an Oncotype Dx® recurrence score < 26. Oncotype Dx® is an expensive test (current list price of $4,620.00). Although the Oncotype Dx® test has gained widespread acceptance, cost may be an impediment to its adoption in certain areas of the world, and it may not be the most cost-effective or cost-efficient option in certain subsets of breast cancer patients. Based on a modification of the new Magee equations, we published an algorithmic approach in 283 ER+ cases using an average modified Magee equation (Turner BM, et al. Mod Pathol. 2015;28(7):921-31), and subsequently published a validation of this algorithmic approach in an additional 620 ER+ cases (Cancer Med. 2019 Jun 14. doi: 10.1002/cam4.2323. [Epub ahead of print]), supporting only reflex Oncotype Dx® testing for certain subsets of breast cancer patients. We now have recurrence data and at least five years of outcome data on 310 cases. METHODS: A total of 310 cases with ER+ invasive breast cancer, all with Oncotype Dx® recurrence scores, were included in this study from the University of Rochester and the University of Louisville. The outcome analysis included all patients who had at least five years of follow-up data, and all patients who had a breast cancer recurrence. In order to calculate the average modified Magee score, information on the Nottingham score, ER, PR, Ki-67, HER-2, and tumor size was extracted from the pathology report. RESULTS: 255/310 cases had an Oncotype Dx® < 26, and 4% (10/255) of these cases were associated with a breast cancer recurrence. 187/310 cases had an average modified Magee score ≤ 18, and 184 of these 187 cases (98%) had an Oncotype Dx® recurrence score < 26. Only 2% (n = 4) of these 184 cases were associated with a breast cancer recurrence. On average, the Nottingham score and Ki-67 were lower for cases with an average modified Magee score ≤ 18 compared to cases with an Oncotype Dx® recurrence score < 26 (Table 1). On average, the modified ER and PR H-scores were higher for cases with an average modified Magee score ≤ 18 compared to cases with an Oncotype Dx® recurrence score < 26 (Table 1). CONCLUSION: If the average modified Magee score is ≤ 18, the patient has a low likelihood of recurrence, and the Oncotype Dx® recurrence score will likely be < 26. Given the recent TAILORx findings, we suggest that patients with an average modified Magee score ≤ 18 not be sent out for Oncotype Dx® testing if an Oncotype Dx® recurrence score < 26 is the basis on which the clinician will use to decide whether or not to give chemotherapy. The information needed to use the average modified Magee equation is already generated by many pathology laboratories during the initial assessment of breast cancer, and the Magee equations are free of charge (https://path.upmc.edu/onlineTools/mageeequations.html). Based on the Genomic Health 2017 Annual Report, our data suggests that not sending out cases with an average modified Magee score ≤ 18 would have resulted in a potential cost savings to the health-care system in 2018 of almost 300 million dollars. Table 1: Average Nottingham score, modified ER H-score, modified PR H-score, and Ki-67, with number and percent recurrence for cases with an average modified Magee score ≤ 18 and cases with an Oncotype Dx® recurrence score < 26AVERAGERecurrence Sore typeNNS**ER***PR***Ki-67Recurrence N (%)Average modified Magee score ≤ 18*1875.2262.3235.410.34 (2)Oncotype Dx® recurrence score < 262555.5253.2192.613.610 (4)*98% (184/187) of patients with an average modified Magee score ≤ 18 had an Oncotype Dx® recurrence score < 26**Nottingham score***modified H-score (Turner BM, et al. Mod Pathol. 2015;28(7):921-31) Citation Format: Hasan Khatib, Mary Ann G Sanders, Andrea Breaux, Armen Soukiazian, Nyrie Soukiazian, David Hicks, Bradley M Turner. ER+ patients with an average modified Magee score ≤ 18 have a low likelihood of breast cancer recurrence and a high likelihood of an Oncotype Dx® recurrence score < 26 [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-07-06.

Background: The gene expression profiling assay OncotypeDx (ODx) prognosticates the risk of estrogen receptor positive (ER+) breast cancer (BC) recurrence and assesses the likely benefit from adjuvant chemotherapy in addition to endocrine therapy. There have been several attempts to develop algorithms that provide similar outcome prediction to the ODx assay with the use of routine clinicopathological characteristics. These models appear to predict high-risk ODx RS but are unable to reliably rule out the presence of patients with intermediate-risk disease. Our objective was to identify the clinicopathological factors associated with intermediate and high-risk categories. Methods: We retrospectively reviewed the electronic medical records of patients with early-stage ER+ BC for whom ODx recurrence score (RS) was available. Patients were diagnosed and treated at 3 specialized cancer centers between 2010 and 2017. Two multinomial logistic regression models (crude and adjusted) were constructed to assess the association between clinicopathological characteristics and ODx RS as a categorical variable. The adjusted model included the following variables: ODx RS, age, tumor size, node status, grade, lymphovascular invasion and hormonal receptors. The reported association measure was the relative prevalence ratio (RPR) with its respective 95%CI. Results: A total of 551 patients were included. Patients had a mean age of 56.2 ± 11.9 (SD) (range: 26-89). 9.6% (n=53) of patients were≤40 years old.The size of the tumors ranged from 0.1 cm to 7.2 cm (median = 1.5 cm; IQR 1.0-2.2cm). A minority of patients had lymph node involvement (5.8%, n=32). By subtype, carcinomas were mostly ductal (83.5%, n=460), followed by lobular (10.0%, n=55) and mucinous (3.5%, n=19). The majority of tumor exhibited an intermediate histological grade (71.6%, n=386). Ki 67 was available in 58.8% patients (n= 324), with a median Ki67 of 20 (IQR 10-30). In the adjusted multinomial logistic regression model, factors associated with ODx intermediate-risk category were grade 3 (RPR=4.78; 95%CI: 2.01-11.39) and having either ER or PR <50 (RPR=2.80; 95%CI: 1.83-4.27). Factors associated with ODx high-risk category were grade 3 (RPR=15.89; 95%CI: 3.23-78.19), having either ER or PR <50 (RPR=4.58; 95%CI: 2.37-8.87), age≤40 (RPR=2.96; 95%CI: 1.20-7.29) and T2-3 (RPR=2.20; 95%CI: 1.13-4.32). Conclusion: Grade 3, ER o PR <50, age ≤40 years and T2-3 are clinicopathological characteristics strongly associated with high-risk ODx RS. The associations with intermediate-risk ODx RS are weaker. The way these factors could be integrated into a clinicopathologic risk prediction model to identify high-risk patients needs further analysis. Citation Format: Ruiz R, Morante Z, Namuche F, Urrunaga D, Leon M, Ziegler G, Aguilar A, Chavez Mac Gregor M, Gomez H. Clinicopathological characteristics associated with intermediate and high-risk ODx RS [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-08-18.

32 Background: The Oncotype DX (ODX) multigene assay has been previously suggested to result in an overall reduction in the use of adjuvant chemotherapy and associated costs for women with early stage, estrogen receptor (ER)-positive breast cancer. However, the association between adoption of ODX and chemotherapy costs has only previously been considered in theoretical models and has not been examined using actual patient outcomes in real world clinical practice. Methods: Retrospective analysis of Surveillance, Epidemiology, and End Results (SEER) -Medicare data of the association between overall and chemotherapy-specific costs associated with adoption of ODX testing in patients diagnosed with invasive, non-metastatic, ER-positive breast cancer between 2005 and 2009. We limited our primary analysis to women ages 66 to75 to include women in which adjuvant chemotherapy would be most likely to be considered. Total Medicare payments were used to calculate direct costs in the year following diagnosis. NCCN guidelines were used to stratify patients on the basis of clinical and pathologic features into low ( < 1.0 cm), intermediate, and high risk (node positive) disease. Results: A total of 21,272 women met study criteria. Average costs in the year following diagnosis were $31,532 in the overall cohort, and was highest for women with NCCN high risk ($45,192) vs. intermediate ($28,642) or low ($23,662) risk disease. Chemotherapy costs followed similar trends ($4,819, $1,157, and $226 respectively). In multivariable analyses, ODX was associated with a relative decrease in chemotherapy costs among high risk women (RR 0.54, 0.37-0.77), but increased costs among low and intermediate risk women (RR1.36, 1.13-1.26 and RR 3.73, 2.13-56.54). Women with high risk disease had significantly lower absolute chemotherapy costs associated with receipt of ODX (-$2,298, -$3,049 to -$1,547; All P < 0.001). Conclusions: Receipt of ODX testing was associated with relative and absolute decreases in chemotherapy costs, but only in women with high NCCN risk disease. Further research is needed to disentangle correlative vs. causative association of ODX testing with patient outcomes and costs.

Oncotype Dx (ODx) is a multigene assay that is prognostic and predictive in estrogen receptor (ER) positive early breast cancer. ODx recurrence score (RS) is reported to be histologic grade dependent. Relationship of RS with breast cancer histologic subtypes is unknown. This study was designed to assess the relationship of histologic subtype with RS. Histologic grade dependence of RS was also investigated. Results of consecutive ODx tests (1/2007-7/2016) from two institutions were reviewed. Histologic subtypes (in: Lakhani et al., WHO classification, IARC Press, Lyon, 2012), combined Nottingham histologic grade, age and tumor size were recorded from pathology reports. Univariate and multivariate analysis was performed to investigate the relationship between RS and ODx risk categories and histologic subtypes, grade, age and tumor size. RS was grade dependent. RS of grade 1 and grade 2 tumors were significantly lower than grade 3 tumors. There was no high-risk grade 1 tumor. In favorable histologic subtypes there was no high-risk tumor. Mean RS of grade 1 lobular tumors was significantly higher than grade 1 ductal tumors. Using newer ODx cut-offs, 5 grade 1 tumors were reclassified as high risk (RS>25) and grade 3 lobular tumors showed significantly higher rate of reclassification as high-risk than grade 3 ductal tumors. In a multivariate analysis, only grade showed a significant positive correlation with RS. Adding dichotomous histologic subtyping (favorable vs. non-favorable) to grade further improved correlation with RS. The Oncotype Dx result is impacted by histologic grade and histologic subtype. Tumors with favorable histologic subtypes and histologic grade 1 tumors do not have high-risk RS. High RS in a grade 1 tumor or in a tumor with favorable histology is unusual that warrants further investigation. Invasive lobular carcinomas rarely show high-risk RS. Histologic grade and histologic subtype should be considered while ordering ODx testing.

Background: Genomic tests are needed in addition to clinical/pathological factors to optimize adjuvant treatment decisions in most patients with HR positive Her-2 negative early BC. We aimed to develop an AI-based model to predict high risk HR positive Her-2 negative early BC by using clinical/pathological data and Oncotype DX scores. Methods: 234 patients who underwent Oncotype Dx test at our center between 2008 and 2023 were included. Clinical/pathologic characteristics and OncotypeDx scores were analysed by an ensemble of XgBoost, Random Forest, LightGBM, and Decision Tree machine learning algorithms to predict high risk disease. The dataset was split into two subsets: 75% for the training and 25% for the test set. The final model is chosen to maximize the precision of high-risk patients by using the data associated with high risk (HR) Oncotype-Dx scores and also the clinical/pathological data from patients who had distant metastases (DM) with low and intermediate recurrence score (RS). Results: Median follow up was 84 months and median age was 47. Among 234 patients % 65 had stage I, % 22 had node positive, % 78 had grade 2 disease. Fourteen %, 68% and 18% of patients had HR, intermediate (IR) and low risk (LR) Oncotype Dx scores respectively, % 72 had clinically LR disease. Six %, 21% and 88% of patients with LR, IR and HR RS received adjuvant chemotherapy. There were 25 events: 13 DM, 2 local/regional relapse, 1 second primary, 9 deaths (7 due to BC). Two patients with HR, 8 patients with IR and 3 patients with LR Oncotype Dx scores had DM. The training (n: 175) and test set (n: 59) had similar risk level distributions. There were 6 DM, 2 deaths due to BC in the test set. Among 6 patients with DM, one patient had LR Oncotype Dx and IR AI-based score, another patient had IR Oncotype Dx and HR AI-based score, 4 patients have concordant scores (3 had IR, 1 had HR in both Oncotype Dx and AI-based model). Oncotype Dx and AI-based predicted scores were concordant in 5/11 (45%) patients with HR, 28/39 (%72) patients with IR and 1/9 (11%) patients with LR Oncotype Dx scores in the test set. There were no DMs in the AI-based low risk group vs one patient with DM in the Oncotype DX LR group. Conclusion: Our model might be used to predict high/intermediate risk patients and its performance needs validation with additional new datasets before its clinical implementation. Citation Format: Gul Basaran, Akif Gunes, Ufuk Kara, Alper Sonkaya, Mehmet Teomete, Alptekin Arıfoglu, Evrim Tezcanlı, Ozge Gumusay, Aykut Soyder. Prediciting high risk hormone receptor (HR) positive Her-2 negative early breast cancer (BC) by artificial intelligence (AI) algorithms [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P4-11-03.

The Oncotype DX® assay is a validated genomic test that predicts the likelihood of breast cancer recurrence, patient survival within ten years of diagnosis and the benefit of chemotherapy in early-stage, node-negative, estrogen receptor-positive breast cancer. Further markers of recurrence include disseminated tumor cells (DTCs) in the bone marrow (BM) and circulating tumor cells (CTCs) in the blood, particularly stemness-like tumor cells (slCTCs). In this study, Oncotype DX, DTCs, CTCs and slCTCs were used to evaluate the risk of recurrence in 68 patients with human epidermal growth factor receptor 2-negative, early-stage breast cancer. Formalin-fixed, paraffin-embedded tissue sections were analyzed for the expression of 16 cancer genes and 5 reference genes by Oncotype DX, yielding a recurrence score (RS). G2 tumors were evaluated for urokinase-type plasminogen activator (uPA)/plasminogen activator inhibitor type 1 (PAI1) and Ki-67. Two BM aspirates were analyzed by immunocytochemistry for DTCs using the pan-cytokeratin antibody A45-B/B3. CTCs and slCTCs in the blood were detected using the AdnaTest BreastCancer, AdnaTest EMT and the AdnaTest TumorStemCell. Oncotype DX was performed in 68 cases, yielding a low RS in 30/68 patients (44%), an intermediate RS in 29/68 patients (43%) and a high RS in 9/68 patients (13%). DTCs were detected in 19/68 patients (28%), CTCs in 13/68 patients (19%) and slCTCs in 26/68 (38%) patients. Moreover, 8/68 patients (12%) with G2 tumors were positive for uPA, 6/68 (9%) for PAI1 and 21/68 (31%) for Ki-67. Ki-67, progesterone receptor (PR) and G3 tumors were significantly correlated with RS (P<0.001; P=0.006; and P=0,002, respectively), whereas no correlation was identified between DTCs, CTCs, slCTCs and RS. Ki-67 may support therapeutic decisions in cases where Oncotype DX is not feasible. Larger patient cohorts are required to estimate the additional detection of DTCs and CTCs for the determination of risk recurrence.

e11092 Background: It was suggested that breast tumors that are ER positive / PR negative carry a worse prognosis than ER /PR positive tumors if treated only by adjuvant hormonal therapy. Most studies did not account for HER 2 expression. PR was assessed by methods that measure protein content. Oncotype Dx measures mRNA expression. In clinical trials the discrepancy between IHC and RT PCR for the expression of PR was about 20%, but this was not assessed in a community setting. The meaning of a high level of mRNA without detectable PR protein is unclear and might be due to technical issues or a biological mechanism such as translational inhibition by miRNA's. We sought to verify the prognostic value of the expression of the progesterone receptor by IHC in women with early stage HER 2 negative, ER positive BC treated with systemic hormonal therapy only and to characterize the results of Oncotype Dx in ER positive, PR negative HER 2 negative tumors. Methods: An historical prospective analysis of patients files. Results: Initial cohort: Files of 286 consecutive patients with ER positive HER 2 negative tumors that were treated by adjuvant hormonal treatment during 2000-2006 were reviewed. 82% of tumors were node negative. 80 were PR negative and 209 were PR positive. There were no significant differences between the groups with respect to nodal status or grade. After a median follow up of 7.24 years (0.78 to 9.5) 8 women with PR negative and 9 women with PR positive tumors experienced disease recurrence (10% vs 4.3%, RR 2.32, p 0.014). Second cohort: 58 PR negative tumors were testes by Oncotype Dx during 2007-2010. 24 (47%) were PR positive by RT PCR. Most (47.3%) were intermediate risk by the test, 36.9 were low risk and 15.8 were high risk. Of note, grade 3 tumors were more likely to be RT PCR negative for progesterone receptor (91% vs 40.5%) and high risk (35% vs 7.5%). Conclusions: In our hands ER positive PR negative HER 2 negative tumors carry a worse prognosis when treated by hormonal adjuvant treatment compared with ER/PR positive tumors. We found a high rate of discrepancy between IHC and RT PCR concerning PR expression. Most PR negative tumors are labeled intermediate risk. The value of the test in this setting is unclear.

Low local recurrence rate without postmastectomy radiation in node-negative breast cancer patients with tumors 5 cm and larger

e13083 Background: This study investigates the concordance between the 21-gene recurrence score assay, Oncotype DX (ODX), and the 70-gene signature MammaPrint (MP) in a single academic center and evaluates their correlation with breast cancer outcomes. We previously reported a discordance between MP and ODX results; here, we present a 10-year follow-up survival data. Methods: This prospective cohort study included eighty-six consecutive patients between 2012-2014 with node-negative, hormone receptor-positive breast cancer whose tumor samples were tested using ODX. Based on ODX recurrence scores (RS), patients were treated with either a chemotherapy plus hormonal therapy group or a hormonal therapy-alone group. All patients were per standard practices at our institution. MP was subsequently performed on all samples, classifying patients as either low-risk (MP-L) or high-risk (MP-H). ODX risk classification was defined as intermediate/high risk (ODX-H) for RS ≥ 26 in postmenopausal and ≥ 16 in premenopausal patients (As defined by the TAILORx trial, the addition of chemotherapy to endocrine therapy resulted in an absolute risk reduction of 1.6% in distant recurrence-free survival), and otherwise low risk (ODX-L) (ODX ≤ 26 in postmenopausal or ≤ 15 in premenopausal). Results: In the overall population, the 10-year overall survival (OS), disease recurrence-free survival (DRFS), and breast cancer-free interval (BCFI) were 93%, 91%, and 86%, respectively. Patients classified as ODX-H had lower survival rates (OS: 88%, DRFS: 86%, BCFI: 84%) compared to MP-H patients (OS: 92%, DRFS: 89%, BCFI: 87%). Patients classified as ODX-L had numerically better outcomes than MP-L patients (OS: 100%, DRFS: 97%, BCFI: 89%) compared to (OS: 94%, DRFS: 92%, BCFI: 86%). Approximately 29% of ODX-L patients were classified as MP-H, yet these patients demonstrated favorable long-term outcomes. Conclusions: Despite the limitations of a small sample size, our findings suggest that for patients with node negative hormonal receptor positive breast cancer ODX predicts better prognosis and response to chemotherapy than MP. Future prospective randomized trials directly comparing different genomic assays are warranted to optimize risk stratification and treatment decisions. Breast cancer outcomes based on genomic assay risk classification. Genomic assay / outcomes ODX MP High (n = 51) Low (n = 35) High (n = 37) Low (n = 49) OS 88% 100% 92% 94% DRFS 86% 97% 89% 92% BCFI 84% 89% 87% 86%

The Oncotype DX (ODX) test is a commercially available molecular test for breast cancer assay that provides prognostic and predictive breast cancer recurrence information for hormone positive, HER2-negative patients. The aim of this study is to propose a novel methodology to assist physicians in their decision-making. A retrospective study between 2012 and 2020 with 333 cases that underwent an ODX assay from three hospitals in the Bourgogne Franche-Comté region (France) was conducted. Clinical and pathological reports were used to collect the data. A methodology based on distributional random forest was developed to predict the ODX score classes (ODX [Formula: see text] and ODX [Formula: see text]) using 9 clinico-pathological characteristics. This methodology can be used particularly to identify the patients of the training cohort that share similarities with the new patient and to predict an estimate of the distribution of the ODX score. The mean age of participants is 56.9 years old. We have correctly classified [Formula: see text] of patients in low risk and [Formula: see text] of patients in high risk. The overall accuracy is [Formula: see text]. The proportion of low risk correct predicted value (PPV) is [Formula: see text]. The percentage of high risk correct predicted value (NPV) is approximately [Formula: see text]. The F1-score and the Area Under Curve (AUC) are of 0.87 and 0.759, respectively. The proposed methodology makes it possible to predict the distribution of the ODX score for a patient. This prediction is reinforced by the determination of a family of known patients with follow-up of identical scores. The use of this methodology with the pathologist's expertise on the different histological and immunohistochemical characteristics has a clinical impact to help oncologist in decision-making regarding breast cancer therapy.

Oncotype DX recurrence score (RS) is a clinically validated assay, which predicts the likelihood of disease recurrence in oestrogen receptor-positive/HER2-negative (ER+/HER2-) breast cancer (BC). In this study we aimed to compare the performance of Oncotype DX against the conventional clinicopathological parameters using a large BC cohort diagnosed in a single institution. A cohort (n= 430) of ER+/HER2- BC patients who were diagnosed at the Nottingham University Hospitals NHS Trust and had Oncotype DX testing was included. Correlation with the clinicopathological and other biomarkers, including the proliferation index, was analysed. The median Oncotype DX RS was 17.5 (range=0-69). There was a significant association between high RS and grade 3 tumours. No grade 1BC or grade 2 tumours with mitosis score 1 showed high RS. Low RS was significantly associated with special tumour types where none of the patients with classical lobular or tubular carcinomas had a high RS. There was an inverse association between RS and levels of ER and progesterone receptor (PR) expression and a positive linear correlation with Ki67 labelling index. Notably, six patients who developed recurrence had an intermediate RS; however, four of these six cases (67%) were identified as high-risk disease when the conventional clinical and molecular parameters were considered. Oncotype DX RS is correlated strongly with the conventional clinicopathological parameters in BC. Some tumour features such as tumour grade, type, PR status and Ki67 index can be used as surrogate markers in certain scenarios.

59 Background: The Oncotype DX (ODX) genomic assay has been used since 2004 to estimate prognosis and guide adjuvant treatment in patients with estrogen receptor-positive, node-negative invasive breast tumors. However, the impact of ODX assay testing in patients with invasive lobular carcinoma (ILC) has not been specifically reported. We describe our experience in patients with ILC who underwent ODX assay testing and how this affected adjuvant treatment. Methods: An IRB approved retrospective review was performed on all patients undergoing treatment for early stage, node-negative ILC from 2006-2011. All cases were reviewed by a single pathologist (GA) for verification of histology and subtype. The impact of ODX assay testing on treatment management was evaluated. Results: Thirty-nine patients underwent ODX genomic assay testing of early stage, node-negative ILC with a median age of 62 years. ILC tumor classification included classic (67%), pleomorphic (23%) and mixed (10%) subtypes with a median tumor size of 2.0 cm (0.6-6.0). Ninety-seven percent of tumors were estrogen receptor-positive, 74% progesterone receptor-positive. Median ODX recurrence score was 15 (0-34) with an ODX 10-yr risk of 10% (3-23). ODX risk classification was: Low (N=26), Intermediate (N=12), and High (N=1). There was no difference in ODX recurrence score or risk classification between ILC subtypes (p=0.52 and p=0.35, respectively). Adjuvant chemotherapy was used in 26% (TC N=8, AC N=1, FEC 100 N=1). Tumor size or ILC tumor subtype were not significant for adjuvant chemotherapy use while tumor grade (p=0.046), ODX recurrence score (22.8 vs 13.6, p&lt;0.0001), ODX risk classification (p=0.009) and ODX 10 year risk (15.9 vs 9.0, p&lt;0.0001) were significant. With a median follow up of 16.5 months, there were no recurrences or tumor-related deaths. Conclusions: ODX testing on early stage node-negative ILC may serve as a useful adjunct when counseling patients on the decision for adjuvant therapy. The long-term impact on recurrence or survival in patients with ILC who receive adjuvant chemotherapy based on ODX recurrence score remains undetermined and warrants further testing.