Oncotype DX Assay Research Articles

Abstract PO3-01-13: A study of the impact of the implementation of the Oncotype Dx 21-gene breast-cancer assay during the SARS-CoV-2 pandemic for 1-3 node positive early breast cancer: The Clatterbridge Cancer Centre Experience

Abstract Background: The 21-gene breast cancer (BC) assay (Oncotype Dx) is used to predict chemotherapy benefit in ER+, HER-2- lymph node (LN) negative BC. The RxPONDER trial demonstrated that post-menopausal patients with ER+, HER2- BC with 1-3 LN and a recurrence score (RS) of < 25 did not benefit from adjuvant chemotherapy. In response to the SARS-CoV-2 pandemic Oncotype Dx was implemented for ER+, HER2- 1-3 LN+ BC patients in Merseyside, UK (population 2.6 million) This study sought to understand the impact of this measure on clinical practice. Methods: All patients whose tumours were assessed with Oncotype Dx assay between Mar 20 - Nov 22 were identified through the Exact Sciences portal. Menopausal status, tumour characteristics, RS and chemotherapy regimen were recorded. The impact on cancer care was estimated using the number of infusions and clinic visits specified by our institution chemotherapy protocol for node positive disease. The financial impact was calculated using the test list price, estimated costs of adjuvant chemotherapy in the UK (Annals of Oncology, Volume 32, suppl 2 S46). The number of patients predicted to receive chemotherapy pre-oncotype DX was calculated using data from the UK PONDx trial data. Results: A total of 242 patients were tested over the 32 month period. Table 1 summarises the clinico-pathological characteristics including by RS score. The median age was 58.5 and 77.3% (n=187) were post-menopausal. The median tumour size was 25mm and median LNs with macrometastasis was 1. 198 (81.9%) patients had an RS < 25. Overall 73 patients (30%) received chemotherapy of which 34 had an RS < 25. Pre-menopausal patients were more likely to receive chemotherapy with RS < 25 than post-menopausal patients, 47.9% (23/48) vs 7.3% (11/150) respectively. In the absence of Oncotype Dx testing 169 patients would have received adjuvant chemotherapy. Therefore, 96 patients avoided chemotherapy. This equates to 1,620 chemotherapy infusions and 324 clinic attendances. The estimated cost saving was ~£50K. Conclusions: This real-world study demonstrates that the use of Oncotype DX assay for ER+,HER2- 1-3 LN+ early BC spared a significant number of patients chemotherapy during the SARS-CoV-2 pandemic, so avoiding the risk of adverse effects. This also resulted in benefits to the healthcare system with a reduction in scheduled cancer care and achieved financial savings. This, therefore, demonstrates the benefit of using the Oncotype Dx assay. Table 1: Clinico-pathological information Citation Format: Alexandra Johnson, Sonia Bathla, Leena Chagla, Joanne Cliff, Allison Hall, Anita Hargreaves, Clare Hart, Zafar Malik, Chandeena Roshanlall, Raj Sripadam, Carlo Palmieri, Helen Innes. A study of the impact of the implementation of the Oncotype Dx 21-gene breast-cancer assay during the SARS-CoV-2 pandemic for 1-3 node positive early breast cancer: The Clatterbridge Cancer Centre Experience [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-01-13.

Abstract PO5-01-08: The impact of Human Epidermal growth factor Receptor 2 (HER2) expression by RT-PCR on tumor characteristics and outcome in estrogen receptor (ER) positive, HER2 negative early-stage breast cancer

Abstract Background: The role of HER2 expression in HER2 negative breast cancer is evolving. Whether HER2 expression has an impact on ER positive, HER2 negative early-stage breast cancer in unclear. Oncotype DX assay Recurrence Score comprises of standardized quantitative HER2 mRNA expression levels by RT-PCR. The objective of this study was to investigate the impact of Oncotype HER2 score on tumor characteristics and outcome. Methods: All women assigned to the Clalit Health Services registry between 2006 to 2011 with ER positive, HER2 negative early-stage breast cancer whose tumor were sent for oncotype were identified. Patients in which HER2 score was reported in the oncotype results were included. Differences in age, tumor characteristics and chemotherapy administration between lower HER2 score (H1), i.e. HER2 score ≤ 9.1, to higher HER2 score (H2), i.e. 9.2-10.7, were analyzed. 5-year Kaplan–Meier for distant recurrence free survival (DRFS) and overall survival (OS) were estimated and differences between H1 and H2 were evaluated. Analyses were repeated separately for node negative and for node positive (N mic and 1-3 lymph nodes) disease. Results: A total of 1535 patients were included, 948 patients with node negative disease and 587 with node positive disease. Of these, 814 and 721 patients were categorized to the H1 and H2 groups, respectively. Median follow-up was 5.4 years. The distribution of oncotype RS was different, with significantly lower proportion of patients with RS between 0-10 and higher proportion of RS above 25 in the H1 group compared to the H2 groups. This difference was identified both for patients with node negative and node positive disease (Table). The distributions of age, tumor size and histological subtype were comparable between the H1 and the H2 groups, both for node negative and node positive disease. Within each RS group, administration of chemotherapy was comparable between H1 and H2 groups, regardless of nodal involvement. DRFS and OS were not associated with HER2 score. For RS 0-25, node negative disease, 5-year DRFS was 96.2% in H1 and 96.4% in H2 groups (p=0.92) and 5-year OS was 97.2% in H1 and 98.1% in H2 groups (p=0.39). For RS >25, node negative disease, 5-years DRFS was 89% in H1 and 90.8% in H2 groups (p=0.33) and 5-year OS was 90.6% in H1 and 90.8% in H2 groups (p=0.7). Similarly, for node positive disease, DRFS and OS were comparable between H1 and H2 group in all subgroups. Conclusions: Low HER2 mRNA expression by Oncotype RT-PCR assay was associated with significantly high Oncotype RS. Patients' outcome was not directly affected by HER2 score and was determined by RS. Further study is desired to clarify the role of HER2 expression in early stage, luminal breast cancer. Table. Comparison of age and tumor characteristics between low (H1) and high (H2) HER2 score Citation Format: Hadar Goldvaser, Raz Mutai, Salomon Stemmer, Iryna Kuchuk, Margarita Tokar, Shani Paluch-Shimon, Rinat Yerushalmi, Karen Drumea, Ella Evron, Amir Sonnenblick, Einav Gal-Yam, Gil Bar Sela, Ayelet Shai, Rotem Merose, Avital Bareket-Samish, Lior Soussan-Gutman. The impact of Human Epidermal growth factor Receptor 2 (HER2) expression by RT-PCR on tumor characteristics and outcome in estrogen receptor (ER) positive, HER2 negative early-stage breast cancer [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-01-08.

Abstract PO3-15-04: Oncotype DX Recurrence Score as an Informative Tool to Optimize Neoadjuvant Therapy in HR-positive, HER2-negative Breast Cancers

Abstract Oncotype DX Recurrence Score as an Informative Tool to Optimize Neoadjuvant Therapy in HR-positive, HER2-negative Breast Cancers Background: Neoadjuvant therapy (NAT) in HR+/HER2- breast cancers rarely leads to pathological complete response (pCR) and often doesn’t achieve substantial tumor shrinkage. Therefore, neoadjuvant therapy in HR+/HER2- breast cancers has the most value in the assessment of in vivo treatment response. For this purpose, biomarkers that can predict clinical and pathological response to systemic therapy could help inform treatment. The role of Oncotype DX testing in the management of early-stage HR+/HER2- breast cancer has been widely recognized in the adjuvant setting. However, no large prospective studies evaluating the Oncotype DX test in the neoadjuvant setting have clearly shown its predictive or prognostic utility to date. In this study, we evaluated the effect of the recurrence score (RS) on systemic treatment recommendations and its correlation to NAT response. Methods: Patients with clinically selected T2-T4 and/or clinically node positive HR+/HER2- breast cancer who were referred to BC Cancer Vancouver for consideration of NAT between September 2021 – April 2023 were screened for study eligibility. Suitable candidates who agreed to participate in the study had Oncotype DX testing on core biopsy specimens and Ki-67 testing on the core biopsy and final resection specimen. Clinico-pathological information, treatment regimens, operative details, and clinical, radiologic, and pathological responses were recorded. Results: Of the 66 patients who met eligibility criteria, 40 enrolled in the study comprising the accrual rate of 60.6 %. There were multiple factors that influenced patients and/or Oncologists participation in the study highlighting the low uptake of NAT in HR+/HER2- tumors. Technical feasibility of obtaining Oncotype DX from core biopsy samples was 95% which correlates with our previously reported data on a smaller cohort. The mean turnaround time from patient consent to RS report was 18.5 calendar days. The mean time from the initial consult to the start of NAT was 18.8 calendar days. Amongst all tumors tested, 28 % had RS equal or greater than 26 while only 2 % of participants had a score less than 10. Overall, 32 % of treatment recommendations were changed based on RS. Approximately 31% of patients who were initially recommended to receive neoadjuvant chemotherapy were spared from this treatment based on low RS. Conversely, 23.8% of patients who were recommended to pursue upfront surgery were escalated to chemotherapy due to a high RS. Approximately 88 % of the participants who were treated with neoadjuvant chemotherapy had a favourable clinical response and 38.5% had a complete clinical response. There were no pCR observed amongst patients who completed surgery after neoadjuvant chemotherapy, but half of them were rendered eligible for less surgery (either spared a mastectomy and/or an axillary dissection). Conclusions: Oncotype DX testing changed systemic treatment recommendations in one third of neoadjuvant patients and presents a unique opportunity for personalized medicine. Pre-operative testing thus influenced clinical decision-making, allowing treatment de-escalation and omission of chemotherapy in those with a low RS and conversely an earlier escalation of systemic therapy for those with a high RS. Turnaround time of the Oncotype DX assay on a core biopsy sample was acceptable and did not significantly delay treatment initiation. The benefit of the Oncotype DX assay prior to NAT was valuable for predicting in vivo response although pCR was not observed in any of the patients enrolled. Analyses are ongoing to further correlate RS to neoadjuvant response and to dynamic Ki-67 and MRI changes. Citation Format: Lidiya Luzhna, Stephen Chia, Mehrnoosh Pauls, Nathalie Levasseur. Oncotype DX Recurrence Score as an Informative Tool to Optimize Neoadjuvant Therapy in HR-positive, HER2-negative Breast Cancers [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-15-04.

Addressing Social Determinants in the Era of Precision Medicine in Breast Cancer: Is It Sufficient to Reduce Disparities?

The Oncotype DX (ODX) assay predicts recurrence risk and demonstrates the benefits of adjuvant therapy in patients with early-stage, hormone receptor (HR)-positive/HER2-negative breast cancer. ODX uptake varies by patients' racial/ethnic backgrounds and socioeconomic status (SES). However, community-level variability remains unknown, and research regarding the association between testing status and receipt of adjuvant chemotherapy is limited. To fill these knowledge gaps, Van Alsten and colleagues found a 6% lower prevalence of ODX uptake among patients residing in high SES-deprived areas than among those residing in low SES-deprived areas. Among patients with low and median ODX recurrence scores, those who underwent testing were 28% and 21% less likely to receive adjuvant chemotherapy than those who did not, respectively. The findings emphasize the role of social determinants of health. However, to further reduce or eliminate racial/ethnic disparities and SES inequities, we would need sufficient and effective multi-level approaches. These involve lower ODX testing costs, health insurance coverage expansion, re-classification and validation of ODX recurrence scores in patients of minority ancestry, and the development of a faster, more accurate, and affordable test. See related article by Van Alsten et al., p. 654.

A cost-consequence model of using the 21-gene assay to identify patients with early-stage node-positive breast cancer who benefit from adjuvant chemotherapy in the Netherlands

Background Patients with early-stage hormone receptor positive, human epidermal growth factor receptor-2 (HER2) negative invasive breast cancer with 1–3 positive lymph nodes (N1) often undergo surgical excisions followed by adjuvant chemotherapy (ACT). Many patients have no benefit from ACT and receive unnecessary, costly treatment often associated with short- and long-term adverse events (AEs). Gene expression profiling (GEP) assays, such as the 21-gene assay (i.e. the Oncotype DX assay), can identify patients at higher risk for recurrence who may benefit from ACT. However, the budgetary consequence of using the Oncotype DX assay versus no GEP testing in the Netherlands is unknown. Our study therefore assessed it using a cost-consequence model. Methods A validated model was used to create the N1 model. The model compared the costs and consequences of using the Oncotype DX assay versus no GEP testing and MammaPrint, and subsequent ACT use with corresponding costs for chemotherapy, treatment of AEs, productivity losses, GEP testing, and treatment of recurrences, according to the Oncotype DX results. The model time horizon was 5 years. Results Costs for the total population amounted to €8.0 million (M), €16.2 M, and €9.5 M, and cost per patient amounted to €13,540, €27,455, and €16,154 for using the Oncotype DX assay, no GEP testing, and MammaPrint, respectively. Total cost savings of using the Oncotype DX assay amounted to €8.2 M versus no GEP testing and €1.5 M versus MammaPrint. Using the Oncotype DX assay would result in fewer patients receiving ACT and thus fewer AEs, sick days, and hospitalizations, leading to overall cost savings compared with no GEP testing and MammaPrint. Conclusions Implementing Oncotype DX testing in this population can prevent unnecessary overtreatment, reducing clinical and economic burden on the patient and Dutch healthcare system.

A UK prospective multicentre decision impact, decision conflict and economic evaluation of the 21-gene assay in women with node+ve, hormone receptor+ve, HER2-ve breast cancer

BackgroundFor a tumour profiling test to be of value, it needs to demonstrate that it is changing clinical decisions, improving clinical confidence, and of economic benefit. This trial evaluated the use of the Oncotype DX Breast Recurrence Score® assay against these criteria in 680 women with hormone receptor-positive (HR+), HER2-negative early breast cancer with 1–3 lymph nodes positive (LN+) in the UK National Health Service (NHS).MethodsPrior to receipt of the Recurrence Score (RS) result, both the physician and the patient were asked to state their preference for or against chemotherapy and their level of confidence on a scale of 1–5. Following receipt of the RS result, the physician and patient were asked to make a final decision regarding chemotherapy and record their post-test level of confidence.ResultsReceipt of the RS result led to a 51.5% (95% CI, 47.2–55.8%) reduction in chemotherapy, significantly increased the relative and absolute confidence for both physicians and patients and led to an estimated saving to the NHS of £787 per patient.ConclusionThe use of the Oncotype DX assay fulfils the criteria of changing clinical decisions, improving confidence and saving money.

A new methodology to predict the oncotype scores based on clinico-pathological data with similar tumor profiles.

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.

Molecular Characterization of HER2-Low Invasive Breast Carcinoma by Quantitative RT-PCR Using Oncotype DX Assay.

HER2 immunohistochemistry (IHC) reproducibility is suboptimal for HER-low cases (IHC 1+ or 2+). The Yale cohort included 214 stages I-II estrogen receptor positive breast cancers with IHC scores 0, 1+, and 2+ and routine Oncotype DX Recurrence Score (RS) results. The Exact Sciences (ES) cohort included 9 57 624 patients who had an Oncotype DX RS assay that assigns HER2-negative, equivocal, or positive status based on HER2 mRNA levels. HER2 mRNA levels varied across IHC categories but with increasing medians of 9.10 (n = 89), 9.20 (n = 71), and 9.45 (n = 54) in IHC 0, 1+, and 2+, respectively. 22.4% of HER2-low (1+/2+) cancer had RS > 25. Over 98% of HER-low cancers were HER2-negative by Oncotype DX assignment. Cancers with higher mRNA levels exist within IHC 0 and low categories, most of the HER2-low patients by IHC have low RS indicating no benefit from current adjuvant chemotherapies.

Predicting Chemotherapy Benefit across Different Races in Early-Stage Breast Cancer Patients Using the Oncotype DX Score.

Oncotype DX assay, a multigene molecular test, has been widely used to stratify relapse risk and guide chemotherapy treatment in breast cancer. However, the optimal threshold of the Oncotype DX score in predicting chemotherapy benefit and its racial variation has not been investigated. In this study, we apply a random forest survival model to the SEER-Oncotype cohort data (Surveillance, Epidemiology, and End Results with Oncotype DX test information for breast cancer patients) and determine chemotherapy benefit thresholds in early-stage, estrogen-receptor-positive (ER+), and HER2-negative (HER2-) patients of different races. Our results indicate that early-stage ER+, HER2-, and LN-/LN+ patients may benefit from receiving chemotherapy at a lower Oncotype DX score than current guidelines (Recurrence Score, RS > 25 or RS > 30) suggest. According to the estimated chemotherapy sensitivity thresholds from our models, 2.05-2.72-fold more lymph-node-negative (LN-) and 2.08-5.02-fold more lymph-node-positive (LN+) patients who may not currently be recommended for chemotherapy by their Oncotype DX test result may actually have the potential to benefit from chemotherapy. Furthermore, our models indicate a racial difference in chemotherapy benefit: white, black, and Asian women with early-stage ER+/LN- tumors benefit from chemotherapy when their Oncotype DX scores are greater than 19.9, 37.2, and 18.0, respectively. Our study provides a method for calibrating multigene molecular tests to help guide treatment decisions in racially and ethnically diverse patients with cancer. Specifically, we identify key chemotherapy sensitivity thresholds for the Oncotype DX recurrence score test in breast cancer patients and provide evidence that certain patients may benefit from receiving chemotherapy at a lower threshold than the current clinical guidelines suggest.

ECOG-ACRIN E2197: Comparison of HER2 gene expression by RT-PCR across all HER2 immunohistochemistry groups with recurrence analysis.

515 Background: Accurate assessment of HER2-low [immunohistochemical (IHC) score of 1+ or 2+ and in-situ hybridization negative] breast cancer is clinically relevant following DESTINY-Breast04. In hormone receptor positive breast cancers, Oncotype DX assay Recurrence Score results are widely used to guide adjuvant chemotherapy selection, and standardized quantitative HER2 mRNA reference-normalized expression levels by RT-PCR are included in patient reports. Previously, a high degree of concordance between central IHC and quantitative HER2 by RT-PCR was demonstrated in E2197. Here, we compare quantitative HER2 gene expression using Oncotype DX assay within all IHC subgroups and risk of recurrence in E2197. Methods: Data analyzed were from a case-control sample from E2197 (no adjuvant anti-HER2 therapy) of 755 patients. Central IHC for HER2 used duplicate 1.0 mm microarrays (HercepTest, Dako) and scored per 2007 ASCO/CAP guidelines. Based on quantitative reference-normalized RT-PCR measures of HER2 expression, cases were assigned: positive ≥11.5 units, equivocal ≥10.7 to <11.5 units, and negative <10.7 units (1 unit increase represents ~2-fold increase in RNA). Spearman’s rank correlation was used to measure strength of association between the 2 methods. Cox proportional hazards regression was used to evaluate the association between quantitative HER2 expression and recurrence (defined as invasive breast cancer in local, regional, or distant sites). Analyses were weighted to account for study design. Results: 48% were IHC 0, 11% were IHC 1+, 23% were IHC 2+, and 18% were IHC 3+. 85% were HER2 negative, 3% equivocal, and 11% positive by RT-PCR. While there was moderately strong correlation between HER2 IHC and HER2 gene expression by RT-PCR [Spearman’s rho: 0.63 (95% CI: 0.59, 0.67); p<0.001], there were wide and overlapping ranges of HER2 gene expression across all IHC categories with almost identical medians and inter-quartile ranges for HER2 IHC 1+ and 2+ (n=248; see table). An exploratory analysis showed 86% of IHC 0, 97% of IHC 1+, and 99% of IHC 2+ had HER2 >8 units. After controlling for baseline clinicopathologic features (age, hormone receptor status, grade, size, and nodal status), continuous quantitative HER2 expression was associated with time to recurrence in HER2 IHC 0 to 2+ [HR 1.17 per 1-unit increase in HER2 (95% CI 1.01, 1.35; p=0.036)]. Conclusions: There was a wide range of quantitative HER2 gene expression using Oncotype DX assay across all IHC categories; continuous HER2 expression is independently associated with recurrence. Quantitative HER2 expression may be useful for identification of HER2-low breast cancer. Further studies are needed. [Table: see text]

The impact of the 4R Oncology model on OncotypeDx turnaround time.

531 Background: The OncotypeDx genomic assay has been widely used to predict recurrence in those with early-stage ER positive breast cancer to determine whether adjuvant chemotherapy is beneficial. Receiving OncotypeDx results in a timely manner prior to medical oncology appointment is critical to inform adjuvant therapy decisions. In this quality improvement project, we implemented a care delivery optimization for patients with breast cancer receiving OncotypeDx assays in an integrated healthcare system. The optimization was conducted as part of implementing the 4R Oncology model (Right Info / Right Care / Right Patient / Right Time), a novel care model that aims to optimize patient-centric care planning, team-based delivery and patient self-management. Herein, we assess the impact of implementing this model on improving OncotypeDx turnaround time. Methods: We examined two patient groups with newly diagnosed early-stage ER positive breast cancer who attended a multidisciplinary clinic and underwent surgery at a single medical center. The historical control (HC) group received care pre-4R implementation, from January 2019 through June 2020, while the 4R Intervention group received care from July 2020 to December 2021. The OncotypeDx-related care optimization was a team effort of 5 medical specialties that involved establishing an OncotypeDx reflex testing practice, updating workflows, working with a testing company to streamline the process and implementing a method to align result timing with scheduling medical oncology appointments. Bivariate analyses were used to compare the HC and 4R groups on demographic characteristics and OncotypeDx turnaround time. Results: Within the 4R group (N = 208), 72 (34.6%) patients received OncotypeDx and 89 (32.5%) patients from the HC cohort (N = 274) received OncotypeDx. The groups were balanced by median age at time of multidisciplinary clinic visit (HC: 59 years vs 4R: 60 years), female gender (HC: 99.6% vs 4R: 99.5%), race, combined race and ethnicity, language spoken, Charlson Comorbidity Index and tumor characteristics. The average turnaround time from surgery date to delivery of OncotypeDx results was significantly reduced in the 4R cohort (19.4 days) compared to that of the HC cohort (23.1 days), a difference of 3.8 days, p = 0.004. The availability of OncotypeDx results by the first medical oncology visit after surgery increased from 40% in the HC cohort to 69% in the 4R cohort, p = 0.0003, helping optimize the transition to adjuvant therapy for the appropriate population. Conclusions: The implementation of the 4R Oncology model within an integrated healthcare system significantly reduced turnaround time for patients with breast cancer awaiting OncotypeDx results, improving patient-centric care and enabling informed discussions at medical oncology visits. The 4R Oncology model is an effective method to optimize delivery of time-sensitive care.

Abstract 5354: Prediction of OncotypeDX high risk group for chemotherapy benefit in breast cancer by deep learning analysis of hematoxylin and eosin-stained whole slide images

Abstract Introduction: Breast cancer is the most common cancer in women, with most patients diagnosed with early-stage disease. While adjuvant therapy should be considered in all fit patients, in many low-risk patients drug toxicity may out-weigh the potential benefit, therefore, identifying these patients is desired. Hematoxylin and eosin (H&E) is the basic staining routinely performed for biopsies. It allows visual examination of the tissue and cells. Such manual examination, however, does not provide information about the molecular profile of the cancer, which is essential for diagnosis and guidance of treatment. The OncotypeDX assay is recommended for patients with node-negative, estrogen receptor-positive invasive breast cancer for determining chemotherapy benefit. It is based on RT-PCR gene expression profiling and provides a recurrence score (RS) that enables patient stratification to non-high risk (RS < 26) and high risk (RS ≥ 26). It was shown that high risk patients are likely to benefit from chemotherapy, while non-high risk are not. Unlike H&E staining, the OncotypeDX assay is costly, time consuming, and inaccessible in low-income countries. Here, we sought to evaluate whether analysis of scanned H&E-stained slides by convolutional neural networks (CNNs) could predict the RS risk group (high versus non-high), which determines eligibility for chemotherapy. Methods: 684 H&E-stained slides were collected from 430 invasive breast cancer patients who were assayed for OncotypeDX between 2014 and 2020 at Sheba medical center, Israel. The slides were scanned at 0.25 micron/pixel, and automatically segmented and split to 256 × 256 non-overlapping tiles, resulting in overall 339,986 tile images containing tissue. The patients were randomly split into training (75%) and test (25%) sets, and a CNN model was trained and validated on the training set to classify each tile to non-high risk (RS < 26) versus high risk (RS ≥ 26), in 5-fold cross-validation. The final model was then applied to the held-out test set, and tile scores were aggregated to produce per-patient prediction scores. The final CNN prediction scores on the test set were compared to the ground truth risk group and the AUC performance was calculated. Results: The AUC performance of the model on the held out test set for high-risk versus non-high-risk classification based on the H&E images alone was high (0.798, 95% CI: 0.689 - 0.875, P value < 0.001), showing that the H&E image analysis could predict the high risk group. Conclusions: These results show, for the first time, that CNN-based analysis of H&E images could predict benefit from chemotherapy, thus implying distinct tumor morphologies differing between RS groups. Utilizing such a system may enable physicians in countries that lack genetic profiling capabilities to refine chemotherapy stratification based on H&E images alone. Citation Format: Gil Shamai, Ran Schley, Ron Kimmel, Nora Balint-Lahat, Iris Barshack, Chen Mayer. Prediction of OncotypeDX high risk group for chemotherapy benefit in breast cancer by deep learning analysis of hematoxylin and eosin-stained whole slide images. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5354.

Abstract 5787: Modelling the spatial heterogeneity of CD45-positive tumor infiltrating lymphocytes in early-stage, estrogen receptor-positive breast cancer

Abstract The frequency of lymphocytes infiltrating tumors is a known prognostic in estrogen receptor (ER) negative cancers. ER+ disease is putatively believed to be immune cold, however, there exists a subset of ER+ tumors with high immune infiltrate and with a significant spatial heterogeneity. The clinic impact of such infiltrate - especially between the Oncotype Dx Recurrence Score risk categories - remains unclear. Moreover, the distribution of tumor and stromal tissues, while noted as significantly heterogenous, is still ill-defined and not yet clinically used prognostically, despite evidence to support its utility. Using a cohort (n=450) of serial sections taken from early-stage, ER+/HER2- breast tumors of the Irish arm of the TAILORx clinical trial, we aimed to investigate the tumor architecture and spatial distribution of tumor immune infiltrate and proliferating tumor cells using digital image analysis. Antibodies against Ki67 (proliferation marker) and CD45 (leukocyte common antigen), and a routine Haematoxylin and Eosin stain were applied to serial sections of 450 full-face tumors via chromogenic immunohistochemistry, as outlined previously [1]. Digital image analysis was performed using open-source software, QuPath [2]. Pixel classifiers were trained and validated against an expert pathologist in order to define observed lymphocytes as tumor or stromal-infiltrating, and to establish a classifier to quantify the tumor-stroma ratio (TSR) and infiltrating tumor area. Distances of CD45-positive cells from tumor were computed, along with autocorrelation statistics [3,4] of CD45 and Ki67 hotspots; firstly in order to quantify spatial heterogeneity, and secondly to examine whether Ki67 as a component gene in the Oncotype Dx assay has a foundation in tumor biology or is being confounded by potentially Ki67-positive lymphocytes. Subdividing by Oncotype Dx risk categories, no significant difference in TSR was observed (p=0.09799), neither for intermediary risk patients receiving hormone therapy alone or in combination with chemotherapy (p=0.3873). While there was an observed trend overall (p=0.092), no significance was found for recurrence between intermediary risk subcategories (HT alone: p=0.393, HT+CT: p=0.288). However, in the cohort as a whole, median TSR was 0.3215 (range 0 - 5.023), with statistically significant differences in recurrence risk observed (cohort high v low by median TSR. HR: 6.356, 95CI: 2.263-17.84, p<0.0001). Citation Format: Zak Kinsella, Anna Blümel, Mairi Lucas, Andreas Lindner, Claudia A. Gonzalez, Arman Rahman, Joanna Fay, Tony O'Grady, Verena Murphy, John Crown, Cathy Kelly, William Gallagher, Darran O'Connor. Modelling the spatial heterogeneity of CD45-positive tumor infiltrating lymphocytes in early-stage, estrogen receptor-positive breast cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5787.

Abstract P6-01-41: Influence of HER2 expression status in the distribution of recurrence score from the OncotypeDx assay among women with early-stage estrogen-receptor-positive/HER2-negative breast cancer

Abstract BACKGROUND: The OncotypeDX assay analyzes the expression of 21 genes, including HER2 and Grb7, to assess the predictive effect of chemotherapy in breast cancer recurrence among women with early-stage estrogen-receptor (ER) positive/HER2-negative breast cancer. In this study, we evaluate the distribution of recurrence score from the OncotypeDX assay based on the HER2 expression status by immunohistochemistry (IHC). METHODOLOGY: Utilizing the National Cancer Database, we identified adult women with a diagnosis of stage I – III ER+/HER2- invasive ductal carcinoma of the breast between the years 2010 and 2018. Recurrence scores were classified into low (< 11), intermediate (11-26), and high risk (>26); and the distribution of these recurrence scores cross-tabulated with different IHC scores of HER2 expression (0, 1+, and 2+) by IHC. Women with an IHC score of 2+ had to be negative for HER2 gene amplification by FISH. Multivariate logistic regression model adjusting for age, race, origin, progesterone receptor status, grade, and cancer stage was used to assess the odds of receiving a high OncotypeDx Score (≥26). RESULT: Among 198,931 women with ER+/HER2 negative breast cancer, 59,632 (30.0%) had HER2 IHC score of 0, 102,170 (51.4%) had IHC score of 1+, and 37,129 (18.6%) women were with IHC score of 2+. The median age at diagnosis of breast cancer among all three categories of HER2 expression was 59 years. The median recurrence score for women with IHC scores of 0,1+, and 2+ was 15, 15, and 16 respectively (p < 0.001). A higher proportion of women with HER2 IHC score of 2+ had a high recurrence score compared to women with an IHC score of 0 (15.2% vs. 13%, p < 0.001). In multivariate analysis, compared to women with HER2 IHC score of 0, women with HER2 IHC score of 2+/FISH negative were observed to have higher Odds (OR: 1.16; 95% CI: 1.11 – 1.20, p < 0.001) of receiving high recurrence score. There was no significant difference in the odds of receiving a high recurrence score between women with IHC 0+ and 1+(OR: 0.99; 95% CI: 0.96 – 1.03, p 0.82). CONCLUSION: Women with a HER2 IHC score of 2+ were observed to have a higher odds of receiving high-risk recurrence scores as compared to women with IHC score of 0. This needs to be further correlated with the response to chemotherapy and the risk of recurrence. Citation Format: Anish Shah, Pravash Budhathoki, Suman Gaire, Utsav Joshi, Siddhartha Yadav. Influence of HER2 expression status in the distribution of recurrence score from the OncotypeDx assay among women with early-stage estrogen-receptor-positive/HER2-negative breast cancer [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-01-41.

Abstract P5-03-15: Application of 21-gene Breast Recurrence Score® assay to evaluate prognosis and benefit of adjuvant chemotherapy in BRCA1 and BRCA2 pathogenic variant carriers with early stage, estrogen receptor positive breast cancer

Abstract Background: Minimal data exists for the utilization of the Oncotype Dx® assay specifically in breast cancers associated with BRCA1/2 pathogenic variants (PVs). It is unknown whether estrogen receptor positive (ER+) breast cancer associated with an inherited BRCA1 or BRCA2 (BRCA1/2) PV is more aggressive than disease seen in patients who do not carry an inherited PV, and whether there are differences between BRCA1 and BRCA2. In prostate cancer patients with inherited cancer predisposition due to a BRCA2 PV, more aggressive cancers are observed, which influences first-line treatment. Limited data exists for the optimal management of early stage ER+ breast cancer in BRCA1/2 PV carriers. Comparing Recurrence Score® (RS) results in ER+ breast cancer patients with an inherited BRCA1/2 PV (cases) versus matched patients who test negative for a PV in BRCA1/2 (controls) may inform whether biologically more aggressive breast cancer is seen in BRCA1/2 carriers and optimal treatment approaches. Methods: A retrospective case control study was performed to compare RS results in women with breast cancer with an inherited BRCA1/2 PV versus patients who tested negative for an inherited BRCA1/2 PV. Female breast cancer patients seen between 2005-2020 at NorthShore University Health System with ER+Her2- early stage invasive breast cancer with 0-3 lymph nodes who completed genetic testing for BRCA1/2 were eligible for enrollment. BRCA1/2 cases were defined as individuals with an inherited PV in BRCA1/2 and controls were negative for BRCA1/2 or other known breast cancer risk gene PVs tested. Subjects were excluded if they had neoadjuvant therapy (hormonal or cytotoxic chemotherapy). Eligible cases were matched to control patients by age, grade, and stage. The Recurrence Score result was obtained by chart review; if not previously evaluated, Oncotype Dx assay was performed by Exact Sciences. Statistical analysis of the primary outcome used the paired t-test to determine mean difference in RS results between BRCA1/2 PV carriers and patients negative for a PV in BRCA1/2 using a 1:1 matched pairs design. Results: A total of 46 matched cases and controls were analyzed. Median age was 50 with a range of 28-74. Of the cases, 18 had a BRCA1 PV and 28 had a BRCA2 PV. Cases and controls were well matched for age (> 50 and ≤ 50); race, grade, stage, and progesterone receptor status. As expected, a higher number of BRCA1/2 carriers were treated with mastectomy while more of the controls received breast-conserving surgery. Chemotherapy was utilized more frequently in the cases (67.4%) versus the controls (54.4%). The average RS result was higher in the cases (27) than the controls (21.3) by a mean difference of 5.7 (p = 0.0195). Using Oncotype Dx cutoffs of low < 18, intermediate 18-30 and high ≥ 31, a statistically significant difference in RS result was noted in the cases versus controls. For cases in the highest risk group (Oncotype Dx ≥ 31), only 20% of their matches also had a score in the highest risk group while 35% had a score in the lowest risk group. Subgroup analysis showed that the cases had the largest difference in RS result from their controls in premenopausal women (age ≤ 50), BRCA1 carriers, and the node negative population. Conclusion We present one of the largest data sets available to date of a well-matched cohort of cases and controls which shows that BRCA1/2 PV carriers are more likely to have a higher Recurrence Score result than their matched controls when matched for age, grade, and stage. These findings suggest ER+ breast cancer in BRCA1/2 PV carriers is biologically more aggressive. Further investigation is warranted to evaluate how this important finding impacts adjuvant therapy recommendations for BRCA1/2 PV carriers. Citation Format: Poornima Saha, Ashley Aller, Amanda Deliere, Peter Hulick, Katharine Yao, Kristine Kuchta, Megan Sullivan, Allison DePersia. Application of 21-gene Breast Recurrence Score® assay to evaluate prognosis and benefit of adjuvant chemotherapy in BRCA1 and BRCA2 pathogenic variant carriers with early stage, estrogen receptor positive breast cancer [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-03-15.

Abstract P2-11-31: FAVORABLE LOCOREGIONAL CONTROL IN CLINICALLY NODE-NEGATIVE HORMONE-RECEPTOR POSITIVE BREAST CANCER WITH LOW 21-GENE RECURRENCE SCORES: A SINGLE INSTITUTIONAL STUDY WITH 10-YEAR FOLLOW-UP

Abstract Background: Recent studies have shown a lower likelihood of locoregional recurrences in patients with low 21-gene recurrence score (RS). In this single institutional study, we investigated any associations between different cut-off values of 21-gene RS and histopathological factors and outcome in patients with long-term follow-up. Methods: Between February 2010 to February 2013, 61 patients with early stage clinically node-negative hormone receptor-positive and HER2-negative breast cancer tested for the 21-gene RS assay, were included into the study. The clinicopathological, treatment and outcome characteristics were analyzed. Results: Median age was 48 (range, 29-72). Of those, 53 patients were diagnosed with Stage 1 (86.9%) and 8 patients were Stage 2 (13.1%) following surgery. Patients with high histologic grade (HG), or with Ki-67>=25% were significantly more likely to have intermediate/high RS based on RS >11 and >18 (Table 1). Based on the 21-gene RS assay, only 19 patients (31%) received adjuvant chemotherapy. At a median follow-up of 112 months, 3 patients developed local recurrences (4.9%) treated with endocrine therapy alone. Among patients treated with endocrine treatment alone (n=42), the clinicopathological characteristics including age < 40, age < 50, high histologic or nuclear grade, high Ki67-scores (>=20%, >=25%, >=30%), presence of lymphovascular invasion, luminal-A type tumor, multifocality, lymph node positivity, tumor size more than 2 cm, RS>=18 or RS>11 were not significantly found to be associated with 10-year locoregional recurrence free survival (LRRFS). However, patients with a RS >=16 have significantly shown a poorer 10-year LRRFS cpmpared to those with RS < 16 (RS < 16; 100% vs RS>=16; 75%, p=0.039, Table 2). Conclusions: These results suggest that Oncotype DX assay may be of little value in patients with high histologic grade or high Ki67 scores (>= 25%) as treatment decision criteria to determine any benefit from chemotherapy. Furthermore, patients with a RS >=16 are more likely to benefit from adjuvant chemotherapies, whereas those with a RS < 16 have an excellent outcome and local control on long term follow-up with endocrine treatment alone. Further prospective studies should be performed to validate these findings in future. Table 1. Correlation of 21-gene Recurrence Score (RS) including different cut-off values with Ki-67 scores and histopathological factors Table 2. Clinicopathologic factors and 21-gene Recurrence Score (=RS) associated with disease free survival (DFS) and locoregional recurrence free survival free survival (=LRRFS). (DFS is equal to LRRFS since none of the patients developed distant metastases) Citation Format: Cihan Uras, Neslihan Cabıoğlu, Fatma Tokat, Halil Kara, Ozlem ER, Taner Korkmaz, Nuran Bese, Umit Ince. FAVORABLE LOCOREGIONAL CONTROL IN CLINICALLY NODE-NEGATIVE HORMONE-RECEPTOR POSITIVE BREAST CANCER WITH LOW 21-GENE RECURRENCE SCORES: A SINGLE INSTITUTIONAL STUDY WITH 10-YEAR FOLLOW-UP [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-11-31.

Abstract P6-01-25: Evaluation of Multigene Assays as Predictors for Response to Neoadjuvant Chemotherapy in Early-Stage Breast Cancer Patients

Abstract Background: Oncotype DX (ODX) and MammaPrint (MP) are gene-expression assays that have been established to predict distant cancer recurrence in the adjuvant chemotherapy setting. However, they have not been validated to predict pathologic complete response (pCR) to neoadjuvant chemotherapy (NACT). We sought to examine the ability of the ODX and MP assays to predict the likelihood of pCR to NACT in early-stage breast cancer patients. Methods: Data from breast cancer patients diagnosed between 2010 and 2019 were obtained from the National Cancer Database. All patients who received NACT (at least 30 days of treatment) and had pathologic response data and ODX or MP results were included. Analysis of ODX was limited to patients with hormone receptor (HR)+/HER2- stage I-III disease, while analysis of MP included both HR+/HER2- and HR-/HER2- stage I-III patients. ODX scores were modeled both as a continuous variable and a categorical variable classified as low (0-25) and high (≥26) per the TAILORx trial cutoff, whereas MP results were modeled as a dichotomous variable (i.e., low risk and high risk) because numeric values were unavailable. Multivariable logistic regression models were used to assess the relationship between pCR (defined as ypT0/Tis ypN0) and ODX or MP results, adjusting for age, race/ethnicity, clinical T and N stages, tumor grade, and progesterone receptor status. Adjusted odds ratios (AOR) and 95% confidence intervals (CI) were calculated. Results: A total of 2,219 patients, treated at 630 institutions, who received NACT with an ODX recurrence score were included in the ODX cohort. Of 1,181 patients with a high ODX score, 11.2% achieved pCR, while only 1.6% of 867 patients with a low ODX score did. In the adjusted model, having a high ODX score was associated with greater odds of pCR (AOR = 4.48, 95% CI: 2.44-8.22). There was a significant monotonic increasing trend of pCR by continuous ODX score. The mean ODX score was 42.5 (SD = 15.5) in patients who achieved pCR, compared to 27.9 (SD = 13.7) in patients who did not; the discriminating capacity of ODX was moderate to strong (area under the ROC curve = 0.767). A total of 1,349 patients, treated at 337 institutions, who received NACT and had MP test results were included in the MP cohort. Of 1,141 patients with MP high risk disease, 11.8% achieved pCR, compared to < 4.8% of 208 patients with MP low risk disease. In the adjusted model, having MP high risk disease was associated with greater odds of pCR (AOR = 2.21, 95% CI: 1.02-4.77). A similar association between MP results and pCR was also found in the subset of patients who were HR+/HER2- (AOR = 2.25, 95% CI: 0.99-5.15). Conclusions: Both ODX and MP were independently associated with likelihood of pCR after NACT for early-stage, high-risk breast cancer. These findings suggest a potential role for ODX or MP testing as a predictive biomarker in the NACT setting, and can facilitate clinical decision making between physicians and patients. Citation Format: Jincong Q. Freeman, Sarah Shubeck, Frederick M. Howard, Nan Chen, Rita Nanda, Dezheng Huo. Evaluation of Multigene Assays as Predictors for Response to Neoadjuvant Chemotherapy in Early-Stage Breast Cancer Patients [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-01-25.

Abstract P6-01-27: Utilization of the Oncotype Dx Assay for Young Patients with Early Stage, Hormone-Receptor Positive, HER2-Negative Breast Cancer in an Integrated Health System

Abstract Background: Oncotype Dx (ODX) is one of the most widely used prognostic multigene expression assays for early-stage, hormone receptor-positive (HR+), HER2-negative (HER2-) breast cancer. Previous studies demonstrated reduction in chemotherapy in patients with a low ODX Recurrence Score (RS). Over recent years, utilization of this tool has increased in young breast cancer (YBC) patients (age≤40 years), a population often underrepresented in randomized clinical trials. The purpose of this retrospective study was to assess the utilization of this assay over time in an integrated health system and whether ODX RS results altered clinical practice. Methods: YBC patients with T1-T2N0 HR+HER2- breast cancer were retrospectively evaluated. Descriptive analysis examined the association between tumor characteristics, year of diagnosis, ODX testing, treatment, and recurrence outcomes. ODX risk categories were defined as: low risk 0-15; intermediate risk 16-25; high risk > 25. Recurrence was determined by the date of confirmation on pathology or imaging. Bivariate analysis compared characteristics between groups using Fisher exact tests for categorical variables and t-tests and nonparametric tests for continuous variables. Results: From 2008-2018, 1,436 Stage I-III YBC patients were diagnosed with invasive breast cancer, and 455 met eligibility criteria for ODX testing. Median follow-time (interquartile range, IQR) was 4.9 (2.8, 7.9) years for the 255 women who were tested and 5.7 (3.5, 8.7) years for the 200 women who were not tested (p< 0.05). Prior to 2018, 52.1% of patients were tested; after 2017, ODX testing rate increased to 88%. Overall, there were 255 patients who underwent ODX testing (55.9%). The median age (IQR) of patients who had an ODX test was 38.0 (35.0, 39.0). Of 225 tested patients, 42.0% (n=107) were White, 6.3% (n=16) Black, 33.7% (n=86) Asian/Pacific Islander, 15.7% (n=40) Hispanic, and 2.3% (n=6) identified as Other. There was no overall difference in testing based on ethnicity (p=0.42). More patients with grade 1 versus grade 3 disease were tested, 61.5% vs 45.2% (p=0.02 from overall Fisher exact test). Adjuvant chemotherapy was administered to 61.0% (122/200) patients who were not tested, whereas 38.4% (98/255) of those tested received chemotherapy (p< 0.001). In tested patients, 6% of low-risk (RS 0-15), 37% of intermediate risk (RS 16-25), and 92% of high risk (RS >25) patient received adjuvant chemotherapy. Among patients with T2 lesions, a higher proportion not tested (90.8% [59/65]) received adjuvant chemotherapy compared with those not tested (57.1% [40/70]). There were no differences in recurrence based on ODX testing, 11.0% (22/200) not tested vs. 9.4% (24/255) tested (p=0.26). Conclusions: Utilization of ODX testing increased after 2017. A significantly lower proportion of women who underwent ODX testing received chemotherapy, compared with women not tested for ODX. A higher percent of women with T2 cancer received chemotherapy if testing was not completed, which may reflect a greater fear of recurrence in younger patients. Further investigation is needed to better understand this potential risk of overtreatment in the YBC population. Citation Format: Ashley Aller, Jeanne A. Darbinian, Raymond Liu, Gillian Kuehner, Alison Savitz, Patience Odele, Laurel Habel, Brooke Vuong. Utilization of the Oncotype Dx Assay for Young Patients with Early Stage, Hormone-Receptor Positive, HER2-Negative Breast Cancer in an Integrated Health System [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-01-27.

Abstract P6-01-11: A UK prospective multicentre decision impact, decision conflict and economic evaluation of the use of Oncotype DX® to guide chemotherapy in 680 women with hormone receptor positive, HER2 negative breast cancer and 1 to 3 nodes involved

Abstract Introduction: For a test to be of value, it needs to demonstrate that it is changing clinical decisions, improving clinical confidence and of economic benefit. This trial looked at the use of Oncotype DX Breast Recurrence Score ® (RS) assay against these criteria in 680 women with hormone receptor positive (HR+), HER2 negative early breast cancer with 1 to 3 lymph nodes positive (LN+) in the UK National Health Service (NHS) (5 teaching and 9 district general hospitals) between 2017 and 2022. Methods: Patients with LN+ early breast cancer who were willing and fit to receive chemotherapy (CT) were consented to join the trial. At the initial oncologists’ appointment, physicians were asked to state their preference for or against CT and their level of confidence in their decision on a scale of 1 to 5. Following receipt of the RS result physicians were asked to make a final decision for or against CT and similarly record their level of confidence. Descriptive analyses were used to characterize (1) patient and tumour characteristics, (2) change in treatment recommendations post-RS testing (by RS result and nodal status), and (3) change in physicians’ level of confidence post-RS testing (by RS result and nodal status). Average cost for chemotherapy and RS test price were used to estimate overall cost savings. Results: A total of 680 patients were recruited. 16 patients were excluded (5 failed samples, 5 withdrew consent, 3 HER2 positives, 2 with advanced disease, and 1 specimen delayed in transit), leaving 664 assessable patients. The median age was 58 years and 77.1% of women were post-menopausal. Most patients had a RS of 0-17 (n=400, 60.2%); while 206 (31%) had a RS of 18-30 and 58 (8.7%) had a RS of 31-100. Using post-RxPONDER cutoffs, 566 (85.2%) had an RS of 0-25; 98 (14.8%) had an RS of 26-100. Decision impact results: The decision impact results broken down by RS result and nodal status are detailed in Table 1. Of the 662 patients with complete decision impact data, in 359 (54.2%) the recommendation by the physician changed from CT+ hormone therapy (HT) to HT alone. In 286 (43.2%) cases the decision was unchanged and in 17 (2.6%) the recommendation changed from HT alone to CT+HT. Overall 342 (51.7%) cases were spared chemotherapy. Decision conflict results: The change in the physicians’ level of confidence by RS result and nodal status are detailed in Table 2. Of the 660 cases with complete decision conflict data, physicians reported an increase in confidence in their recommendations after receiving the RS in 363 (55.0%), confidence was unchanged in 219 (33.2%) and decreased confidence in 78 (11.8%) cases. Economic analysis: Using the estimates of Burdanov et al, the average costs of a course of chemotherapy in the UK is £6,000 to £7,000. An estimate of the overall cost saving of 342 courses is £2,064,00 to £2,408,000 and the overall cost of 664 RS assays at the list price of about £2580 (although an undisclosed discount applies to the NHS) is £1,713,120. This suggests that the use of RS assay represents a significant saving to the NHS. Conclusion: The use of Oncotype DX assay in node positive early breast cancer leads to about half of women being spared chemotherapy, a significant improvement in clinical confidence for oncologists and an economic saving to the health care system. Table 1 Pre- vs Post-Oncotype DX Treatment Recommendation by Physician According to Recurrence Score and Nodal Status Table 2 Change in Physicians’ Level of Confidence Post-Oncotype DX Testing According to Recurrence Score and Nodal Status Citation Format: Simon D. Holt, Priyadharshini Sai-Giridhar, Mark Verrill, Laura Pettit, Anna Rigg, Tamas Hickish, Caroline D. Archer, Anshu Wadhawan, Marianne Dillon, Jo Dent, Mark R. Nathan, Ludger Barthelmes, Shazza rehman, Paige Innis, Saira Khawaja. A UK prospective multicentre decision impact, decision conflict and economic evaluation of the use of Oncotype DX® to guide chemotherapy in 680 women with hormone receptor positive, HER2 negative breast cancer and 1 to 3 nodes involved [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-01-11.

Abstract P2-11-08: Multimodal Prediction of Breast Cancer Recurrence Assays and Risk of Recurrence

Abstract Background: Hormone receptor positive breast cancer constitutes about 70% of newly diagnosed early-stage disease in the United States, and gene-expression based recurrence assays such as Oncotype DX (ODX) are strongly recommended by guidelines to aid in treatment decisions. However, recurrence assays are costly, time-consuming, underutilized in low resource settings, and unavailable in developing countries. Deep Learning (DL) using hematoxylin and eosin (H&E) stained digital pathology has been shown to approximate gene expression patterns for multiple cancer types, and may provide a cost-effective, fast, and scalable method to predict risk of recurrence in community settings. Methods: We first developed a model for ODX using only DL on pathology, comprised of two Xception-based modules, trained on 1,039 slides from The Cancer Genome Atlas (TCGA) tessellated into 10x magnification image tiles. The first module predicts tumor likelihood, and was trained using pathologist annotations for tumor regions versus normal stroma. The second module was trained to predict ODX score, estimated from gene expression data within TCGA. Patient-level predictions were calculated by weighting the predicted recurrence score by tumor likelihood for all tiles within a slide. Separately, ODX score was predicted from clinical variables using the University of Tennessee Nomogram, which incorporates grade, progesterone receptor, size, age, and histologic subtype. Finally, we developed a combined model by fitting a logistic regression to the DL pathologic model and the clinical nomogram predictions. Performance of the clinical nomogram, pathologic, and combined models were then compared in a single-institution external validation cohort of patients diagnosed with breast cancer between 2006 and 2020, all of whom had the commercial ODX assay run. Results: We identified 428 cases for our diverse validation cohort (69% White, 24% Black, 6% Asian, and 3% Hispanic) with mean ODX score of 18. Chemotherapy was administered for 104 (24.3%) of patients, the remaining 323 (75.4%) received endocrine therapy alone. Area under the receiver operating characteristic curve (AUROC) for prediction of high ODX score (≥ 26) of the combined model was 0.83 (95% confidence interval [CI] 0.78 – 0.89) in the validation cohort, which was higher than either the DL pathology model (AUROC 0.80, 95% CI 0.75 – 0.85, p = 0.026) or the Tennessee nomogram (AUROC 0.77, 85% CI 0.70 – 0.83, p = 0.003). Performance was similar in Black (AUROC 0.86, 95% CI 0.78 – 0.94) and White (AUROC of 0.81, 95% CI 0.74 – 0.88) subgroups. The combined model was more accurate in prediction of recurrence-free interval in patients receiving endocrine therapy (hazard ratio [HR] 2.02 per standard deviation [SD], 95% CI 1.16 – 3.52, p = 0.013, Concordance [C]-index 0.75) than the clinical nomogram (HR 1.75 per SD, 95% CI 1.09 – 2.81, p = 0.021, C-index 0.68). No model was prognostic in patients receiving chemotherapy. Pathologist review of heatmaps of DL model predictions identified lymphovascular invasion, necrosis, high grade, and infiltrative borders as features contributing to model prediction of high risk. Conclusions: DL can improve on existing clinical prediction of breast cancer with low recurrence risk. This approach could improve the speed at which treatment decisions are made due to the time-consuming nature of genomic testing and simultaneously reduce the cost of care. Given the equal performance in racial subgroups, this approach has promise for application in global health settings where genomic assays are not widely available or are prohibitively expensive. Citation Format: Frederick M. Howard, James M. Dolezal, Sara Kochanny, Galina Khramtsova, Jasmine Vickery, Andrew Srisuwananukorn, Anna Woodard, Nan Chen, Rita Nanda, Charles Perou, Olufunmilayo I. Olopade, Dezheng Huo, Alexander Pearson. Multimodal Prediction of Breast Cancer Recurrence Assays and Risk of Recurrence [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-11-08.