Node-negative Early-stage Breast Cancer Research Articles

Abstract PO1-01-10: Comparison of the cost-effectiveness of multigene assays for HR+/HER2- node-negative early-stage breast cancer in the US

Abstract Background Traditional clinico-pathologic features, such as tumor grade and size, age, and menopausal status, may provide incomplete prognostic information, and have limited ability to predict chemotherapy benefit for patients with hormone-receptor-positive (HR+), human epidermal growth factor receptor 2-negative (HER2)-, axillary lymph node-negative (N0) early-stage breast cancer, leading to both under- and over-treatment with adjuvant chemotherapy. Multigene expression profile tests have been used in clinical practice to complement clinico-pathologic information to optimize treatment decisions. To date, the Oncotype DX Breast Recurrence ScoreÒ test is the only multigene assay with the ability to predict chemotherapy benefit validated by evidence from randomized clinical trials and supported by NCCN and ASCO guidelines. Other commercially available assays in the US include MammaPrintÒ, EndoPredictÒ and Prosigna RORÒ, which offer prognostic information beyond traditional clinico-pathologic features to improve treatment decisions. The relative cost-effectiveness of the four multigene assays compared to using clinico-pathologic risk alone to guide chemotherapy decisions was assessed from a societal perspective in the US. Methods A cost-effectiveness model built in Microsoft Excel estimated the cost-effectiveness of the Oncotype DXÒ test, MammaPrint, EndoPredict and Prosigna ROR compared to using clinical-pathological risk alone over a lifetime horizon. The model structure consisted of a decision-tree portion informed using published studies for the distribution of distant recurrence risk categories as defined by each test (low, intermediate, and high risk for the Oncotype DX test and Prosigna ROR, low and high risk for MammaPrint and EndoPredict). Decision impact studies and clinical expert opinion informed the proportion of patients allocated to adjuvant chemotherapy conditional on test result category. A Markov discrete health state model then simulated the future risk of distant recurrence, and long-term adverse effects of chemotherapy: acute myeloid leukemia (AML), and congestive heart failure (CHF), and aggregated the costs and outcomes of treatment decisions over a patient’s lifetime. The probability of distant recurrence of breast cancer conditional on test result category was derived from the most up-to-date studies assessing the performance of tests: TAILORx for the Oncotype DX test, MINDACT for MammaPrint, and TransATAC for EndoPredict and Prosigna ROR. Costs were estimated from a US societal perspective in 2020 US dollars (including Medicare healthcare costs, lost productivity and patient out-of-pocket chemotherapy costs from literature), and outcomes were expressed in terms of life-years, quality-adjusted life-years (QALYs), and net monetary benefit (NMB). The cost-effectiveness of each multigene assay compared to using clinico-pathologic risk alone was presented using the incremental cost-effectiveness ratio (ICER). Results Using the Oncotype DX test resulted in the largest lifetime cost savings (-$13,395) and generated the largest QALY gain (0.25) out of the four multigene assays included in the model, when compared to using clinico-pathologic risk alone (Table 1). All four multigene assays represented a cost-effective use of resources from a societal perspective (based on the accepted willingness-to-pay threshold of $100,000 per QALY gained), with a low net lifetime cost or net cost savings with positive QALY gain. Conclusion These results support the use of multigene assays to guide treatment decisions in node-negative HR+/HER2- early-stage breast cancer to improve treatment outcomes, and also result in substantial net cost savings from a societal perspective. The prediction of chemotherapy benefit using the Oncotype DX test resulted in substantially higher cost savings and improved breast cancer outcomes. Citation Format: Vladislav Berdunov, Gebra Cuyun Carter, Christy Russell, Sara Campbell, Jennifer Racz, Yara Abdou. Comparison of the cost-effectiveness of multigene assays for HR+/HER2- node-negative early-stage breast cancer in the US [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 PO1-01-10.

Abstract PO1-02-05: An exploratory comparison of MINDACT and TAILORx genomic risk proportions and outcomes in patients with HR+/HER2-, node-negative early-stage breast cancer, stratified by clinical risk

Abstract Introduction: Genomic tests provide physicians with critical information to evaluate the risk of distant metastasis and inform treatment plans by identifying those patients who may safely avoid chemotherapy (CT). The 70-gene signature (MammaPrint, MP) through the MINDACT trial and the 21-gene signature (Oncotype DX, ODx) through the TAILORx trial are the only tests validated with level 1A evidence that can be used for CT de-escalation decisions. Differences in trial design have made direct comparison of results difficult. We provide a comparison of MammaPrint and ODx genomic risk distributions and associated survival in patients with HR+/HER2-, node-negative (N0) early-stage breast cancer (EBC), stratified by clinical risk. Methods: Data from the MINDACT 2021 publication in The Lancet Oncology and landmark TAILORx 2019 publication in The New England Journal of Medicine were reviewed, in addition to publicly available data submitted to the Dutch Health Technology Assessment (HTA) authorities regarding survival data in patients with clinical High Risk in TAILORx. In MINDACT, patients were randomized based on both clinical (as per MINDACT definition) and genomic risk (clinical high/MP low, randomized to treatment according to clinical risk [i.e. CT] vs treatment according to MammaPrint risk [i.e. no CT]), while in TAILORx randomization was based on genomic risk alone. In TAILORx, enrollment for the population in the RS 11–25 group was enriched by 73% to account for the larger than expected non-adherence to randomized treatment plan based on the ODx results, as described in the original publication. This enrichment was only done for the RS 11–25 group. In this analysis, this enrollment enrichment was removed by applying the same proportional 73% corrective factor to the RS 11–25 group. Outcome data are reported for Overall Survival (OS), the only comparable endpoint used in both trials. Results: This explorative comparison focuses on patients with HR+/HER2- and N0 EBC in MINDACT (n=4,225) and TAILORx (n=6,686, without population enrichment). Similar proportions of patients identified as candidates for CT de-escalation (MP Low Risk or ODx RS≤25) were observed in this population of MINDACT (75%) and TAILORx (80%) when removing the RS 11–25 population enrichment in TAILORx (Table). In clinical high risk patients, the proportion with genomic low risk of distant metastasis were also comparable between MINDACT (55%) and TAILORx (63%) (Table). In the original clinical high risk but MammaPrint Low Risk population treated without CT in MINDACT the reported 8-year OS was 93.9%. In TAILORx, the comparable population of patients (ie. clinical high risk, low genomic risk, not treated with CT) had a 9-year OS of 89.2% in the RS 0–10 group and 91.6% in the RS 11–25 group (Table). Conclusions: Accounting for differences in trial design and population, comparable proportions of patients with genomic low risk results were observed between MP and ODx, justifying similar rates of CT de-escalation. Patients with clinical high risk tumors who are MammaPrint Low Risk had an excellent survival at 8 years. Genomic risk distributions and associated OS in HR+HER2- N0 patients in MINDACT and TAILORx MP, MammaPrint; OS, Overall Survival; RS, Recurrence Score. Number of patients in the TAILORx RS 11 – 25 group are adjusted to remove the proportional enrichment in the trial. OS data presented per original MINDACT and TAILORx populations. Citation Format: Patrick Neven, Josephine Lopes-Cardozo, Fatima Cardoso, Laura Van ’t Veer. An exploratory comparison of MINDACT and TAILORx genomic risk proportions and outcomes in patients with HR+/HER2-, node-negative early-stage breast cancer, stratified by clinical risk [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 PO1-02-05.

Carbon nanoparticles versus patent blue dye for detection of sentinel lymph node in patients with early breast cancer

Background Sentinel lymph nodes biopsy (SLNB) has replaced axillary lymph node dissection (ALND) in a considerable percentage of patients with early-stage breast cancer which was a great advance in preventing many surgical complications and enhancing their health welfare. Although there are different sentinel lymph nodes (SLNs) tracers with different identification rates, there is no agreement about the idealistic method. Aim The study was designed to compare carbon nanoparticles and patent blue v dye regarding SLNs detection rate, number of SLNs, time of detection, metastatic SLNs, cost, and safety in patients with early breast cancer and clinically node-negative axilla. Patients and methods A total of 40 patients with axillary lymph node-negative early-stage breast cancer patients were divided into two groups and subjected to carbon nanoparticles and patent blue V dye in group A and group B, respectively. Patients who were pregnant or lactating had node-positive axilla (N1-3) or metastatic breast cancer (M1) or had neoadjuvant chemotherapy were excluded. Results The mean age was 48.3±9.5 and 47 ±8.9, while the mean BMI of 33.3±4.8 and 32.834±4.862, the SLN detection rates were 95% and 90% in groups A and B, respectively. A total of 128 sentinel lymph nodes (SLNs) were removed from patients in the two groups (65 with Carbon Nanoparticles and 63 with patent blue dye). The mean number of SLNs was 3.4±0.7 (range, 2–5) and 3.5±1.2 (range, 2–7), mean time of SLNs detection was 13.5±4.5 (range, 7–22) and 12.7± 3.6 (range, 7–18 min) between group A and B, respectively. Conclusion There is no significant difference between carbon nanoparticles and patent blue dye regarding axillary SLNs in early breast cancer regarding identification rate, number of SLNs, time of detection, metastatic SLNs, cost, and safety with slight preference to carbon nanoparticles regarding postoperative skin staining and Egyptian market availability.

Long-term outcomes and predictors of recurrence in node-negative early stage breast cancer patients.

We evaluated the outcomes, and risk factors for recurrence in patients with early stage node-negative breast cancer in this study. Retrospective data analysis was done on patient treatment records from 1988 to 2018. The patient's demographic, clinical, pathological, and therapeutic characteristics were noted. To evaluate survival analysis and predictors of recurrence, we employed Kaplan-Meier analysis with the log-rank test. A total of 357 patients in all were enrolled in the research. At the time of diagnosis, the median age was 50 (with a range of 18-81). A total of 85.5% of patients had undergone a lumpectomy, while 14.5% had a mastectomy. 78.7% of patients had sentinel lymph node biopsy, and 21.3% had axillary lymph node dissection. In addition, the patients received adjuvant radiotherapy (88.7%), adjuvant endocrine therapy (82.1%), and adjuvant chemotherapy (48.5%). Recurrence of the tumor occurred in 31 (8.7%) patients (local recurrence 45.2% and metastatic disease 54.8%). Ten- and twenty-year recurrence-free survival rates were 92% and 77%. 19 (5.3%) patients had also developed contralateral breast cancer. Ten-year survival rates were 91.6%, and 20-year survival rates were 76.6%, respectively. Aged over 65years (p = 0.004), necrosis (p = 0.002), mitosis (p = 0.003), and nuclear pleomorphism (p = 0.049) were found as statistically significant factors for recurrence in univariate analysis. In the ROC analysis, the largest size of the tumor (over 1.45cm, p = 0.07) remained outside the statistical significance limit in terms of recurrence. Thirty-year outcomes in individuals with early stage, node-negative breast cancer were shown in this study. We found that the recurrence ratios between 10 and 20 years were more frequent than the first 10 years during the follow-up. Despite the small number of patients who experienced a recurrence, we demonstrated that, in univariate analysis, being older than 65 and having some pathological characteristics (nuclear pleomorphism, mitosis, and necrosis) were statistically significant factors for disease recurrence.

Treatment pattern of early-stage premenopausal breast cancer patients with low Oncotype DX score: Single center experience.

e12510 Background: The predictive value of Oncotype DX in early-stage node negative breast cancer had been previously well established. As per TAILORx, patients with a low Oncotype DX Recurrence Score RS (0–25) did not benefit from chemotherapy. Premenopausal patients are a crucial subgroup for which recommendations based on Oncotype DX are imprecise. Those with Oncotype DX RS 16-25 are thought to have a small benefit from chemotherapy. Our goal was to study the treatment pattern for this group of patients managed at our institution. Methods: We conducted a retrospective analysis to assess premenopausal patients with early-stage breast cancer who had Oncotype DX testing done. All eligible patients were compared based on whether chemotherapy was given or not. The primary endpoint evaluated was the rate of local recurrence. Univariate comparisons were assessed using t-tests or ANOVA for continuous variable & chi-square for categorical variable. Results: A total of 506 patients younger than 50 years diagnosed between 01/2018 and 12/2020 who had an Oncotype DX available were identified. Of these, 344 were White, 137 were Black and 23 identified as another race. The average age of patients was 42.7 ± 5.5 years. Patients treated with chemotherapy were found to be on average younger than those who did not receive chemotherapy (42 years old vs 43.6 years old, P < 0.0001). The average tumor size for patients who got chemotherapy was 3.26 cm compared to an average size of 2.78 cm for patients who avoided chemotherapy. 72.1% of patients with stage 2 disease had received chemotherapy compared to 38.1 % of patient with stage 1. There were 462 patients with RS < 16, 28 patients with RS 16-25, and 16 patients with RS > 25 as illustrated. Patients were followed from a range of 1.8 months to 30.7 months. There were 6 patients who had local recurrences identified, all of which had RS scores < 16. Conclusions: In our study, 91.3% had RS < 16 and generally should have avoided chemotherapy, however 35% of these patients continued to receive chemotherapy. Only 5.5% of our population had RS 16-25, a quarter of which (n = 7) received chemotherapy. Of those 7 patients, 3 had positive lymph node disease. None of these patients had disease recurrence to date. Patients with RS >25 made up 3.2% of the population and would generally benefit from chemotherapy, however 12.5% of these patients did not end up receiving it. Using Oncotype DX in early-stage breast cancer helps avoid chemotherapy safely in majority of patients. [Table: see text]

Helical Tomotherapy of Lymph Node-negative Early-stage Breast Cancer After Breast-conserving Surgery: Long-term Results.

Adjuvant radiotherapy is an integral part of the interdisciplinary curative treatment of breast cancer. We aimed to examine the long-term clinical results of helical tomotherapy in female patients with local restricted, lymph node negative breast cancer after breast-conserving surgery. In this single-centre analysis, 219 female patients with early-stage breast cancer (T1/2) and no lymph node metastasis (N0) following breast-conserving surgery and sentinel-node biopsy were treated with adjuvant fractionated whole breast radiation therapy using helical tomotherapy. When boost irradiation was indicated, it was administered sequentially or using the simultaneous-integrated boost technique. Local control (LC), metastasis and survival rates, acute toxicity, late toxicity, and secondary malignancy rates were analysed retrospectively. The mean follow-up time was 71 months. The 5- and 8-year overall survival (OS) rates were 97.7% and 92.1%, respectively. The 5- and 8-year LC rates were 99.5% and 98.2%, while the 5- and 8-year metastasis-free survival (MFS) rates of 97.4% and 94.3%, respectively. Patients with G3 grading or negative hormone receptor status did not show significantly different results. Acute erythema occurred in 79% (grade 0-2) and 21% (grade 3) of the patients. Lymphedema of the ipsilateral arm and pneumonitis occurred in 6.4% and 1.8% of the treated patients. None of the patients developed >grade 3 toxicities during follow-up, while 1.8% developed a secondary malignancy during follow-up. Helical tomotherapy showed excellent long-term results and low toxicity rates. The incidence rates of secondary malignancy were relatively low and correlated with pre-existing data on radiotherapy, suggesting wider implementation of helical tomotherapy in adjuvant radiotherapy for breast cancer patients.

Abstract B042: Sentinel lymph node detection in breast cancer: Comparison of blue dye and filtered Technetium 99m sulphur colloid radiotracer injection

Abstract Aim: This prospective study was performed to compare the identification rate, sensitivity and specificity of blue dye and filtered Technetium 99m labelled sulphur colloid (Tc99m SC), individually and in combination for the detection of sentinel lymph nodes in clinically node negative early-stage breast cancer. Materials and methods: Thirty-two women with biopsy-proven breast cancer (cT1N0M0-T2N0M0) were included in this prospective study over a period of 18 months (2019-2020). Four peri-areolar intradermal injections of filtered Tc99m SC were administered 2-3 hours before surgery. Sequential dynamic anterior images of the chest were acquired on a gamma camera, followed by anterior and anterior oblique static images and hybrid single photon emission computed tomography / computed tomography (SPECT/CT) of the affected breast and ipsilateral axilla. Pre-operatively, the sites of detected sentinel lymph nodes (SLNs) were marked on the skin surface using a hand-held gamma probe. SLNs were then identified intra-operatively after peri-tumoral methylene blue injection (blue nodes) and with the gamma probe (hot nodes). These SLNs were removed and sent for frozen section histopathology separately, labelled as hot only, blue only and hot and blue. All patients underwent axillary lymph node dissection (ALND). The final histopathology of all lymph nodes excised during ALND and SLN biopsy were analysed. Identification rate, sensitivity and specificity of both the methods for SLN detection individually and in combination were compared. Results: SLN data of 32 female patients aged 54.7±10.6 years with cT1N0M0-T2N0M0 was analysed. In 87.5% of patients (n=28) the same lymph node was identified by blue dye and filtered Tc99m SC and discordance was seen in 12.5% (n=3) patients. No SLN was detected by any method in one patient. Identification rate of filtered Tc99m SC alone, blue dye alone and combination techniques was 96.9% (n=31), 93.8% (n=30) and 96.9% (n=31) respectively. Of total 593 dissected lymph nodes including 78 SLNs (average 2.4±1.7 per patient), metastasis was found in 198 lymph nodes (12/32 patients). The sensitivity and specificity of the procedure considering only filtered Tc99m SC, blue dye alone and combination of the two tracers was found to be 83.3% (95% CI of 51.6-97.9) and 100% (95% CI of 82.4-100), 75% (95% CI of 42.8- 94.5) and 100% (95% CI of 81.5 -100) and 83.3% (95% CI of 51.6-97.9) and 100% (95% CI of 82.4-100) respectively. Conclusion: SLNB with filtered Tc99m SC using the intra operative gamma probe has better identification rate and sensitivity than blue dye in detecting SLN, while preoperative SPECT/CT helped the surgeon to locate the SLN during surgery. Blue dye has a complementary role in identifying the SLN due to ease of visualisation while operating. Citation Format: Anish Bhattacharya, Yamini Mathur, Ishita Laroiya, Amanjeet Bal, Sarika Sharma, Bhagwant R. Mittal. Sentinel lymph node detection in breast cancer: Comparison of blue dye and filtered Technetium 99m sulphur colloid radiotracer injection [abstract]. In: Proceedings of the AACR Special Conference: Cancer Metastasis; 2022 Nov 14-17; Portland, OR. Philadelphia (PA): AACR; Cancer Res 2022;83(2 Suppl_2):Abstract nr B042.

EE375 Cost-Effectiveness Analysis of the Oncotype DX® Test to Reduce Under-Treatment in Patients With HR+/HER2- Node Negative Early-Stage Breast Cancer and Grade 2 T1C Tumours

EE375 Cost-Effectiveness Analysis of the Oncotype DX® Test to Reduce Under-Treatment in Patients With HR+/HER2- Node Negative Early-Stage Breast Cancer and Grade 2 T1C Tumours

Cost and Clinical Benefits Associated with Oncotype DX® Test in Patients with Early-Stage HR+/HER2- Node-Negative Breast Cancer in the Netherlands

Objectives Patients with early-stage HR+/HER2- N0 breast cancer may receive adjuvant chemotherapy in combination with surgery. However, chemotherapy does not always lead to improved survival and incurs high healthcare costs and increased adverse events. To support decision-making regarding adjuvant chemotherapy, genomic profile testing performed with tests such as the Oncotype DX® test can help healthcare practitioners decide whether chemotherapy provides any benefit to these patients. As such, a cost-consequence model was developed with the aim to estimate the economic impact of using different gene expression tests or no testing, in patients with node-negative early-stage breast cancer. Methods A cost-consequence model was developed to estimate the economic impact of three different scenarios in the Dutch setting: (1) Oncotype DX® test, (2) MammaPrint®, and (3) and no genomic profile testing. The model included chemotherapy costs, administration costs, short- and long-term adverse event costs, productivity loss, genomic profiling testing costs, cost of cancer recurrence, and hospitalization costs. Results A treatment paradigm with Oncotype DX resulted in average savings per patient of €6,768 vs. a paradigm with MammaPrint and €13,125 vs. a paradigm with no genomic testing. Furthermore, due to less patients receiving adjuvant chemotherapy through better targeting by the Oncotype DX test, fewer adverse events, sick days, practice visits, and hospitalizations were required compared to MammaPrint and no genomic profiling. Conclusions Testing with Oncotype DX test in Dutch clinical practice in patients with early-stage breast cancer proved to be cost-saving versus MammaPrint and no genomic profiling tests. Introducing the Oncotype DX test to the Dutch setting will likely reduce the economic resources that are required.

Axillary sentinel lymph node identification using superparamagnetic iron oxide versus radioisotope in early stage breast cancer: The UK SentiMag trial (SMART study)

Background & ObjectivesSentinel lymph node biopsy (SLNB) is an accurate and reliable method for staging the axilla in early breast cancer. The gold standard technique for localizing the sentinel lymph node (SLN) is the use of radioisotope with or without blue dye. However, this technique has its limitations. Various alternatives have been explored to overcome the disadvantages of the standard SLNB technique and superparamagnetic iron oxide mapping agents have garnered significant attention. The SMART study aims to compare the magnetic technique using the superparamagnetic iron oxide particles (SPIO, Sienna+®) to the radioisotope technique (Tc99) +/− blue dye, for SLN identification in patients with early breast cancer. MethodsA prospective, multicenter study was done that recruited 109 clinically node-negative early-stage breast cancer patients from five centres in the United Kingdom (UK). The patients received radioisotope ± blue dye injections, followed by intraoperative injection of magnetic tracer prior to SLNB. The sentinel node identification rate was compared between the magnetic and standard techniques to evaluate detection rate (per patient and per node), non-inferiority and concordance. ResultsData was analysed for 107 patients. The per patient detection rate was 98.13% (105/107) when using the magnetic tracer and 92.26% (103/107) when using the standard technique. The nodal detection rate was 93.07% (188/202 nodes) when using the magnetic tracer and 96.53% (195/202) when using the standard technique. Of the 31 patients with positive sentinel lymph nodes (SLNs), all 31 (100%) were detected by both techniques. ConclusionOur study demonstrates that the magnetic technique is a feasible method for SLNB, with an identification rate that is not inferior to the standard technique. The magnetic technique offers a suitable alternative to the standard technique thereby avoiding the need for the complexities of nuclear medicine, the hazards of radiation and the anaphylaxis risk of blue dye.

Behavioral and health outcomes from the NRG Oncology/NSABP B-36 trial comparing two different adjuvant therapy regimens for early-stage node-negative breast cancer.

The NSABP B-36 compared four cycles of doxorubicin and cyclophosphamide (AC) with six cycles of 5-fluorouracil, epirubicin, and cyclophosphamide (FEC-100) in node-negative early-stage breast cancer. A sub-study within B-36, focusing on symptoms, quality of life (QOL), menstrual history (MH), and cardiac function (CF) was conducted. Patients completed the QOL questionnaire at baseline, during treatment, and every 6months through 36months. FACT-B Trial Outcome Index (TOI), symptom severity, and SF-36 Vitality and Physical Functioning (PF) scales scores were compared between the two groups using a mixed model for repeated measures analysis. MH was collected at baseline and subsequently assessed if menstrual bleeding occurred within 12months prior to randomization. Post-chemotherapy amenorrhea outcome was examined at 18months and was defined as lack of menses in the preceding year. Logistic regression was used to test for association of amenorrhea and treatment. CF assessment was done at baseline and 12months. Correlation analysis was used to address associations between changes in baseline and 12-month PF and concurrent CF changes measured by LVEF. FEC-100 patients had statistically significantly lower TOI scores during chemotherapy (P = 0.02) and at 6months (P < 0.001); lower Vitality score at 6months (P < 0.01), and lower PF score during the first year than AC patients. There were no statistically significant QOL score differences between the two groups beyond 12months. No significant differences in symptom severity between the two groups were observed. Rates of amenorrhea were significantly different between FEC-100 and AC (67.4% vs. 59.1%, P < 0.001). There was no association between changes in LVEF and PF (P = 0.38). Statistically significant QOL differences between the two groups favored AC; however, the magnitude was small and unlikely to be clinically meaningful. There was a clinical and statistically significant difference in risk for amenorrhea, favoring AC. NCT00087178; Date of registration: 07/08/2004.

Recent Advances in the Tracer Technology Used for Sentinel Lymph Node Biopsy in Breast Cancer

The high incidence of breast cancer poses one of the greatest risks to female health worldwide. Sentinel lymph node biopsy (SLNB) is the standard of treatment for patients with axillary lymph node-negative early-stage breast cancer. Herein, the precise use of tracers is the key to ensuring the success of SLNB. However, owing to select-few limitations of traditional tracers, their clinical application is limited. New tracer techniques, such as the near-infrared fluorescent dye method (using indocyanine green), contrast-enhanced ultrasound, and superparamagnetic iron oxide nanoparticles are being applied in clinical practice. In this paper, we review the recent progress in SLNB tracer technology.

A Comparison of Outcomes in Early-Stage Breast Cancer After External Beam Radiotherapy, Endocrine Therapy, Both, or Neither: A Retrospective Cohort Study Across a Large Health System

Several trials have compared endocrine therapy (ET) alone vs. ET and radiotherapy (RT) in low-risk early-stage ER+ breast cancer and have shown that the addition of RT to ET modestly reduces the rate of loco-regional recurrence (LRR) without altering survival. ET alone has thus become a standard strategy in managing these patients. Data evaluating the efficacy of RT alone is scarce. However, the toxicity of ET is not trivial, so as RT courses become ever shorter and safer, a re-evaluation of the strategy of RT alone is warranted. This analysis reports on outcomes of early-stage node-negative breast cancer treated with RT, ET, both, or neither at a multi-hospital health system. We hypothesized that RT alone yields comparable LRR to ET alone in ER+ positive early-stage breast cancer.The shared electronic health record of multiple facilities within a university health system was queried to identify women with pathologic T1-2 N0 breast cancer treated with breast-conserving surgery between 2007 and 2016. Data collection included demographic features, tumor characteristics, treatment type, and evidence of LRR. Analysis was restricted to those patients with ER+ disease. Five-year LRR were compared across groups using the Kaplan-Meier method. A multivariate analysis was performed to identify factors associated with LRR.1340 patients met inclusion criteria. Of these, 84 were treated with RT alone, 155 with ET alone, 1042 with both RT and ET, and 59 received breast conserving surgery only. Demographic and tumor variables are listed below. The 5-year rates of LRR for RT, ET, both, and neither, respectively, were 4.2% (95% CI 1.4-12.7), 5.6% (95% CI 2.6-11.7), 0.6% (95% CI 0.2-1.4), and 31.9% (95% CI 19.8-48.6). While not receiving any adjuvant therapy was associated with significantly higher rate of LRR (P < 0.001), the difference between RT and ET alone was not significant (P = .395). On multivariate analysis, age < 50 (HR 3.3; 95% CI 1.5-7.1), high grade (HR 2.5; 95% CI 1.1-5.8), and observation (HR 9.8; 95% CI 3.2-29.6) were associated with higher LRR, whereas receipt of both RT and ET was associated with lower LRR (HR 0.2; 95% CI 0.1-0.5).RT yielded LRR equivalent to ET in this population, while the combination was superior to either treatment alone. For patients with early-stage ER+ breast cancer, a prospective randomized trial comparing RT to ET and evaluating quality of life (QoL) is needed. In light of new trial data, a five-day course of RT can be considered; this treatment may yield equivalent LRR and better QoL than a five-year course of ET.

Risk evaluation of early-stage hormone receptor-positive and human epidermal growth factor receptor 2-negative breast cancer patients: a population-based study from Taiwan

To assess the prognostic value of the Dutch criteria for patients with early-stage hormone receptor-positive and human epidermal growth factor receptor 2-negative breast cancer from the Taiwan Cancer RegistryDatabase. We included 8295 patients with early-stage node-negative breast cancer who underwent surgery during January 2008-December 2012. Patients were stratified into low- and high-risk groups based on the Dutch criteria. The Kaplan-Meier method and log-rank test were used to estimate the difference in breast cancer-specific survival (BCSS) and overall survival (OS) between groups. Multivariable analysis was used to evaluate the prognostic value of the Dutch criteria. Overall, the low-risk and high-risk groups comprised 5375 and 2920 patients, respectively. In the low- and high-risk groups, the 5-year BCSS rate was 99.6% and 98.2% (P < 0.0001) and the 5-year OS rate was 98.3% and 96.8% (P < 0.0001), respectively. The hazard ratio for BCSS was 4.18 (95% confidence interval [CI] 2.63-6.63, P < 0.0001), and the hazard ratio for OS was 1.94 (95% CI 1.48-2.55); both were significantly poorer in the high-risk group than in the low-risk group. In the low-risk group, the 5-year BCSS and OS of patients who did and did not receive adjuvant chemotherapy were similar (99.5% versus 99.6% [P = 0.927] and 98.8% and 98.1% [P = 0.0683], respectively). The prognosis of low-risk patients as classified using the Dutch criteria is excellent with or without adjuvant chemotherapy. The benefit of multi-gene testing for chemotherapy decision-making might be minimal in these patients.

Long-term outcomes and predictors of recurrence in patients with early-stage node negative breast cancer.

e12597 Background: In this study, we aimed to assess the outcomes, and predictors of recurrence in patients with early-stage node-negative breast cancer. Methods: We evaluated data of the patients who were treated between 1988 and 2018 years retrospectively. Demographical, clinical, pathological, and treatment features of the patients were recorded. SPSS 25 version was used for statistical analysis. We used Kaplan-Meier and Cox regression analysis to assess survival analysis. Also, we performed logistic regression and ROC analysis for recurrence predictors. Results: In total, 347 patients were included in the study.The median age was 50 (range, 18-81) at diagnosis. The percent of the patients who had stage 1 and 2 were 86.6% and 13.4%, respectively. The most common histopathological type was invasive ductal carcinoma (71.6%). Estrogen receptor and progesterone receptor positivity were 80.3% and 62.1%, respectively. Her2 receptor positivity was 15.9%. The number of patients who had undergone lumpectomy and mastectomy was 85.5% and 14.5%, respectively. Sentinel lymph node biopsy was performed on 78.7% of the patients, and axillary lymph node dissection 21.3%. Also, the patients received adjuvant radiotherapy (88.7%), adjuvant chemotherapy (48.5%), and adjuvant hormonotherapy (82.1%). Tumor recurrence was occurred in 31 (8.7%) patients (local recurrence-45.2% and metastasis-54.8%). Five-, ten- and twenty-years recurrence ratios were 4.3%, 8%, and 23%. Also, contralateral breast cancer has occurred in 19 (5.3%) patients. During the study period, 30 (8.4%) patients died. Ten-years and twenty-years survival ratios were 91.6% and 76.6%, respectively. In univariate analysis, aged over 65 years (p = 0.004), nuclear pleomorphism (p = 0.049), mitosis (p = 0.003), and necrosis (p = 0.002) were statistically significant for recurrence. In ROC analysis, the tumor's longest size was not statistically significant for recurrence (for over 1.45 cm, p = 0.07). Conclusions: In this study, we determined thirty-years outcomes in patients with early-stage node-negative breast cancer. In the follow-up, we detected the recurrences ratios, between ten and twenty years, were more common than the first ten-years. Despite a small number of patients who had a recurrence, we showed that being age over 65 years and pathological features (nuclear pleomorphism, mitosis, and necrosis) were statistically significant for disease recurrence in univariate analysis.

Application of RSClin to guide treatment recommendations for premenopausal patients with early-stage hormone-positive breast cancer and intermediate risk oncotype recurrence scores.

e12511 Background: The RSClin model, which incorporates the Oncotype Recurrence Score (RS) and clinicopathologic features, was recently developed to further tailor prognosis and prediction of chemotherapy benefit for patients with early-stage hormone positive (HR+) breast cancer (BC) (Sparano et al, 2020). The RSClin calculator is available online to assist treatment planning for situations where chemotherapy benefit is uncertain. Covariates include Oncotype RS, tumor grade, tumor size and patient age. The risk calculator generates a 10-year distant recurrence risk and absolute chemotherapy benefit. This tool may be especially helpful to determine treatment management for premenopausal patients with early-stage HR+ BC with intermediate risk (IR) Oncotype RS (16-25). We retrospectively applied RSClin to this patient population to determine if it would have changed treatment recommendations. Methods: We identified premenopausal women with node-negative early-stage BC with IR RS (16-25) within our large Oncotype database. Using the RSClin model, we selected &gt;5% absolute chemotherapy benefit as a reasonable cutoff to recommend chemotherapy. We compared the treatment recommendation based on RSClin with the treatment previously recommended by breast oncologists at our large academic medical center in New York City. Results: There were 86 patients who met criteria with a median age of 46 years. Of these, 26 patients (30%) were recommended chemotherapy plus endocrine therapy (ET) and 60 (70%) were recommended ET alone. After applying the RSClin model (data available for 83/86 patients), 19 (23%) would have resulted in a change in treatment recommendation and 64 (77%) would have remained unchanged. Overall, 8 (10%) would have withheld chemotherapy when it was previously offered and 11 (13%) would have recommended chemotherapy when it was previously excluded. There were 8 (9%) secondary invasive breast events in this population, with 2 (2%) being ipsilateral, 3 (3%) being contralateral and 3 (3%) metastatic at a median follow up of 46.9 months. Conclusions: The RSClin model would have changed management of premenopausal patients with IR RS in 23% of patients. This model, although not yet prospectively validated, may help individualize therapy for patients with less definitive treatment plans. Using RSClin, we can aim to minimize recurrence rates and avoid unnecessary chemotherapy in selected patients. This model is easy to apply and will have important clinical utility moving forward.

Can preoperative axillary ultrasound and biopsy of suspicious lymph nodes be an alternative to sentinel lymph node biopsy in clinical node negative early breast cancer?

The aim of this study was to assess the efficacy of preoperative axillary ultrasonography (AUS) and preoperative axillary fine-needle aspiration biopsy (FNAB) from suspicious lymph nodes in clinically node-negative breast cancer to compare with radiologically positive and sentinel lymph node biopsy (SLNB) positive involvement. Clinically node-negative early-stage breast cancer patients were included in the study. These patients underwent preoperative AUS examination, suspicious lymph nodes were evaluated with FNAB. AUS-FNAB results were compared with those of SLNB or axillary dissection. Of 181 patients undergoing AUS, 32 were reported to have axillary metastasis, 25 suspicious, and 124 benign nodes. The suspicious group underwent FNAB examination and metastasis was found in 9 of them. The sensitivity of AUS-FNAB was found to be 64.06%, specificity 100%, positive predictive value 100%, and negative predictive value (NPV) 83.5%. The false negativity rate (FN) of this method was 16,4%. Lymphovascular invasion and tumour size were found statistically significant factors for false negativity. It was concluded that axillary AUS-FNAB with its high NPV, low FN rate, may be a clinical alternative to SLNB for early-stage breast cancer patients.

Abstract PS8-38: Development of random forest classifier method to predict breast cancer mortality among early stage breast cancer patients

Abstract Background: Among women diagnosed with early stage breast cancer, breast cancer mortality was conditional on the onset of distant recurrence. Gene expression profiling (GEP) tests (e.g., Oncotype Dx) have been commonly used to determine the risk level of distant recurrence, but its predictive ability is often questioned. The objective of this study aims to explore the potential of random forest machine learning to predict the risk of breast cancer mortality.Method: 63,104 patients with node-negative early stage breast cancer in the SEER-GHI dataset were randomly sampled into two groups, for the purpose of training and testing the model where each group consisted of 67% and 33% of all patients, respectively. The patient characteristics and tumor features in the study included age, tumor size, grade, chemotherapy use, Oncotype DX test recurrence score, and more. Further feature selection in the construction of the predictive model was conducted via the Boruta algorithm. As the dataset is highly imbalanced with only a very small percentage of samples being fatal, Synthetic Minority Over-Sampling Technique (SMOTE) was employed to address this issue. A cross-validated grid search finds the best combination of hyper-parameters to improve the model’s performance.Results: The mean and standard deviation of follow-up duration in the cohort of patients were 32.4 and 14.3 months, respectively. Of 37,043 patients, there were 158 breast cancer deaths. Within up to a 59-month follow-up period, the breast cancer specific mortality rate was 0.43%. Based on the limited available data, the outcome of the random forest classifier method in predicting breast cancer mortality shows sensitivity 0.712, specificity 0.773, overall accuracy 0.773, and AUC (area under ROC curve) 0.742. Conclusion: The use of the random forest ML algorithm to predict breast cancer mortality is promising when the hyper-parameters are fine-tuned. However, the model's positive predictive value is highly associated with sufficient data sources. Citation Format: Arwen Chandler. Development of random forest classifier method to predict breast cancer mortality among early stage breast cancer patients [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS8-38.

Abstract 4631: Pre-diagnosis comorbidity class, health-related quality of life, and treatment selection among older women with breast cancer

Abstract Purpose of the study: The NCCN guidelines state that adjuvant radiotherapy (RT) can be omitted in older women with estrogen receptor (ER)-positive, node-negative early-stage (i.e., Tis and T1) breast cancer (ESBC). However, there is no specific algorithm to select a subgroup of women who can safely forgo RT. Older women commonly present with multiple comorbid conditions with different levels of functional limitations and health-related quality of life (HRQOL), which challenge the RT-decision process. This study aimed to understand pre-diagnosis comorbidity burdens in this patient group and to examine its relationship with HRQOL and treatment selection in a retrospective cohort study design. Methods: From the SEER-MHOS database, we included 985 older women who were diagnosed at age ≥ 70 years, with ER-positive, node-negative ESBC between 1998 and 2013. SEER provides primary treatments (surgery and radiation), and MHOS provides self-reported comorbidity, functionality, and HRQOL. We used latent class analysis (LCA) to identify clinically distinct comorbidity burden classes. The associations between the comorbidity burdens with HRQOL and treatment selection (lumpectomy only, lumpectomy + RT, mastectomy, mastectomy + RT) were examined using generalized linear models. Results: A mean age at diagnosis was 77 years, 81 % of patients were whites, and 43 % were married. A majority had stage I disease (80 %) with 20 % stage 0. About one fourth (26 %) had ≤ 1 comorbid condition, whereas 28 % had ≥ 4 comorbidities. LCA identified four distinct comorbidity burden classes; healthy (n=577, 58 %), low comorbidity with limited functionality (n=64, 7 %), moderate comorbidity with symptoms (n=247, n=25 %), and high comorbidity with highly limited functionality and symptoms (n=97, 10 %). Classes were distinguished by the presence or absence of comorbid conditions and functional limitations as well as symptoms. Comorbidity burdens were independently associated with HRQOL mental component summary (MCS-12) score (p &amp;lt; .0001) and physical component summary (PCS-12) score (p &amp;lt; .0001) but not with treatment selection (p = 0.096). In stage 0, age (p=0.025) and education level (p=0.024) were the independent predictors of treatment selection while age (p&amp;lt;.0001), education level (p=0.0013), income (p=0.044), and obesity (p=0.026) determined the treatment selection in stage I. Conclusions: Older women with ESBC can be described by four distinct comorbidity burdens that are independent predictors of HRQOL. However, it seems that the comorbidity burdens were not considered in the treatment selection process. Future studies need to address treatment outcomes (such as changes in HRQOL and survival) by comorbidity burdens to better guide treatment decisions. Citation Format: Eunkyung Lee, Jianbin Zhu, Robert Hines, Eunji Nam, Cassie Odahowski. Pre-diagnosis comorbidity class, health-related quality of life, and treatment selection among older women with breast cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4631.

The 21-Gene Recurrence Score Assay and Prediction of Chemotherapy Benefit: A Propensity Score-Matched Analysis of the SEER Database.

Background: To evaluate the performance of the 21-gene recurrence score (RS) assay in predicting chemotherapy benefit in the Surveillance, Epidemiology, and End Results population, we aimed to assess breast cancer-specific mortality (BCSM) by chemotherapy use within each of the RS categories. Methods: Data on breast cancer (BC) cases diagnosed between 2004 and 2015 with available RS results were released. Our analysis included patients with hormone receptor-positive, node-negative early-stage BC (n = 89,402), and three RS groups were defined; RS < 11, low; RS 11–25, intermediate; RS > 25, high. A propensity score matched-analysis was performed to assess and compare BCSM. Results: Chemotherapy was significantly associated with a reduced risk of BC death among patients in the high RS group (hazard ratio = 0.782; 95% CI, 0.618–0.990; p = 0.041). However, in the low and intermediate RS groups, there were no significant differences in BCSM between patients who received chemotherapy and those who did not. Among those with RS 11–25, chemotherapy benefit varied with tumor size (p = 0.001). Conclusions: Our findings provide real-world evidence that the 21-gene RS assay is predictive of chemotherapy benefit among patients in clinical practice. More refined risk estimates would be needed for patients with an intermediate RS.