Comparative Evaluation of a 28-Gene versus 70-Gene Panel for Recurrence Risk Prediction in Early-Stage HR-Positive/HER2-Negative Breast Cancer in Chinese Women

Comparative Evaluation of a 28-Gene versus 70-Gene Panel for Recurrence Risk Prediction in Early-Stage HR-Positive/HER2-Negative Breast Cancer in Chinese Women | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Comparative Evaluation of a 28-Gene versus 70-Gene Panel for Recurrence Risk Prediction in Early-Stage HR-Positive/HER2-Negative Breast Cancer in Chinese Women L. Lei, W. Cao, Y. Yu, J. Luo, G. Qiao, X. Xie, C. Du, Y. Tan, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6832631/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background : Multigene expression assays help guide adjuvant therapy decisions in early-stage hormone receptor-positive, HER2-negative (HR+/HER2-) breast cancer by predicting recurrence risk. While the 21-gene Oncotype DX and 70-gene MammaPrint tests, developed in Western populations, allow low-risk patients to safely avoid chemotherapy, their validity in Asian cohorts remains understudied. A 28-gene signature (RecurIndex), developed using Asian patient data, predicts distant and locoregional recurrence. This study compares the 28-gene RecurIndex with the 70-gene MammaPrint in Chinese women with early HR+/HER2- breast cancer. Patients and methods : We retrospectively analyzed 99 Chinese patients (median age 52 years) with stage pT1-3N0-1 HR+/HER2- breast cancer who underwent MammaPrint testing post-surgery (2019-2022). Formalin-fixed paraffin-embedded tumor samples were retested using the 28-gene RecurIndex. Clinical-pathological risk was defined as ≥2 high-risk factors (age ≤40 years, tumor ≥T2, N1 nodal status, lymphovascular invasion, grade 3, Ki-67 ≥20%). Genomic risk stratification (high vs low) by each assay and concordance with clinical risk were assessed using Cohen’s kappa. Median follow-up was 56 months. Results : By clinical criteria, 48.5% (48/99) of patients were high-risk. The 28-gene assay classified 26% as high genomic risk versus 34% by the 70-gene MammaPrint. Overall concordance between assays was 72% (71/99 cases; kappa = 0.50). The 28-gene panel showed stronger alignment with clinical risk (kappa = 0.51) than the 70-gene (kappa = 0.39). Among clinically low-risk patients, 98% (50/51) were classified as low-risk by the 28-gene test compared to 84% (43/51) by the 70-gene. For clinically high-risk patients, both assays identified similar proportions as genomic high-risk (52% vs 54%, respectively). No recurrences or deaths occurred during follow-up. Conclusion : The 28-gene RecurIndex demonstrated comparable performance to the 70-gene MammaPrint in stratifying recurrence risk in Chinese HR+/HER2− breast cancer patients. It classified fewer patients as high-risk, identified a larger low-risk subgroup (74% vs 66%), and aligned more closely with traditional clinical-pathological factors. These findings support its potential as an ethnicity-tailored prognostic tool. Larger studies with extended follow-up are needed to confirm predictive accuracy and chemotherapy benefit in this population. 28-gene RecurIndex panel 70-gene MammaPrint Chinese HR+/HER2 − breast cancer Risk stratification Clinicopathologic concordance Figures Figure 1 Figure 2 Introduction Breast cancer heterogeneity drives the use of multigene assays (e.g., 21-gene Oncotype DX[ 1 ], 70-gene MammaPrint[ 2 ]) to refine risk stratification in early HR+/HER2 − disease[ 3 ]. Landmark trials (TAILORx[ 4 ], MINDACT[ 5 ]) validated their prognostic utility: TAILORx showed ~ 70% of early HR+/HER2 − patients (especially intermediate-risk) avoided chemotherapy safely with endocrine therapy alone, while MINDACT demonstrated 94–95% 5-year metastasis-free survival in clinically high-risk/gnomically low-risk patients spared chemotherapy. These assays guide chemotherapy de-escalation for genomic low-risk patients while identifying high-risk subgroups for intensified treatment. Most widely used multigene signatures (e.g., Oncotype DX) were developed and validated in Western populations, raising questions about their generalizability for risk stratification in Asian patients. Studies highlight ethnic differences in breast cancer biology and epidemiology, including distinct gene expression profiles and clinical outcomes in Asian versus Western populations[ 6 – 9 ]. To address this, the RecurIndex, a 28-gene signature derived from Asian cohorts, was developed. It integrates tumor gene expression with clinicopathological variables (age, lymphovascular invasion, etc.) to generate separate risk indices as the recurrence index for distant recurrence (RI-DR) and the recurrence index for local recurrence (RI-LR), providing binary risk stratification (low vs. high) [ 10 , 11 ]. Validation studies demonstrated > 80% concordance with Oncotype DX in recurrence risk prediction, with both tools effectively stratifying 4-year relapse outcomes[ 12 ]. In 2022, RecurIndex became the first China-developed multigene assay endorsed by the Chinese Society of Clinical Oncology (CSCO) guidelines[ 13 ]. A national expert consensus recommends its use alongside established assays (Oncotype DX, MammaPrint) to guide adjuvant therapy decisions in early-stage HR+/HER2 − breast cancer[ 13 ]. Despite advances, comparative data between the 28-gene RecurIndex and Western assays (e.g., 70-gene MammaPrint) in Asian populations remain scarce. We retrospectively compared both tests in Chinese women with early-stage HR+/HER2 − breast cancer, applying RecurIndex to MammaPrint-tested tumor samples. Analyses included risk stratification concordance, alignment with clinicopathologic factors, and clinical implications of discordant results. This head-to-head study evaluates whether RecurIndex offers equivalent risk assessment to MammaPrint, potential advantages (e.g., enhanced low-risk identification), and ethnic-specific applicability. Findings address critical gaps in interpreting conflicting risk classifications and guide confidence in using indigenous genomic tools for therapy decisions in Chinese patients. Materials and Methods S tudy Design and Patient Population This study was a retrospective comparative analysis of genomic risk assays in a cohort of Chinese breast cancer patients. Eligible patients were women with early-stage (pathologic stage T1-T3, N0-N1, M0) HR-positive, HER2-negative breast cancer who had undergone genomic testing with the 70-gene MammaPrint panel after primary surgical treatment. All patients had surgery between January 2019 and October 2022 at participating centers. Tumor samples and clinical data were collected under institutional review board approval, and all patients gave informed consent for genetic testing and data use. We identified a total of 99 patients meeting these criteria. By design, all 99 patients had available MammaPrint results (high or low risk) from their initial post-operative evaluation. For this study, formalin-fixed, paraffin-embedded (FFPE) tissue blocks from the primary tumor of these patients were retrieved and subjected to testing with the 28-gene RecurIndex assay in a central accredited laboratory. The RecurIndex test was performed on FFPE sections using the validated platform and algorithm provided by the test developer (Simcere Diagnostics, Nanjing, China), blinded to patients’ MammaPrint results. Figure S1 showed the 28-gene testing laboratory related operations and quality control requirements. The output of the 28-gene assay is a numeric score: RI-DR (Recurrence Index for distant recurrence; low-risk <29.3, high-risk ≥29.3) and RI-LR (Recurrence Index for locoregional recurrence; low-risk <35.3, high-risk ≥35.3). The 70-gene MammaPrint results were available from patients’ records (MammaPrint categorizes tumors as “Low-Risk” or “High-Risk” of distant metastasis based on the proprietary 70-gene expression algorithm). Figure 1 shows the patient selection and assessment of this study. Clinical Data and Risk Definitions We collected clinicopathological variables including patient age at diagnosis, tumor size (pathological T stage), nodal status (N stage), histologic grade, lymphovascular invasion (LVI) status, estrogen receptor (ER), and progesterone receptor (PR) expression, Ki-67 proliferation index, and adjuvant treatment received. All tumors were ER-positive (immunohistochemistry ER>1%) and HER2-negative (immunohistochemistry 0-1+ or non-amplified by in situ hybridization) as part of the inclusion criteria. We defined a clinicopathologic risk score based on six traditional risk factors associated with recurrence in HR+ breast cancer. The following factors were considered adverse (high-risk) features in our analysis: Risk Factor High-Risk Criteria Age ≤40 years at diagnosis Pathological tumor size ≥T2 (≥2 cm) Nodal status N1 (1-3 positive axillary nodes) Lymphovascular invasion (LVI) Positive Histologic grade Grade 3 Ki-67 proliferation index ≥20% Each patient’s tumor was evaluated for these factors; patients with two or more high-risk clinicopathologic features were categorized as “clinical high-risk,” whereas those with 0-1 of these features were categorized as “clinical low-risk.” This dichotomous clinical risk classification was used to examine how well each genomic test agreed with conventional risk assessment. We note that our definition is analogous to but not the same as criteria used in trials like MINDACT; it was tailored to identify patients who, based on traditional parameters, might be recommended chemotherapy in the absence of genomic information. S tatistical Analysis The primary analysis compared risk stratification (high/low) between the 28-gene and 70-gene panels using percentage agreement and Cohen’s kappa coefficient ( κ > 0.75: excellent; 0.4-0.75: moderate; <0.4: poor). Concordance with clinical risk categories (high-risk defined as ≥2 clinicopathologic factors) was similarly assessed. Chi-square/Fisher’s exact tests evaluated differences in classification distributions (e.g., genomic vs clinical risk), with McNemar’s test for paired discordant cases. Due to zero recurrence/death events, survival analysis was omitted; focus remained on risk reclassification metrics. Descriptive statistics summarized patient characteristics (medians/ranges for continuous variables; frequencies/percentages for categorical data). Analyses used SPSS v26 (IBM), with two-tailed p <0.05 as significance threshold. F ollow-up and Outcome Collection Patients were followed according to standard practice with clinical exams and imaging as indicated. We recorded any instances of breast cancer recurrence (locoregional or distant) or breast cancer-specific death after surgery. The median follow-up duration was 56 months (approximately 4.7 years), with a follow-up cut-off as of December 30, 2024. Follow-up time was calculated from the date of surgery. At the last follow-up, we noted the survival status and disease status (no evidence of disease, alive with recurrence, or deceased) for each patient. This outcome data was used to provide preliminary observations, although, with no events, no comparative survival analysis between tests could be performed. Results Pa tient Characteristics A total of 99 patients were included, all of whom were Chinese women diagnosed with early-stage (stage I-II) HR-positive, HER2-negative breast cancer. Key clinicopathologic characteristics of the cohort are summarized in Table 1. The median age was 52 years (range 29-74 years), with 11% of patients ≤40 years old, reflecting that a minority were very young at diagnosis. Tumor size was T1 in 61% of cases, T2 in 38%, and there was one patient with a T3 tumor (3 cm) but node-negative. Nodal status was positive (N1, 1-3 involved nodes) in 36% of patients, while 64% had node-negative disease. Most tumors (77%) were histologic grade 2; only 11% were grade 3, and 11% were grade 1 (grade was unspecified for one case). Lymphovascular invasion was present in 20% of tumors. All 99 tumors were ER-positive; PR was positive in 82% and negative in 13% (5% unknown). Ki-67 proliferation index was reported for 89 patients: among those, 45 tumors (approximately half of the evaluable cases) had high Ki-67 (≥20%). Overall, using our predefined clinical risk criteria, 48 patients (48.5%) were categorized as clinical high-risk (having ≥2 high-risk factors), and the remaining 51 (51.5%) were clinical low-risk (0-1 factors). Table 1. Baseline characteristics Characteristic Distribution (n = 99) Age Median 52 years (range 29–74) Age > 40 88 patients (89%) Age ≤ 40 11 patients (11%) Tumor size T1 (≤2 cm) 60 (61%) T2 (>2-5 cm) 38 (38%) T3 (>5 cm, N0 only) 1 (1%) Lymph node status N0 63 (64%) N1 (1–3 positive nodes) 36 (36%) Histologic Grade Grade 1 11 (11%) Grade 2 76 (77%) Grade 3 11 (11%) Unknown 1 (1%) LVI Present 20 (20%) Absent 79 (80%) Estrogen Receptor Positive 99 (100%) Negative 0 Progesterone Receptor Positive 81 (82%) Negative 13 (13%) Unknown 5 (5%) Ki-67 Index < 20% 44 (44%) ≥ 20% 45 (46%) Unknown 10 (10%) Clinical Risk Category Low (0–1 risk factors) 51 (52%) High (≥2 risk factors) 48 (48%) Chemotherapy No 42 (42%) Yes 39 (40%) Unknown 18 (18%) Surgery Mastectomy 31 (31%) BCS 68 (69%) Radiation therapy No 31 (31%) Yes 52 (53%) Unknown 16 (16%) Endocrine therapy No 0 Yes 96 (97%) Unknown 3 (3%) BCS, breast-conserving surgery; LVI, lymphovascular invasion. All patients received primary surgery. In terms of adjuvant therapies, by being HR+, all patients were treated with endocrine therapy (tamoxifen and/or aromatase inhibitor as appropriate). Chemotherapy usage in this cohort was not uniform: treating oncologists integrated the 70-gene test results with clinical factors to decide on chemotherapy. Overall, 39 out of 99 patients (39.4%) received adjuvant chemotherapy, while 42 (42.4%) did not and 18 (18.2%) reflected the intermediate risk nature of many cases. Notably, during a median follow-up of 56 months, no patient experienced a recurrence or cancer-related death. Two patients were lost to follow-up before 3 years, but among the 97 patients monitored, none had developed distant metastasis or locoregional recurrence at the last contact. While the follow-up duration is relatively short and longer observation is needed, this absence of events is consistent with the fact that many patients were low genomic risk and received appropriate therapy. It precludes any direct comparison of survival outcomes between risk groups at this time. Risk Stratification by 28-Gene vs 70-Gene Panels We next compared the recurrence risk categorization provided by the 28-gene and 70-gene assays for each patient in Table S1-3. The 70-gene MammaPrint test had classified 34 of the 99 patients (34%) as High Risk and the remaining 65 (66%) as Low Risk of distant recurrence. In comparison, the 28-gene RecurIndex assay classified only 26 patients (26%) as High Risk, while 73 (74%) were Low Risk. Thus, the 28-gene panel identified a larger proportion of patients as low-risk (about three-quarters of the cohort) and correspondingly fewer as high-risk, relative to the 70-gene panel. Figure 2 illustrates the distribution of risk results across the two genomic assays and the clinical criteria. The clinical risk assessment (by traditional factors) labeled the highest percentage of patients as high-risk (48.5%), whereas both genomic tests categorized fewer patients as high-risk, especially the 28-gene signature. We found an overall concordance of 72% (71 out of 99 patients) between the 28-gene and 70-gene tests in binary risk assignment. In 71 cases, the two assays agreed on whether the tumor was low-risk or high-risk. In the remaining 28 cases (28% of the cohort), the assays disagreed: 18 patients were called high-risk by 70-gene but low-risk by 28-gene, while 10 patients were high-risk by 28-gene but low-risk by 70-gene. This pattern indicates that most discordances (18 vs 10 cases) were instances where the 70-gene test flagged risk that the 28-gene did not. Cohen’s kappa coefficient for the agreement was calculated to be 0.54 (approximately), consistent with moderate agreement beyond chance. This kappa falls in the upper-moderate range, underscoring that while the two assays often concur, there is a non-negligible subset of tumors that they stratify differently. Importantly, among those 28 discordant cases, none have recurred to date, so it remains unclear which assay was “correct” in identifying risk that longer follow-up is needed to determine if one test was more prognostically accurate. Comparison with Clinicopathologic Risk and Consistency Analysis Our analysis compared the alignment of the 28-gene RecurIndex and 70-gene MammaPrint assays with clinicopathologic risk categories in Chinese HR+/HER2− breast cancer patients in Table 2. The 28-gene panel demonstrated stronger consistency with clinical risk stratification than the 70-gene panel. Among 51 clinically low-risk patients, 98% (50/51) were classified as low-risk by the 28-gene test, while the 70-gene assay reclassified 16% (8/51) as high-risk, a statistically significant discrepancy ( p < 0.001). Conversely, in the 48 clinically high-risk patients, both tests classified approximately half as genomically high-risk (52% by 28-gene vs. 54% by 70-gene; p = 0.80). Cohen’s kappa values confirmed moderate agreement with clinical risk for both assays but higher concordance for the 28-gene panel ( κ = 0.51 vs. κ = 0.39). Table 2. Comparative analysis of the performance of risk-prediction by the 28-gene and the 70-gene panels. Characteristic 28-gene (%) Total Kappa 70-gene (%) Total Kappa High-risk Low-risk High-risk Low-risk Clinical risk High 25(25) 23(23) 48(48) 0.51 26(26) 22(22) 48(48) 0.39 Low 1(1) 50(51) 51(52) P<0.001 8(8) 43(43) 51(51) P<0.001 Total 26(26) 73(74) 99(100) - 34(34) 65(66) 99(100) - The 28-gene test produced fewer high-risk classifications overall (26% vs. 34% by 70-gene; p = 0.24), potentially reducing chemotherapy recommendations. Notably, the 70-gene assay frequently discorded with clinical profiles: it labeled 16% of clinical low-risk patients as high-risk (vs. 2% by 28-gene) and downgraded 46% of clinical high-risk patients to low-risk (vs. 48% by 28-gene). These reclassifications mirror findings from trials like MINDACT, where genomic low-risk patients safely avoided chemotherapy. In our cohort, none of the genomically low-risk patients (regardless of clinical risk) experienced recurrence during 56-month follow-up. However, the 70-gene test identified 8 clinical low-risk patients as high-risk, raising questions about potential overestimation or detection of occult aggressive biology. While both tests reclassified ~50% of clinical high-risk patients to low genomic risk-suggesting chemotherapy, the 28-gene panel aligned more closely with traditional risk factors. For example, 9 of 10 patients uniquely labeled high-risk by the 28-gene test had multiple clinical risk factors, whereas 7 of 8 patients uniquely high-risk by the 70-gene test had favorable clinical profiles. This divergence implies the 28-gene signature may better reflect clinician concerns, while the 70-gene test could flag unexpected risk in clinically low-risk cases. In summary, the 28-gene and 70-gene assays showed moderate overall agreement (72%) in risk stratification. Key differences lie in their alignment with clinical risk: the 28-gene panel is more conservative and clinicopathologically concordant, whereas the 70-gene test exhibits greater discordance, particularly in upgrading clinically low-risk patients. Both tools enable chemotherapy de-escalation in ~50% of clinical high-risk cases, but long-term outcome data are critical to validate their prognostic accuracy. The absence of recurrences precludes definitive conclusions, underscoring the need for extended follow-up to clarify whether observed discrepancies reflect overestimation or improved risk detection. These findings highlight the importance of integrating population-tailored genomic tools with clinical judgment in precision oncology. Discussion In this study, we performed the first head-to-head comparison of a 28-gene Asian genomic signature (RecurIndex) against the well-known 70-gene MammaPrint assay for recurrence risk prediction in a cohort of Chinese women with early HR+/HER2 − breast cancer. Our findings demonstrate that the two multigene assays have largely similar performance in stratifying patients into low vs high-risk categories but with some notable differences in classification tendencies and concordance with clinical factors. The overall agreement of 72% ( κ ~ 0.5) indicates that in about three-quarters of cases the 28-gene and 70-gene tests would lead to the same treatment recommendation, which is encouraging in terms of interchangeability. In the remaining one-quarter of cases, however, discordant results emerged that could potentially alter management depending on which assay is followed. This highlights the importance of understanding each test’s characteristics and the context of their development when interpreting results for clinical decision-making. The 28-gene RecurIndex, derived from Asian data, demonstrated stronger alignment with clinicopathologic risk factors in Chinese HR+/HER2 − breast cancer patients compared to the 70-gene MammaPrint ( κ = 0.51 vs. 0.39). This likely stems from RecurIndex’s design as a clinical-genomic hybrid model integrating gene expression and clinical variables[ 14 ], validated with 93% accuracy for 5-year locoregional recurrence prediction in Chinese cohorts[ 15 ]. In contrast, MammaPrint, developed using Western populations[ 16 ], prioritizes proliferation genes without clinical inputs, potentially capturing "hidden" high-risk biology[ 17 ]. Notably, MammaPrint classified 16% of clinical low-risk patients as high-risk vs. 2% by RecurIndex ( p < 0.001), raising concerns about overestimation in Asian populations where genomic risk profiles may differ[ 18 ]. A recent study found higher average genomic risk scores in Western vs. Eastern cohorts using RecurIndex (10.5% vs. 6.6% high-risk) [ 18 ], aligning with our observation of more MammaPrint high-risk calls in Chinese patients. RecurIndex’s high-risk group showed worse 4-year metastasis-free survival in Asian cohorts (52%) compared to Oncotype/PAM50[ 19 ], suggesting sharper risk discrimination. While no recurrences occurred in our cohort, MammaPrint-flagged patients (n = 8) might reflect overestimation, whereas the single RecurIndex-high/MammaPrint-low case could indicate undetected aggressive biology. These findings underscore population-specific genomic calibration: RecurIndex’s Asian-centric design may better balance clinical-genomic concordance, while Western-derived tools like MammaPrint risk misclassification in non-European populations. Using the 28-gene RecurIndex versus the 70-gene MammaPrint in Chinese HR+/HER2 − breast cancer reduces chemotherapy recommendations (26% vs 34%), primarily sparing clinically low-risk patients. While RecurIndex aligns closely with clinicopathologic factors, MammaPrint identifies 8 clinical low-risk patients as high-risk. However, MINDACT trial data suggest clinical low-risk patients (even genomically high-risk) have excellent outcomes without chemotherapy (5-year DMFS ~ 95–97%; no significant chemo benefit) [ 5 ]. In our cohort, none of these discordant cases recurred, implying MammaPrint’s additional high-risk calls may represent false positives. The RecurIndex's conservative alignment with favorable clinical profiles may reduce overtreatment risks, while MammaPrint's genomic risk stratification may lack clinical relevance in Asian populations. Both assays effectively de-escalate therapy for ~ 50% of clinical high-risk patients. Prospective studies are needed to validate whether RecurIndex’s restraint or MammaPrint’s broader risk classification optimizes outcomes in Chinese patients, particularly for discordant cases. Our study has limitations. First, the small sample size (n = 99), selected from patients with prior MammaPrint testing, risks selection bias, as genomic testing is typically ordered for chemotherapy-uncertain cases (often intermediate clinical risk). Second, the median 4.7-year follow-up is insufficient for HR + breast cancer, where recurrences may emerge beyond 5–10 years; the absence of events precludes direct prognostic comparison. Third, we excluded Oncotype DX, PAM50, and EndoPredict, though a recent analysis suggested RecurIndex may outperform other assays in Asian cohorts[ 20 ]. Prospective multi-assay comparisons (including Oncotype DX, used in China at high cost) are needed to optimize test selection. Finally, our cohort’s balanced clinical risk distribution (50% high/low) may not reflect real-world populations. Extended follow-up will clarify whether genomic high-risk groups (by either test) develop recurrences, particularly in discordant cases. While our focus on MammaPrint reflects clinical availability, broader comparisons would better inform precision oncology strategies in Chinese patients. Our study supports the clinical adoption of the 28-gene RecurIndex (CSCO-endorsed) in Chinese HR+/HER2- breast cancer, demonstrating risk stratification comparable to MammaPrint while reducing chemotherapy recommendations (26% vs 34%). RecurIndex offers logistical advantages over MammaPrint, including lower cost and faster local processing. Its alignment with clinical factors may spare patients unnecessary treatment toxicity, particularly for borderline cases where RecurIndex-low results align with MammaPrint (or where MammaPrint-high calls lack prognostic urgency). However, caution remains for discordant cases (e.g., 23 patients with RecurIndex-low but clinical high-risk), necessitating multidisciplinary review. Clinicians occasionally sought Oncotype DX for conflicting results, mirroring international scenarios where MammaPrint and Oncotype disagree[ 21 ]. While RecurIndex’s reliability in Chinese patients is strengthened by its guideline integration and cost-effectiveness, long-term outcome validation is critical. Real-world evidence will clarify the management of discordances, balancing genomic guidance with clinical judgment. Looking forward, further validation of the 28-gene panel in larger Asian cohorts and prospective trials is crucial. An ideal study would randomize patients to use one test vs another for decision-making and compare outcomes, or at least follow a large cohort where multiple assays are applied. Thus far, we have retrospective data like ours and the Taiwanese study by Yang et al. (2019) which showed the 18-gene RI-DR model was prognostically on par with the Oncotype in an Asian cohort[ 12 ]. We also have registry data indicating RecurIndex low-risk patients have excellent survival (> 98% 4-year RFS) [ 22 ]. These are promising, but a prospective trial in China (perhaps similar in concept to TAILORx or MINDACT) would provide the highest level of evidence. It is worth noting that the Chinese clinical community is rapidly embracing these tests – as mentioned, CSCO guidelines have incorporated them, and cost-effectiveness analyses are underway. In time, it will be important to assess not just prognostic ability but also predictive utility (i.e., does chemotherapy benefit the high-genomic-risk patients as expected?). Western trials showed that Oncotype high-risk (RS ≥ 26) derive clear chemotherapy benefits while low-risk do not[ 23 ]. For MammaPrint, high-risk patients did better with chemo in MINDACT (though that trial wasn't powered to show it definitively). For RecurIndex, this remains to be demonstrated prospectively (We are conducting relevant research and have completed enrollment (NCT04972448|| https://www.clinicaltrials.gov/ ) with the Clinical Trial Registry (May 2024). One ongoing effort is to see if RecurIndex can predict benefits from radiotherapy in addition to chemotherapy (given it has a RI-LR component for local recurrence. ClinicalTrials.gov Identifier NCT04517266). Conclusion Our study shows comparable prognostic performance between the 28-gene RecurIndex and 70-gene MammaPrint in Chinese patients with early-stage, hormone receptor-positive, HER2-negative breast cancer, demonstrating ~ 72% agreement. RecurIndex aligned more closely with clinicopathologic risk factors, classifying fewer patients as high-risk, which may reduce overtreatment. These findings support integrating RecurIndex into Chinese clinical guidelines as a cost-effective alternative to MammaPrint or Oncotype DX. Long-term survival outcomes are needed to confirm prognostic equivalence. Ethnicity-specific genomic tools like RecurIndex challenges universal "one-size-fits-all" approaches, underscoring the importance of population-tailored assays in precision oncology. Further validation of long-term outcomes and exploration of multi-signature integration are warranted to optimize risk prediction globally. Declarations ACKNOWLEDGEMENTS We thank all the participants of the study. Consent to Publish declaration All the patients included in this study had given informed consent for publication, and all the procedures performed were in accordance with relevant guidelines/regulations, as well as the 1964 Helsinki Declaration and its later amendments. Ethics Declaration All participants provided written informed consent for both participation in this research and the publication of its anonymized results. The study was approved by the Institutional Review Board of Zhejiang Cancer Hospital (Approval No.: IRB-2020-261). Declaration of conflicting interests The author(s) declared no potential conflicts of interest for the research, authorship, and/or publication of this article. Funding The project was supported by funding from the National Key Research and Development Program / International Cooperation in science and technology innovation (2019YFE0196500). Data availability The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Author Contribution X.W. had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: X.W.. Acquisition, analysis, or interpretation of data: W.C., Y.Y., J.L., G. Q., X.X., C.D., Y.T., and X.W.. Drafting of the manuscript: L.L.. Critical review of the manuscript for important intellectual content: L.L., Y.T., and X.W.. Statistical analysis: C.D., Y.T.. Funding acquisition: X.W.. Administrative, technical, or material support: C.D., Y.T., and X.W.. References Syed YY: Oncotype DX breast recurrence score®: a review of its use in early-stage breast cancer . Molecular diagnosis & therapy 2020, 24 (5):621-632. Slodkowska EA, Ross JS: MammaPrint™ 70-gene signature: another milestone in personalized medical care for breast cancer patients . Expert review of molecular diagnostics 2009, 9 (5):417-422. Curigliano G, Dent R, Llombart-Cussac A, Pegram M, Pusztai L, Turner N, Viale G: Incorporating clinicopathological and molecular risk prediction tools to improve outcomes in early HR+/HER2–breast cancer . NPJ Breast Cancer 2023, 9 (1):56. Sparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS, Hayes DF, Geyer CE, Jr., Dees EC, Goetz MP, Olson JA, Jr. et al : Adjuvant Chemotherapy Guided by a 21-Gene Expression Assay in Breast Cancer . N Engl J Med 2018, 379 (2):111-121. Cardoso F, van't Veer LJ, Bogaerts J, Slaets L, Viale G, Delaloge S, Pierga JY, Brain E, Causeret S, DeLorenzi M et al : 70-Gene Signature as an Aid to Treatment Decisions in Early-Stage Breast Cancer . N Engl J Med 2016, 375 (8):717-729. Chlebowski RT, Chen Z, Anderson GL, Rohan T, Aragaki A, Lane D, Dolan NC, Paskett ED, McTiernan A, Hubbell FA et al : Ethnicity and Breast Cancer: Factors Influencing Differences in Incidence and Outcome . JNCI: Journal of the National Cancer Institute 2005, 97 (6):439-448. Chen CH, Lu YS, Cheng AL, Huang CS, Kuo WH, Wang MY, Chao M, Chen IC, Kuo CW, Lu TP et al : Disparity in Tumor Immune Microenvironment of Breast Cancer and Prognostic Impact: Asian Versus Western Populations . Oncologist 2020, 25 (1):e16-e23. Lin C-H, Yap YS, Lee K-H, Im S-A, Naito Y, Yeo W, Ueno T, Kwong A, Li H, Huang S-M et al : Contrasting Epidemiology and Clinicopathology of Female Breast Cancer in Asians vs the US Population . JNCI: Journal of the National Cancer Institute 2019, 111 (12):1298-1306. Iqbal J, Ginsburg O, Rochon PA, Sun P, Narod SA: Differences in breast cancer stage at diagnosis and cancer-specific survival by race and ethnicity in the United States . Jama 2015, 313 (2):165-173. Huang E, Cheng SH, Dressman H, Pittman J, Tsou MH, Horng CF, Bild A, Iversen ES, Liao M, Chen CM et al : Gene expression predictors of breast cancer outcomes . Lancet 2003, 361 (9369):1590-1596. Lei L, Wang XJ, Mo YY, Cheng SH, Zhou Y: DGM-CM6: A New Model to Predict Distant Recurrence Risk in Operable Endocrine-Responsive Breast Cancer . Front Oncol 2020, 10 :783. Yang PS, Lee YH, Chung CF, Chang YC, Wang MY, Lo C, Tsai LW, Shih KH, Lei J, Yu BL et al : A preliminary report of head-to-head comparison of 18-gene-based clinical-genomic model and oncotype DX 21-gene assay for predicting recurrence of early-stage breast cancer . Jpn J Clin Oncol 2019, 49 (11):1029-1036. Jiang Z, Li J, Chen J, Liu Y, Wang K, Nie J, Wang X, Hao C, Yin Y, Wang S et al : Chinese Society of Clinical Oncology (CSCO) Breast Cancer Guidelines 2022 . Transl Breast Cancer Res 2022, 3 :13. Huang T-T, Lei L, Chen C-HA, Lu T-P, Jen C-W, Cheng SH-C: A new clinical-genomic model to predict 10-year recurrence risk in primary operable breast cancer patients . Sci Rep 2020, 10 (1):4861. Van't Veer LJ, Dai H, Van De Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, Van Der Kooy K, Marton MJ, Witteveen AT: Gene expression profiling predicts clinical outcome of breast cancer . Nature 2002, 415 (6871):530-536. Brandão M, Pondé N, Piccart-Gebhart M: Mammaprint™: a comprehensive review . Future oncology 2018, 15 (2):207-224. Shieh Y, Eklund M, Sawaya GF, Black WC, Kramer BS, Esserman LJ: Population-based screening for cancer: hope and hype . Nature Reviews Clinical Oncology 2016, 13 (9):550-565. Xing F, Chen T-H, Wang H, Ma X, Ding R, Lu Z: Differences of prediction performances by multigene assays in patients with breast cancer with distinct genetic backgrounds . In . : American Society of Clinical Oncology; 2023. Huang TT, Lei L, Chen CA, Lu TP, Jen CW, Cheng SH: A new clinical-genomic model to predict 10-year recurrence risk in primary operable breast cancer patients . Sci Rep 2020, 10 (1):4861. Chen K, Wu J, Fang Z, Shao X, Wang X: The Clinical Research and Latest Application of Genomic Assays in Early-Stage Breast Cancer . Technology in Cancer Research & Treatment 2022, 21 :15330338221117402. Socoteanu MP, O'Shaughnessy J, Hoskins K, Brufsky A, Graham CL, Vukelja SJ, Misleh JG, Tedesco KL, Layeequr Rahman R, Lee J: Clinical implications for patients with discordant oncotype and MammaPrint results . In . : American Society of Clinical Oncology; 2022. Wang H, Ma L, Zhang Y, Wang O, Wei Z, Xie X, Zha X, Zeng J, Lv Q, Ren Y et al : RecurIndex assay as an aid for adjuvant chemotherapy decisions in HR-positive HER2-negative breast cancer patients . Front Oncol 2022, 12 :896431. Schaafsma E, Zhang B, Schaafsma M, Tong CY, Zhang L, Cheng C: Impact of Oncotype DX testing on ER+ breast cancer treatment and survival in the first decade of use . Breast Cancer Res 2021, 23 (1):74. Additional Declarations No competing interests reported. Supplementary Files FigureS1.docx TableS1.docx TableS2.docx TableS3.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6832631","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":476335658,"identity":"929bad77-8541-485f-990a-d339fb81c88e","order_by":0,"name":"L. Lei","email":"","orcid":"","institution":"Zhejiang Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"L.","middleName":"","lastName":"Lei","suffix":""},{"id":476335660,"identity":"62375ee1-a7ca-49e8-817a-f626ca69e449","order_by":1,"name":"W. Cao","email":"","orcid":"","institution":"Zhejiang Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"W.","middleName":"","lastName":"Cao","suffix":""},{"id":476335661,"identity":"1a6a5c98-e185-4b14-a522-27f73715c29a","order_by":2,"name":"Y. Yu","email":"","orcid":"","institution":"Zhejiang Cancer Hospital","correspondingAuthor":false,"prefix":"","firstName":"Y.","middleName":"","lastName":"Yu","suffix":""},{"id":476335662,"identity":"69260b9c-7fd2-4c2b-91d9-7f3e93e6cb9a","order_by":3,"name":"J. Luo","email":"","orcid":"","institution":"Sichuan Academy of Medical Sciences \u0026 Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China","correspondingAuthor":false,"prefix":"","firstName":"J.","middleName":"","lastName":"Luo","suffix":""},{"id":476335667,"identity":"d1296e6c-d1a4-4a11-84ba-1af99a3eeae7","order_by":4,"name":"G. Qiao","email":"","orcid":"","institution":"Yantai Yuhuangding Hospital","correspondingAuthor":false,"prefix":"","firstName":"G.","middleName":"","lastName":"Qiao","suffix":""},{"id":476335668,"identity":"63c5b3f8-6bef-4586-a35e-4da124360fd1","order_by":5,"name":"X. Xie","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhejiang Chinese Medical University","correspondingAuthor":false,"prefix":"","firstName":"X.","middleName":"","lastName":"Xie","suffix":""},{"id":476335670,"identity":"2345c203-b337-4aca-8dd6-144c1100d8d4","order_by":6,"name":"C. Du","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"C.","middleName":"","lastName":"Du","suffix":""},{"id":476335671,"identity":"2c3e70ff-40ea-4d54-86b0-5e1101b0dc26","order_by":7,"name":"Y. Tan","email":"","orcid":"","institution":"State Key Laboratory of Neurology and Oncology Drug Development, Jiangsu Simcere Diagnostics Co., Ltd","correspondingAuthor":false,"prefix":"","firstName":"Y.","middleName":"","lastName":"Tan","suffix":""},{"id":476335672,"identity":"bdb0aca4-8aaa-4f3e-880c-731744ef17f7","order_by":8,"name":"Xiao-Jia Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYFACxgYGBh4JBjb2xuYHHwxs7IjXws9z+JjhjIK0ZOItk5yRliDN8+EQyAT8wOBGcuOHDzIW8gZnzhgY2xgcYGZgP3x0A14tZw42S87gkTDccLzH4HGOwR0+Bp60tBv4tJgdb2yQ5uGRYNwAsiXH4BkzgwSPGX4thxmbfwO12G+4kWMgbWFwmLGBoJbjjW0gWxJngrzPQIwW+zMH2yyBfknuBwVyj0FaMhshv0jOSH9842NPnW0bKCp//LGx42c/fAyvFjBg7EHisBFUDgY/iFM2CkbBKBgFIxQAADhCTfV25wcpAAAAAElFTkSuQmCC","orcid":"","institution":"Zhejiang Cancer Hospital","correspondingAuthor":true,"prefix":"","firstName":"Xiao-Jia","middleName":"","lastName":"Wang","suffix":""}],"badges":[],"createdAt":"2025-06-06 01:38:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6832631/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6832631/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":85743972,"identity":"0edd509a-5dc0-425f-b583-25ebfb49a65d","added_by":"auto","created_at":"2025-07-01 09:13:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":87843,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFlow chart for patient selection and assessment.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/100a2e6b9de06a226cea6552.png"},{"id":85743969,"identity":"d971d2e4-41c4-403d-a9e4-1258191e1e3a","added_by":"auto","created_at":"2025-07-01 09:13:15","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":224261,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRisk classification results by clinical criteria vs. multigene assays. \u003c/strong\u003eEach bar shows the percentage of patients categorized as low-risk or high-risk according to: clinicopathologic risk factors (left), the 28-gene RecurIndex genomic assay (center), and the 70-gene MammaPrint genomic assay (right). The clinical criteria designated 48.5% of patients as high-risk, whereas the 28-gene and 70-gene assays identified 26% and 34% as high-risk, respectively. The 28-gene panel thus classified a greater percentage of patients as low-risk compared to the 70-gene panel. This suggests that the 28-gene signature may be more conservative in assigning high-risk status, potentially sparing additional patients from chemotherapy recommendations.\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/2dea89dc62e5e9d8f521a44f.jpeg"},{"id":89486552,"identity":"c3a0f391-eabc-46a7-a849-f64d1b9f1963","added_by":"auto","created_at":"2025-08-20 12:54:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2094793,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/8aa1f227-6e50-44e8-a5a3-263ec6cc4144.pdf"},{"id":85742266,"identity":"d2812cbe-66bb-4b79-8928-c3a3839b80ea","added_by":"auto","created_at":"2025-07-01 09:05:15","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":91126,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/b7eef145f272e6c0311d01c8.docx"},{"id":85742270,"identity":"fcb26a24-bef9-498c-86c0-4e760a51a42d","added_by":"auto","created_at":"2025-07-01 09:05:15","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":19475,"visible":true,"origin":"","legend":"","description":"","filename":"TableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/437564838430f0a80c91b8bd.docx"},{"id":85745026,"identity":"d9bc25dc-4904-474f-bd58-3233e76a2263","added_by":"auto","created_at":"2025-07-01 09:21:15","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":19786,"visible":true,"origin":"","legend":"","description":"","filename":"TableS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/454fb30a724349e7fadb79e6.docx"},{"id":85742267,"identity":"34749bfb-0bea-4241-bfd8-3f4421fddd30","added_by":"auto","created_at":"2025-07-01 09:05:15","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":19063,"visible":true,"origin":"","legend":"","description":"","filename":"TableS3.docx","url":"https://assets-eu.researchsquare.com/files/rs-6832631/v1/92b39ae202f70c4b8411e633.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Comparative Evaluation of a 28-Gene versus 70-Gene Panel for Recurrence Risk Prediction in Early-Stage HR-Positive/HER2-Negative Breast Cancer in Chinese Women","fulltext":[{"header":"Introduction","content":"\u003cp\u003eBreast cancer heterogeneity drives the use of multigene assays (e.g., 21-gene Oncotype DX[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], 70-gene MammaPrint[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]) to refine risk stratification in early HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;disease[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Landmark trials (TAILORx[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], MINDACT[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]) validated their prognostic utility: TAILORx showed\u0026thinsp;~\u0026thinsp;70% of early HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;patients (especially intermediate-risk) avoided chemotherapy safely with endocrine therapy alone, while MINDACT demonstrated 94\u0026ndash;95% 5-year metastasis-free survival in clinically high-risk/gnomically low-risk patients spared chemotherapy. These assays guide chemotherapy de-escalation for genomic low-risk patients while identifying high-risk subgroups for intensified treatment.\u003c/p\u003e \u003cp\u003eMost widely used multigene signatures (e.g., Oncotype DX) were developed and validated in Western populations, raising questions about their generalizability for risk stratification in Asian patients. Studies highlight ethnic differences in breast cancer biology and epidemiology, including distinct gene expression profiles and clinical outcomes in Asian versus Western populations[\u003cspan additionalcitationids=\"CR7 CR8\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. To address this, the RecurIndex, a 28-gene signature derived from Asian cohorts, was developed. It integrates tumor gene expression with clinicopathological variables (age, lymphovascular invasion, etc.) to generate separate risk indices as the recurrence index for distant recurrence (RI-DR) and the recurrence index for local recurrence (RI-LR), providing binary risk stratification (low vs. high) [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Validation studies demonstrated\u0026thinsp;\u0026gt;\u0026thinsp;80% concordance with Oncotype DX in recurrence risk prediction, with both tools effectively stratifying 4-year relapse outcomes[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. In 2022, RecurIndex became the first China-developed multigene assay endorsed by the Chinese Society of Clinical Oncology (CSCO) guidelines[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. A national expert consensus recommends its use alongside established assays (Oncotype DX, MammaPrint) to guide adjuvant therapy decisions in early-stage HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;breast cancer[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDespite advances, comparative data between the 28-gene RecurIndex and Western assays (e.g., 70-gene MammaPrint) in Asian populations remain scarce. We retrospectively compared both tests in Chinese women with early-stage HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;breast cancer, applying RecurIndex to MammaPrint-tested tumor samples. Analyses included risk stratification concordance, alignment with clinicopathologic factors, and clinical implications of discordant results. This head-to-head study evaluates whether RecurIndex offers equivalent risk assessment to MammaPrint, potential advantages (e.g., enhanced low-risk identification), and ethnic-specific applicability. Findings address critical gaps in interpreting conflicting risk classifications and guide confidence in using indigenous genomic tools for therapy decisions in Chinese patients.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cp\u003e\u003cstrong\u003eS\u003c/strong\u003e\u003cstrong\u003etudy\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDesign\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eand\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ePatient\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ePopulation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis\u0026nbsp;study\u0026nbsp;was\u0026nbsp;a\u0026nbsp;retrospective\u0026nbsp;comparative\u0026nbsp;analysis\u0026nbsp;of\u0026nbsp;genomic risk assays in a cohort of Chinese breast cancer patients. Eligible patients were women with early-stage (pathologic stage T1-T3, N0-N1, M0) HR-positive, HER2-negative breast cancer who had undergone genomic testing with the 70-gene MammaPrint panel after primary surgical treatment. All patients had surgery between January 2019 and October 2022 at participating centers. Tumor samples and clinical data were collected under institutional review board approval, and all patients gave informed consent for genetic testing and data use. We identified a total of 99 patients meeting these criteria. By design, all 99 patients had available MammaPrint results (high or low risk) from their initial post-operative evaluation. For this study, formalin-fixed, paraffin-embedded (FFPE) tissue blocks from the primary tumor of these patients were retrieved and subjected to testing with the 28-gene RecurIndex assay in a central accredited laboratory. The RecurIndex test was performed on FFPE sections using the validated platform and algorithm provided by the test developer (Simcere Diagnostics, Nanjing, China), blinded to patients\u0026rsquo; MammaPrint results. Figure S1 showed the 28-gene testing laboratory related operations and quality control requirements. The output of the 28-gene assay is a numeric score: RI-DR (Recurrence Index for distant recurrence; low-risk \u0026lt;29.3, high-risk \u0026ge;29.3) and RI-LR (Recurrence Index for locoregional recurrence; low-risk \u0026lt;35.3, high-risk\u0026nbsp;\u0026ge;35.3). The 70-gene MammaPrint results were available from patients\u0026rsquo; records (MammaPrint categorizes tumors as \u0026ldquo;Low-Risk\u0026rdquo; or \u0026ldquo;High-Risk\u0026rdquo; of distant metastasis based on the proprietary 70-gene expression algorithm). Figure 1 shows the patient selection and assessment of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eData\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eand\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eRisk\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eDefinitions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe\u0026nbsp;collected\u0026nbsp;clinicopathological\u0026nbsp;variables\u0026nbsp;including\u0026nbsp;patient\u0026nbsp;age\u0026nbsp;at diagnosis, tumor size (pathological T stage), nodal status (N stage), histologic grade, lymphovascular invasion (LVI) status, estrogen receptor (ER), and progesterone receptor (PR) expression, Ki-67 proliferation index, and adjuvant treatment received. All tumors were ER-positive (immunohistochemistry ER\u0026gt;1%) and HER2-negative (immunohistochemistry 0-1+ or non-amplified by in situ hybridization) as part of the inclusion criteria. We defined a clinicopathologic risk score based on six traditional risk factors associated with recurrence in HR+ breast cancer. The following factors were considered adverse (high-risk) features in our analysis:\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRisk Factor\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHigh-Risk Criteria\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003e\u0026le;40 years at diagnosis\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003ePathological tumor size\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003e\u0026ge;T2 (\u0026ge;2 cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003eNodal status\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003eN1 (1-3 positive axillary nodes)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003eLymphovascular invasion (LVI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003eHistologic grade\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003eGrade 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 274px;\"\u003e\n \u003cp\u003eKi-67 proliferation index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 350px;\"\u003e\n \u003cp\u003e\u0026ge;20%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eEach\u0026nbsp;patient\u0026rsquo;s\u0026nbsp;tumor\u0026nbsp;was\u0026nbsp;evaluated\u0026nbsp;for\u0026nbsp;these\u0026nbsp;factors;\u0026nbsp;patients\u0026nbsp;with\u0026nbsp;two\u0026nbsp;or\u0026nbsp;more\u0026nbsp;high-risk clinicopathologic features were categorized as \u0026ldquo;clinical high-risk,\u0026rdquo; whereas those with 0-1 of these features were categorized as \u0026ldquo;clinical low-risk.\u0026rdquo; This dichotomous clinical risk classification was used to examine how well each genomic test agreed with conventional risk assessment. We note that our definition is analogous to but not the same as criteria used in trials like MINDACT; it was tailored to identify patients who, based on traditional parameters, might be recommended chemotherapy in the absence of genomic information.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eS\u003c/strong\u003e\u003cstrong\u003etatistical\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eAnalysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary analysis compared risk stratification (high/low) between the 28-gene and 70-gene panels using percentage agreement and Cohen\u0026rsquo;s kappa coefficient (\u003cem\u003e\u0026kappa;\u003c/em\u003e \u0026gt; 0.75: excellent; 0.4-0.75: moderate; \u0026lt;0.4: poor). Concordance with clinical risk categories (high-risk defined as \u0026ge;2 clinicopathologic factors) was similarly assessed. Chi-square/Fisher\u0026rsquo;s exact tests evaluated differences in classification distributions (e.g., genomic vs clinical risk), with McNemar\u0026rsquo;s test for paired discordant cases. Due to zero recurrence/death events, survival analysis was omitted; focus remained on risk reclassification metrics. Descriptive statistics summarized patient characteristics (medians/ranges for continuous variables; frequencies/percentages for categorical data). Analyses used SPSS v26 (IBM), with two-tailed \u003cem\u003ep\u003c/em\u003e\u0026lt;0.05 as significance threshold. \u0026nbsp;\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eF\u003c/strong\u003e\u003cstrong\u003eollow-up\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eand\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eOutcome\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eCollection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePatients were followed according to standard practice with clinical exams and imaging as indicated. We recorded any instances of breast cancer recurrence (locoregional or distant) or breast cancer-specific death after surgery. The median follow-up duration was 56 months (approximately 4.7 years), with a follow-up cut-off as of December 30, 2024. Follow-up time was calculated from the date of surgery. At the last follow-up, we noted the survival status and disease status (no evidence of disease, alive with recurrence, or deceased) for each patient. This outcome data was used to provide preliminary observations, although, with no events, no comparative survival analysis between tests could be performed.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePa\u003c/strong\u003e\u003cstrong\u003etient\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eCharacteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 99 patients were included, all of whom were Chinese women diagnosed with early-stage (stage I-II) HR-positive, HER2-negative breast cancer. Key clinicopathologic characteristics of the cohort are summarized in Table 1. The median age was 52 years (range 29-74 years), with 11% of patients \u0026le;40 years old, reflecting that a minority were very young at diagnosis. Tumor size was T1 in 61% of cases, T2 in 38%, and there was one patient with a T3 tumor (3 cm) but node-negative. Nodal status was positive (N1, 1-3 involved nodes) in 36% of patients, while 64% had node-negative disease. Most tumors (77%) were histologic grade 2; only 11% were grade 3, and 11% were grade 1 (grade was unspecified for one case). Lymphovascular invasion was present in 20% of tumors. All 99 tumors were ER-positive; PR was positive in 82% and negative in 13% (5% unknown). Ki-67 proliferation index was reported for 89 patients: among those, 45 tumors (approximately half of the evaluable cases) had high Ki-67 (\u0026ge;20%). Overall, using our predefined clinical risk criteria, 48 patients (48.5%) were categorized as clinical high-risk (having \u0026ge;2 high-risk factors), and the remaining 51 (51.5%) were clinical low-risk (0-1 factors).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1. Baseline characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eCharacteristic\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistribution (n = 99)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003eMedian 52 years (range 29\u0026ndash;74)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eAge \u0026gt; 40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e88 patients (89%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eAge \u0026le; 40\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e11 patients (11%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTumor size\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eT1 (\u0026le;2 cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e60 (61%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eT2 (\u0026gt;2-5 cm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e38 (38%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eT3 (\u0026gt;5 cm, N0 only)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e1 (1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLymph node status\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eN0\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e63 (64%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eN1 (1\u0026ndash;3 positive nodes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e36 (36%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eHistologic Grade\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eGrade 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e11 (11%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eGrade 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e76 (77%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eGrade 3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e11 (11%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e1 (1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLVI\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003ePresent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e20 (20%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eAbsent\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e79 (80%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEstrogen Receptor\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e99 (100%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eProgesterone Receptor\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003ePositive\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e81 (82%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eNegative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e13 (13%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e5 (5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eKi-67 Index\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003e\u0026lt; 20%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e44 (44%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003e\u0026ge; 20%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e45 (46%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e10 (10%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClinical Risk Category\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eLow (0\u0026ndash;1 risk factors)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e51 (52%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eHigh (\u0026ge;2 risk factors)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e48 (48%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eChemotherapy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e42 (42%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e39 (40%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e18 (18%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSurgery\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eMastectomy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e31 (31%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eBCS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e68 (69%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRadiation therapy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e31 (31%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e52 (53%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e16 (16%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\" style=\"width: 435px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eEndocrine therapy\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e96 (97%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 183px;\"\u003e\n \u003cp\u003eUnknown\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 252px;\"\u003e\n \u003cp\u003e3 (3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eBCS, breast-conserving surgery; LVI, lymphovascular invasion.\u003c/p\u003e\n\u003cp\u003eAll patients received primary surgery. In terms of adjuvant therapies, by being HR+, all patients were treated with endocrine therapy (tamoxifen and/or aromatase inhibitor as appropriate). Chemotherapy usage in this cohort was not uniform: treating oncologists integrated the 70-gene test results with clinical factors to decide on chemotherapy. Overall, 39 out of 99 patients (39.4%) received adjuvant chemotherapy, while 42 (42.4%) did not and 18 (18.2%) reflected the intermediate risk nature of many cases.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNotably, during a median follow-up of 56 months, no patient experienced a recurrence or cancer-related death. Two patients were lost to follow-up before 3 years, but among the 97 patients monitored, none had developed distant metastasis or locoregional recurrence at the last contact. While the follow-up duration is relatively short and longer observation is needed, this absence of events is consistent with the fact that many patients were low genomic risk and received appropriate therapy. It precludes any direct comparison of survival outcomes between risk groups at this time.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRisk\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eStratification\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eby\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e28-Gene\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003evs\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e70-Gene\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ePanels\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe next compared the recurrence risk categorization provided by the 28-gene and 70-gene assays for each patient in Table S1-3. The 70-gene MammaPrint test had classified 34 of the 99 patients (34%) as High Risk and the remaining 65 (66%) as Low Risk of distant recurrence. In comparison, the 28-gene RecurIndex assay classified only 26 patients (26%) as High Risk, while 73 (74%) were Low Risk. Thus, the 28-gene panel identified a larger proportion of patients as low-risk (about three-quarters of the cohort) and correspondingly fewer as high-risk, relative to the 70-gene panel. Figure 2 illustrates the distribution of risk results across the two genomic assays and the clinical criteria. The clinical risk assessment (by traditional factors) labeled the highest percentage of patients as high-risk (48.5%), whereas both genomic tests categorized fewer patients as high-risk, especially the 28-gene signature.\u003c/p\u003e\n\u003cp\u003eWe found an overall concordance of 72% (71 out of 99 patients) between the 28-gene and 70-gene tests in binary risk assignment. In 71 cases, the two assays agreed on whether the tumor was low-risk or high-risk. In the remaining 28 cases (28% of the cohort), the assays disagreed: 18 patients were called high-risk by 70-gene but low-risk by 28-gene, while 10 patients were high-risk by 28-gene but low-risk by 70-gene. This pattern indicates that most discordances (18 vs 10 cases) were instances where the 70-gene test flagged risk that the 28-gene did not. Cohen\u0026rsquo;s kappa coefficient for the agreement was calculated to be 0.54 (approximately), consistent with moderate agreement beyond chance. This kappa falls in the upper-moderate range, underscoring that while the two assays often concur, there is a non-negligible subset of tumors that they stratify differently. Importantly, among those 28 discordant cases, none have recurred to date, so it remains unclear which assay was \u0026ldquo;correct\u0026rdquo; in identifying risk that longer follow-up is needed to determine if one test was more prognostically accurate.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eComparison\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003ewith\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eClinicopathologic\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eRisk\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eand\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eConsistency\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eAnalysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur analysis compared the alignment of the 28-gene RecurIndex and 70-gene MammaPrint assays with clinicopathologic risk categories in Chinese HR+/HER2\u0026minus; breast cancer patients in Table 2. The 28-gene panel demonstrated stronger consistency with clinical risk stratification than the 70-gene panel. Among 51 clinically low-risk patients, 98% (50/51) were classified as low-risk by the 28-gene test, while the 70-gene assay reclassified 16% (8/51) as high-risk, a statistically significant discrepancy (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001). Conversely, in the 48 clinically high-risk patients, both tests classified approximately half as genomically high-risk (52% by 28-gene vs. 54% by 70-gene; \u003cem\u003ep\u003c/em\u003e = 0.80). Cohen\u0026rsquo;s kappa values confirmed moderate agreement with clinical risk for both assays but higher concordance for the 28-gene panel (\u003cem\u003e\u0026kappa;\u003c/em\u003e = 0.51 vs. \u003cem\u003e\u0026kappa;\u0026nbsp;\u003c/em\u003e= 0.39). \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. Comparative analysis of the performance of risk-prediction by the 28-gene and the 70-gene panels.\u003c/strong\u003e\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\u0026nbsp;\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"605\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" rowspan=\"2\" style=\"width: 113px;\"\u003e\n \u003cp\u003eCharacteristic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 132px;\"\u003e\n \u003cp\u003e28-gene (%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003eKappa\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 142px;\"\u003e\n \u003cp\u003e70-gene (%)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 47px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd rowspan=\"2\" style=\"width: 57px;\"\u003e\n \u003cp\u003eKappa\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003eHigh-risk\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003eLow-risk\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003eHigh-risk\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 76px;\"\u003e\n \u003cp\u003eLow-risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003eClinical risk\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003eHigh\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e25(25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e23(23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e48(48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e26(26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e22(22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e48(48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.39\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 47px;\"\u003e\n \u003cp\u003eLow\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e1(1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e50(51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e51(52)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eP\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e8(8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e43(43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e51(51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eP\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" style=\"width: 113px;\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e26(26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e73(74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e99(100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e- \u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 66px;\"\u003e\n \u003cp\u003e34(34)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e65(66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 66px;\"\u003e\n \u003cp\u003e99(100)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e- \u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eThe 28-gene test produced fewer high-risk classifications overall (26% vs. 34% by 70-gene; p = 0.24), potentially reducing chemotherapy recommendations. Notably, the 70-gene assay frequently discorded with clinical profiles: it labeled 16% of clinical low-risk patients as high-risk (vs. 2% by 28-gene) and downgraded 46% of clinical high-risk patients to low-risk (vs. 48% by 28-gene). These reclassifications mirror findings from trials like MINDACT, where genomic low-risk patients safely avoided chemotherapy. In our cohort, none of the genomically low-risk patients (regardless of clinical risk) experienced recurrence during 56-month follow-up. However, the 70-gene test identified 8 clinical low-risk patients as high-risk, raising questions about potential overestimation or detection of occult aggressive biology. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhile both tests reclassified ~50% of clinical high-risk patients to low genomic risk-suggesting chemotherapy, the 28-gene panel aligned more closely with traditional risk factors. For example, 9 of 10 patients uniquely labeled high-risk by the 28-gene test had multiple clinical risk factors, whereas 7 of 8 patients uniquely high-risk by the 70-gene test had favorable clinical profiles. This divergence implies the 28-gene signature may better reflect clinician concerns, while the 70-gene test could flag unexpected risk in clinically low-risk cases. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn summary, the 28-gene and 70-gene assays showed moderate overall agreement (72%) in risk stratification. Key differences lie in their alignment with clinical risk: the 28-gene panel is more conservative and clinicopathologically concordant, whereas the 70-gene test exhibits greater discordance, particularly in upgrading clinically low-risk patients. Both tools enable chemotherapy de-escalation in ~50% of clinical high-risk cases, but long-term outcome data are critical to validate their prognostic accuracy. The absence of recurrences precludes definitive conclusions, underscoring the need for extended follow-up to clarify whether observed discrepancies reflect overestimation or improved risk detection. These findings highlight the importance of integrating population-tailored genomic tools with clinical judgment in precision oncology.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we performed the first head-to-head comparison of a 28-gene Asian genomic signature (RecurIndex) against the well-known 70-gene MammaPrint assay for recurrence risk prediction in a cohort of Chinese women with early HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;breast cancer. Our findings demonstrate that the two multigene assays have largely similar performance in stratifying patients into low vs high-risk categories but with some notable differences in classification tendencies and concordance with clinical factors. The overall agreement of 72% (\u003cem\u003eκ\u003c/em\u003e\u0026thinsp;~\u0026thinsp;0.5) indicates that in about three-quarters of cases the 28-gene and 70-gene tests would lead to the same treatment recommendation, which is encouraging in terms of interchangeability. In the remaining one-quarter of cases, however, discordant results emerged that could potentially alter management depending on which assay is followed. This highlights the importance of understanding each test\u0026rsquo;s characteristics and the context of their development when interpreting results for clinical decision-making.\u003c/p\u003e \u003cp\u003eThe 28-gene RecurIndex, derived from Asian data, demonstrated stronger alignment with clinicopathologic risk factors in Chinese HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;breast cancer patients compared to the 70-gene MammaPrint (\u003cem\u003eκ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.51 vs. 0.39). This likely stems from RecurIndex\u0026rsquo;s design as a clinical-genomic hybrid model integrating gene expression and clinical variables[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], validated with 93% accuracy for 5-year locoregional recurrence prediction in Chinese cohorts[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. In contrast, MammaPrint, developed using Western populations[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], prioritizes proliferation genes without clinical inputs, potentially capturing \"hidden\" high-risk biology[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Notably, MammaPrint classified 16% of clinical low-risk patients as high-risk vs. 2% by RecurIndex (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), raising concerns about overestimation in Asian populations where genomic risk profiles may differ[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. A recent study found higher average genomic risk scores in Western vs. Eastern cohorts using RecurIndex (10.5% vs. 6.6% high-risk) [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], aligning with our observation of more MammaPrint high-risk calls in Chinese patients. RecurIndex\u0026rsquo;s high-risk group showed worse 4-year metastasis-free survival in Asian cohorts (52%) compared to Oncotype/PAM50[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], suggesting sharper risk discrimination. While no recurrences occurred in our cohort, MammaPrint-flagged patients (n\u0026thinsp;=\u0026thinsp;8) might reflect overestimation, whereas the single RecurIndex-high/MammaPrint-low case could indicate undetected aggressive biology. These findings underscore population-specific genomic calibration: RecurIndex\u0026rsquo;s Asian-centric design may better balance clinical-genomic concordance, while Western-derived tools like MammaPrint risk misclassification in non-European populations.\u003c/p\u003e \u003cp\u003eUsing the 28-gene RecurIndex versus the 70-gene MammaPrint in Chinese HR+/HER2\u0026thinsp;\u0026minus;\u0026thinsp;breast cancer reduces chemotherapy recommendations (26% vs 34%), primarily sparing clinically low-risk patients. While RecurIndex aligns closely with clinicopathologic factors, MammaPrint identifies 8 clinical low-risk patients as high-risk. However, MINDACT trial data suggest clinical low-risk patients (even genomically high-risk) have excellent outcomes without chemotherapy (5-year DMFS\u0026thinsp;~\u0026thinsp;95\u0026ndash;97%; no significant chemo benefit) [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In our cohort, none of these discordant cases recurred, implying MammaPrint\u0026rsquo;s additional high-risk calls may represent false positives. The RecurIndex's conservative alignment with favorable clinical profiles may reduce overtreatment risks, while MammaPrint's genomic risk stratification may lack clinical relevance in Asian populations. Both assays effectively de-escalate therapy for ~\u0026thinsp;50% of clinical high-risk patients. Prospective studies are needed to validate whether RecurIndex\u0026rsquo;s restraint or MammaPrint\u0026rsquo;s broader risk classification optimizes outcomes in Chinese patients, particularly for discordant cases.\u003c/p\u003e \u003cp\u003eOur study has limitations. First, the small sample size (n\u0026thinsp;=\u0026thinsp;99), selected from patients with prior MammaPrint testing, risks selection bias, as genomic testing is typically ordered for chemotherapy-uncertain cases (often intermediate clinical risk). Second, the median 4.7-year follow-up is insufficient for HR\u0026thinsp;+\u0026thinsp;breast cancer, where recurrences may emerge beyond 5\u0026ndash;10 years; the absence of events precludes direct prognostic comparison. Third, we excluded Oncotype DX, PAM50, and EndoPredict, though a recent analysis suggested RecurIndex may outperform other assays in Asian cohorts[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Prospective multi-assay comparisons (including Oncotype DX, used in China at high cost) are needed to optimize test selection. Finally, our cohort\u0026rsquo;s balanced clinical risk distribution (50% high/low) may not reflect real-world populations. Extended follow-up will clarify whether genomic high-risk groups (by either test) develop recurrences, particularly in discordant cases. While our focus on MammaPrint reflects clinical availability, broader comparisons would better inform precision oncology strategies in Chinese patients.\u003c/p\u003e \u003cp\u003eOur study supports the clinical adoption of the 28-gene RecurIndex (CSCO-endorsed) in Chinese HR+/HER2- breast cancer, demonstrating risk stratification comparable to MammaPrint while reducing chemotherapy recommendations (26% vs 34%). RecurIndex offers logistical advantages over MammaPrint, including lower cost and faster local processing. Its alignment with clinical factors may spare patients unnecessary treatment toxicity, particularly for borderline cases where RecurIndex-low results align with MammaPrint (or where MammaPrint-high calls lack prognostic urgency). However, caution remains for discordant cases (e.g., 23 patients with RecurIndex-low but clinical high-risk), necessitating multidisciplinary review. Clinicians occasionally sought Oncotype DX for conflicting results, mirroring international scenarios where MammaPrint and Oncotype disagree[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. While RecurIndex\u0026rsquo;s reliability in Chinese patients is strengthened by its guideline integration and cost-effectiveness, long-term outcome validation is critical. Real-world evidence will clarify the management of discordances, balancing genomic guidance with clinical judgment.\u003c/p\u003e \u003cp\u003eLooking forward, further validation of the 28-gene panel in larger Asian cohorts and prospective trials is crucial. An ideal study would randomize patients to use one test vs another for decision-making and compare outcomes, or at least follow a large cohort where multiple assays are applied. Thus far, we have retrospective data like ours and the Taiwanese study by Yang et al. (2019) which showed the 18-gene RI-DR model was prognostically on par with the Oncotype in an Asian cohort[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. We also have registry data indicating RecurIndex low-risk patients have excellent survival (\u0026gt;\u0026thinsp;98% 4-year RFS) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. These are promising, but a prospective trial in China (perhaps similar in concept to TAILORx or MINDACT) would provide the highest level of evidence. It is worth noting that the Chinese clinical community is rapidly embracing these tests \u0026ndash; as mentioned, CSCO guidelines have incorporated them, and cost-effectiveness analyses are underway. In time, it will be important to assess not just prognostic ability but also predictive utility (i.e., does chemotherapy benefit the high-genomic-risk patients as expected?). Western trials showed that Oncotype high-risk (RS\u0026thinsp;\u0026ge;\u0026thinsp;26) derive clear chemotherapy benefits while low-risk do not[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. For MammaPrint, high-risk patients did better with chemo in MINDACT (though that trial wasn't powered to show it definitively). For RecurIndex, this remains to be demonstrated prospectively (We are conducting relevant research and have completed enrollment (NCT04972448||\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.clinicaltrials.gov/\u003c/span\u003e\u003cspan address=\"https://www.clinicaltrials.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) with the Clinical Trial Registry (May 2024). One ongoing effort is to see if RecurIndex can predict benefits from radiotherapy in addition to chemotherapy (given it has a RI-LR component for local recurrence. ClinicalTrials.gov Identifier NCT04517266).\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003e Our study shows comparable prognostic performance between the 28-gene RecurIndex and 70-gene MammaPrint in Chinese patients with early-stage, hormone receptor-positive, HER2-negative breast cancer, demonstrating\u0026thinsp;~\u0026thinsp;72% agreement. RecurIndex aligned more closely with clinicopathologic risk factors, classifying fewer patients as high-risk, which may reduce overtreatment. These findings support integrating RecurIndex into Chinese clinical guidelines as a cost-effective alternative to MammaPrint or Oncotype DX. Long-term survival outcomes are needed to confirm prognostic equivalence. Ethnicity-specific genomic tools like RecurIndex challenges universal \"one-size-fits-all\" approaches, underscoring the importance of population-tailored assays in precision oncology. Further validation of long-term outcomes and exploration of multi-signature integration are warranted to optimize risk prediction globally.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank all the participants of the study.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Publish declaration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll the patients included in this study had given informed consent for publication, and all the procedures performed were in accordance with relevant guidelines/regulations, as well as the 1964 Helsinki Declaration and its later amendments.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Declaration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll participants provided written informed consent for both participation in this research and the publication of its anonymized results. The study was approved by the Institutional Review Board of Zhejiang Cancer Hospital (Approval No.: IRB-2020-261).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of conflicting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author(s) declared no potential conflicts of interest for the research, authorship, and/or publication of this article.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe project was supported by funding from the National Key Research and Development Program / International Cooperation in science and technology innovation (2019YFE0196500). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eX.W. had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. Concept and design: X.W.. Acquisition, analysis, or interpretation of data: W.C., Y.Y., J.L., G. Q., X.X., C.D., Y.T., and X.W.. Drafting of the manuscript: L.L.. Critical review of the manuscript for important intellectual content: L.L., Y.T., and X.W.. Statistical analysis: C.D., Y.T.. Funding acquisition: X.W.. Administrative, technical, or material support: C.D., Y.T., and X.W..\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSyed YY: \u003cstrong\u003eOncotype DX breast recurrence score\u0026reg;: a review of its use in early-stage breast cancer\u003c/strong\u003e. \u003cem\u003eMolecular diagnosis \u0026amp; therapy \u003c/em\u003e2020, \u003cstrong\u003e24\u003c/strong\u003e(5):621-632.\u003c/li\u003e\n\u003cli\u003eSlodkowska EA, Ross JS: \u003cstrong\u003eMammaPrint\u0026trade; 70-gene signature: another milestone in personalized medical care for breast cancer patients\u003c/strong\u003e. \u003cem\u003eExpert review of molecular diagnostics \u003c/em\u003e2009, \u003cstrong\u003e9\u003c/strong\u003e(5):417-422.\u003c/li\u003e\n\u003cli\u003eCurigliano G, Dent R, Llombart-Cussac A, Pegram M, Pusztai L, Turner N, Viale G: \u003cstrong\u003eIncorporating clinicopathological and molecular risk prediction tools to improve outcomes in early HR+/HER2\u0026ndash;breast cancer\u003c/strong\u003e. \u003cem\u003eNPJ Breast Cancer \u003c/em\u003e2023, \u003cstrong\u003e9\u003c/strong\u003e(1):56.\u003c/li\u003e\n\u003cli\u003eSparano JA, Gray RJ, Makower DF, Pritchard KI, Albain KS, Hayes DF, Geyer CE, Jr., Dees EC, Goetz MP, Olson JA, Jr.\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eAdjuvant Chemotherapy Guided by a 21-Gene Expression Assay in Breast Cancer\u003c/strong\u003e. \u003cem\u003eN Engl J Med \u003c/em\u003e2018, \u003cstrong\u003e379\u003c/strong\u003e(2):111-121.\u003c/li\u003e\n\u003cli\u003eCardoso F, van\u0026apos;t Veer LJ, Bogaerts J, Slaets L, Viale G, Delaloge S, Pierga JY, Brain E, Causeret S, DeLorenzi M\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003e70-Gene Signature as an Aid to Treatment Decisions in Early-Stage Breast Cancer\u003c/strong\u003e. \u003cem\u003eN Engl J Med \u003c/em\u003e2016, \u003cstrong\u003e375\u003c/strong\u003e(8):717-729.\u003c/li\u003e\n\u003cli\u003eChlebowski RT, Chen Z, Anderson GL, Rohan T, Aragaki A, Lane D, Dolan NC, Paskett ED, McTiernan A, Hubbell FA\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eEthnicity and Breast Cancer: Factors Influencing Differences in Incidence and Outcome\u003c/strong\u003e. \u003cem\u003eJNCI: Journal of the National Cancer Institute \u003c/em\u003e2005, \u003cstrong\u003e97\u003c/strong\u003e(6):439-448.\u003c/li\u003e\n\u003cli\u003eChen CH, Lu YS, Cheng AL, Huang CS, Kuo WH, Wang MY, Chao M, Chen IC, Kuo CW, Lu TP\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eDisparity in Tumor Immune Microenvironment of Breast Cancer and Prognostic Impact: Asian Versus Western Populations\u003c/strong\u003e. \u003cem\u003eOncologist \u003c/em\u003e2020, \u003cstrong\u003e25\u003c/strong\u003e(1):e16-e23.\u003c/li\u003e\n\u003cli\u003eLin C-H, Yap YS, Lee K-H, Im S-A, Naito Y, Yeo W, Ueno T, Kwong A, Li H, Huang S-M\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eContrasting Epidemiology and Clinicopathology of Female Breast Cancer in Asians vs the US Population\u003c/strong\u003e. \u003cem\u003eJNCI: Journal of the National Cancer Institute \u003c/em\u003e2019, \u003cstrong\u003e111\u003c/strong\u003e(12):1298-1306.\u003c/li\u003e\n\u003cli\u003eIqbal J, Ginsburg O, Rochon PA, Sun P, Narod SA: \u003cstrong\u003eDifferences in breast cancer stage at diagnosis and cancer-specific survival by race and ethnicity in the United States\u003c/strong\u003e. \u003cem\u003eJama \u003c/em\u003e2015, \u003cstrong\u003e313\u003c/strong\u003e(2):165-173.\u003c/li\u003e\n\u003cli\u003eHuang E, Cheng SH, Dressman H, Pittman J, Tsou MH, Horng CF, Bild A, Iversen ES, Liao M, Chen CM\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eGene expression predictors of breast cancer outcomes\u003c/strong\u003e. \u003cem\u003eLancet \u003c/em\u003e2003, \u003cstrong\u003e361\u003c/strong\u003e(9369):1590-1596.\u003c/li\u003e\n\u003cli\u003eLei L, Wang XJ, Mo YY, Cheng SH, Zhou Y: \u003cstrong\u003eDGM-CM6: A New Model to Predict Distant Recurrence Risk in Operable Endocrine-Responsive Breast Cancer\u003c/strong\u003e. \u003cem\u003eFront Oncol \u003c/em\u003e2020, \u003cstrong\u003e10\u003c/strong\u003e:783.\u003c/li\u003e\n\u003cli\u003eYang PS, Lee YH, Chung CF, Chang YC, Wang MY, Lo C, Tsai LW, Shih KH, Lei J, Yu BL\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eA preliminary report of head-to-head comparison of 18-gene-based clinical-genomic model and oncotype DX 21-gene assay for predicting recurrence of early-stage breast cancer\u003c/strong\u003e. \u003cem\u003eJpn J Clin Oncol \u003c/em\u003e2019, \u003cstrong\u003e49\u003c/strong\u003e(11):1029-1036.\u003c/li\u003e\n\u003cli\u003eJiang Z, Li J, Chen J, Liu Y, Wang K, Nie J, Wang X, Hao C, Yin Y, Wang S\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eChinese Society of Clinical Oncology (CSCO) Breast Cancer Guidelines 2022\u003c/strong\u003e. \u003cem\u003eTransl Breast Cancer Res \u003c/em\u003e2022, \u003cstrong\u003e3\u003c/strong\u003e:13.\u003c/li\u003e\n\u003cli\u003eHuang T-T, Lei L, Chen C-HA, Lu T-P, Jen C-W, Cheng SH-C: \u003cstrong\u003eA new clinical-genomic model to predict 10-year recurrence risk in primary operable breast cancer patients\u003c/strong\u003e. \u003cem\u003eSci Rep \u003c/em\u003e2020, \u003cstrong\u003e10\u003c/strong\u003e(1):4861.\u003c/li\u003e\n\u003cli\u003eVan\u0026apos;t Veer LJ, Dai H, Van De Vijver MJ, He YD, Hart AA, Mao M, Peterse HL, Van Der Kooy K, Marton MJ, Witteveen AT: \u003cstrong\u003eGene expression profiling predicts clinical outcome of breast cancer\u003c/strong\u003e. \u003cem\u003eNature \u003c/em\u003e2002, \u003cstrong\u003e415\u003c/strong\u003e(6871):530-536.\u003c/li\u003e\n\u003cli\u003eBrand\u0026atilde;o M, Pond\u0026eacute; N, Piccart-Gebhart M: \u003cstrong\u003eMammaprint\u0026trade;: a comprehensive review\u003c/strong\u003e. \u003cem\u003eFuture oncology \u003c/em\u003e2018, \u003cstrong\u003e15\u003c/strong\u003e(2):207-224.\u003c/li\u003e\n\u003cli\u003eShieh Y, Eklund M, Sawaya GF, Black WC, Kramer BS, Esserman LJ: \u003cstrong\u003ePopulation-based screening for cancer: hope and hype\u003c/strong\u003e. \u003cem\u003eNature Reviews Clinical Oncology \u003c/em\u003e2016, \u003cstrong\u003e13\u003c/strong\u003e(9):550-565.\u003c/li\u003e\n\u003cli\u003eXing F, Chen T-H, Wang H, Ma X, Ding R, Lu Z: \u003cstrong\u003eDifferences of prediction performances by multigene assays in patients with breast cancer with distinct genetic backgrounds\u003c/strong\u003e. In\u003cem\u003e.\u003c/em\u003e: American Society of Clinical Oncology; 2023.\u003c/li\u003e\n\u003cli\u003eHuang TT, Lei L, Chen CA, Lu TP, Jen CW, Cheng SH: \u003cstrong\u003eA new clinical-genomic model to predict 10-year recurrence risk in primary operable breast cancer patients\u003c/strong\u003e. \u003cem\u003eSci Rep \u003c/em\u003e2020, \u003cstrong\u003e10\u003c/strong\u003e(1):4861.\u003c/li\u003e\n\u003cli\u003eChen K, Wu J, Fang Z, Shao X, Wang X: \u003cstrong\u003eThe Clinical Research and Latest Application of Genomic Assays in Early-Stage Breast Cancer\u003c/strong\u003e. \u003cem\u003eTechnology in Cancer Research \u0026amp; Treatment \u003c/em\u003e2022, \u003cstrong\u003e21\u003c/strong\u003e:15330338221117402.\u003c/li\u003e\n\u003cli\u003eSocoteanu MP, O\u0026apos;Shaughnessy J, Hoskins K, Brufsky A, Graham CL, Vukelja SJ, Misleh JG, Tedesco KL, Layeequr Rahman R, Lee J: \u003cstrong\u003eClinical implications for patients with discordant oncotype and MammaPrint results\u003c/strong\u003e. In\u003cem\u003e.\u003c/em\u003e: American Society of Clinical Oncology; 2022.\u003c/li\u003e\n\u003cli\u003eWang H, Ma L, Zhang Y, Wang O, Wei Z, Xie X, Zha X, Zeng J, Lv Q, Ren Y\u003cem\u003e et al\u003c/em\u003e: \u003cstrong\u003eRecurIndex assay as an aid for adjuvant chemotherapy decisions in HR-positive HER2-negative breast cancer patients\u003c/strong\u003e. \u003cem\u003eFront Oncol \u003c/em\u003e2022, \u003cstrong\u003e12\u003c/strong\u003e:896431.\u003c/li\u003e\n\u003cli\u003eSchaafsma E, Zhang B, Schaafsma M, Tong CY, Zhang L, Cheng C: \u003cstrong\u003eImpact of Oncotype DX testing on ER+ breast cancer treatment and survival in the first decade of use\u003c/strong\u003e. \u003cem\u003eBreast Cancer Res \u003c/em\u003e2021, \u003cstrong\u003e23\u003c/strong\u003e(1):74.\u003cu\u003e\u003c/u\u003e\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"28-gene RecurIndex panel, 70-gene MammaPrint, Chinese HR+/HER2 − breast cancer, Risk stratification, Clinicopathologic concordance","lastPublishedDoi":"10.21203/rs.3.rs-6832631/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6832631/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e: Multigene expression assays help guide adjuvant therapy decisions in early-stage hormone receptor-positive, HER2-negative (HR+/HER2-) breast cancer by predicting recurrence risk. While the 21-gene Oncotype DX and 70-gene MammaPrint tests, developed in Western populations, allow low-risk patients to safely avoid chemotherapy, their validity in Asian cohorts remains understudied. A 28-gene signature (RecurIndex), developed using Asian patient data, predicts distant and locoregional recurrence. This study compares the 28-gene RecurIndex with the 70-gene MammaPrint in Chinese women with early HR+/HER2- breast cancer.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePatients and methods\u003c/strong\u003e: We retrospectively analyzed 99 Chinese patients (median age 52 years) with stage pT1-3N0-1 HR+/HER2- breast cancer who underwent MammaPrint testing post-surgery (2019-2022). Formalin-fixed paraffin-embedded tumor samples were retested using the 28-gene RecurIndex. Clinical-pathological risk was defined as ≥2 high-risk factors (age ≤40 years, tumor ≥T2, N1 nodal status, lymphovascular invasion, grade 3, Ki-67 ≥20%). Genomic risk stratification (high vs low) by each assay and concordance with clinical risk were assessed using Cohen’s kappa. Median follow-up was 56 months.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e: By clinical criteria, 48.5% (48/99) of patients were high-risk. The 28-gene assay classified 26% as high genomic risk versus 34% by the 70-gene MammaPrint. Overall concordance between assays was 72% (71/99 cases; kappa = 0.50). The 28-gene panel showed stronger alignment with clinical risk (kappa = 0.51) than the 70-gene (kappa = 0.39). Among clinically low-risk patients, 98% (50/51) were classified as low-risk by the 28-gene test compared to 84% (43/51) by the 70-gene. For clinically high-risk patients, both assays identified similar proportions as genomic high-risk (52% vs 54%, respectively). No recurrences or deaths occurred during follow-up.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e: The 28-gene RecurIndex demonstrated comparable performance to the 70-gene MammaPrint in stratifying recurrence risk in Chinese HR+/HER2− breast cancer patients. It classified fewer patients as high-risk, identified a larger low-risk subgroup (74% vs 66%), and aligned more closely with traditional clinical-pathological factors. These findings support its potential as an ethnicity-tailored prognostic tool. Larger studies with extended follow-up are needed to confirm predictive accuracy and chemotherapy benefit in this population.\u003c/p\u003e","manuscriptTitle":"Comparative Evaluation of a 28-Gene versus 70-Gene Panel for Recurrence Risk Prediction in Early-Stage HR-Positive/HER2-Negative Breast Cancer in Chinese Women","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-01 09:05:10","doi":"10.21203/rs.3.rs-6832631/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a457f5aa-01ee-4c39-8b69-ace59e5797fe","owner":[],"postedDate":"July 1st, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-08-20T12:53:52+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-01 09:05:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6832631","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6832631","identity":"rs-6832631","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}