Abstract
Background/Aim: The 21-gene assay recurrence score (RS) can guide the use of chemotherapy for the management of patients with pN0- 1 [1-3 positive lymph nodes (LNs)] breast cancer. However, practice patterns based on this assay, as well as associated outcomes, have not been evaluated for patients with pN2 (4-9 positive LNs) disease.
Patients and Methods: The National Cancer Database (NCDB) was queried for patients with newly-diagnosed, non-metastatic, hormone receptor-positive, Her2-negative, pN2 breast cancer who underwent adjuvant endocrine therapy and had a known RS. Kaplan–Meier analysis was used to evaluate overall survival (OS); Cox proportional hazards modeling determined variables associated with OS.
Results: Of 1,658 patients, 1,109 (67%) received chemotherapy and 549 (33%) did not. Chemotherapy was administered to 54% of patients with a low-risk recurrence score (RS), 67% with intermediate-risk RS, and 75% with high-risk RS. Chemotherapy was associated with improved 5-year OS in low-risk RS (95.5% vs. 87.4%), intermediate-risk RS (91.9% vs. 83.5%), and high-risk RS (81.3% vs. 50.2%) (p≤0.001 for all). On Cox multivariable analysis, chemotherapy and the RS risk group significantly associated with OS (p<0.05 for both). Qualitatively, patients over 70 years of age appeared to benefit comparatively less from chemotherapy.
Conclusion: Despite the underutilization of chemotherapy for hormone receptor-positive, Her2-negative, pN2 patients, it was associated with improved OS for all 21-gene panel risk groups. These results support the existing standard of chemotherapy for this population, although omission could be considered in patients over 70 years of age.
Introduction
Molecular profiling of neoplastic disease in efforts to personalize cancer care is considered the next great frontier of oncology (1). At the current time, molecular stratification of tumors is most mainstream for breast cancer. There are multiple gene assays that have been developed for this purpose, such as the 21-gene recurrence score (RS) for hormone receptor-positive, Her2-negative disease (Oncotype DX®) (2) as well as the 70-gene RS for a broader population (Mammaprint®) (3). These assays involve RNA extraction from tumor tissue, followed by reverse-transcribing that RNA into DNA, followed by polymerase chain reaction-mediated amplification of that DNA (4). Depending on the expression of the given number of genes compared to standardized controls, a numeric RS is generated (and is generally divided based on pre-specified cutoffs, e.g., low-risk, intermediate-risk, or high-risk).
Initial trials of the 21-gene RS (2, 5) excluded lymph node (LN)-positive cases, and those of the 70-gene RS underrepresented this population (3). More recently, the randomized RxPONDER trial of the 21-gene RS in women with 1-3 pathologically positive LNs (pN1) and a low/intermediate RS (≤25) have been reported, showing that chemotherapy benefitted premenopausal but not postmenopausal women (6).
Despite these important findings in pN1 patients, the 21-gene RS has not been evaluated for pN2 (4-9 positive LNs) disease. Because pN2 cases have significantly more nodal involvement, the risk of distant metastasis is considerably higher, and it has largely been presumed that chemotherapy for all pN2 cases should be delivered. In this study we sought to examine current patterns of care and management of these patients, as well as to examine the expected outcomes with vs. without chemotherapy for pN2 cases by RS.
Patients and Methods
The National Cancer Database (NCDB) is a joint project of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society, which consists of de-identified information regarding tumor characteristics, patient demographics, and patient survival for approximately 70% of the US population (7). All pertinent cases are reported regularly from CoC-accredited centers and compiled into a unified dataset, which is then validated. The data used in the study were derived from a de-identified NCDB file (2009-2017). The American College of Surgeons and the CoC have not verified and are neither responsible for the analytic or statistical methodology employed nor the conclusions drawn from these data by the investigators. As all patient information in the NCDB database is de-identified, this study was exempt from institutional review board evaluation.
Inclusion criteria for this study were patients with newly-diagnosed non-metastatic estrogen receptor-positive, Her2-negative breast cancer with a record of chemotherapy use, a known 21-gene RS, and 4-9 pathologically positive LNs. This study used the following criteria to classify risk groups of the 21-gene RS: 0-10 low RS, 11 to 25 intermediate RS, and 26+ high RS.
In accordance with the variables in the NCDB files, information collected on each patient broadly included demographic, clinical, and treatment data. All statistical tests were two-sided, with a threshold of p<0.05 for statistical significance, and were performed using STATA (version 14, College Station, TX, USA). Multivariable logistic regression modeling was utilized to determine characteristics that were predictive for receipt of chemotherapy. The variables used include age, race (White, African American, Other), T-stage. Charlson-Deyo score (for medical comorbidities), insurance status (Medicare, Medicaid, Private, Not Insurance), median income, facility type, extent of lymph node dissection, oncotype score, type of surgery, and use of radiation. The Kaplan–Meier method was used for survival analysis, and comparisons between groups were performed with the log-rank test. Overall survival (OS) was defined as the interval between the date of diagnosis and the date of death or censored at last contact. Univariate analysis was performed to determine factors associated with OS, and subsequently Cox multivariate analysis was performed including variables that were either significant or showed a strong trend to statistical significance on univariate analysis.
Results
A complete flow diagram of patient selection is provided in Figure 1. In total, 1,658 patients met study criteria (Table I). Of these, 1,109 (67%) received chemotherapy and 549 (33%) did not. Of note, most patients had an intermediate RS, underwent mastectomy, and had pT1-2 disease. Additionally, the rate of use of chemotherapy was 54%, 669%, and 75% for patients in the low, intermediate, and high RS groups (p<0001).
Patient selection diagram.
Clinical and demographic characteristics of the study population.
Multivariable logistic regression analysis revealed that independent associated factors of withholding chemotherapy were older age, greater comorbidities, low RS, and no radiotherapy (p<0.05 for all). Of note, the strongest predictor for chemotherapy use was a high RS [odds ratio (OR)=6.92, 95% confidence interval (CI)=4.49-10.68], whereas the OR for intermediate RS was 1.62 (95%CI=1.25-2.10), implying that clinicians may be managing intermediate-risk cases like low-risk (rather than high-risk).
The median follow-up was 49.9 months (interquartile range=33.6-71.2 months). Chemotherapy was associated with improved 5-year OS in all patients (90.3% vs. 82.5%, p<0.001), as well in those with low RS (95.5% vs. 87.4%, p<0.001), intermediate RS (91.9% vs. 83.5%, p=0.001), and high RS (81.3% vs. 50.2%, p=0.001). The analysis was repeated when including only patients with a complete axillary dissection (at least 10 LNs removed), and the results were similar (p<0.001), as well as for patients who did not undergo a complete axillary LN dissection (ALND) (p= 0.001) A sensitivity analysis was also performed when removing patients who did not receive radiotherapy (which could represent biases, since radiotherapy is recommended for all patients in this population), and in this subset of patients as well, an improved OS was observed with the use of chemotherapy (5 year OS 90.9% vs. 84.4%, p<0.001). On multivariable analysis, receipt of chemotherapy significantly associated with OS, along with age, increasing comorbidities, RS risk group, and the extent of dissection (p<0.05 for all). A forest plot was constructed to examine several covariates, in a univariate fashion, for which chemotherapy might be differentially beneficial (Figure 2). The benefit of chemotherapy appeared to be independent of the particular RS and whether or not a complete LN dissection was performed. The only potential difference was that patients over 70 years of age as well as patients with a low LN ratio (percentage of dissected LNs that were pathologically positive) may have benefitted comparatively less from chemotherapy.
Forest plot showing the effect of chemotherapy based on several covariates.
Discussion
Our study of a contemporary national database most notably demonstrates that chemotherapy was associated with improved outcomes for all risk groups based on the 21-gene RS in patients with hormone receptor-positive, Her2-negative, pN2 breast cancer. These data support the existing standard of care to deliver chemotherapy in all pN2 cases.
A significant finding herein was that about half of pN2 patients with a low RS did not receive chemotherapy. There were statistically significant differences in utilization of chemotherapy after stratifying patients by RS. The rate of chemotherapy use in patients with a low RS was 54%, compared to 75% of patients in the high RS cohort who received chemotherapy. This suggests that despite not being the standard of care, physicians may be using the RS in patients with N2 disease to select for patients in whom to omit chemotherapy. Since the relative benefit of chemotherapy was greatest in patients with a high RS, RS could potentially help to serve as a risk stratifying tool in frail patients when the physician is uncertain whether or not the patient may be able to tolerate chemotherapy.
An often-overlooked factor that could have impacted this study’s results is the choice of axillary management. Most patients herein had ALND, which has historically been the standard of care in this population. However, sentinel LN biopsy (SLNB) has been increasingly used in the modern era. A minority of patients in this dataset had less than 10 LNs removed, which could refer to SLNB (the NCDB does not record the type of axillary LN procedure). pN2 patients with a limited number of LNs removed could harbor occult pN3 disease, as the risk of occult LN involvement is known to increase with the number of pathologically positive LNs (8). The finding of our Cox multivariable analysis that having <5 LNs removed (implying SLNB) was independently associated with considerably poorer OS hints at this possibility, since having pN2 disease with only 4-5 removed LNs implies a high risk of additional involved unremoved LNs. Of note, patients with four pathologically involved nodes were included in the AMAROS trial, which suggested that no further axillary surgery was required in this patient population if they received axillary radiotherapy (9). Also, a repeat analysis of patients with at least 10 LNs removed showed similar results, implying that “true” pN2 patients drove the outcome differences in this study.
Based on the findings of the forest plot, we do believe that it is worth exploring in a randomized manner whether older women with pN2 disease and a low RS benefit from chemotherapy. A subset analysis of the PACS-01 trial, whose randomization was based on chemotherapy agents, illustrated that outcomes of patients with a low RS and pN2 disease were similar to those for patients with an intermediate/high RS and pN1 disease (10). Because chemotherapy is of no appreciable benefit to post-menopausal patients with low/intermediate RS and pN1 disease (6), it is theoretically possible that the same could be true for postmenopausal low RS pN2 cases.
Although randomized trials to corroborate this study are required, future research on gene panels for breast cancer is also necessary. For instance, the use of neoadjuvant chemotherapy for locally advanced breast cancer is rising, and a key question in this setting is whether gene panels can predict response to neoadjuvant chemotherapy (11). One NCDB investigation has suggested this could be the case, having illustrated that patients with a high RS are more likely to have a pathologic complete response (12), and it likely would not be difficult to accrue prospective trials evaluating this question.
Although the NCDB provides a unique platform to study the predictive impact of the RS in the pN2 patient population, this investigation is not without additional limitations to those discussed elsewhere (13). First, the NCDB does not keep track of several other variables, including chemotherapy cycles/agents, endocrine duration/agents, radiotherapy fields/volumes, or nature of preoperative workup. Second, there is also no information regarding technical details of surgical procedures (i.e., axillary assessment). Third, the NCDB does not allow for an assessment of subsequent lines of treatment (e.g., re-irradiation, further systemic and/or targeted therapy), which could influence overall survival. Lastly, propensity matching was attempted but could not be completed for this study owing to sample size related issues. Nevertheless, the known shortcomings of a national, large-volume database do not undermine the necessity for prospective investigation.
Conclusion
Our study of a contemporary national database demonstrates that chemotherapy was associated with improved outcomes for all risk groups based on the 21-gene RS in hormone receptor-positive, Her2-negative, pN2 patients. These data support the existing standard of care to deliver chemotherapy in all pN2 cases, though omission of chemotherapy should be further explored in older patients, age ≥70 years.
Footnotes
Authors’ Contributions
WH: Conception of study, data utilization, statistics work and drafting the manuscript. VV: Drafting the manuscript, data utilization. NM: Data utilization, statistics work. MS: Drafting the manuscript. EBB: Data utilization, drafting the manuscript. KS: Drafting the manuscript, data utilization. DJ: Drafting the manuscript. CA: Drafting the manuscript. AR: Drafting the manuscript. BST: Drafting the manuscript, data utilization.
Conflicts of Interest
All Authors declare no conflicts of interest in relation to this study.
Funding
There was no research support for this study.
- Received December 29, 2024.
- Revision received February 6, 2025.
- Accepted February 10, 2025.
- Copyright © 2025 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.