Background
The Oncotype DX Breast Recurrence Score (RS) test is a 21-gene expression assay that became commercially available for clinical use in 2004. In patients with early-stage, hormone receptor−positive (HR+), HER2-negative breast cancer, RS has been validated to be prognostic for distant recurrence and breast cancer−specific survival,1,2 and predictive of chemotherapy benefit in retrospective studies of archival tissue.1,3,4 Multiple studies have shown that the RS assay influences physician treatment recommendations for women with lymph node−positive (LN+) disease,3,5–14 and that those who receive treatment based on RS results have low rates of distant recurrence15 and excellent 5-year breast cancer–specific survial.16–20
In 2015, the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Breast Cancer included use of the 21-gene assay for select patients with LN+ breast cancer (1–3 positive nodes),21 although its use in this population preceded its inclusion. Patients with LN+ disease who underwent RS testing tended to be older and have lower grade and smaller sized tumors,4 and black women were significantly less likely to receive the 21-gene assay than nonblack women.3,4
However, previous studies of the 21-gene assay have not been population-based,3,4 have not been adequately powered to more deeply characterize racial and socioeconomic disparities over time,4 and have not included a reliable measure for whether the test was ordered (vs RS received and reported).3,4 Furthermore, examining uptake over time may elucidate how it varies according to key covariates (eg, race) during diffusion of the assay among women with LN+ disease, which may have implications for understanding adoption of tumor gene profiling tests more broadly. Thus, this study aimed to assess uptake of the assay over time and to characterize which sociodemographic and clinical factors are associated with its uptake for the first time in a population-based study of women with LN+, HR+, HER2-negative breast cancer.
Methods
We performed a secondary data analysis to examine uptake of the 21-gene assay among women with LN+ disease through descriptive analyses and multivariable modeling.
Dataset
21-gene assay results from Genomic Health’s Clinical Laboratory (2004–2014) were linked to the NCI’s SEER breast cancer cases (diagnosed 2004–2013). SEER registries have been required to collect results from multigene assays, including Oncotype DX for breast cancer, since 2010. However, due to a fragmented US healthcare system, this information was not always available to the cancer registrars, resulting in underreporting (40% of Oncotype DX tests provided through the linkage were not captured by the cancer registrars). The linkage was performed to obtain more complete, population-based data. The success rate of the linkage was estimated to be >93% and was based on comparison with data collected from medical records by cancer registrars. After SEER approval, the study investigators used deidentified data.
Sample
Women with LN+, HR+, HER2-negative invasive breast cancer who had no distant metastases were included in the study, and only those diagnosed in 2010 through 2013, because reporting of HER2 receptor status into the SEER registries began in 2010.
Measures
The dependent variable was uptake of the 21-gene assay, which was defined as whether a patient had the assay ordered by a clinician. Sociodemographic and tumor covariates that may be associated with assay uptake were included in descriptive statistics and the primary multivariable analysis. Covariates of interest included age at diagnosis (<45, 45–54, 55–64, 65–74, ≥75 years), year of breast cancer diagnosis (2010–2013), number of positive LNs (1, 2, 3, ≥4, unknown), tumor grade (I, II, III, unknown), tumor size (≤1, >1 and ≤2, >2 and ≤5, >5 cm, unknown), race (black, white, other, unknown), area-level socioeconomic status (SES; Yost quintile 1 [low] to 5 [high], unknown), marital status (married, divorced/separated, single/unmarried, widowed, unknown), insurance status (insured, insured/nonspecific, Medicaid, uninsured, unknown), and state of cancer registry reporting. Our analyses also included the assay orders, the 21-gene assay RS (0–100), and risk group (low: RS <18; intermediate: RS 18–30; high: RS ≥31) from Genomic Health.
Statistical Analysis
The proportion of women who had Oncotype DX testing in 2010 through 2013 was plotted. We stratified these plots by key demographic variables, including race, SES, and age, given the need to more deeply characterize test uptake in these groups over time. Population characteristics were analyzed using chi-square tests. Among those who received the 21-gene assay, kernel density estimates of the RS were plotted by the number of positive LNs. Notably, for a number of women (n=336; 7.9%), we received indication that a test was ordered but assay results were not provided to SEER due to a canceled test, failed test (eg, inadequate tissue sample), multiple tests for a patient, or sample outside analysis criteria (HR-negative and/or HER2-positive by reverse-transcription PCR); these individuals were excluded from the kernel density plots and mean RS calculations. Finally, our primary analysis identified factors associated with test uptake using a multivariable logistic regression model. We used modified Poisson regression to estimate unadjusted and adjusted risk ratios. Data were analyzed using Stata (StataCorp LLC).
Results
Sociodemographic Characteristics
Overall uptake of the 21-gene assay increased over time (Figure 1; Tables 1 and 2). Overall, 14% of women with LN+ breast cancer had an Oncotype DX test ordered. Odds of uptake were higher among patients who were white versus black (unadjusted odds ratio [uOR], 1.35; P<.001), those living in higher (Yost quintile 3, 4, and 5) versus low (Yost quintile 1) SES areas (P<.001), and older women (all age categories ≥45 years; P<.001). Conversely, odds of uptake were lower among patients with Medicaid (uOR, 0.59; P<.001) or who were uninsured (uOR, 0.73; P<.01), and those who were widowed versus married (uOR, 0.73; P<.001).
Oncotype DX testing uptake by (A) race, (B) area-level socioeconomic status quintile, and (C) age among patients with lymph node–positive breast cancer.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 17, 6; 10.6004/jnccn.2018.7266
Oncotype DX testing uptake by (A) race, (B) area-level socioeconomic status quintile, and (C) age among patients with lymph node–positive breast cancer.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 17, 6; 10.6004/jnccn.2018.7266
Oncotype DX testing uptake by (A) race, (B) area-level socioeconomic status quintile, and (C) age among patients with lymph node–positive breast cancer.
Citation: Journal of the National Comprehensive Cancer Network J Natl Compr Canc Netw 17, 6; 10.6004/jnccn.2018.7266
Demographic and Tumor Characteristics
Unadjusted Odds of Oncotype DX Testing Uptake
When adjusting for covariates in multivariable analysis (Table 3), race was no longer associated with uptake of the 21-gene assay (white vs black: adjusted odds ratio [aOR], 1.12; P=.07). However, living in a high SES area (quintile 5 vs 1: aOR, 1.60; P<.001) was associated with higher odds of receiving the assay after controlling for covariates. Conversely, having Medicaid (vs private insurance: aOR, 0.86; P=.02) and being widowed (vs married: aOR, 0.79; P<.001) were associated with lower odds of uptake when adjusted for covariates, including age. Rates of uptake greatly varied by state, even when controlling for covariates.
Multivariable Logistic Regression Model of Oncotype DX Testing Uptake
Tumor Characteristics
On average, tumor characteristics of women for whom the 21-gene assay was ordered differed from those of women who did not have an assay ordered; for example, those who did not receive testing, on average, had higher grade and larger sized tumors and greater LN involvement than those who did receive the test (P<.001) (Table 1). The 21-gene assay was used in 7.3% of patients with ≥4 positive nodes.
These associations between clinical covariates and uptake held when controlling for covariates in multivariable analysis (Table 3). Greater LN involvement (≥4 vs 1 positive LNs: aOR, 0.14; P<.001), higher tumor grade (III vs I: aOR, 0.42; P<.001), and larger tumor size (>5 vs ≤1 cm: aOR, 0.36; P<.001) were associated with lower odds of uptake.
Among women who received the 21-gene assay, distribution of RS results did not vary greatly by number of positive LNs (supplemental eAppendix 1 and eFigure 1, available with this article at JNCCN.org).
Discussion
Uptake of the 21-gene assay increased over time and preceded its inclusion in NCCN Guidelines for LN+ breast cancer. Differences in uptake by race, SES, and age have persisted during diffusion of the assay during the 4-year study period, and in multivariable models, SES, demographic, and clinical characteristics were associated with assay uptake.
Insurance status was found to be associated with higher odds of 21-gene assay uptake, consistent with previous literature.3 Furthermore, we found that high area-level SES was associated with an increased odds of uptake. Interestingly, when clinical and SES variables were controlled for, racial differences in the uptake of the test were no longer observed, suggesting that these variables may partially explain disparities in the uptake of this genomic innovation. Similar findings were reported in a North Carolina cohort.4
In addition to SES and clinical factors, demographic factors such as age and being widowed or having unknown marital status influenced uptake of the test. Similar to previous findings among women with LN-negative disease, those who were widowed were less likely to receive the assay, as were the oldest and youngest women with LN+ breast cancer.22 A previous qualitative study found that low test uptake in the youngest and oldest age groups may result from clinician and patient preferences regarding chemotherapy, including providing compulsory chemotherapy in the youngest group and avoiding chemotherapy due to fragility in the oldest age group.23 Finally, test use varied greatly by geographic region. In a qualitative study, local practice patterns and peer use of the test were found to influence clinicians’ adoption of the assay.23 Similarly, a recent Kaiser Permanente study found that differences in oncologist factors (eg, sex) and organizational factors (eg, use patterns among practice leaders) explained more variation in testing rates than patient-level characteristics,24 and another study (conducted in a single state) found that the likelihood of testing decreased for women when a clinician’s patient volume decreased.25 Thus, clinician- and organizational-level factors may partially explain geographic variation in test uptake. Further investigations should explore this hypothesis explicitly.
We found that mean risk scores did not vary by the number of positive LNs. Recent studies show that women with 1 to 3 positive LNs who received treatment based on RS results had good health outcomes.16,19 For example, 5-year breast cancer−specific mortality was 1% among women with RS <18 who had micrometastases and up to 3 positive nodes.19 Outcomes worsen among women with ≥4 positive nodes16; uptake of the 21-gene assay was low in this group, which aligns with current outcomes data and the lack of clinical guidelines for using the assay in this subgroup. Our findings, along with those showing lower use in higher grade and larger tumors, suggests that clinicians may choose to not order a test in more advanced cases for which chemotherapy has historically been indicated.
Limitations
Because we do not have information on clinician characteristics and patient−clinician communication, we could not determine whether the test was discussed and declined by the patient or whether clinician characteristics influenced test uptake. However, recent studies suggested that clinician- and organizational-level factors may drive differential uptake. For example, patients seen within comprehensive cancer centers appear to be more likely to receive the 21-gene assay than those seen in tertiary centers, explaining observed disparities in assay uptake among patients with LN-negative disease.26 We were unable to examine these factors, and future studies should focus on this area. Uptake results to date reflect behavior through 2013. During this study period, use of the 21-gene assay was not included in clinical guidelines. Understanding how clinical guideline inclusion has influenced assay uptake and observed disparities in uptake will be important to monitor across diffusion of the innovation. We plan to perform annual updates to gain additional information regarding uptake and patient outcomes. Finally, the ongoing RxPONDER trial will provide prospective data regarding whether patients with LN+, HR+, HER2-negative breast cancer and RS <25 benefit from adjuvant chemotherapy, and will provide further guidance on the use of RS for patients with LN+ disease.
Conclusions
Future research should continue to document practice patterns related to the 21-gene assay and longer term outcomes among women who undergo breast cancer treatment based on RS. Given variation in testing associated with area-level SES, insurance coverage, and geographic region, interventions to understand and reduce differential uptake are needed to ensure equitable access to this genomic test.
References
- 1.↑
Albain KS, Barlow WE, Shak S, et al.. Prognostic and predictive value of the 21-gene recurrence score assay in postmenopausal women with node-positive, oestrogen-receptor-positive breast cancer on chemotherapy: a retrospective analysis of a randomised trial. Lancet Oncol 2010;11:55–65.
- 2.↑
Dowsett M, Cuzick J, Ingle J, et al.. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010;28:509–518.
- 3.↑
Jasem J, Fisher CM, Amini A, et al.. The 21-gene recurrence score assay for node-positive, early-stage breast cancer and impact of RxPONDER trial on chemotherapy decision-making: have clinicians already decided? J Natl Compr Canc Netw 2017;15:494–503.
- 4.↑
Roberts MC, Weinberger M, Dusetzina SB, et al.. Racial variation in the uptake of Oncotype DX testing for early-stage breast cancer. J Clin Oncol 2016;34:130–138.
- 5.↑
de Boer RH, Baker C, Speakman D, et al.. The impact of a genomic assay (Oncotype DX) on adjuvant treatment recommendations in early breast cancer. Med J Aust 2013;199:205–208.
- 6.↑
Yamauchi H, Nakagawa C, Takei H, et al.. Prospective study of the effect of the 21-gene assay on adjuvant clinical decision-making in Japanese women with estrogen receptor-positive, node-negative, and node-positive breast cancer. Clin Breast Cancer 2014;14:191–197.
- 7.↑
Bargallo JE, Lara F, Shaw-Dulin R, et al.. A study of the impact of the 21-gene breast cancer assay on the use of adjuvant chemotherapy in women with breast cancer in a Mexican public hospital. J Surg Oncol 2015;111:203–207.
- 8.↑
Eiermann W, Rezai M, Kümmel S, et al.. The 21-gene recurrence score assay impacts adjuvant therapy recommendations for ER-positive, node-negative and node-positive early breast cancer resulting in a risk-adapted change in chemotherapy use. Ann Oncol 2013;24:618–624.
- 9.↑
Gligorov J, Pivot XB, Jacot W, et al.. Prospective clinical utility study of the use of the 21-gene assay in adjuvant clinical decision making in women with estrogen receptor-positive early invasive breast cancer: results from the SWITCH study. Oncologist 2015;20:873–879.
- 10.↑
Lee MH, Han W, Lee JE, et al.. The clinical impact of 21-gene recurrence score on treatment decisions for patients with hormone receptor-positive early breast cancer in Korea. Cancer Res Treat 2015;47:208–214.
- 11.↑
Oratz R, Kim B, Chao C, et al.. Physician survey of the effect of the 21-gene recurrence score assay results on treatment recommendations for patients with lymph node-positive, estrogen receptor-positive breast cancer. J Oncol Pract 2011;7:94–99.
- 12.↑
Torres S, Trudeau M, Gandhi S, et al.. Prospective evaluation of the impact of the 21-gene recurrence score assay on adjuvant treatment decisions for women with node-positive breast cancer in Ontario, Canada. Oncologist 2018;7:768–775.
- 13.↑
Sun J, Fenton D, Baxter S, et al.. A prospective clinical utility study of the impact of the 21-gene recurrence score (RS) assay in the treatment of estrogen receptor positive (ER+), HER2 negative (HER2-), 1-3 node positive (N+) breast cancer (BC). Presented at the 2017 San Antonio Breast Cancer Symposium; December 5−9, 2017; San Antonio, Texas.
- 14.↑
Kurian AW, Bondarenko I, Jagsi R, et al.. Recent trends in chemotherapy use and oncologists’ treatment recommendations for early-stage breast cancer. J Natl Cancer Inst 2018;110:493–500.
- 15.↑
Stemmer SM, Steiner M, Rizel S, et al.. Clinical outcomes in ER+ HER2 -node-positive breast cancer patients who were treated according to the Recurrence Score results: evidence from a large prospectively designed registry. NPJ Breast Cancer 2017;3:32.
- 16.↑
Roberts MC, Miller DP, Shak S, et al.. Breast cancer-specific survival in patients with lymph node-positive hormone receptor-positive invasive breast cancer and Oncotype DX Recurrence Score results in the SEER database. Breast Cancer Res Treat 2017;163:303–310.
- 17.↑
Nitz U, Gluz O, Christgen M, et al.. Reducing chemotherapy use in clinically high-risk, genomically low-risk pN0 and pN1 early breast cancer patients: five-year data from the prospective, randomised phase 3 West German Study Group (WSG) PlanB trial. Breast Cancer Res Treat 2017;165:573–583.
- 18.↑
Gluz O, Nitz U, Christgen N, et al.. Prospective WSG phase III PlanB trial: clinical outcome at 5-year follow up and impact of 21 Gene Recurrence Score result, central/local-pathological review of grade, ER, PR and Ki67 in HR+/HER2- high risk node-negative and –positive breast cancer. Presented at the 2016 European Breast Cancer Conference; March 9−11, 2016; Amsterdam, the Netherlands.
- 19.↑
Petkov VI, Miller DP, Howlader N, et al.. Breast-cancer-specific mortality in patients treated based on the 21-gene assay: a SEER population-based study. NPJ Breast Cancer 2016;2:16017.
- 20.↑
Shak SM, Petkov VI. Breast cancer-specific mortality (BCSM) in patients with node-positive (N+) breast cancer (BC) treated based on the 21-gene assay in clinical practice. Presented at the 2017 San Antonio Breast Cancer Symposium; December 5−9, 2017; San Antonio, Texas.
- 22.↑
Cronin KPV, Howlader N, Howe W, et al.. Dissemination of 21-gene assay testing among female breast cancer patients in the US. Presented at the 2016 American Society of Clinical Oncology Annual Meeting; June 3−7, 2016; Chicago, Illinois.
- 23.↑
Roberts MC, Bryson A, Weinberger M, et al.. Oncologists’ barriers and facilitators for Oncotype Dx use: qualitative study. Int J Technol Assess Health Care 2016;32:355–361.
- 24.↑
Lieu TA, Ray GT, Prausnitz SR, et al.. Oncologist and organizational factors associated with variation in breast cancer multigene testing. Breast Cancer Res Treat 2017;163:167–176.
- 25.↑
Reeder-Hayes KE, Wheeler SB, Baggett CD, et al.. Influence of provider factors and race on uptake of breast cancer gene expression profiling. Cancer 2018;124:1743–1751.
- 26.↑
Guth AA, Fineberg S, Fei K, et al.. Utilization of Oncotype DX in an inner city population: race or place? Int J Breast Cancer 2013;2013:653805.
