Olaya W, Morgan JW, Lum SS. Unnecessary Axillary Surgery for Patients With Node-Negative Breast Cancer Undergoing Total Mastectomy. Arch Surg. 2011;146(9):1029-1033. doi:10.1001/archsurg.2011.196
Author Affiliations: Division of Surgical Oncology, Department of Surgery, School of Medicine (Drs Olaya and Lum), School of Public Health (Dr Morgan), and California Cancer Registry, Desert Sierra Cancer Surveillance Program, Medical Center (Drs Morgan and Lum), Loma Linda University, Loma Linda, California.
Objective To identify factors associated with the use of axillary lymph node dissection (ALND) as the initial axillary staging in node-negative breast cancer patients undergoing total mastectomy.
Design California Cancer Registry study.
Setting Academic research.
Patients Women treated with total mastectomy for Tis, T1, or T2 node-negative breast carcinoma treated between January 1, 2004, and December 31, 2008.
Main Outcome Measures Proportions of patients who underwent ALND without prior sentinel lymph node dissection were compared by demographic characteristics.
Results Of 18 238 women treated with total mastectomy for Tis, T1, or T2 node-negative breast carcinoma, 35.1% underwent initial axillary staging by ALND without prior sentinel lymph node dissection. On multivariable analyses, patients were significantly more likely to undergo ALND if they had T2 disease or were 65 years or older, were hormone receptor negative, of Hispanic or Asian/Pacific Islander race/ethnicity, of lower socioeconomic quintile, operated on during earlier years of the study period, and not treated by a hospital cancer program approved by the American College of Surgeons.
Conclusions More than one-third of patients in California who underwent total mastectomy for treatment of early-stage node-negative breast carcinoma received ALND without prior sentinel lymph node dissection; furthermore, certain subsets of patients have higher odds of undergoing ALND alone. To avoid the unnecessary morbidity of ALND in early-stage breast carcinoma, further research is required to elucidate how tumor, patient, and system factors can be modified to improve delivery of optimal breast cancer care.
Sentinel lymph node dissection (SLND) has become the standard of care for staging the axilla in clinically node-negative breast carcinoma. When performed by experienced surgeons, SLND is highly success in predicting metastatic disease1- 3 in lymph nodes, with little morbidity compared with axillary lymph node dissection (ALND).3,4 The American Society of Clinical Oncology5 guidelines state that SLND is an accurate method for identifying early-stage breast cancer and cite SLND as an alternative to routine staging ALND for patients with clinically negative axillary nodes. The National Comprehensive Cancer Network6 also recommends SLND staging for patients with clinically negative nodes and further suggests referral to an experienced SLND team if one is unavailable at the initial treating facility.
Although SLND has been accepted clinical practice for years, less than 70% of patients with early-stage breast cancer undergo SLND.7,8 In addition, patients undergoing total mastectomy for treatment of their primary breast carcinoma more commonly receive ALND rather than SLND for staging of the axilla.7,8 We sought to identify factors associated with the use of ALND as initial axillary staging in patients with node-negative breast cancer undergoing total mastectomy.
Included in the study were all patients treated in California hospitals between January 1, 2004, and December 31, 2008, with records in the California Cancer Registry (part of the Surveillance, Epidemiology, and End Results Program) who underwent total mastectomy with initial diagnoses of Tis, T1, or T2 node-negative breast carcinoma. Patients were excluded if they had incomplete surgical or demographic data, did not have pathologic stage at diagnosis data, were stage T3 or higher, had node-positive disease, had a second or subsequent cancer diagnosis, had a tumor not originating in the breast and/or of nonepithelial origin, or if none of the first course therapy was performed at the reporting facility.
The main outcome measure was the proportion of patients undergoing ALND without prior SLND. Variables analyzed included age, race/ethnicity, year of diagnosis, hormone receptor status, human epidermal growth factor receptor 2 (HER-2) positivity, socioeconomic status (SES) quintile, type of initial axillary staging procedure, American Joint Commission on Cancer T stage at diagnosis, and hospital cancer program American College of Surgeons (ACOS) approval. Age groups were defined as younger than 40 years, 40 to 64 years, or 65 years or older. Racial/ethnic groups were categorized as Hispanic, non-Hispanic (NH) Asian/Pacific Islander, NH black, NH white, or other. Spanish surname was used to define Hispanic if race/ethnicity was not noted in the medical record. As previously reported by Yost et al,9 SES is an ecologic index based on the census block group of residence at diagnosis computed using data from the 2000 US census, with SES 1 referring to the lowest and SES 5 to the highest quintile.
Using Pearson χ2 and logistic regression analyses, proportions of patients who underwent initial axillary staging by ALND alone vs SLND with or without subsequent ALND were compared by age, T stage, hormone receptor status, HER-2 status, race/ethnic group, SES quintile, year of diagnosis, and hospital cancer program ACOS approval. Statistical analyses were performed using commercially available software (NCSS 2007 for Windows; NCSS, Kaysville, Utah). This study was approved by the Loma Linda University Institutional Review Board.
Between January 1, 2004, and December 31, 2008, a total of 18 238 patients in the California Cancer Registry underwent total mastectomy for Tis, T1, or T2 node-negative breast cancer and had demographic and treatment data available for review. Initial axillary staging by ALND without prior SLND was performed in 35.1% of patients. The mean age of patients was 59.5 years (age range, 18-100 years). Table 1 gives additional demographic data for the study population.
Table 2 compares patients who underwent axillary staging by ALND alone vs SLND with or without subsequent ALND. Univariate analysis confirmed that patients were significantly more likely to undergo ALND alone if they had T2 disease, had tumors that were hormone receptor negative or HER-2 positive, were operated on during earlier years of the study period, were not treated by a hospital cancer program approved by the ACOS, or were older, Hispanic, or of lower SES quintile.
On multivariable analysis, patients had statistically lower odds of undergoing ALND alone if they had lower T stage, had hormone receptor–positive disease, were treated in more recent years, or were younger than 65 years. Hispanic and NH Asian/Pacific Islander women were more likely than NH white patients to undergo axillary staging by ALND alone without prior SLND after total mastectomy. Patients in the lowest SES quintile were almost 3 times more likely to undergo ALND alone than those in the highest SES quintile. Patients treated at a cancer center not approved by the ACOS were more likely to undergo ALND alone than patients treated at an ACOS-approved center. On multivariable analysis, HER-2 status was not significantly associated with receipt of ALND alone after total mastectomy. Table 2 lists odds ratios (95% confidence intervals) for factors associated with receipt of axillary staging by ALND alone vs SLND with or without subsequent ALND.
Sentinel lymph node dissection has been validated as a sensitive and accurate method to detect breast carcinoma involvement of the axilla.1 Compared with SLND, ALND is associated with significantly increased morbidity, including increased arm size, numbness, muscle weakness, restricted work and social activity, and impaired quality of life.10,11 Although pathologic axillary examination provides important prognostic information and can be performed by SLND or ALND, SLND has been shown to be an oncologically safe procedure in clinically node-negative patients, with no difference in disease-free and overall survival using ALND vs SLND for initial axillary staging.12,13
Most women with early-stage breast cancer undergo breast-conserving therapy for treatment of their primary breast tumor; however, almost 30% of patients with early-stage breast cancer undergo total mastectomy.14 Previous studies8,15 of patients with early-stage breast cancer undergoing total mastectomy have shown that this population is more likely to receive ALND than SLND. We sought to further investigate factors associated with initial axillary staging by ALND vs SLND among patients with early-stage node-negative breast cancer who underwent total mastectomy.
Older patients were significantly more likely to receive ALND without prior SLND. In contrast, a 2010 study by Schonberg et al16 showed that treatments for older individuals (>80 years) are usually less aggressive than those for younger patients. Our study did not evaluate older women who received no type of axillary staging procedure, nor could it assess potential reasons for disparate care in older women, such as impaired health literacy or patient-physician communication.
Patients with hormone receptor–negative tumors were more likely to undergo initial axillary staging by ALND alone. Evidence has shown a decrease in survival among patients with hormone receptor–negative cancers that involve lymph nodes.17 In addition, an increase in axillary recurrence rates associated with estrogen receptor negativity has been demonstrated after negative SLND.12,18 Yi et al19 demonstrated that patients were more likely to undergo SLND alone without completion ALND if they had hormone-positive tumors. Physicians treating patients with hormone receptor–negative tumors may have a bias toward a more aggressive surgical approach for these tumors.
Patients with T2 tumors were more likely to receive ALND alone than those with T1 tumors. Because the size of T2 tumors ranges from greater than 2 cm to 5 cm or less, patients with larger T2 tumors may not have been offered initial staging by SLND. Based on final pathologic findings, high-risk patients undergoing total mastectomy for Tis (ductal carcinoma in situ [DCIS]) may have rates of upstaging to invasive cancer as high as 33%.20 Because SLND performed after mastectomy has been shown to be less effective,21 it is recommended for patients undergoing total mastectomy for DCIS.22 In the present series, almost 30% of patients undergoing total mastectomy for DCIS underwent primary ALND. Although DCIS is not an invasive cancer, these patients were still subjected to the morbidity of ALND.
Lower SES was associated with higher rates of axillary staging by ALND alone. Studies7- 9 have shown that patients of lower SES are more likely to have disparities in their breast cancer care. Uninsured patients have been reported to have a higher incidence of no axillary staging whatsoever in early-stage breast cancer.23 Differences in receipt of axillary staging could be attributed to education level, access to care, or other perceived barriers to follow-up adjuvant therapy. Although indigent patients are more likely to be initially seen with advanced stages of disease,24 our results showed that SES predicted receipt of ALND independent of T stage on multivariable analysis.
Race/ethnicity was also found to be an independent predictor of ALND alone. Hispanic and Asian/Pacific Islander patients were more likely than NH white patients to undergo ALND alone. The large Hispanic and Asian/Pacific Islander populations in California may provide the basis for future studies to elucidate whether language barriers or cultural differences have a role in determining surgical and systemic therapy for breast cancer. Although previous investigations have shown that NH black women have a greater likelihood of receiving no axillary staging,23 the receipt of axillary staging by ALND alone was equivalent among NH black patients and NH white patients.
Patients were half as likely to undergo ALND alone if treated in 2008 compared with 2004. A 2009 study25 showed that the use of ALND as a primary axillary staging modality decreased from 94% in 1998 to 36% in 2005. However, more than 25% of patients with early-stage node-negative breast cancer in the most recent period of the present study were still receiving primary ALND.
In this study, patients were more commonly treated at cancer centers not approved by the ACOS and were significantly more likely to undergo axillary staging with ALND alone. Previous studies have shown differences in treatment for breast cancer based on where a patient receives therapy7- 9,14 and from what type of surgeon.26 Despite the stringent ACOS accreditation process, 30.1% of patients in ACOS-approved centers underwent ALND alone as primary axillary staging.
A limitation of our study that is inherent in the use of registry data is less than robust reporting of clinical data. Although patients may have had no axillary disease on final pathologic examination, clinicians might have palpated axillary lymphadenopathy on physical examination and recommended ALND rather than SLND. Because clinical axillary examination has been demonstrated to be inaccurate more than 40% of the time for moderate to highly suspicious nodes27 and because physical examination has been shown to overestimate and underestimate lymph node size in cancer,28 we evaluated patients by pathologic nodal stage. Although factors like physician training or hospital nuclear medicine capabilities may have a role in determining uptake of SLND, such data are not specifically recorded in the registry database; we used hospital cancer program ACOS approval status as a surrogate for overall system expertise.
Sentinel lymph node dissection in early-stage breast cancer has proven to be an accurate and safe method for staging the clinically negative axilla.13 Patients continue to undergo total mastectomy for early-stage breast cancer, and disparities in recommended axillary staging by SLND among these patients exist in California. Axillary lymph node dissection is a morbid procedure that many patients undergo unnecessarily. To avoid the morbidity of ALND in early-stage breast cancer, further research is required to elucidate how tumor, patient, and system factors can be modified to improve delivery of optimal breast cancer care.
Correspondence: Sharon S. Lum, MD, Division of Surgical Oncology, Department of Surgery, Loma Linda University, 11175 Campus St, Coleman Pavilion Room 21111, Loma Linda, CA 92350 (firstname.lastname@example.org).
Accepted for Publication: April 19, 2011.
Author Contributions:Study concept and design: Lum. Acquisition of data: Morgan and Lum. Analysis and interpretation of data: Olaya, Morgan, and Lum. Drafting of the manuscript: Olaya. Critical revision of the manuscript for important intellectual content: Olaya, Morgan, and Lum. Statistical analysis: Morgan and Lum. Administrative, technical, and material support: Olaya. Study supervision: Lum.
Financial Disclosure: None reported.
Funding/Support: This study was supported in part by contracts N01-PC-35136, N01-PC-35139, and N02-PC-15105 from the National Cancer Institute and the Surveillance, Epidemiology, and End Results Program; by contract U58DP000807-01 from the Centers for Disease Control and Prevention and the National Program for Cancer Registries; and by the California Department of Public Health Cancer Surveillance Branch.
Previous Presentation: This paper was presented at the 82nd Annual Meeting of the Pacific Coast Surgical Association; February 20, 2011; Scottsdale, Arizona, and is published after peer review and revision.