Fax: (212) 639-2417
Favorable prognosis in patients with T1a/T1bN0 triple-negative breast cancers treated with multimodality therapy
Version of Record online: 5 MAR 2012
Copyright © 2012 American Cancer Society
Volume 118, Issue 20, pages 4944–4952, 15 October 2012
How to Cite
Ho, A. Y., Gupta, G., King, T. A., Perez, C. A., Patil, S. M., Rogers, K. H., Wen, Y. H., Brogi, E., Morrow, M., Hudis, C. A., Traina, T., McCormick, B., Powell, S. N. and Robson, M. E. (2012), Favorable prognosis in patients with T1a/T1bN0 triple-negative breast cancers treated with multimodality therapy. Cancer, 118: 4944–4952. doi: 10.1002/cncr.27480
- Issue online: 5 OCT 2012
- Version of Record online: 5 MAR 2012
- Manuscript Accepted: 11 JAN 2012
- Manuscript Revised: 9 DEC 2011
- Manuscript Received: 10 NOV 2011
- triple negative;
- breast cancer;
- breast-conserving therapy
The authors evaluated the clinical characteristics, natural history, and outcomes of patients who had ≤1 cm, lymph node-negative, triple-negative breast cancer (TNBC).
After excluding patients who had received neoadjuvant therapy, 1022 patients with TNBC who underwent definitive breast surgery during 1999 to 2006 were identified from an institutional database. In total, 194 who had lymph node-negative tumors that measured ≤1 cm comprised the study population. Clinical data were abstracted, and survival outcomes were analyzed.
The median follow-up was 73 months (range, 5-143 months). The median age at diagnosis was 55.5 years (range, 27-84 years). Tumor (T) classification was microscopic (T1mic) in 16 patients (8.2%), T1a in 49 patients (25.3%), and T1b in 129 patients (66.5%). Most tumors were poorly differentiated (n = 142; 73%), lacked lymphovascular invasion (n = 170; 87.6%), and were detected by screening (n = 134; 69%). In total, 129 patients (66.5%) underwent breast-conserving surgery, and 65 patients (33.5%) underwent mastectomy. One hundred thirteen patients (58%) received adjuvant chemotherapy, and 123 patients (63%) received whole-breast radiation. The patients who received chemotherapy had more adverse clinical and disease features (younger age, T1b tumor, poor tumor grade; all P < .05). Results from testing for the breast cancer (BRCA) susceptibility gene were available for 49 women: 19 women had BRCA1 mutations, 7 women had BRCA2 mutations, and 23 women had no mutations. For the entire group, the 5-year local recurrence-free survival rate was 95%, and the 5-year distant metastasis-free survival rate was 95%. There was no difference between patients with T1mic/T1a tumors and patients with T1b tumors in the distant recurrence rate (94.5% vs 95.5%, respectively; P = .81) or in the receipt of chemotherapy (95.9% vs 94.5%, respectively; P = .63).
Excellent 5-year locoregional and distant control rates were achievable in patients with TNBC who had tumors ≤1.0 cm, 58% of whom received chemotherapy. These results identified a group of patients with TNBC who had favorable outcomes after early detection and multimodality treatment. Cancer 2012. © 2012 American Cancer Society.
Advents in mammographic screening have resulted in increased detection of stage I breast cancer over the past decade,1, 2 including triple-negative breast cancer (TNBC), a distinct clinical and molecular subtype of breast cancer defined by the lack of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). The lack of a molecular target renders TNBC ineligible for conventional targeted therapies in breast cancer, such as hormone therapy or trastuzumab. Compared with other biologic subtypes, TNBC has a predilection for aggressive clinical behavior manifested by early onset, distant metastasis (DM), and a poor prognosis, even after adjustment for other known poor prognostic factors.3-6
In contrast, unselected, subcentimeter, lymph node-negative (N0) breast cancers generally carry an excellent prognosis. According to a large, population-based cohort study, the 10-year breast cancer-specific mortality rate was 5% for patients with T1a, T1b, lymph node-negative breast cancers, including all molecular subtypes. However, the presence of negative ER and PR status elevated the risk of breast cancer-specific mortality.7 Others have reported that patients who have ER-negative/PR-negative, T1a,b, lymph node-negative breast cancer have a higher risk of local recurrence compared with their ER-positive/PR-positive counterparts.8 In 2 independent studies of lymph node-negative, operable breast cancers ≤1 cm, triple-negative tumors were identified as prognostic for an increased risk of recurrence.9, 10 Taken together, these data suggest that triple-negative receptor status may pose a meaningful risk of both local and distant recurrence, even among patients with otherwise low-risk features, such as small primary tumor size and pathologically negative lymph nodes.
To date, studies that have evaluated outcomes of patients with ≤1 cm, lymph node-negative breast cancers have been limited by small subsets of patients with TNBC. Under representation of this population in TNBC studies may be attributable to the rarity of this stage, because TNBCs tend to present with advanced-stage disease at diagnosis.3, 11 Given the known aggressive course of TNBC, clinicians wrestle with therapeutic decisions because of the paucity of clinical data on small, lymph node-negative TNBCs. Currently, there is a lack of supportive information regarding incremental gains with chemotherapy in this specific population of breast cancer patients.
We undertook the current study to address the uncertainty of treatment recommendations regarding this population of patients with TNBC. We describe the clinical features and evaluate the risk of recurrence in a cohort of patients with lymph node-negative TNBC who had a primary tumor size ≤1 cm (microscopic T1 [T1mic], T1a, and T1b) and who were treated in an era of contemporary surgery and adjuvant therapies.
MATERIALS AND METHODS
The study design and data collection were approved by the Institutional Review Board at Memorial Sloan-Kettering Cancer Center (MSKCC). From an institutional breast cancer database, 1022 consecutive patients with stage I to III TNBC who underwent definitive breast surgery with or without adjuvant therapy at our center from 1999 to 2006 were identified. Patients who had primary tumors ≥1 cm; pathologic lymph node-positive disease; a history of previous or concurrent, invasive secondary malignancy; and bilateral, invasive breast cancer were excluded, which left 194 patients with lymph node-negative disease who had tumors ≤1 cm (T1mic, T1a, T1b) available for the current analysis.
Patient, disease, and treatment characteristics for the study cohort were collected retrospectively from electronic medical records. Information on breast cancer susceptibility gene (BRCA1/BRCA2) mutation status was obtained from the MSKCC Clinical Genetics Service database. Pathologic evaluation of all 194 tumors was performed by MSKCC breast pathologists at the time of diagnosis for the following variables: tumor size, histologic subtype, histologic grade, and lymphovascular invasion. Immunohistochemistry was used to determine ER, PR, and her-2-neu status. Quantification of ER and PR staining was performed when slides were available for review. Negative HER2 status was defined as 0 to 1+ staining by immunohistochemistry and/or nonamplification (HER2/chromosome enumeration probe ratio, <2) by fluorescence in situ hybridization. Primary tumors and lymph nodes were staged according to the sixth edition of the American Joint Committee on Cancer's AJCC Cancer Staging Manual.12
Local therapy consisted of breast-conserving surgery or total mastectomy. Reasons for mastectomy were abstracted from the medical records, when documented. The axilla was staged pathologically with a sentinel lymph node biopsy in all patients. No patients underwent an axillary dissection. Negative surgical margins were defined as no tumor at the inked margin. Close margins were defined as tumor present ≤2 mm from the inked margin. Adjuvant radiation consisted of whole-breast radiation only. A separate supraclavicular or axillary radiation field was not used in any of the patients. The median dose delivered to the tumor bed was 6040 centigrays delivered in 28-30 fractions.
The primary endpoints were locoregional recurrence (LRR)-free survival and distant recurrence-free survival. These endpoints were analyzed separately, because the clinical implications and management of isolated recurrences local versus distant recurrences are distinct. LRR was defined as biopsy-proven, triple receptor-negative recurrence in the ipsilateral breast or chest wall or the axillary, supraclavicular, or internal mammary lymph nodes. The time to recurrence or death was defined from the date of definitive surgery. Patients who died before recurrence or did not have a recurrence were censored at the date of death or last follow-up, respectively. The last follow-up date was defined as the last breast cancer evaluation by a physician or a mammogram. Overall survival was estimated from the date of definitive surgery to death. An analysis of event-free survival also was performed, and an event was defined as LRR, DM, or death, whichever occurred first.
Differences in demographic and clinical variables by tumor size (T1mic + T1a vs T1b) and by receipt of chemotherapy were tested using the Pearson chi-square test. The Kaplan-Meier method was used to analyze the rates of LRR, DM recurrence, event-free survival, and overall survival. The log-rank test was used to compare differences in Kaplan-Meier estimates for each demographic and clinical variable. Multivariate analyses were not performed because of the low number of events.
Table 1 delineates the baseline clinical and pathologic characteristics for the 194 patients grouped according to tumor size (T1mic/T1a vs T1b). The median age at diagnosis for the entire cohort was 55.5 years (range, 27-84 years). The majority of patients were postmenopausal (68%) and Caucasian (86%). Tumor classification was T1mic in 16 patients (8.2%), T1a in 49 patients (25.3%), and T1b in 129 patients (66.5%). Sixty-nine percent of patients had tumors discovered initially by imaging studies (117 mammogram, 7 sonogram, 10 magnetic resonance imaging), and 60 patients (31%) had tumors discovered initially by clinical (patient or physician) detection. The majority of tumors, as expected, were poorly differentiated (73%) and had invasive ductal histology (95%). Eighty-eight percent of tumors lacked lymphovascular invasion. One hundred sixty-one patients (83%) had ductal carcinoma in situ (DCIS) associated with invasive tumor, 16 (10%) in T1mic tumors, 45 (28%) in T1a tumors, and 100 (62%) in T1b tumors.
|T1mic and T1 Tumors (n = 65)a||T2 Tumors (n = 129)|
|Characteristic||Total No. of Patients||No.||%||No.||%||P|
|Age at diagnosis, y|
|Final margin status|
|Wide local excision||129||41||31.8||88||68.2|
|Type of chemotherapy|
|Anthracycline and taxane||28||3||10.7||25||89.3|
|BRCA mutation (n = 49 tested)|
One hundred twenty-nine patients underwent wide local excision (WLE). One hundred twenty-two patients (95%) who underwent WLE received adjuvant whole-breast radiation. The remaining 65 patients underwent total mastectomy and sentinel lymph node biopsy, and none of those patients received radiation. Twenty-five of the patients (38%) who underwent mastectomy had multicentric disease. Multifocal tumor was present in 21 patients (16%) who underwent WLE. In patients with multicentric or multifocal disease, tumor classification was assigned according to the largest focus of invasive disease. No patients in the entire cohort had positive margins. Sixteen patients (13 who underwent WLE, 3 who underwent mastectomy) had close margins.
Forty-nine patients (25%) in the study cohort were tested for a BRCA1 or BRCA2 deleterious mutation. Among these, 26 patients had positive results (19 had BRCA1 mutations, 7 had BRCA2 mutations), and 23 had negative results. Of the 26 patients who had positive BRCA1 or BRCA2 deleterious mutations, 18 underwent mastectomy, and 19 received chemotherapy.
Table 2 examines characteristics of the cohort categorized according to the receipt of systemic therapy. No patients received adjuvant endocrine therapy, and 58%113 received chemotherapy. The most common regimens delivered were combined cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) or sequential methotrexate, 5-fluorouracil, and leucovorin (MFL) in 64 patients (56.6%), followed by anthracycline and taxane (28 patients; 24.8%), anthracycline without taxane (14 patients; 12.4%), and other regimens, such as cyclophosphamide plus docetaxel (3 patients), capecitabine (2 patients), carboplatin plus paclitaxel (1 patient), and an unspecified regimen in 1 patient who received chemotherapy at an outside institution. Patients who received chemotherapy tended to be younger (P = .009) and had more adverse pathologic features, such as larger tumors (P = .0001) and poor tumor grade (P = .02). A significantly greater proportion of patients who underwent mastectomy received chemotherapy compared with the patients who underwent WLE (P = .012).
|Chemotherapy (n = 113)||No Chemotherapy (n = 81)|
|Characteristic||Total No. of Patients||No.||%||No.||%||P|
|Age at diagnosis, y|
|Tumor classification, cma|
|Wide local excision||129||67||51.9||62||48.1|
|BRCA mutation (n = 49 tested)|
With a median follow-up of 73 months (range 5-143) among survivors as of March 14, 2011, there have been 7 local recurrences, 1 lymph node recurrence, 11 distant recurrences, and 12 deaths. The 5-year LRR-free, DM-free, event-free, and overall survival rates were 96%, 95%, 91%, and 97%, respectively. Kaplan Meier survival analyses were performed to examine 5-year locoregional and distant outcomes according to tumor size (Fig. 1). There was no difference in 5-year LRR-free survival (96.6% vs 95.9%; nonsignificant P value) or 5-year distant recurrence-free survival (94.5% vs 95.5%; nonsignificant P value) between patients with T1mic/T1a tumors and patients with T1b tumors, respectively. Similarly, these endpoints were compared between patients according to the receipt of chemotherapy (Fig. 2), and there was no difference in 5-year LRR-free survival (96.2% vs 96%; nonsignificant log-rank P value) or distant recurrence-free survival (95.9% vs 94.5%, log-rank P value nonsignificant) between patients who did and did not receive chemotherapy, respectively. The 5-year DM-free survival rates were analyzed among 4 subgroups of patients according to tumor size and receipt of chemotherapy, and there were no differences in outcomes (Table 3).
To address the possibility that misclassification of ER and PR status may have a differential impact on prognosis, DM-free survival also was analyzed according to the quantification of ER and PR staining (Fig. 3). Sixty-eight percent (n = 131) of the study cohort had 0% ER and PR staining, and the remaining patients had either 1% to 10% ER and/or PR staining (n = 28) or unquantified ER/PR staining (n = 35). There was no difference in 5-year DM-free survival between the groups (0% ER/PR staining, 94.8%; 1%-10% ER and/or PR staining, 100%; unquantified ER/PR staining, 94.46%; P = .42; log-rank test).
|No. of Patients (%) [95% CI]|
|Treatment Group||T1mic/T1a Tumors||T1b Tumors||Total No.|
|Chemotherapy||24 (100)||89 (94.7) [0.86-0.98]||13|
|No chemotherapy||41 (91.6) [0.7-0.97]||40 (97.2) [0.81-99]||1|
The clinicopathologic characteristics of patients who developed recurrent disease are outlined in Table 4. None of the 3 patients who underwent mastectomy and experienced LRR had features that were predictive of local failure, such as multicentric disease or close margins. All 7 local recurrences developed before 3 years (mean, 15 months; range, 5-30 months). Eight of 11 distant recurrences (73%) developed before 5 years (mean, 41 months; range, 1.55-100 months). No patients who developed a local recurrence experienced a subsequent distant recurrence. Of the 11 women who had distant recurrences, 6 have died. The initial sites of metastatic involvement in these patients were lung (5 patients), liver (3 patients), and brain (3 patients). The mean survival from distant recurrence to death was 10.3 months (range, 6-20 months).
|Patient No.||Age at Diagnosis, y||Tumor Classification||Local Treatment||Chemotherapy||Site of Disease Recurrence||Recurrence-Free Survival, mo|
|1||44||T1b||BCS, RT||None||Breast, supraclavicular lymph nodes||11.44|
On univariate analysis, none of the studied variables (age, tumor size, lymphovascular invasion, receipt of chemotherapy, tumor grade) had a significant association with 5-year LRR-free, DM-free, or event-free survival (all P > .05). Given the lack of statistical significance achieved and the small number of events, multivariate analyses were not performed.
In this single-institution, retrospective study, we studied the outcomes of 194 patients with tumors ≤1 cm who had lymph node-negative TNBC and received multimodality therapy. There were 7 local recurrences, 1 LLR, and 11 distant recurrences, resulting in 5-year LLR-free and DM-free survival rates of 96% and 95%, respectively. These results represent an optimistic departure from the poor prognosis associated with TNBC, demonstrating that excellent short-term clinical outcomes are achievable in a select subset of patients with low-risk TNBC.
Immunohistochemically defined subtype as a prognostic factor specifically in patients who have ≤1 cm, lymph node-negative breast cancers has been explored in 2 recently published studies. In a Korean study, 56 women in a cohort of 378 women with T1mic, T1a, and T1b lymph-node-negative breast cancer had triple-negative receptor status. After a median follow-up of 60 months, TNBC was identified as an independent prognostic factor for recurrence (hazard ratio, 4.93; 95% confidence interval, 1.312-18.519; P = .018). Similarly, an Italian study evaluated the patterns of recurrence in 1691 women with ≤1 cm, lymph node-negative, invasive breast cancer, including 95 women who had TNBC. At a median follow-up of 6 years, the triple-negative subtype was correlated with a significantly increased risk of distant relapse, LRR, and breast cancer-related survival in multivariate analysis. Our study did not include a comparison group of patients with breast cancer subtypes other than TNBC. However, our observation of excellent outcomes in patients with subcentimeter TNBCs suggests that differences in prognosis in this group may not be clinically meaningful.
Treatment assignment was not random in our study patients, and it is possible that adjuvant treatment may have mitigated an adverse prognostic impact of the TNBC phenotype. The majority of patients with T1b tumors (79%) and the minority of patients with T1mic/T1a tumors (21%) in our study population received chemotherapy. The patterns of chemotherapy administration were largely in accordance with the recommendations set forth by consensus guidelines. The National Comprehensive Cancer Network13 and the St. Gallen panel14 do not recommend chemotherapy for T1aN0 TNBC, but they support its consideration in T1bN0 TNBC. The use of prognostic variables to guide the recommendation for chemotherapy in patients with T1b tumors was not specified. In our study, well established prognostic features for recurrence, such as young age, poorly differentiated tumor grade, and larger tumor size, were associated with the receipt of chemotherapy. A greater proportion of patients who underwent mastectomy received chemotherapy (40% vs 20% no chemotherapy; P = .012), although this trend probably was secondary to the greater prevalence of multifocal/multicentric disease and BRCA mutations in the group.
Among the 7 patients in our study who developed a local recurrence, 3 received breast-conserving therapy (BCT), and 3 underwent mastectomy. Only 38% of patients who underwent mastectomy had multicentric disease that precluded BCT. For the remaining 40 patients who underwent mastectomy, the reasons driving the decision for mastectomy were not clearly identifiable given the retrospective nature of data collection. Irrespective of the type of local therapy received, the risk of local recurrence was similarly low. Nevertheless, this risk was not negligible in the patients who underwent mastectomy, suggesting that mastectomy does not obviate the risk of local recurrence in patients with small TNBCs. Increased rates of LLR after modified radical mastectomy without adjuvant radiotherapy versus BCT were observed recently in a study of 768 patients with T1/T2N0 TNBC. In that study the 5-year LRR-free survival rate was 96% versus 90% in the BCT and modified radical mastectomy groups, respectively (P = .027), highlighting the effectiveness of BCT as treatment for patients with TNBC.15 The characteristics of our study population differ in subtle but important ways from the demographic and clinical profile of the patients with TNBC described by others. The prevalence of TNBC among younger women and African American women has been reported in large cohort studies.3, 11, 16 In contrast, in our study, approximately 66% of women were aged >50 years, and black race represented only 5.6% of our study population. Sixty percent of patients presented with mammographically detected tumors, whereas TNBCs in other series commonly present as “interval cancers” detected by palpation.3, 16 In our study, the index tumor was associated with DCIS in 83% of patients, contrary to the low incidence of associated DCIS in TNBC tumors described by others.17 The clinical and pathologic features of the TNBCs observed in our study may reflect a slow growth rate or a less aggressive pattern of carcinogenesis, consisting of a stepwise pattern of progression from DCIS to invasive breast cancer. These data underscore the importance of screening in populations identified at risk for high prevalence of TNBC, because early detection and treatment results in an excellent prognosis. Thus, as the detection of small breast cancers continues to improve with advances in breast imaging, the incidence of TNBCs ≤1 cm will increase, further enhancing the applicability of these data to large populations of patients with breast cancer.
The various limitations of our current study require discussion. Selection bias was unavoidable given its retrospective design. More adverse clinical and pathologic features predicted that patients would receive chemotherapy, which made it impossible to distinguish whether or not the favorable prognosis observed in our study population was indicative of the low intrinsic metastatic potential of small, lymph node-negative TNBC or the efficacy of chemotherapy. Second, because all patients were treated at a single institution, the effect of referral bias and the general applicability of these results to larger populations of patients with TNBC must be considered. Given the vastly different clinical implications and management of local and distant recurrences, we chose to analyze these both independently and as a single entity in the form of event-free survival. For each endpoint, the small number of events limited the statistical significance and power of the data obtained. On univariate analysis, no factors were associated with decreased LLR-free survival, distant recurrence-free survival, or event-free survival.
The potential for misclassification of tumors based on ER and PR staining quantification also was present. According to recent American Society of Clinical Oncology guidelines, negative ER/PR status is defined as <1% immunohistochemically reactive tumor cell nuclei by immunohistochemistry for ER or PR.18 During the early years of the study, the definition of TNBC included patients with ≤10% ER and PR expression, and the percentage of ER and PR expression was not routinely quantified. Although approximately 33% (n = 63) of our study population had ER and PR staining of 1% to 10% or had unquantified staining, concern regarding the effect of positive receptor status on treatment response can be mitigated by the finding that none of our study patients received hormone therapy. Furthermore, an analysis of 5-year distant relapse rates between patients with confirmed ER 0%/PR 0% staining versus those with ER/PR staining from 1% to 10% or unquantified ER and PR staining demonstrated no difference between the groups.
We regarded the median follow-up of 73 months as sufficient for the evaluation of clinical outcomes in this population, because the mean time to recurrence in patients with TNBC typically is brief and peaks at 1 to 3 years after diagnosis, followed by a decline over the next 5 years.3 In the current study, the mean time from definitive surgery to distant recurrence was 41 months, which was more prolonged than what was reported previously for DM events in patients with TNBC.17 In contrast, the mean time to local recurrence (15 months) was significantly shorter. Moreover, patients who experienced local recurrences did not develop subsequent DM, suggesting that the pathogenesis of local and distant recurrences may be independent in these tumors. Patterns of recurrence and the clinical course after recurrence corroborated those previously described in TNBC.19 Visceral sites, including lung, liver, and brain, represented sites of DM; whereas metastases to bones were absent. Distant recurrence resulted in rapid clinical deterioration and death in approximately 50% (6 of 11) of patients.
In summary, excellent 5-year locoregional and distant control rates were achieved in patients with ≤1 cm, lymph node-negative TNBC treated at our institution, a substantial number of whom received chemotherapy and BCT. Early detection of these tumors is critical given the excellent short-term prognosis attainable after treatment. Our findings reflect 1 aspect of the continuum of natural histories, clinicopathologic features, and outcomes observed within the TNBC group. Whether or not these results translate into long-term cure will require longer observation and validation by other similar studies.
In the absence of clinical trials evaluating the role of chemotherapy exclusively in this subset of patients with TNBC, the current data provide reassurance that reliance on standard patient and tumor-related characteristics to guide the selection of patients for chemotherapy in this population is useful and should remain standard practice.
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 11Descriptive analysis of estrogen receptor (ER)-negative, progesterone receptor (PR)-negative, and HER2-negative invasive breast cancer, the so-called triple-negative phenotype: a population-based study from the California Cancer Registry. Cancer. 2007; 109: 1721-1728., , , , .
- 12AJCC Cancer Staging Manual, 6th ed. New York: Springer-Verlag; 2002., , , et al, eds.
- 13North Center Cancer Network (NCCN). NCCN Clinical Practical Guidelines in Oncology Breast Cancer Guidelines. Version 2.2009. Fort Washington, PA: NCCN; 2009. Available from: http://www.nccn.org. Accessed May 11, 2011.