Long-Term Survival of Women with Locally Advanced Breast Cancer with ≥10 Involved Lymph Nodes at Diagnosis

Background: Axillary lymph node status at diagnosis remains the strongest predictor of long-term survival in breast cancer. Patients with more than ten axillary lymph nodes at diagnosis have a poor long-term survival. In this single institutional study, we set out to evaluate the prognosis of this high-risk group in the era of multimodality therapy. Materials and Methods: In this retrospective study, we looked at all breast cancer patients with greater than ten axillary lymph nodes diagnosed at Mount Sinai Medical Center (MSMC) from January 1st 1990 to December 31st 2007 (n=161). In the univariate analysis, descriptive frequencies, median survival, and 5- and 10-year survival rates were estimated for common prognostic factors. A multivariate prognostic analysis for time-to-event data, using the extended Cox regression model was carried out. Results: With a median and mean follow-up of 70 and 89.9 months, respectively, the overall median survival was estimated to be 99 months. The five-year disease-free survival (DFS) was 59.3% and the ten-year DFS was 37.9%, whereas the five- and ten-year overall survival (OS) was 66.6% and 43.9%, respectively. Multivariate analysis revealed a significant improvement in DFS among black patients compared to whites (p=0.05), improved DFS and OS among young patients (ages 21-45) compared to elderly patients (age greater than 70) (p=0.00176, p=0.0034, respectively), and improved DFS and OS among patients whose tumors were ER positive (p=0.049, p=0.0034). Conclusions: In this single institution study of patients with greater than 10 positive axillary nodes, black patients had a significantly improved DFS compared with white patients. Young age and ER tumor positivity was associated with improved outcomes. Using multivariate analysis, there were no other variables associated with statistically significant improvements in DFS or OS including date of diagnosis. Further work is needed to improve breast cancer survival in this subgroup of patients.


Introduction
With an estimated 226,870 diagnoses and 39,510 deaths in 2012, breast cancer remains the most commonly occurring and second most lethal cancer among women in the United States (Howlader et al., 2012). Although the prognosis is variable with a five-year survival ranging from 18-95 percent (Ries et al., 2001;Colleoni et al., 2005), axillary lymph node status at diagnosis remains the strongest predictor of long-term survival (Ragaz et al., 1997). Locally advanced breast cancer (LABC) makes up approximately 6.5 % of newly diagnosed breast cancer (VanderWalde et al., 2012). In patients with LABC, the rate of recurrence and associated disease related mortality is high, with an estimated 10-year overall survival (OS) between 24-50% (Woodward et al., 2003;Montero et al., 2005), with inflammatory breast cancer (IBC) having the

Materials and Methods
In this retrospective study, we looked at all breast cancer patients with greater than ten axillary lymph nodes diagnosed at Mount Sinai Medical Center (MSMC) from January 1st 1990 to December 31 st 2007 (n=161), with a follow-up until December 31 st 2012. The minimum follow-up time was 5 years, while the maximum follow-up was 23 years. The mean follow-up time was 89 months. Excluded patients included those with de-novo metastatic disease, men, secondary cancers, bilateral breast cancers, 2 nd primary breast cancers, inflammatory breast cancers, those lost to follow-up, and those who only received a onetime consult. The number of involved lymph nodes was determined by pathologic analysis after definitive surgery. The results were presented in descriptive frequencies, using means and standard deviation for continuous variables while for the other types of variables we used frequencies and percentages. The study consisted in two parts: the univariate and the multivariate analyzes. In the univariate section, the studied factors were analyzed through the time-to-event endpoints, in two ways: first, overall survival (OS) where deaths from any cause were included, while all other events were considered censored; second, disease-free survival (DFS) where the relapses and deaths from any cause were included, while all other events were censored. For both types of endpoints the median survival, the 5-and 10-years survival rates were estimated for all analyzed factors (i.e., adjuvant treatment types, race, ethnicity, age at diagnosis, year at diagnosis, pathologic subtype, tumor size, number of metastatic nodes, hormone receptor status, HER2-neu status, etc.). The survival rates were estimated using the Kaplan Meier method and the survivorships were compared using nonparametric survival comparisons. For both endpoints, the survival time was calculated from the date of diagnosis to the time-to-event (i.e. relapse, or death). The simple Cox proportional hazards models were also use to estimate the crude hazard ratios (HR) along with their corresponding 95% confidence intervals and the Wald test for significance (Suciu et al., 2004). The multivariate section consisted in the analysis of all prognostic factors significant at 0.25 level in the univariate analysis. We used the multiple Cox proportional hazards regression for this purpose, were the estimated hazard ratios were considered significant at the level of 0.10. We increased the error margins since the sample size was small.

Discussion
Numerous studies have shown that even after accounting for health access disparities, young African American women tend to have worse outcomes in breast cancer (Balakrishnan and Rao, 2002;Palmer et al., 2003). Our study demonstrated a paradoxical effect and showed that black patients had improved DFS in this N3 population even when accounting for all other variables in the multivariate model. There are several possible explanations for this unusual finding. First of all, the number of black patients in our study was small (n=23), and thus it is likely that a few long-term survivors skewed the survival results for the entire group. Upon sub-group analysis based on race, a greater percentage of black patients were younger (21-45: 26.1% vs 18.5%), diagnosed in the later time frame (2000-2007: 69.6% vs 37.0%), had fewer involved nodes (>15: 30.4% vs 44.9%), and received an anthracycline as part of neo-adjuvant/ adjuvant chemotherapy (92.3% vs 86.2%). Taxanes were more commonly used in white patients (70.6% vs 57.1%). All of the other clinical variables and treatment regimens were similar among the cohorts. This subgroup analysis suggests the black patients in our cohort had more favorable prognostic variables than their white counterparts. This hypothesis is supported by the fact that only 34.8% of black patients in our cohort relapsed compared to 60.9% and of white patients.
There has been conflicting data on the outcomes of Hispanic patients with breast cancer compared to non-Hispanics Banegas and Li, 2012;Anaya-Ruiz, 2014). Most studies have suggested that socioeconomic status and patient access to care are greater determinants of long-term survival than ethnicity alone (Livaudais et al., 2012). In our study, there was a paradoxical effect showing a trend toward an improved overall survival in Hispanics compared to Non-Hispanics, even when controlling for all other variables in the multivariate analysis. The Hispanic paradox has been well described (Turra and Elo, 2008;Blue and Fenelon, 2011).
We observed a significant improvement in DFS among patient aged 21-45, but not in OS. This observed difference was likely secondary to the selection of fit, healthy patients who were able to complete multi-modality therapy compared with those older patients who had associated comorbidities, were unable to tolerate standard treatment, and were likely to die due to secondary causes. However, the magnitude of this effect was likely tempered, and failed to reach significance on multivariate analysis, by the fact that younger patients often have tumors that have inherently more aggressive biology (Anders et al., 2008) and have associated poor prognostic factors (Fan et al., 2006, Keramatinia et al., 2014. In our study, ER tumor status was found to be significant for both DFS and OS in the multivariate analysis. This finding supports the fact that ER positive tumors have less aggressive biology, are more likely to metastasize later, and are more likely to metastasize to bone as opposed to visceral organs. These findings support previous studies suggesting different inherent genetic profiles between ER positive and ER negative tumors (Carey et al., 2006;Hu et al., 2006;Parker et al., 2009;Voduc et al., 2010;Livi et al., 2012). Although we only had data for a small minority of patients, our cohort had high rates of HER2 positive disease (47%) compared with the 25-30% incidence seen in the general breast cancer population. This finding is not surprising given the fact that HER2 positive tumors are known to be more locally aggressive (essentially all of our patients) and have a greater potential to metastasize early (Koca et al., 2013).
Not surprisingly, there was no difference in DFS or OS based on histopathologic subtype. Interestingly, the size of the initial tumor had no significant effect on DFS or OS. This finding is in contrast to numerous large studies and TNM staging, which dictate that larger primary tumors have a higher recurrence rate and a worse overall survival (Akhsan and Aryandono, 2010;Rezaianzadeh et al., 2012). Our study suggests that tumor size may not be of great importance in this select group of high-risk patients. Within our cohort, there was no difference in survival based on number of involved lymph nodes at diagnosis (10-15 versus >15). This finding suggests that 10 involved axillary lymph nodes remains a critical threshold at which more involved nodes may not confer a worse prognosis. These findings are consistent with current TNM breast cancer staging, which designates ten involved axillary lymph nodes as the highest nodal stage (pN3a), beyond which prognosis remains unchanged. Meanwhile, several studies have shown alternative and possibly improved prognostication with use of lymph node ratio (LNR), or the number of positive axillary lymph nodes to the number of total nodes removed (Vinh-Hung et al., 2004;Danko et al., 2010). Vinh-Hung et al proposed a new node staging system in which patients were divided based on risk (low: ≤0.2; moderate: >0.2, ≤0.65;high: >0.65;Vinh-Hung et al., 2004). To date, this proposed staging system has not been studied in large randomized studies.
Our study only had a median follow-up of 70 months. Therefore, it is likely that there was not enough follow-up for the patients diagnosed in the later cohort to show a significant improvement in survival. We suspect, with 10 years of follow-up, the difference between the two groups would be readily apparent, especially among patients with ER negative tumors, who are known to recur earlier.
The use of anthracyclines as part of standard adjuvant chemotherapy dates back to the late 1980's-early 1990's (Ambrosini et al., 1988). Therefore, most of the patients, for whom we had data, received this type of chemotherapy (87.7%). However, taxanes, currently part of the standard of care, did not appear in widespread practice until the early 2000's and thus fewer patients received this class of chemotherapy (68%; Nabholtz et al., 2001). Tamoxifen (55% of ER positive patients in our study) had been used in widespread clinical practice since the late 1980s, but aromatase inhibitors (47% of ER positive patients in our study), another hormonal therapy used in hormone receptor positive patients, was not used in widespread clinically practice until the late 1990's (Buzdar et al., 1996). Her-2 testing and trastuzumab became part of the standard of care in the early 2000s, and although we don't have all of the Her-2 data on our population sample, we suspect many of the patients diagnosed in the later timeframe, who were Her-2 positive, did in fact receive this therapy (Mieog et al., 2007). As previously mentioned, there was a lack of improvement in DFS or OS in the final multivariate model, with the use of an adjuvant anthracycline, with or without the use of a taxane. Although ER positive patients had a significantly improved survival, patients who received hormonal therapy did not have significantly improved DFS or OS in our final multivariate model. This finding, likely accounting for the lack of improvement in survival endpoints between the time frame cohorts, suggests that chemotherapy and hormonal therapy may only improve long-term survival in a select subset of very high-risk patients. There was no significant improvement in DFS or OS in patients who were HER2 positive. For reasons that are unclear, there was a low rate of adjuvant radiation among the patients in our cohort (59%), but surprisingly, there was no difference in DFS or OS in those patients who received this therapy. This finding suggests that adjuvant radiation may only improve local disease control in a subset of patients and is in contrast to studies showing that those patients who do not receive radiation have high local-regional recurrence (LRR). Not surprisingly, patients receiving neoadjuvant chemotherapy had a worse overall survival. This finding is likely due to a selection bias, as patients who received neoadjuvant chemotherapy had larger, more aggressive tumors than those patients who received adjuvant chemotherapy (Duman et al., 2012;Prajoko and Aryandono, 2014). Type of surgery performed was not found to have a significant impact on overall survival, which is in contrast to previous studies (Zeichner et al., 2014).
One of the main limitations with our study was our inability to make a definitive statement regarding the impact of HER2 status and trastuzumab therapy on long term-outcomes. Although there was no significant difference seen among patients based on HER2 status, we only had this information for less than 25% of our cohort. In this single institution study of patients with greater than 10 positive axillary nodes, black patients had a significantly improved DFS compared with white patients. Young age and ER tumor positivity was associated with improved outcomes. Using multivariate analysis, there were no other variables associated with statistically significant improvements in DFS or OS including date of diagnosis. Further work is needed to improve breast cancer survival in this subgroup of patients.