PDF
ePub
Citation
Print
INTRODUCTION
Breast cancer is the most common cancer and the leading cause of cancer-related mortality in women, both worldwide and in Thailand. Locally advanced breast cancer (LABC), the most advanced stage of nonmetastatic breast cancer, has a substantial risk of recurrence, metastasis, and death, with a 5-year overall survival (OS) rate of approximately 57% [1]. LABC includes patients with any tumor >5 cm, or that involves the skin or chest wall, and also those with fixed axillary lymph nodes, or ipsilateral supraclavicular, infraclavicular, or internal mammary nodal involvement [2]. LABC accounts for only 5% to 7% of all breast cancer in the United States [1], whereas it represents 24% to 27% of all newly diagnosed breast cancer cases in Thailand [3]. Despite multimodality treatment using systemic chemotherapy, surgery, and radiotherapy, the majority of patients develop metastases; therefore, LABC remains a clinical challenge.
Neoadjuvant chemotherapy is the standard treatment for locally advanced and inflammatory breast cancer, with the aim of achieving tumor resectability, as well as for patients with early breast cancer who are considering breast-conserving surgery (BCS). Neoadjuvant chemotherapy is advantageous because it shrinks tumors, thereby rendering inoperable tumors resectable; increases rates of BCS; enables early treatment of micrometastasis; and facilitates in vivo assessment of chemotherapy-sensitivity [4]. However, well-validated accurate pathological tumor staging cannot be performed after neoadjuvant therapy. As tumor progression occurs very rarely (1%-2%) after neoadjuvant therapy, operable tumors rarely become unresectable.
Patients treated with neoadjuvant chemotherapy achieve a clinical response in 50% to 80% of cases, with a clinical complete response (cCR) rate of 10% to 20% and a pathological complete response (pCR) rate to chemotherapy of 10%-30% [5,6,7,8,9,10,11]. As patients who achieve pCR have superior long-term outcomes, pCR is a potential surrogate marker of survival [12].
We performed a retrospective analysis to determine the rate of pCR, clinicopathological factors associated with pCR, and clinical outcomes in breast cancer patients treated with neoadjuvant chemotherapy.
METHODS
In this retrospective study, medical records of patients with nonmetastatic breast cancer, treated with neoadjuvant chemotherapy at Siriraj Hospital between January 2007 and September 2011, were reviewed. Patients' medical records were selected from the hospital database using ICD-10 coding. Only patients who had been treated with neoadjuvant chemotherapy and had undergone subsequent surgery at Siriraj Hospital, and had received postoperative chemotherapy, radiotherapy, and hormonal treatment, if indicated, were included in the present study. The study protocol was approved by Siriraj Institutional Review Board (protocol number: 222/2556[EC4]), Siriraj Hospital Faculty of Medicine, Mahidol University, Thailand.
Invasive breast cancer was diagnosed from core biopsies, and staging was performed according to the American Joint Committee on Cancer/Union for International Cancer Control (AJCC/UICC) tumor-node-metastasis (TNM) staging criteria (v.3 2010). Initial workups for distant metastases included chest radiography, liver ultrasonography, and bone scans.
Neoadjuvant and adjuvant chemotherapy regimens administered included: (1) AC (60 mg/m2 doxorubicin and 600 mg/m2 cyclophosphamide intravenously on day 1, every 3 weeks, for four cycles); (2) EC (90 mg/m2 epirubicin and 600 mg/m2 cyclophosphamide intravenously on day 1, every 3 weeks, for four cycles); (3) FAC (500 mg/m2 fluorouracil, 50 mg/m2 doxorubicin, and 500 mg/m2 cyclophosphamide intravenously on day 1, every 3 weeks, for six cycles); (4) FEC (500 mg/m2 fluorouracil, 90 mg/m2 epirubicin, and 500 mg/m2 cyclophosphamide intravenously on day 1, every 3 weeks, for six cycles); (5) CMF (100 mg/m2/day cyclophosphamide orally on days 1-14, and 40 mg/m2 methotrexate and 500 mg/m2 fluorouracil intravenously on days 1 and 8, every 4 weeks, for six cycles); (6) GC (1,000 mg/m2 gemcitabine intravenously on days 1 and 8, and carboplatin AUC5 intravenously on day 1, every 3 weeks, for six cycles, as part of a clinical study [13]); and (7) D-FEC (75 mg/m2 docetaxel intravenously on day 1, every 3 weeks, for three cycles, followed by 600 mg/m2 fluorouracil, 90 mg/m2 epirubicin, and 600 mg/m2 cyclophosphamide intravenously on day 1, every 3 weeks, for three cycles). Trastuzumab (8 mg/kg loading dose, followed by 6 mg/kg every 3 weeks, for 1 year) was administered to patients with human epidermal growth factor receptor 2 (HER2) overexpressed.
Surgical procedures consisted of mastectomy or BCS. Adjuvant breast radiotherapy was administered to patients who had undergone BCS, as well as to patients with initial clinical stage cT3-T4 and cN2-3 disease. Adjuvant endocrine therapy was administered to all patients with hormone receptor-positive tumors for 5 years.
Assessment of clinical response
The clinical response was assessed following administration of the final neoadjuvant chemotherapy cycle. The following definitions were used [6]: cCR was defined as the absence of clinically evident tumor on palpation; clinically partial response (cPR) was defined as a reduction of 50% or more in the two maximum perpendicular diameters of the tumor; clinically progressive disease (cPD) was defined as an increase of >25% in the two maximum perpendicular diameters of the tumor; and clinically stable disease (cSD) was defined as a clinical breast response that does not meet the definitions of cCR, cPR, or cPD.
Pathological assessment
Expression of estrogen receptor (ER), progesterone receptor (PR), and HER2 was determined on pretreatment biopsies (preferentially) or on surgical specimens if immunohistochemistry (IHC) had not previously been performed. Hormonal receptor (HR) status was considered positive if ≥1% of tumor cells stained for ER and/or PR. HER2 status was considered positive if an IHC score of 3+ was recorded, or if there was positive gene amplification using in situ hybridization testing. As Ki-67 assessment had not been routinely performed, it was not possible to define breast cancer intrinsic subtypes according to the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer [14]. Accordingly, histological grade was used to rank cell proliferation. The following definitions of tumor types were used [15]: (1) luminal A like-tumors, defined as ER-positive and/or PR-positive, HER2-negative, grade 1 or 2; (2) luminal B like-tumors, defined as ER-positive and/or PR-positive, HER2-negative, grade 3; or ER-positive and/or PR-positive, HER2-positive, all grades; (3) HER2-like tumors, defined as ER- and PR-negative, HER2-positive, all grades; and (4) triple-negative tumors: ER-, PR-, and HER2-negative, all grades.
pCR was evaluated according to the criteria [15] of the National Surgical Adjuvant Breast and Bowel Project (NSABP), the MD Anderson Cancer Center (MDACC), and the German Breast Group (GBG), as no invasive cancer in the breast (ypT0/is ypN0/+), no invasive cancer in the breast and lymph nodes (ypT0/is ypN0), and no invasive or in situ cancer in the breast and lymph nodes (ypT0 ypN0), respectively.
Statistical assessment
The primary endpoint of this study was to determine the rate of pCR. The secondary endpoints were to determine the clinical factors associated with pCR, the clinical response rate, disease-free survival (DFS; defined as the interval between the date of diagnosis and the date of disease recurrence or death), and OS (defined as the interval between the date of diagnosis and the date of death from any cause). The required sample size was calculated based on an estimated proportion of one group method, using 13% for pCR in accordance with previous NSABP studies [5,7], with a 95% confidence interval (CI) and 5% error; this calculation resulted in a required sample size of 174 patients. On univariate analysis, the relationships between clinical factors and pCR were assessed using Pearson chi-square or Fisher exact test, as appropriate, and binary logistic regression was used to calculate the odds ratio.
Survival was analyzed using the Kaplan-Meier method, and comparisons made using the log-rank or Breslow test. The Cox proportional hazard model for survival was used for univariate and multivariate analyses. Median follow-up time was calculated using the reverse Kaplan-Meier method. A two-sided level of significance of 0.05 was applied to all statistical tests. SPSS Statistical software version 19.0 (IBM Corp., Armonk, USA) was used for all statistical analyses.
RESULTS
Patients and tumor characteristics
The total number of patients treated with chemotherapy followed by surgery between January 1, 2007 and September 31, 2011 was 237. Data of 179 patients met the inclusion criteria and were used in this study (Figure 1). The median patient age was 48 years (range, 24-75 years), and 59.8% of patients were premenopausal. LABC was diagnosed in 93.3% of cases. The mean tumor size on palpation was 7.9 cm. The majority of patients (69.3%) had node-positive disease. Histological examination revealed that 94.4% of tumors were invasive ductal carcinoma. Tumors were HR-positive in 63.1% of cases (ER-positive, 58.1%; PR-positive, 50.8%). Tumors were HER2-positive in 39.7%, HER2-negative in 50.8%, and equivocal in 9.5%. IHC results were reported from preoperative specimens in 71.5%, and postoperative specimens in 28.5%. Of the 179 tumors, 166 (92.7%) could be classified into one of the four intrinsic subtypes: luminal A like (25.1%), luminal B like (34.6%), HER2 (18.4%), and triple-negative (14.5%).
In all, 152 (89.4%) received anthracycline-based chemotherapy, AC, EC, FAC, or FEC regimens. Of the 71 patients with HER2-positive tumors, 19 (26.8%) received neoadjuvant trastuzumab. At least four cycles of neoadjuvant chemotherapy were received by 66.5% patients. Total mastectomy was performed on 162 (90.5%) of the patients, and the remaining 17 (9.5%) underwent BCS. Patients and tumor characteristics are shown in Table 1.
pCR and clinical objective response
cCR and cPR were observed in seven (3.9%) and 103 (57.5%) patients, resulting in a total clinical response rate of 61.5% (Table 2). Clinical progression was seen in two patients (1.1%). Among the 69 nonresponders (cSD and cPD), seven tumors remained inoperable, and these patients received additional neoadjuvant therapy. Five of these patients were treated with a second neoadjuvant chemotherapy regimen (mostly taxane-based regimens), and one patient received preoperative radiotherapy. The remaining patient received both a second neoadjuvant chemotherapy regimen and preoperative breast radiation. All seven patients eventually underwent total mastectomy. The absence of invasive cancer in the breast (ypT0) was found in 19 patients (10.6%), and the absence of cancer in the lymph nodes (ypN0) was found in 57 (31.8%). pCR occurred in 21 (11.7%), 20 (11.2%), and 17 patients (9.5%), according to the NSABP, MDACC, and GBG criteria, respectively (Table 2).
Association between baseline clinicopathological factors and pCR using the NSABP criteria
According to univariate analysis, pCR was significantly associated with HR status and intrinsic breast cancer subtype (Table 3). The pCR rate was lower in HR-positive tumors (defined as ER- and/or PR-positive) than in HR-negative tumors (7.1% vs. 19.7%; 95% CI, 1.26-8.25; p=0.015). However, only ER status had a statistically significant association with pCR rate (pCR was 21.3% for ER-negative tumors and 4.8% for ER-positive tumors; 95% CI, 1.87-15.42; p=0.002). Analysis of the four intrinsic breast cancer subtypes found that pCR rates were 4.4%, 9.7%, 24.2%, and 19.2% for luminal A like, luminal B like, HER2, and triple-negative tumors, respectively. The pCR rate was significantly higher in the HER2 subtype than that in the luminal A like subtype (95% CI, 1.35-34.97; p=0.020). However, on multivariate analysis, only ER-negative tumors were significantly associated with pCR (95% CI, 1.32-48.2; p=0.024) (Table 4).
Treatment following neoadjuvant chemotherapy and surgery
For 62 patients (35%), postoperative chemotherapy regimens were changed to taxane-based (60 patients) or CMF regimens (two patients), whereas 117 patients (65%) continued with the same chemotherapy regimen that they had received preoperatively (predominantly anthracycline-based). Adjuvant endocrine therapy and trastuzumab were administered to 113 (63%) and 30 patients (17%), respectively. The majority of patients (94%) also received postoperative radiotherapy (Supplementary Table 1).
Survival
The median follow-up duration of the 179 patients was 54.2 months. At the cutoff date for follow-up (September 20, 2013), 59 patients (33%) had disease recurrence, and 55 patients had died (30.7%). Of the 59 patients with recurrent disease, 47 (80%) had distant metastases, seven (11%) had local relapse, and five (9%) had contralateral breast cancer occurrence. The 5-year DFS and OS rates of all patients in this study were 56% and 57%, respectively.
Univariate analysis was performed using the Cox regression method to evaluate whether relevant clinicopathological variables, and known prognostic factors had significant associations with DFS and OS (Table 5). Factors that were significantly associated with improved DFS and OS were low-to-moderate tumor grade, ER positivity of >50%, PR positivity, luminal A like subtype, ypT0/is, ypN0-1, and absence of angiolymphatic invasion. pCR was predictive of longer DFS (95% CI, 0.09-0.95; p=0.041), and trended towards longer OS (95% CI, 0.06-1.05; p=0.059).
Compared with luminal A like subtype, luminal B like tumors were significantly associated with a shorter DFS (unadjusted hazard ratio [HR], 2.74; p=0.012) and nonsignificantly with a shorter OS (unadjusted HR, 2.04; p=0.104), whereas HER2 and triple-negative tumors were significantly associated with shorter DFS and OS (HER2 tumors: unadjusted HR for DFS, 3.05, p=0.012; unadjusted HR for OS, 2.86, p=0.025; triple-negative: unadjusted HR for DFS, 3.42, p=0.006; unadjusted HR for OS, 2.96, p=0.026) (Figure 2, which shows Kaplan-Meier survival curves).
Kaplan-Meier survival curves according to pCR are shown in Figure 3. Patients who achieved pCR showed significant positive associations with DFS and OS compared to those without a pCR (5-year DFS: 80% vs. 53%, log-rank test, p=0.030; 5-year OS: 86% vs. 54%, log-rank test, p=0.042). Subgroup analysis according to ER status demonstrated that pCR was significantly associated with longer DFS (p=0.007) and OS (p=0.004) in patients with ER-negative tumors. In contrast, there was no difference in survival outcome between patients with ER-positive tumors with or without pCR (Supplementary Figure 1).
In this study, ER status (ER-positive vs. ER-negative) was not significantly associated with survival outcome. However, ER-negative tumors showed a general trend towards shorter DFS and OS than ER-positive tumors, as can be seen from the Kaplan-Meier survival curves. The potential influence of the strength of ER-positivity was therefore explored, and patients with ER positivity <50% had worse prognoses (DFS, p=0.028; OS, p=0.023) as the same as ER-negative tumors (DFS, p=0.019; OS, p=0.015). Conversely, tumors with ER positivity >50% had significantly better prognoses (Supplementary Figure 2).
According to multivariate Cox proportional hazard regression analysis, the independent risk factors that were significantly associated with prolonged DFS and OS were pCR and intrinsic breast cancer subtypes; HER2 and triple-negative subtypes were independent risk factors for poorer outcomes (Table 6).
DISCUSSION
Neoadjuvant chemotherapy is the standard treatment for locally advanced and inflammatory breast cancer, and is also offered to patients with early breast cancer who are considering BCS. In developed countries, BCS is used in most cases of breast cancer. However, in Thailand, the majority of patients present with locally advanced disease, making the main aim of initial treatment to downstage the disease and render inoperable tumors resectable. In the present study, the pCR rate following neoadjuvant chemotherapy was 11.7%. pCR was more frequently observed in HER2 and triple-negative breast tumor subtypes. Patients who achieved pCR had significantly improved survival.
At present, the most widely accepted criteria to measure response to chemotherapy are the Response Evaluation Criteria for Solid Tumors (RECIST). However, in this study, we used World Health Organization (WHO) criteria to enable comparison of our findings with those of previously published studies of neoadjuvant chemotherapy, most of which have used WHO criteria [5,6,7,8,9]. Anthracycline-based regimens are highly effective in breast cancer and have showed to result in pCR rates of 10% to 15% [5,8]. The addition of taxane to anthracycline-based regimens was shown to increase the pCR rate to 25% to 30%, but did not have an impact on DFS or OS [7,9,16]. In the two large randomized NSABP studies, NSABP B-18 [5], and B-27 [7], four cycles of an AC regimen achieved a pCR rate of 13%; our pCR rate of 11% was slightly lower. This may be attributable to the greater proportion of patients with locally advanced stage disease in our study; most patients in the NSABP studies had earlier, operable breast cancer (Supplementary Table 2) [5,6,7,8,9,10,11,17,18,19,20]. In a retrospective MDACC study [17], 372 patients with LABC were treated with four cycles of neoadjuvant doxorubicin-containing regimens, mainly FAC, and this resulted in a pCR rate of 12%, which is also consistent with our current study. In contrast, patients with operable breast cancer in the European Organisation for Research and Treatment of Cancer 10902 trial [8] received four cycles of neoadjuvant FEC regimens and achieved only a 4% pCR rate. Another clinical series [18] of 110 LABC patients received between three and eight cycles (mean, 4) of neoadjuvant anthracycline and/or taxane-containing chemotherapy, and reported a pCR rate of 5.5%. The lower pCR rates seen in these trials may be explained by the higher proportion of patients with ER-positive tumors, which are considered to be less responsive to chemotherapy.
Previous studies have addressed clinical and biological factors associated with pCR following neoadjuvant chemotherapy and demonstrated that breast cancer patients with high tumor grade or ER-negative disease [17,21,22,23], and HER2 breast cancer treated with trastuzumab combined with chemotherapy [24,25,26], were more likely to achieve pCR. In the present study, we calculated the pCR rate using NSABP criteria so that we could compare this rate with other clinical and biological factors. We also reported the pCR rates using all three established sets of criteria, to allow comparison of these rates reported in other studies that may have utilized different criteria. ER-negative tumor status, but not tumor grade, was found to be a predictor of pCR. HER2-positive tumor status was not associated with a higher pCR rate, and this may be because the majority of these patients (75%) did not received anti-HER2 therapy. When only patients treated with neoadjuvant chemotherapy combined with anti-HER2 therapy were analyzed, a higher pCR rate was observed.
Patients achieving pCR in this study had 5-year DFS and OS rates of 80% and 86%, respectively, compared to those of 53% and 54% for those with non-pCR. The significantly improved survival in patients with pCR is in concordance with most neoadjuvant chemotherapy trials [5,8,10,16,17]. In our study, some of the patients were also treated with adjuvant chemotherapy, which might have an effect on survival outcome. Therefore, we further explored the effect of adjuvant chemotherapy on survival, but did not find any significant effect.
According to univariate analysis, clinicopathological factors associated with prolonged survival outcome in this study were low-to-moderate tumor grade, absence of angiolymphatic invasion, positive PR status, ER-positivity of >50%, and luminal-type cancer. Patients with luminal A like tumors had better prognoses than patients with luminal B like, HER2, or triple-negative tumors, despite having the lowest pCR rate (4%). Among the four intrinsic breast subtypes, the pCR rate was higher in the triple-negative (19%) and HER2 (24%) groups than in the luminal subtypes.
Although pCR was significantly associated with prolonged DFS, there was only borderline significance of association with OS. This could be explained by the finding that pCR mainly predicted survival outcome in ER-negative breast cancer, but not in ER-positive tumor. Therefore, pCR may be a good predictor of survival for nonluminal (ER-negative) disease following neoadjuvant chemotherapy, rather than for those with luminal breast cancer. Consistent with the findings of the present study, a recent meta-analysis of 6,377 breast cancer patients from seven randomized trials, demonstrated that pCR was associated with improved DFS in luminal B/HER2-negative, nonluminal HER2-positive, and triple-negative disease, but not in luminal A disease [15].
The limitations of this study are its retrospective nature and inclusion of patients from a single center. However, few randomized studies have focused on LABC. This may be attributable to its low incidence (5%-6%) in developed countries compared to developing countries, where LABC accounts for 30% of breast cancer.
The present retrospective study has demonstrated that treatment of Thai LABC patients with anthracycline-based neoadjuvant chemotherapy yields pCR rates comparable to those reported by randomized trials of patients with operable breast cancers in developed countries. pCR may be used as an positive prognostic indicator following neoadjuvant chemotherapy.
XML Download