Oncologist. The. Breast Cancer

Transcription

1 The Oncologist The Oncologist CME Program is located online at To take the CME activity related to this article, you must be a registered user. Breast Cancer Long-Term Outcomes in Stage IIIB Breast Cancer Patients Who Achieved Less Than a Pathological Complete Response ( pcr) After Primary Chemotherapy MARIA TERESA IONTA, a FRANCESCO ATZORI, a MARIA CRISTINA DEIDDA, a VALERIA PUSCEDDU, a SERGIO PALMERI, c BARBARA FRAU, a MONICA MURGIA, a MICHELA BARCA, a LUIGI MINERBA, b BRUNO MASSIDDA a a Department of Medical Oncology and b Department of Public Health, Azienda Ospedaliero-Universitaria, Cagliari, Italy; c Department of Medical Oncology, Policlinico Universitario, Palermo, Italy Key Words. Stage IIIB breast cancer Neoadjuvant chemotherapy Pathological response Long-term outcomes Disclosures Maria Teresa Ionta: None; Francesco Atzori: None; Maria Cristina Deidda: None; Valeria Pusceddu: None; Sergio Palmeri: None; Barbara Frau: None; Monica Murgia: None; Michela Barca: None; Luigi Minerba: None; Bruno Massidda: None. Section editors Gabriel Hortobagyi and Kathleen Pritchard have disclosed no financial relationships relevant to the content of this article. The content of this article has been reviewed by independent peer reviewers to ensure that it is balanced, objective, and free from commercial bias. LEARNING OBJECTIVES After completing this course, the reader will be able to: 1. Summarize the main risk factors for relapse in patients with T4 breast cancer after neoadjuvant chemotherapy. 2. Evaluate the role of hormone receptors and HER-2 as determinants of risk of relapse after neoadjuvant treatment. 3. Compare the difference in outcomes between patients who achieve less than pcr in relation to receptor status. CME This article is available for continuing medical education credit at CME.TheOncologist.com. ABSTRACT Purpose. Pathological complete response (pcr) to primary chemotherapy is the main determinant for improved disease-free survival (DFS) and overall survival (OS). The primary endpoints of our study were the longterm DFS and OS rates in homogeneously treated stage IIIB breast cancer patients who failed to achieve a pcr (<pcr), in relation to residual tumor burden. The secondary endpoint was the prognostic relevance of hormone receptor (HR) and human epidermal growth factor receptor (HER)-2 status. Methods. We analyzed 58 of 74 consecutive stage IIIB patients treated between 1996 and 21 who achieved <pcr following a primary cisplatin, epirubicin, and vinorelbine regimen for up to six cycles. At the time of patient accrual, trastuzumab was not available. After definitive surgery, pathological residual disease re- Correspondence: Maria Teresa Ionta, M.D., Department of Medical Oncology, Azienda Ospedaliero-Universitaria, Cagliari, Italy. Telephone: ; Fax: ; mtionta@yahoo.it Received April 14, 29; accepted for publication October 6, 29; first published online in The Oncologist Express on November 6, 29. AlphaMed Press /29/$3./ doi: / theoncologist The Oncologist 29;14:

2 152 Long-Term Outcomes in Stage IIIB Breast Cancer mained in 4 (69%) patients in both the breast and axilla, in 14 (24%) patients in only the breast, and in four (7%) patients in only the axilla. Results. Fifty-eight (78%) of 74 patients achieved <pcr and 16 (22%) had pcr both in the breast and axilla. After a median follow-up of 99 months (range, months), in patients with <pcr the estimated 1-year DFS and OS rates were 37.6% and 5.3%, respectively, significantly worse than in the pcr group (p.3 and p.8, respectively). Patients with four or more axillary nodes involved had a significantly worse 1-year DFS rate (28.9% versus 62.7%; p.36). Patients with HR tumors had significantly lower 1- year DFS (17.3% versus 46.4%; p.18) and OS (17.3% versus 7%; p.2) rates. Overall, the triple-negative (TN) group showed only a marginally significantly worse OS rate (p.48). HER-2 status alone, in the absence of trastuzumab, did not appear to significantly affect outcomes. Conclusions. Our data suggest that, in stage IIIB patients who achieve <pcr, the number of residual nodes and HR status are strong predictors of poor outcomes. After a long follow-up time, HER-2 expression does not appear to significantly affect DFS and OS. TN patients showed a trend toward early recurrence and death. The Oncologist 29;14: INTRODUCTION Stage IIIB breast cancer includes noninflammatory T4a,b,c, and inflammatory T4d forms of the disease, according to the criteria of the recently established American Joint Committee on Cancer staging system [1]. The treatment of this disease is multidisciplinary, including induction chemotherapy, surgery, radiation, adjuvant chemotherapy, and hormone therapy. As a result, either local control or overall survival (OS) have been significantly improved [2]. It is well known that the achievement of a complete pathological response (pcr) in both breast tissue and the axilla after treatment with induction chemotherapy leads to higher disease-free survival (DFS) and OS rates, presumably via eradication of distant micrometastatic disease [3]. However, the rates of pcr in patients with locally advanced breast cancer are low, in the range of 3% 1% after treatment with an anthracycline alone [4]. Despite the recent addition of novel agents such as trastuzumab and the taxanes [5 7] or more aggressive forms of treatment [8], the majority of patients fail to achieve a pcr ( pcr) in breast tissue or in the axilla. Such patients are at a higher risk for early recurrence and death, with 5-year DFS and OS rates of 4% and 5%, respectively [4, 9]. Furthermore, the presence of residual disease in the axillary lymph nodes is a powerful prognostic factor, because prognosis is inversely related to the number of involved nodes. The 5-year DFS and OS rates are 2% when 1 lymph nodes are involved [3, 1]. Most recurrences occur within 3 years of initial treatment, with particularly high rates of recurrence around months after surgery [11, 12]. The risk for relapse and death are threefold and 2.5-fold higher in patients with pcr than in patients who achieve a pcr, respectively [13]. In several studies, the achievement of a pcr has been shown to correlate with longer survival irrespective of hormone receptor (HR) status [13], human epidermal growth factor receptor (HER)-2 status [14], or chemotherapy regimen [15]. By contrast, emerging evidence suggests that the failure to achieve a pcr ( pcr) identifies a heterogeneous group of patients with a different risk for recurrence and death, irrespective of the type of neoadjuvant chemotherapy received. A large number of studies have investigated prognostic factors in the neoadjuvant setting. Molecular markers, such as the estrogen receptor (ER), progesterone receptor (PgR), and HER-2 status, are considered to be predictive or prognostic factors in the neoadjuvant setting [16]. However, these markers are often contradictory and not conclusive because of the heterogeneity of the patient population in terms of initial stage and chemotherapeutic regimen. Moreover, only a few studies have investigated the relationship between clinical outcome and common prognostic features in pcr patients [17]. Based on this background, we analyzed 58 of 74 consecutive stage IIIB patients who achieved pcr following primary chemotherapy. The primary endpoints of the present study were the long-term DFS and OS rates in homogeneously treated stage IIIB breast cancer patients who had achieved pcr following primary chemotherapy, in relation to residual tumor burden in the breast and axilla. The secondary endpoint was the influence of HR and HER-2 status on long-term prognosis in these patients. PATIENTS AND METHODS Patients were eligible for this study if they initially had stage III B cancer (i.e., T4a,b,c,d), as assessed by physical examination and mammography and confirmed via core needle biopsy. Among this group of patients, we considered only those who had evidence of pathological residual disease in the breast, the axilla, or both ( pcr) following primary chemotherapy and definitive surgery. We identified 58 of 74 consecutive patients treated between 1996 and 21 (median follow-up, 99 months; range, months) who met these criteria. Residual disease

3 Ionta, Atzori, Deidda et al. 153 was confined to only the breast in 14 patients (24%), only the axilla in four patients (7%), and both the breast and the axilla in 4 patients (69%). The median number of involved lymph nodes was six (range, 1 ). In more detail, one to three nodes were involved in 11 patients (19%), four to 1 nodes were involved in 23 patients (4%), and 1 nodes were involved in 1 patients (17%); 14 patients (24%) had no axillary lymph node involvement (pln). All patients completed the multimodality treatment plan including primary chemotherapy, surgery, radiation therapy, adjuvant chemotherapy, and hormone therapy, when indicated. Patients with HER-2 tumors did not receive trastuzumab as part of their neoadjuvant or adjuvant therapy, or for metastatic disease. The median age was 53 years (range, 33 7 years); patients (43%) were 5 years old. Thirteen patients (22%) initially had inflammatory breast carcinoma (T4d) and 45 (78%) initially had noninflammatory cancer (T4a,b,c). Thirty-nine patients (67%) were ER and 19 (33%) were ER ; 23 patients (4%) were PgR and 35 (6%) were PgR ; 18 patients (31%) were ER PgR (i.e., HR ), 22 (38%) were ER PgR (i.e., HR ), 17 (29%) were ER PgR, and one (2%) was ER PgR ; 6 patients (1%) were HER-2, 4 (69%) were HER-2, and 12 (21%) had unknown HER-2 status; 12 patients (21%) were HER-2 ER PgR (i.e., triple negative [TN]), three (5%) were ER PgR HER-2 unknown, and 43 (74%) were non- TN. Among non-tn group, there were eight subgroups according to ER, PgR, and HER-2 expression: 16 patients (37%) were ER PgR HER-2, 11 (26%) were ER PgR HER-2, one (2%) was ER PgR HER-2, four (9%) were ER PgR HER-2 unknown, two (5%) were ER PgR HER-2, one (2%) was ER PgR HER-2, three (7%) were ER PgR HER-2, and five (12%) were ER PgR HER-2 unknown. Baseline characteristics are summarized in Table 1. Table 1. Clinical and pathological characteristics of patients who fail to achieve a pathological complete response n of Characteristic patients (58) % Median age 53 Range years 43% 5 years 33 57% T4a,b,c/T4d 45/13 78%/22% ER /ER 39/19 67%/33% PgR /PgR 23/35 4%/6% grade 2/grade 3 4/18 69%/31% HER % HER-2 6 1% HER-2 unknown 12 21% HR (ER PgR ) 22 38% HR (ER PgR ) 18 31% ER PgR 17 29% ER PgR 1 2% TN/non-TN 12/43 21%/74% HR HER-2 unknown 3 5% pln /pln 14/44 24%/76% pln % pln % pln % B A 4 69% B A 14 24% B A 4 7% Abbreviations: A, axilla; B, breast; ER, estrogen receptor; HER, human epidermal growth factor receptor; PgR, progesterone receptor; pln, pathological lymph nodes; TN, triple negative. Pathological Assessment The pathologists involved in the study agreed on the following criteria in order to define pathological characteristics. A complete response was defined as the absence of residual invasive disease in both the breast and the axilla upon final pathological assessment of the surgical specimen. Gross invasive residual disease in breast tissue or the presence of cancer-positive lymph nodes in the axilla were defined as pcr [18]. Residual tumor burden was classified as residual disease both in the breast and in the axilla (B A ), residual disease only in the breast (B A ), residual disease only in the axilla (B A ), and residual disease in the axilla with or without any residual disease in the breast (B B A ). ER and PgR status was assessed by standard immunohistochemical (IHC) analyses. Nuclear staining of 1% was considered positive. HR was defined as ER and PgR, and HR was defined as ER and PgR. HER-2 status was assessed by IHC or fluorescence in situ hybridization (FISH) in breast cancer tissue. HER-2 tumors were defined as 3 on the IHC test. HER-2 tumors were defined as or 1 on the IHC test; IHC 2 required the FISH test. The TN subtype was defined as ER, PgR, and HER- 2 ; non-tn was defined as tumor with any receptor positivity or HER-2 unknown but ER PgR. Treatment Plan Primary treatment consisted of cisplatin (5 mg/m 2 ), epirubicin (1 mg/m 2 ), and vinorelbine ( mg/m 2 ) (PEV). This regimen was administered i.v. for up to six cycles every 2 weeks. Patients underwent surgery (modified radical

4 154 Long-Term Outcomes in Stage IIIB Breast Cancer 5 Hazard >1 Years From Study Entry Figure 1. Hazard of relapsing for patients who fail to achieve a pathological complete response. A 1. B p = A- / B+ A+ / B+ B- 1 Disease-Free Survival (months) mastectomy or conservative surgery) 3 4 weeks after the last dose of primary chemotherapy. Postoperative adjuvant chemotherapy (six cycles of i.v. cyclophosphamide, methotrexate, and fluorouracil [CMF] on days 1 8 every 28 days) was initiated 3 4 weeks after surgery. Locoregional radiotherapy was given during the fourth CMF cycle. After the completion of adjuvant chemotherapy, all HR patients were treated with tamoxifen (2 mg daily) for 5 years, in combination with 2 years of concurrent goserelin, if premenopausal. At the time of patient accrual, trastuzumab was not available. Initially, patient follow-up was performed every 3 months, including clinical exams and laboratory tests. Instrumental examinations (i.e., mammography, liver ultrasound, chest x-ray, bone scan, and echocardiogram) were performed every 6 months for the first 2 years and every 12 months thereafter for 5 years. At the time of writing, no patient included in this study has been lost to follow-up. Statistical Analyses Follow-up data were analyzed in September 27. DFS and OS were calculated based on the date of initial primary chemotherapy and were analyzed using the Kaplan Meier p =.479 A+ / B+ B- 5 method. Statistical comparisons between groups were performed using two-sided log rank tests. All tests were twotailed and p-values.5 were considered to be significant. DFS was defined as the time from study entry to the time of local, regional, or distant treatment failure, occurrence of contralateral breast cancer or other second primary cancer, or death without evidence of breast cancer or a second primary cancer. The OS time was defined as the time from study entry to the time of death from any cause. RESULTS The study population has included 58 (78%) of 74 consecutive homogeneously treated patients who had been diagnosed initially with stage T4 cancer and who had achieved a pcr in the breast, axillary lymph nodes, or both. The remaining 16 patients (22%) achieved a pcr in both the breast and axilla. As expected, the 1-year DFS and OS rates were significantly better in the pcr group than in the pcr group (87.5%; p.3 and 93.8%; p.8, respectively). The overall 1-year DFS and OS rates in the pcr group were 37.6% (95% confidence interval [CI] % 5%) and 5.3% (95% CI, 37% 63%), respectively. At the A- / B+ Overall Survival (months) Figure 2. Kaplan Meier estimates of disease-free survival (A) and overall survival (B) rates according to axillary (A) and breast (B) status at pathological examination in patients who fail to achieve a pathological complete response. 1 1

5 Ionta, Atzori, Deidda et al. 155 A B HR- ER+ PgR- HR ER+ PgR- HR- HR+. p =.18 5 time of the present analysis, 31 (53%) patients had relapsed and 23 (4%) had died. The highest overall risk for relapse was found within the first 3 years after primary chemotherapy because 21 (68%) of the 31 relapses were observed within this time period (Fig. 1). The median times from the start of primary chemotherapy to the first recurrence and death were 17 months (range, months) and 4 months (range, months), respectively. Patients with residual disease in the axilla, irrespective of residual disease in the breast, had shorter DFS and OS times than patients with residual disease in the breast only (Fig. 2). Of note, these differences were statistically significant at the 5-year follow-up (p.5 and p.5, respectively), although they lost statistical significance at the 1-year follow-up. The 1-year DFS and OS rates in patients with residual disease in the breast and in axilla (B A ) were 37.3% and 1 Disease-Free Survival (months) 1. p = %, respectively, whereas in patients with residual disease only in the breast (B A ), the DFS and OS rates were 55.7% and 62.3% respectively. The DFS and OS rates were similar between patients with residual disease in the axilla only (B A ) and those with residual disease in both the breast and axilla. Of note, the DFS and OS rates decreased with increasing number of involved axillary lymph nodes; this effect was more evident in patients who had 1 involved nodes. In more detail, among patient with zero to three and four or more involved axillary nodes, the 1-year DFS rates were 62.7% and 28.9%, respectively (p.36). In the subgroup with 1 involved nodes, the DFS rate was 16.7% (p.46). Of note, among the subgroup with four or more involved nodes, there was a significantly worse OS rate at the 5-year follow-up but this was lost after longer follow up. ER status was associated with significantly worse 1 Overall Survival (months) Figure 3. Kaplan Meier estimates of disease-free survival (A) and overall survival (B) rates according to hormone receptor status in patients who fail to achieve a pathological complete response. Abbreviations: ER, estrogen receptor; HR, hormone receptor; PgR, progesterone receptor. A p = TN Disease-Free Survival (months) 1 non -TN p =.48 5 Overall Survival (months) Figure 4. Kaplan Meier estimates of disease-free survival (A) and overall survival (B) rates according to triple negative (TN) and non-tn status in patients who fail to achieve a pathological complete response. B 1 TN non-tn 1 1

6 156 Long-Term Outcomes in Stage IIIB Breast Cancer A 1. B HER2 + HER HER2 + HER2 -. p = DFS and OS rates (p.3 and p.2, respectively). PgR status was associated with lower DFS and OS rates, although the difference was statistically significant only for OS (p.9). The estimated 1-year DFS rates in the HR (ER PgR ) and HR (ER PgR ) groups were 46.4% and 17.3%, respectively (p.12) and the 1-year OS rates were 7% and 16.3% (p.1), respectively. The ER PgR subgroup had similar 1-year outcomes as the HR group and significantly better DFS and OS rates than the HR group (57.4%; p.3 and 62.7%; p.45, respectively) (Fig. 3). Among patients with TN tumors, there was a nonsignificant worse 1-year DFS rate than in those with non-tn tumors (23.3% versus 44.4%; p.117) and a marginally significant difference in OS rate (44.% versus 64.5%; p.48) (Fig. 4). The median times from the start of primary chemotherapy to first recurrence and death were 13 months (range, 9 8 months) and 21 months (range, months) in the TN group, respectively, and 17 months (range, months) and 4 months (range, months) in the non-tn group, respectively. Among the non-tn group, the HR HER-2 subgroup had the best prognosis, showing a trend toward better DFS (51.6% versus 23.3%; p.79) and a significantly better OS rate (68.3% versus 44%; p.12) than the TN group. HER-2 status alone, in the absence of trastuzumab, did not appear to significantly affect outcomes. In more detail, among patients with HER-2 and HER-2 tumors, the 1- year DFS rates were 37.4% and 33.4% (p.345), respectively; the 1-year OS rates were 66.7% and 54.1% (p.5), respectively (Fig. 5). Among patients with unknown HER-2 status, the 1-year DFS and OS rates were 45.8% and 45.8%, respectively. No significant differences were 1 Disease-Free Survival (months) 1. p =.5 5 found among HER-2, HER-2, and HER-2 unknown patients (DFS, p.817 and OS, p.227). All these results are summarized in Table 2. DISCUSSION This study explored the long-term survival of a homogeneously treated group of patients with stage IIIB breast cancer who achieved pcr at the primary tumor site or in the axilla after completing multidisciplinary treatment. The primary endpoints of this study were DFS and OS rates in relation to the presence of residual tumors in the breast and axilla. The secondary endpoint was the prognostic relevance of HR and HER-2 status. It is widely assumed that patients who achieve pcr have significantly better DFS and OS rates than patients with pcr, regardless of their HR status [13], HER-2 status [14], or the treatment regimen [15]. Although most clinical trials have demonstrated high rates of clinical response to primary chemotherapy, pcr occurs in only a minority of patients with stage IIIB breast cancer. Patients with pcr at the time of definitive surgery are at high risk for early recurrence and death because of the presence of viable micrometastases. Emerging data suggest that failure to achieve a pcr identifies a heterogeneous group of patients with different risks for recurrence and death, even if they receive the same neoadjuvant chemotherapy. Several studies have investigated numerous prognostic factors in the neoadjuvant setting. Residual tumor burden and, especially, the number of involved nodes have been considered powerful independent predictors of DFS and OS [9]. Molecular markers, including ER, PgR, and HER-2, are considered to be predictive and prognostic factors in the neoadjuvant setting Overall Survival (months) Figure 5. Kaplan Meier estimates of disease-free survival (A) and overall survival (B) rates according to human epidermal growth factor receptor (HER)-2 status in patients who fail to achieve a pathological complete response. 1 1

7 Ionta, Atzori, Deidda et al. 157 Table 2. Analysis of pathological features predicting DFS and OS in patients who fail to achieve a pathological complete response n of 1-Yr DFS estimate 1-Yr OS estimate patients % 95% CI p-value % 95% CI p-value Overall % 5 5.3% ER % % ER % % 5 44 PgR % % PgR % % HR % % HR % % 33 ER PgR % % HR % % 33 HER % % HER % % 39 7 HER-2 unknown % % TN % % non-tn % % 5 79 TN % % HR HER % % B A % % B A % % B B A % % B A % % pln % % pln % % pln % % pln % % 14 Abbreviations: A, axilla; B, breast; CI, confidence interval; DFS, disease-free survival; ER, estrogen receptor; HER, human epidermal growth factor receptor; OS, overall survival; PgR, progesterone receptor; pln, pathological lymph nodes; TN, triple negative. [16, 19]. Carey et al. [14] reported that, among the TN and HR patients who achieved pcr, there was a higher rate of early relapse. Bidard et al. [2] reported that only PgR status, but not ER status, was an independent factor predicting OS, whereas HER-2 status did not predict DFS or OS. Keam et al. [21] found that the TN phenotype was associated with shorter survival. Colleoni et al. [22] reported that HER-2 overexpression and HR status were significant predictors of a worse 5-year DFS rate. However, these markers are often contradictory and not conclusive because of heterogeneous patient populations in terms of initial stage, response to treatment, and the different chemotherapeutic regimens employed [22 ]. Several trials have evaluated the role of pcr, but few studies have investigated the relationship between outcome and clinical-pathological prognostic features in pcr patients [14, 17, 19, 22]. In our study, focused on a homogeneously treated population achieving pcr, the overall 1-year DFS and OS rates were 37.6% and 5.3%, respectively. As expected, pcr patients had significantly lower DFS and OS rates than pcr patients. The median times from the start of primary chemotherapy to first recurrence and death were 17 months (range, months) and 4 months (range, months), respectively. These results are consistent with other studies [12, 26, 27]. Patients with residual disease in the axilla with or without residual disease in the breast had lower DFS and OS rates than patients with residual disease in the breast only. This difference was statistically significant at the 5-year follow-up, but lost its significance at the 1-year follow-up. Nevertheless, the DFS and OS rates decreased with increasing numbers of axillary lymph nodes involved; this effect was most dramatic in patients with

8 158 Long-Term Outcomes in Stage IIIB Breast Cancer 1 involved nodes. These findings are consistent with those of other trials [3, 9, 27, 28]. Despite the significantly higher incidence of pcr achieved in patients with HR, HR HER-2 positive, and TN disease [14, 21, 29], this population has a poor prognosis, using gene expression profiling [14 3] as well as IHC assessment [31]. This paradoxic effect might be explained by the high rates of relapse and death among patients with residual disease after primary chemotherapy [31]. In our study, involving only pcr patients, HR status was the most powerful predictor of worse DFS and OS, whereas patients with the TN phenotype had a nonsignificantly lower DFS rate and only a marginally significant lower OS rate. These findings might be attributable to the long-term follow-up in our study. In fact, as reported by several authors [14, 31], patients with TN and residual disease after neoadjuvant chemotherapy had significantly worse DFS and OS than non-tn patients only in the first 3 4 years. TN patients were much more likely to experience recurrence and death during the first 3 4 years following primary chemotherapy, with rapid declines thereafter. Similar results were reported by Dent et al. [32] in a large series of patients treated in the adjuvant setting: TN patients had a significantly greater likelihood of distant recurrence and death within 5 years of diagnosis, but not thereafter. No recurrence occurred after 8 years of followup. In the TN patients in our series, the median time to first recurrence was precocious (13 months) and no recurrence occurred after 8 months of follow-up, identifying a very rapid aggressive phenotype. Dent et al. [32] also reported that, among the non-tn group, the recurrence risk seemed to be constant over the entire follow-up period (17 years). In the non-tn group in our series, the risk for recurrence was steady and continued for 122 months after primary chemotherapy, consistent with the results of Dent et al. [32]. Moreover, the lack of a significant difference between the TN and non-tn groups might be attributable to the extremely heterogeneous composition of the non-tn group, which included several subgroups according to ER, PgR, and HER-2 expression, with different prognoses. In our series, the HR HER-2 subgroup had the best outcome, but we cannot exclude what other subgroups might have had the worst outcome, as demonstrated by Carey et al. [14] and Liedtke et al. [31]. HER-2 status alone, in the absence of trastuzumab, did not appear to significantly affect clinical outcomes. In fact, no statistically significant difference was observed in terms of DFS and OS between patients with HER-2 and HER-2 tumors. These findings are consistent with those reported by other authors [31, 33]. Liedtke et al. [31], in a larger series of patients, reported that patients with HER-2 overexpression showed a trend toward a shorter progression-free survival time (p.71), but there was no significant effect in terms of OS (p.89). Dawood et al. [33], in a larger series of stage IIIB breast cancer patients, observed no statistically significant differences for either DFS (p.) or OS (p.24) in the absence of trastuzumab between HER-2 and HER-2 patients. Our data suggest that HR status is a more powerful predictor of outcome than HER-2 status, as was also reported by Kaplan and Malmgren [34]. The results emerging from our analysis confirm that, on the basis of the most common prognostic factors, patients achieving pcr can be divided into two different populations one with a low risk and one with a high risk for recurrence and death. Expression array analysis of samples from the National Surgical Adjuvant Breast and Bowel Project B-27 trial demonstrated segregation of the non-pcr group into low-risk and high-risk categories, with differential survival outcomes [35]. It is interesting to note that our findings were obtained by standard IHC analyses. In our series, patients with four or more involved axillary nodes and HR tumors represented the groups at highest risk for recurrence and for recurrence and death, respectively. The failure of the breast tumor tissue or involved lymph nodes to respond to primary chemotherapy might reflect both primary pharmacological resistance and transient kinetic resistance in occult distant micrometastases, which correlates with the presence of unresponsive, nongrowing cells (i.e., viable dormant cells) [36]. Indeed, tumor recurrence may arise from resistant cancer cells that survive previous neoadjuvant and adjuvant regimens and continue to progress. Most recurrences occur within 3 5 years of initial treatment, with an early peak of recurrence around months following surgery [11, 12]. Recurrence synchronization has been shown to occur after surgery when it is followed by chemotherapy [11] or when surgery follows chemotherapy in the neoadjuvant setting [26]. Currently, after optimal multimodal treatment, no further standard therapy exists for patients who fail to achieve a pcr. In a trial from the M.D. Anderson Cancer Center, 11 patients with pcr after primary vincristine, doxorubicin, cyclophosphamide, and prednisone (VACP) therapy and surgery were randomized to further treatment with VACP or noncrossresistant treatment with vinblastine, methotrexate, and prednisone (VbMP). There was a trend toward better outcomes in the VbMP arm than in the VACP arm (OS rate, 65% versus 47%; p.6) [37]. An ongoing multicenter pilot trial at the Dana-Farber Cancer Institute is evaluating the feasibility and toxicity of 1 year of angiogenesis inhibitor treatment with or without chemotherapy for breast cancer patients with a pcr after neoadjuvant chemotherapy [38]. The present study represents one of the few works focusing on pcr patients and their outcomes in relation to tumor burden and HR and HER-2 status. Although there

9 Ionta, Atzori, Deidda et al. 159 REFERENCES 1 Singletary SE, Allred C, Ashley P et al. Revision of the American Joint Committee on Cancer staging system for breast cancer. J Clin Oncol 22; 2: Valero V, Buzdar AU, Hortobagyi GN. Locally advanced breast cancer. The Oncologist 1996;1: Kuerer HM, Newman LA, Smith TL et al. Clinical course of breast cancer patients with complete pathological primary tumor and axillary lymph node response to doxorubicin-based neoadjuvant chemotherapy. J Clin Oncol 1999;17: Buzdar AU, Singletary SE, Booser DJ et al. Combined modality treatment of stage III and inflammatory breast cancer. M.D. Anderson Cancer Center experience. Surg Oncol Clin N Am 1995;4: Estévez LG, Gradishar WJ. Evidence-based use of neoadjuvant taxane in operable and inoperable breast cancer. Clin Cancer Res 24;1: Smith IC, Heys SD, Hutcheon AW et al. Neoadjuvant chemotherapy in breast cancer: Significantly enhanced response with docetaxel. J Clin Oncol 22;2: Arnould L, Arveux P, Couturier J et al. Pathologic complete response to trastuzumab-based neoadjuvant therapy is related to the level of HER-2 amplification. Clin Cancer Res 27;13: Roche H, Viens P, Biron P et al. High-dose chemotherapy for breast cancer: The French PEGASE experience. Cancer Control 23;1: Abrial SC, Penault-Llorca F, Delva R et al. High prognostic significance of residual disease after neoadjuvant chemotherapy: A retrospective study in 71 patients with operable breast cancer. Breast Cancer Res Treat ;94: McCready DR, Hortobagyi GN, Kau SW et al. The prognostic significance of lymph node metastases after preoperative chemotherapy for locally advanced breast cancer. Arch Surg 1989;124: Demicheli R, Miceli R, Brambilla C et al. Comparative analysis of breast was a small number of cases reported in this study, it is extremely important to note that the entire population in our study was composed of stage IIIB T4 patients who received the same chemotherapy regimen. This is the main difference between our study and the other larger series reported above [13, 14, 17, 22, 31, 33], whose patient populations were heterogeneous in terms of stage (I, II, IIIA, IIIB) and treatment (anthracycline based, anthracycline plus taxane based, and others). Moreover, our results were reported after a long-term follow-up approaching 1 years, whereas the results of the majority of the other studies were reported after a shorter follow-up ( 5 years). Interestingly, although the patient number was low in our study, the results appear consistent with those of larger series. In summary, our findings suggest that, in pcr stage IIIB patients, HR status was the strongest predictor of long-term poor outcome, in contrast to patients with HR and ER PgR tumors, who appear to be the group with the best prognosis. This finding might be attributable to the benefit derived by adjuvant endocrine therapy. HER-2 status, in the absence of trastuzumab, doesn t appear to affect outcome. Although the TN phenotype does not seem to strongly affect outcome, it has an aggressive behavior with very early relapse and death. This might be a result of the absence of effective targeted therapies and the intrinsic nature of greater aggressiveness [14]. On the basis of the findings presented here, we believe that pcr stage IIIB patients with HR and TN disease should be considered for innovative therapeutic approaches following primary treatment. AUTHOR CONTRIBUTIONS Conception/Design: Maria Teresa Ionta, Bruno Massidda Provision of study material or patients: Maria Teresa Ionta, Bruno Massidda Collection and/or assembly of data: Maria Teresa Ionta, Bruno Massidda, Francesco Atzori, Maria Cristina Deidda, Valeria Pusceddu, Sergio Palmeri, Barbara Frau, Monica Murgia, Michela Barca Data analysis and interpretation: Maria Teresa Ionta, Bruno Massidda, Luigi Minerba Manuscript writing: Maria Teresa Ionta, Bruno Massidda, Francesco Atzori Final approval of manuscript: Maria Teresa Ionta, Bruno Massidda, Francesco Atzori, Luigi Minerba, Maria Cristina Deidda, Valeria Pusceddu, Sergio Palmeri, Barbara Frau, Monica Murgia, Michela Barca The authors thank Dr. Ettore Bichisao, Medical Officer, Italian Trials in Medical Oncology and Facchetti Foundation, Istituto Nazionale Tumori, for his assistance with the language and style. cancer recurrence risk for patients receiving or not receiving adjuvant cyclophosphamide, methotrexate, fluorouracil (CMF). Data supporting the occurrence of cures. Breast Cancer Res Treat 1999;53: Saphner T, Tormey DC, Gray R. Annual hazard rates of recurrence for breast cancer after primary therapy. J Clin Oncol 1996;14: Guarneri V, Broglio K, Kau SW et al. Prognostic value of pathologic complete response after primary systemic chemotherapy in relation to hormone receptor status and other factors. J Clin Oncol 26;24: Carey LA, Dees EC, Sawyer L et al. The triple negative paradox: Primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res 27; 13: Bear HD, Anderson S, Brown A et al. The effect on tumor response of adding sequential preoperative docetaxel to preoperative doxorubicin and cyclophosphamide: Preliminary results from National Surgical Adjuvant Breast and Bowel Project Protocol B-27. J Clin Oncol 23;21: Bonnefoi H, Diebold-Berger S, Therasse P et al. Locally advanced/inflammatory breast cancers treated with intensive epirubicin-based neoadjuvant chemotherapy: Are there molecular markers in the primary tumour that predict for 5-year clinical outcome? Ann Oncol 23;14: Hennessy BT, Hortobagyi GN, Rouzier R et al. Outcome after pathologic complete eradication of cytologically proven breast cancer axillary node metastases following primary chemotherapy. J Clin Oncol ;23: Sataloff DM, Mason BA, Prestipino AJ et al. Pathologic response to induction chemotherapy in locally advanced carcinoma of the breast: A determinant of outcome. J Am Coll Surg 1995;18: Symmans WF, Peintinger F, Hatzis C et al. Measurement of residual breast cancer burden to predict survival after neoadjuvant chemotherapy. J Clin Oncol 27;: Bidard FC, Matthieu MC, Chollet P et al. p53 status and efficacy of primary anthracyclines/alkylating agent-based regimen according to breast cancer molecular classes. Ann Oncol 28;19:

10 16 Long-Term Outcomes in Stage IIIB Breast Cancer 21 Keam B, Im SA, Kim HJ et al. Prognostic impact of clinicopathologic parameters in stage II/III breast cancer treated with neoadjuvant docetaxel and doxorubicin chemotherapy: Paradoxical features of the triple negative breast cancer. BMC Cancer 27;7: Colleoni M, Bagnardi V, Rotmensz N et al. A risk score to predict diseasefree survival in patients not achieving a pathological complete remission after preoperative chemotherapy for breast cancer. Ann Oncol 29;2: Honkoop AH, Van Diest PJ, De Jong JS et al. Prognostic role of clinical, pathological and biological characteristics in patients with locally advanced breast cancer. Br J Cancer 1998;77: Allred DC, Harvey JM, Berardo M et al. Prognostic and predictive factors in breast cancer by immunohistochemical analysis. Mod Pathol 1998;11: Trock BJ, Yamauchi H, Brotzman M et al. C-erbB2 as a prognostic factor in breast cancer: A meta-analysis [abstract 372]. Proc Am Soc Clin Oncol 2;19:97a. 26 Kimmick GG, Cirrincione C, Duggan DB et al. Fifteen-year median follow-up results after neoadjuvant doxorubicin, followed by mastectomy, followed by adjuvant cyclophosphamide, methotrexate, and fluorouracil (CMF) followed by radiation for stage III breast cancer: A phase II trial (CALGB 8944). Breast Cancer Res Treat 29;113: Rouzier R, Extra JM, Klijanienko J et al. Incidence and prognostic significance of complete axillary downstaging after primary chemotherapy in breast cancer patients with T1 to T3 tumors and cytologically proven axillary metastatic lymph nodes. J Clin Oncol 22;2: Klauber-DeMore N, Ollila DW, Moore DT et al. Size of residual lymph node metastasis after neoadjuvant chemotherapy in locally advanced breast cancer patients is prognostic. Ann Surg Oncol 26;13: Colleoni M, Viale G, Zahrieh D et al. Expression of ER, PgR, HER1, HER2, and response: A study of preoperative chemotherapy. Ann Oncol 28;19: Rouzier R, Perou CM, Symmans WF et al. Breast cancer molecular subtypes respond differently to preoperative chemotherapy. Clin Cancer Res ;11: Liedtke C, Mazouni C, Hess KR et al. Response to neoadjuvant therapy and long-term survival in patients with triple-negative breast cancer. J Clin Oncol 28;26: Dent R, Trudeau M, Pritchard KI et al. Triple-negative breast cancer: Clinical features and patterns of recurrence. Clin Cancer Res 27;13: Dawood S, Broglio K, Gong Y et al. Prognostic significance of HER-2 status in women with inflammatory breast cancer. Cancer 28;112: Kaplan HG, Malmgren JA. Impact of triple negative phenotype on breast cancer prognosis. Breast J 28;14: Paik S. Importance of obtaining tissue for research a case study in NSABP B27. Presented at the National Cancer Institute meeting Preoperative Therapy in Invasive Breast Cancer: Reviewing the State of the Science and Exploring New Research Directions, Bethesda, MD, March 26 27, Demicheli R. Tumour dormancy: Findings and hypotheses from clinical research on breast cancer. Semin Cancer Biol 21;11: Thomas E, Holmes FA, Smith TL et al. The use of alternate, non-crossresistant adjuvant chemotherapy on the basis of pathologic response to a neoadjuvant doxorubicin-based regimen in women with operable breast cancer: Long-term results from a prospective randomized trial. J Clin Oncol 24;22: Mayer EL, Carey LA, Burstein HJ. Clinical trial update: Implications and management of residual disease after neoadjuvant therapy for breast cancer. Breast Cancer Res 27;9: