HER-2, p53, p21 and hormonal receptors proteins expression as predictive factors of response and prognosis in locally advanced breast cancer treated with neoadjuvant docetaxel plus epirubicin combination

Background Neoadjuvant chemotherapy has been considered the standard care in locally advanced breast cancer. However, about 20% of the patients do not benefit from this clinical treatment and, predictive factors of response were not defined yet. This study was designed to evaluate the importance of biological markers to predict response and prognosis in stage II and III breast cancer patients treated with taxane and anthracycline combination as neoadjuvant setting. Methods Sixty patients received preoperative docetaxel (75 mg/m2) in combination with epirubicin (50 mg/m2) in i.v. infusion in D1 every 3 weeks after incisional biopsy. They received adjuvant chemotherapy with CMF or FEC, attaining axillary status following definitive breast surgery. Clinical and pathologic response rates were measured after preoperative therapy. We evaluated the response rate to neoadjuvant chemotherapy and the prognostic significance of clinicopathological and immunohistochemical parameters (ER, PR, p51, p21 and HER-2 protein expression). The median patient age was 50.5 years with a median follow up time 48 months after the time of diagnosis. Results Preoperative treatment achieved clinical response in 76.6% of patients and complete pathologic response in 5%. The clinical, pathological and immunohistochemical parameters were not able to predict response to therapy and, only HER2 protein overexpression was associated with a decrease in disease free and overall survival (P = 0.0007 and P = 0.003) as shown by multivariate analysis. Conclusion Immunohistochemical phenotypes were not able to predict response to neoadjuvant chemotherapy. Clinical response is inversely correlated with a risk of death in patients submitted to neoadjuvant chemotherapy and HER2 overexpression is the major prognostic factor in stage II and III breast cancer patients treated with a neoadjuvant docetaxel and epirubicin combination.


Background
Neoadjuvant or primary chemotherapy in large primary breast cancer has been used with the purpose of reducing tumor volume and promoting less aggressive surgery [1]. The disease's downstaging promotes higher conservative breast surgery rates and [2], in patients with objective clinical and pathological response, an increase in disease free and overall survival rates has been observed [3,4]. Otherwise, about 10% -35% of the patients do not benefit from this clinical approach [5,6]. The identification of this group of patients could avoid the chemotherapy side effects and point them toward other therapeutic alternatives.
Clinical and pathological criteria neither predict response to neoadjuvant chemotherapy nor accurately define tumor biology. Tumors with the same histological type, grade and clinical staging behave differently [7]. The current hypothesis is that morphologically similar tumors have distinct gene expression with distinct phenotypic profiles [8]. The development of new therapies directed at specific molecular expression has lead researchers to evaluate the usefulness of tumor protein expression as predictive factors of response.
Docetaxel, an agent that promotes microtubule stabilization and prevents depolymerization, can cause cell cycle arrest in the mitotic phase [9], and epirubicin, an anthracycline that interacts with topoisomerase II and blocks DNA transcription [10], have both currently been used in adjuvant and neoadjuvant breast cancer treatment. In a neoadjuvant setting, the combination of these two agents is among the most actively described chemotherapy regimens [11]. The study of the molecular tumor characteristics and effects of the taxane and anthracycline combination on protein modulation can provide information about the mechanism of action and play an important role in the individualization of neoadjuvant therapy.
We evaluated the expression of p53, p21 and HER-2 protein expression by immunohistochemistry (IHC) before and after neoadjuvant docetaxel plus epirubicin in locally advanced breast cancer patients and their relationship with clinical response and outcome.

Characteristics of patients and description of neoadjuvant therapy
Sixty women with a histological diagnosis of invasive breast cancer in incisional biopsy specimens were enrolled consecutively in the study. Inclusion criteria included patients with locally advanced epithelial breast cancer (AJCC clinical stage IIa, if T> 3 cm; IIb and III), Karnofski performance status greater than 70% and age under 75 years. From February 2000 to December 2002, patients received treatment with taxane and anthracycline (docetaxel 75 mg/m 2 and epirubicin 50 mg/m 2 , i.v. infusion, for one day every three weeks). No dose modification was allowed. The median cycle of chemotherapy was 3 cycles (2 -5) and the patients' median age was 50.5 years (29 -68). Thirty-one patients were at stage II and twenty-nine were at stage III. Presence of primary inflammatory breast carcinoma was an exclusion factor and invasive ductal carcinoma was the most common histological type (82.7%). Sixteen tumors were grade I (23%) and grade II and III was observed in 55.7% and 21.3% of the tumors, respectively. The median follow up time was 48 months (24 -69). Table 1 summarizes the patients' characteristics.
None of these patients had distant metastasis nor had they received prior therapy for cancer. Hepatic and renal function tests and blood cell count were carried out at each chemotherapy cycle, and no toxicity greater than 1 or 2 was observed. Table 2 describes the toxicity effects due to neoadjuvant chemotherapy. Chemotherapy was continued until tumors became operable in primary inoperable lesions or until breast conserving surgery could be performed in primary operable tumors. The patients with primary inoperable lesions who had no clinical response, defined after at least two cycles of chemotherapy, were subjected to a new incisional biopsy and given a second line of treatment consisting of breast radiation therapy (total dose of 50 Gy) combined with a CMF chemotherapy regimen (cyclophosphamide 500 mg/m 2 , methotrexate 50 mg/m 2 and 5-fluoruracil 500 mg/m 2 i.v. infusion, for 1 day, each 21/21 days) before definitive breast surgery. All procedures were carried out with the prior informed consent of the patient.

Adjuvant treatment
Adjuvant chemotherapy was offered to all patients. Regarding axillary lymph node status. The FEC regimen (5-fluoruracil 500 mg/m 2 , epirubicin 50 mg/m 2 and cyclophosphamide 500 mg/m 2 i.v. infusion, for 1 day, each 21/21 days) was administered when more than three positive axillary lymph nodes (ALN+) were present after neoadjuvant treatment. The CMF regimen (cyclophosphamide 500 mg/m 2 , methotrexate 50 mg/m 2 and 5-fluoruracil 500 mg/m 2 i.v. infusion, for 1 day, each 21/21 days) was used if there were three or less ALN+. All patients received a total of nine chemotherapy cycles. Hormonal adjuvant therapy was offered to all patients with estrogen or progesterone receptor positive tumors. Patients received tamoxifen at 20 mg PO for 60 months or until disease recurrence.
Adjuvant radiotherapy was performed in the residual breast parenchyma of all breast conserving surgery

Assessment of response to neoadjuvant chemotherapy
The pathological and clinical assessment of response to neoadjuvant chemotherapy was taken. All surgical specimens from the definitive breast procedure were submitted for pathological evaluation and were classified as having a complete response (pCR) when no residual invasive carcinoma was seen.
The clinical stage and size of the primary tumor was recorded before treatment. The measurement of the primary tumor consisted of the product of its greatest diameter and its perpendicular diameter. The clinical response was evaluated at each cycle of chemotherapy and prior to definitive surgery on day 21 of the last cycle of chemotherapy according to the product of primary tumor diameters and the axillary clinical status, and was classified as a complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD) according to standard UICC (International Union Against Cancer) criteria [12]. CR was defined as the disappearance of all clinical evidence of the tumor including the axillary site; PR was defined as a reduction of 50% or more in the sum of the products of measured lesions or an estimated decrease in tumor size of at least 50%, without the appearance of new lesions. Clinical down staging in axillary status was necessary; SD was defined as a decrease of less than 50% in the sum of the products of measured lesions, or an estimated decrease of less than 50% in lesion size or an increase of less than 25%, without the appearance of new lesions. Any measured or estimated increase greater than 25% or the appearance of new lesions was defined as PD. No change in clinical axillary status was classified as SD regardless the change in primary tumor diameter. The patients were classified into two groups: the objective response group (OR), into which all patients classified as CR or PR were placed, and the no response group (NR), containing all patients classified as SD or PD.

Immunohistochemistry
All tissue samples had been routinely fixed in 4% neutral formalin and embedded in paraffin. Briefly, 3-μm-thick sections were cut from paraffin blocks containing representative tumor samples. Paraffin sections were de-waxed in xylene, rehydrated through a series of graded alcohols, placed in 10 mM citrate buffer and submitted to heat retrieval using a vapor lock for 40 minutes. After heating, the slides were allowed to cool to room temperature and briefly washed with Tris-buffered saline. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 5 minutes. Normal serum (Novostain Super ABC kit, Novocastra, Newcastle upon Tyne, UK) was used for 30 minutes in order to block non-specific immunoassaying. Immunohistochemical staining was performed using an avidin-biotin peroxidase system (Novostain Super ABC kit, Novocastra, UK Cases of invasive ductal carcinoma, previously known to be positive for ER, PR, p21, p53 or HER-2, were used as positive controls. Negative controls for immunostaining were prepared by omission of the primary antibody.

Scoring methods
The cases were interpreted as positive for p21 if more than 5% of the tumors cells showed nuclear staining [13] and were considered positive to ER, PR and p53 protein expression if more than 10% of the tumors cells showed nuclear staining. HER-2 expression was scored according to the degree and the proportion of membrane staining according to HercepTest protocol [14,15]. HER-2 expression was negative with a score of 0 or 1+. A score of 0 was defined as no staining or membrane staining in less than 10% of tumor cells. A score 1+ comprised faint or partially-stained membrane in more of 10% of tumor tissue. Overexpression of HER-2 was scored as 2+ when weak to moderate complete membrane staining was present in more than 10% of tumor cells. A score of 3+ was interpreted as strong, complete membrane staining in more than 10% of the tumor.

Study endpoints
The primary goal was to test the correlation among p53, p21 and HER-2 protein expression and clinical response to neoadjuvant chemotherapy with docetaxel and epiru-bicin combination and to evaluate the impact of immunohistochemical and clinic-pathological parameters in disease free and overall survivals. A secondary goal was to study the ability of neoadjuvant taxane and anthracycline based chemotherapy to change tumor protein expression. We also studied the association between protein changes and the impact in disease free and overall survivals. The third objective was to evaluate the efficacy and toxicity of a neoadjuvant docetaxel and epirubicin combination. To test the treatment efficacy, we examined a clinical parameter of response and the type of surgery performed.

Statistical analysis
The McNemar test was used to evaluate the paired correlation between p53, p21 and HER-2 protein expression before and after neoadjuvant chemotherapy. Spearman's coefficient of rank correlation [16] was used to determine the correlation between axillary lymph node status and clinical response and to evaluate the correlation between clinical response and the risk of death in disease progression. The relationship between the clinical response with the patients' characteristics and histopathological patterns was evaluated with Fisher exact test and chi-squared test.
The relationship between the number of chemotherapy cycles was analyzed with the Mann-Whitney test. The disease free survival (DFS) and overall survival (OS) were analyzed by the Kaplan-Meyer curve. MedCalc version 6.16.000 and Graphpad Prism version 4 software were used for the above statistical analyses. Multivariate analysis, to test the correlation between HER-2 protein expression, clinical response, axillary lymph node status and type of surgery performed with the risk of recurrence and death to disease progression, was performed using a nominal logistic fit with JMP 6.0 SAS software. The statistical significance was established as p < 0.05.

Neoadjuvant treatment, clinical response rate and surgery
The objective clinical response (OR) was observed in 46 patients (76.6%) treated with neoadjuvant chemotherapy. The clinical complete response rate (cCR) was 15% (nine patients) and the pathological complete response rate (pCR), defined as the absence of residual invasive carcinoma, was 5% (three patients). Patients with stable and progressive disease were defined as no response patients (NR). The NR rate was 23.3% (14 patients) and progressive disease was observed in only two patients (3.3%).
Surgery could be performed in fifty-nine patients (98.3%). One patient with no response to the primary docetaxel plus epirubicin combination was submitted to second line chemo-radiation therapy. The tumor was resistant to second line treatment and she died due to metastatic disease progression before surgery could be performed. Breast conserving surgery could be performed in forty patients (66.7%). Modified Radical Mastectomy was performed in patients for whom breast-conserving therapy was not indicated. Patients with OR had a higher breast conserving surgery rate than patients with NR (73.9% vs. 35.7%; p = 0.03, chi-square test).
Axillary lymph nodes were assessed histologicaly in 59 patients undergoing axillary node dissection. The median number of positive lymph nodes in patients with OR was 0 (0 -19) and in patients with NR was 3 (0 -25). There was a positive correlation between the lack of clinical response and the number of lymph node metastases (p = 0.03; Sperman r). The response to neoadjuvant chemotherapy, the type of surgery performed, and the lymph node status in patients with OR and NR are summarized in Table 3.

Clinical and histopathological characteristics and their relationship with clinical response
We evaluated the association among clinical response, patients' characteristics (patient's age, menopausal status, number of chemotherapy cycles and clinical stage), and histological (histological grade) and immunohistochemical patterns (ER, PR, p53, p21 and HER-2 protein expression). The median age in patients with OR was 51.5 years (range 35 -67) and in the group of patients NR, the median age was 47.5 years (p = 0.3, Student t test). In OR  Table 4 shows the relationship between the clinical response and clinical, pathological and immunohistochemical patterns.

Neoadjuvant chemotherapy and p53, p21 and HER-2 proteins expression change
The expression of p53, p21 and HER-2 proteins was evaluated in biopsy specimens from 60 breast cancer patients before neoadjuvant chemotherapy. Because no or scarce residual tumor remained after chemotherapy, the evaluation of p53 and p21 protein expression was not possible in 6 patients specimens and, HER-2 protein expression was not carried out in 4 of the patient specimens. Positive HER-2 protein overexpression was observed in 11 patients (18.3%) before chemotherapy. After neoadjuvant treatment, the overexpression of HER-2 protein was observed in seven patients (12.5%). Neoadjuvant chemotherapy was not able to change HER-2 protein expression levels (p = 0.12; McNemar test). Table 5 shows immunohistochemical protein expression levels (p53, p21 and HER-2) before and after neoadjuvant chemotherapy.

P21 protein expression and its relationship with p53 and HER-2 protein expression
We analyzed the relationship between p21 expression and p53 and HER-2 protein expression in pre-chemotherapy biopsy specimens. In forty patients with p53 negative tumors, we found 16 patients with p21 positive protein expression (40%). In patients with positive p53 expression (n = 16), we identified 4 patients with p21 positive tumors (25%). The difference was not significant (Fisher's exact test; p = 0.2).
In eleven patients with HER-2 overexpression, we observed p21 positive protein expression in four patients (36.3%) and, in 49 HER-2 negative patients, the p21 expression was positive in sixteen patients (32.6%). There was no difference in the expression of p21 in such patients (Fisher's exact test; p = 1.0).

Disease free and overall survival rates
We evaluated the impact of clinical (age, clinical stage, clinical response, type of surgery), pathological (histological grade, axillary lymph node status) and immunohistochemical (ER, PR, p53, p21 and HER-2 protein expression) features on disease free and overall survival. We observed an inverse correlation between clinical response and the risk of death to disease (p = 0.02; Spearman r). In a univariate analysis, only Her-2 overexpression and two other factors related to clinical response were related to overall survival: the type of surgery (HR = 0.41, 95% IC = 0.12 to 0.96; p = 0.042) and the presence of extensive (more than six positive lymph nodes) residual disease in axillary lymph nodes (HR = 0.35, 95% IC = 0.08 to 0.78; p = 0.017).
In a multivariate analysis, only HER-2 overexpression was a predictive factor of DFS and OS (p = 0.0007 and p = 0.003, respectively). The median time to DFS was 31 months (HR = 0.35, 95% IC = 0.08 to 0.75; p = 0.014) and the median time to OS was 40 months (HR = 0.33, 95% IC = 0.06 -0.75; p = 0.016). Figures 1 and 2 show the Kaplan-Meyer curve for DFS and OS according to HER-2 protein expression, respectively.

Discussion
Neoadjuvant or induction chemotherapy is the current standard of care in locally advanced breast cancer [17]. Such treatment can reduce tumor volume, leading to less aggressive surgery and providing favorable local conditions for surgery in inoperable tumors [1,2]. About 80% of patients achieve a local benefit due to adequate tumor downstaging, however, the overall survival, when comparing neoadjuvant and adjuvant chemotherapy, is not influenced by the primary cytotoxic therapy [17].
We evaluated the response to neoadjuvant chemotherapy in 60 locally advanced breast cancer patients. The objective clinical response rate was 76.6%, with a complete pathological response in 5% of patients. Breast conserving therapy was possible in forty patients (66.6%) and in only one patient could the surgery not be performed due to the lack of response to neoadjuvant chemotherapy. Other studies reported a higher complete pathological response rate ranging from 8% to 10.5%, using similar chemotherapy scheme [16,18]. However, the breast conserving surgery rate was higher in our study. They reported a breast conserving surgery rate of 16% and 15%, respectively, in contrast to the study of Ganem et al who reported a breast conserving surgery rate of 69% [19]. This discordant result should be attributed to the enrollment of different number of patients with primary inoperable tumors among these studies.
The clinical and pathological response to neoadjuvant therapy in patients with large breast tumors has been described as an independent prognostic factor [3,4]. We observed similar results. In our study, patients who achieved an objective response had a better prognostic than patients with stable or progressive disease after induction chemotherapy. The observation that patients with extensive lymph node metastasis after neoadjuvant therapy are more propitious to develop metastatic disease and die due to disease progression than patients with no extensive axillary metastasis supports the hypothesis that residual loco regional disease after neoadjuvant chemotherapy is a clinical marker of chemotherapy resistance.
The neoadjuvant therapy, considered to be an in vivo model to test chemotherapy sensitivity, is helpful to assess biological tumor behavior and to evaluate chemotherapy mechanisms of function and its relationship to drug resistance. In a previous study, we observed that the induction of apoptosis is an important predictor of response to chemotherapy in breast cancer and that such an event is influenced by protein tumor expression patterns [7]. In the present study, we evaluated the correlation among patients' characteristics, immunohistochemical expression of hormonal receptors (ER and PR), p53, p21 and HER-2 protein expression and the clinical and pathological response to a neoadjuvant combination of docetaxel and epirubicin. No parameter studied was able to predict response to neoadjuvant chemotherapy.
Changes in protein expression after neoadjuvant treatment have been reported [20,21]. Such observation suggests that chemotherapy can alter the tumor gene expression and that this event is a important point in drug sensitivity [7]. In the present study, we confirmed the reduction in p53 protein expression, as well in p21 pro-

Conclusion
We have concluded that the clinical response to neoadjuvant chemotherapy is a relevant prognostic factor in patients with locally advanced breast cancer treated with a neoadjuvant docetaxel and epirubicin combination. The value of biological makers as predictive factor of response to neoadjuvant therapy is still not clear. On the other hand, HER-2 overexpression is the major relevant phenotype in breast cancer progression regardless the clinical response to neoadjuvant chemotherapy. HER-