- Email: [email protected]
Apoptosis induced by neoadjuvant chemotherapy in breast cancer
Pathology (February 2006) 38(1), pp. 21–27
ANATOMICAL PATHOLOGY
Apoptosis induced by neoadjuvant chemotherapy in breast cancer DANIEL GUIMARA˜ES TIEZZI*{, JURANDYR MOREIRA DE ANDRADE*, FRANCISCO JOSE´ CAˆNDIDO DOS REIS*, HEITOR RICARDO COSISKI MARANA*, ALFREDO RIBEIRO-SILVA{, MARCELO GUIMARA˜ES TIEZZI§ AND ANTOˆNIO PLA´CIDO PEREIRA Departments of *Gynecology and Obstetrics and {Pathology, School of Medicine of Ribeira˜o Preto, University of Sa˜o Paulo, {Department of Gynecology and Obstetrics, School of Medicine of Presidente Prudente – UNOESTE, and §Presidente Prudente Laboratory of Pathology and Cytopathology, S/A, Brazil
Summary Aim: To evaluate the relationship between apoptosis induced by chemotherapy and clinical response in breast cancer. Methods: Apoptosis index (AI), mutant p53 and Bcl-2 protein expression were evaluated in 44 breast tumour samples from patients submitted to neoadjuvant chemotherapy. Objective response (OR) to primary chemotherapy was observed in 37 patients (84%) and no response (NR) in seven. AI was measured by the rate of apoptotic cells identified using morphological criteria. p53 and Bcl-2 protein expression were evaluated using an immunoperoxidase staining technique. Results: The median AI change observed between prechemotherapy AI and post-chemotherapy AI was 0.84 in the OR group and 0.01 in the NR group, (rho50.4; p50.006). There was no change in Bcl-2 protein expression following chemotherapy. In the OR group, p53 protein expression was positive in 41.6% of patients before and in 22.2% after chemotherapy (difference516.6%; p50.03). No change was detected in the NR group. Conclusion: A positive correlation was found between the increase in AI and clinical response to neoadjuvant chemotherapy in locally advanced breast cancer.
apparent that the response to neoadjuvant chemotherapy is dependent on the molecular biology of the tumour, this correlation still remains to be defined.4 Repeat biopsy has enabled investigators to assess the biological effects of neoadjuvant chemotherapy on breast cancer tissue and this methodology was used as an investigational model of tumour sensitivity that facilitated the biological study of cancer. Several authors have studied apoptosis induction; however, when fine-needle aspiration is the methodology used to obtain samples of tumour cells, this may lead to errors, such as the sampling of necrotic areas.5,6 The aim of this study was to obtain information on intratumoural apoptosis and biological markers, to study their relationship with the response to induction chemotherapy, and to test the hypothesis that inhibition of chemotherapy-induced apoptosis by Bcl-2 and p53 mutation is a mechanism of resistance in locally advanced breast cancer.
Key words: Apoptosis, breast cancer, immunohistochemical, neoadjuvant chemotherapy, p53 protein.
Forty-four breast cancer patients who had been submitted to neoadjuvant chemotherapy were admitted consecutively to one of the two separate phases of the study. From January 1994 to December 1996, 22 patients received treatment with FEC (5-fluorouracil 500 mg/m2, epirubicin 50 mg/ m2 and cyclophosphamide 500 mg/m2, IV infusion, for 1 day every 3 weeks; FEC group). The other 22 patients who were submitted to treatment from January 2000 to December 2001 received anthracycline and taxane (docetaxel 75 mg/m2 and epirubicin 50 mg/m2, IV infusion, for 1 day every 3 weeks; DE group). None of these patients had distant metastasis nor had they received prior therapy for breast 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. Chemotherapy was continued until tumours became operable in locally advanced disease or until breast conserving surgery could be performed in primary operable tumours. The patients who had no clinical response, defined after at least two cycles of chemotherapy, were submitted to a new incisional biopsy. This procedure was carried out with the prior informed consent of the patient. This study was approved by the local Ethics Committee in accordance with the ethical guidelines of the 1975 Declaration of Helsinki, revised 1983. Patients gave their written informed consent prior to enrollment.
Received 23 July, revised 27 September, accepted 4 October 2005
INTRODUCTION Neoadjuvant or induction chemotherapy has been used in the management of patients with large tumours (T.3 cm) and locally advanced breast cancer. Such treatment is administered to try to reduce the size of the primary tumour, increase the likelihood of breast conservation and abolish occult systemic metastasis, thereby improving survival rates.1 The response (both clinical and pathological) of the breast cancer to neoadjuvant chemotherapy is correlated with survival, and predictive factors need to be established in order to identify those groups of patients that would benefit from this kind of treatment.1 The assessment of clinical factors such as age, menopausal status and tumour size, has proved to be an inadequate predictor of the patient’s response to chemotherapy.2,3 Although it is
PATIENTS AND METHODS Characteristics of patients and description of therapy
Assessment of response to neoadjuvant chemotherapy Clinical staging and size of the primary tumour were recorded before treatment. The measurement of the primary tumour consisted of the
ISSN 0031-3025 printed/ISSN 1465-3931 # 2006 Royal College of Pathologists of Australasia DOI: 10.1080/00313020500465315
22
Pathology (2006), 38(1), February
TIEZZI et al.
product of its greatest diameter 6 its perpendicular diameter. The clinical response of the tumour was evaluated at each cycle of chemotherapy and prior to definitive surgery on day 21 of the last cycle of chemotherapy. The clinical response to treatment, measured and evaluated two-dimensionally, was classified as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD) according to World Health Organization (WHO) criteria. CR was defined as the disappearance of all clinical evidence of the tumour; PR was defined as a reduction of 50% or more in the sum of the products of measured lesions or an estimated decrease in tumour size of at least 50%, without the appearance of new lesions; 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. 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. Clinical response Thirty-seven patients (84%) were classified as OR. In the FEC group, four patients presented stable disease and one presented progressive disease. In the DE group, only two patients presented stable disease (p50.41). The patients’ characteristics are summarised in Table 1. Tissue processing Tissue was collected during incisional biopsy and definitive locoregional surgery following neoadjuvant treatment. Tissue samples were fixed in 10% formaldehyde for 24 h and then embedded in paraffin blocks. Histological diagnosis was made using light microscopy examination of tissue sections stained with H&E. The histological diagnosis of breast carcinoma was confirmed by the Department of Pathology of the same institution in which the study was being carried out. Immunohistochemistry for Bcl-2 protein All tissue samples were routinely fixed in 4% neutral formalin and embedded in paraffin. Briefly, 3-mm-thick sections were cut from paraffin blocks containing representative tumour 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 vapour lock for 40 min. After heating, the slides were allowed to cool to room temperature and were briefly washed with Tris-buffered saline. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 5 min. Normal serum (Novostain Super ABC kit; Novocastra, UK) was used for 30 min in order to block non-specific immunostaining. Immunohistochemical staining was performed using an avidin–biotin peroxidase system (Novostain Super ABC kit; Novocastra). Sections were then incubated overnight at room temperature with a Bcl-2 primary antibody (clone Bcl2/100/D5; 1:50; Novocastra). Following washes in PBS, biotinylated universal secondary antibody (Novostain Super ABC kit; Novocastra) was applied for 30 min. TABLE 1 Patients’ characteristics according to tumour response
No. cases Age, median (range) Type of chemotherapy FEC DE Cycles, median (range) Tumour size T2 T3 T4
OR
NR
37 49 (32–63)
7 44 (29–60)
17 20 3 (2–5)
5 2 3 (2–4)
8 20 9
1 2 4
p
0.12 0.41 0.73 0.22
OR, objective response; NR, no response; FEC, fluorouracil/epirubicin/ cyclophosphamide; DE, docetaxel/epirubicin.
The sections were incubated with the avidin–biotin complex reagent (Novostain Super ABC kit; Novocastra) for 30 min and developed with 3,39-diaminobenzidine tetrahydrochloride (DAB) in PBS, pH 7.5, containing 0.036% hydrogen peroxide for 5 min. Light Mayer’s haematoxylin was applied as a counterstain. The slides were then dehydrated in a series of ethanols and mounted with Permount (Fischer, USA). A normal tonsil was used as the positive control for Bcl-2. Negative controls for immunostaining were prepared by omission of the primary antibody. The percentage of positive cells rather than staining intensity was used for analysis of data. To assess the distribution of Bcl-2 staining in tumours, four immunohistochemical (IHC) scoring categories were used for classification according to the percentage of Bcl-2-positive cells: 0, none; 1, ,10%; 2, 11–50%; 3, 51–90%; and 4, .90%. Staining intensity was divided into four score categories: 0, no cytoplasmic staining; 1, weak cytoplasmic staining intensity; 2, moderate cytoplasmic staining intensity; 3, strong cytoplasmic staining intensity. The categories of intensity and proportion of staining were then added together to obtain a total score (range 0–7). The total score (percentage of staining + intensity of staining) was then divided into three groups: 0–2, no immunostaining; 3–5, weak immunostaining (a score of 5 indicated a cytoplasmic staining intensity of 1 and 4 percentage of staining); and 5–7, strong immunostaining (a score of 5 indicated a cytoplasmic staining intensity of 2 and 3 percentage of staining). For the purpose of statistical analysis, tumours were considered to be positive for Bcl-2 if the total score indicated strong immunostaining.7 Figure 1 shows different Bcl-2 protein staining patterns. Immunohistochemistry for p53 protein Immunoassay was carried out using the avidin–biotin peroxidase complex system (ABC). In brief, 4-mm sections were cut, mounted on poly-L-lysinecoated slides, deparaffinised in xylol and cleaned in absolute ethanol. After treatment with 0.3% hydrogen peroxide in methanol to diminish endogenous peroxidase activity, the slides were washed several times in PBS. Immunostaining was preceded by incubation in 10 mM citrate buffer (200 mL/20 slides) for 7 min at the highest potency of a domestic microwave. The primary monoclonal antibody used was mouse anti-human p53 (clone DO-7; 1:100; Dako, Denmark). Subsequently, the slides were washed in PBS overlaid with biotinylated anti-mouse secondary antibody for 30 min at room temperature. After three additional washes, the slides were incubated with ABC complex for 60 min at room temperature. The counterstain used was Harris’ haematoxylin. Positive control was a breast cancer tissue sample with known p53 mutation and immunoreactivity. Negative control was carried out by omission of the primary antibody. Positivity was indicated by the presence of dark brown nuclear staining in more than 10% of neoplastic cells.8 Figure 2 shows an invasive ductal carcinoma of the breast with positive p53 protein expression. Assessment of the apoptotic index (AI) Assessment of the AI was made by light microscopy in slides stained with H&E using standard morphological criteria: condensation of nuclear chromatin, nuclear fragmentation, separation of the surrounding cells and intense cytoplasmic eosinophilia.9,10 One thousand tumour cells were counted in each slide. The AI was expressed as the ratio of the number of apoptotic cells and the total number of tumour cells examined, as previously described.11–14 Figure 3 shows the morphology of apoptosis in H&E stained slides. Statistical analysis Pretreatment characteristics (age, type of chemotherapy, number of chemotherapy cycles and tumour size) were analysed as possible predictors of response using Fisher’s exact test and the Mann–Whitney U test. McNemar’s test was used to evaluate paired correlation between Bcl-2 and p53 protein expression before and after chemotherapy. The change in AI (AI before and after chemotherapy) was analysed using the paired t-test. Spearman’s coefficient of rank correlation (rho) and Kendall’s tau correlation were used to establish the relationship between AI and Bcl-2 and p53 protein expression and clinical response. The change in AI (AIC) was defined as the difference between AI before and after chemotherapy.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
23
Fig. 1 Immunoexpression of Bcl-2 protein in invasive ductal carcinoma of the breast. The slides show different IHC scores based both on the intensity of staining and the percentage of cells stained. (a) Weak cytoplasmic staining is observed in 100% of the tumour cells. The scoring category is 4, weak immunostaining (6100). (b) Moderate cytoplasmic staining. Note that the staining is observed in less than 50% of tumour cells. The scoring category is 4, weak immunostaining (6100). (c) Moderate cytoplasmic staining is observed in 100% of tumour cells. The scoring category is 6, strong immunostaining (6100). (d) Strong cytoplasmic staining intensity in 100% of tumour cells. The scoring category is 7, strong immunostaining (6400). When a score of 5 or 6 was obtained, this was considered Bcl-2 protein over-expression.
The correlation between AIC and clinical response was analysed using Spearman’s coefficient of rank correlation. The capacity of the change in AI to predict clinical response was assessed using the receiver operating characteristic (ROC) curve. The MedCalc software program version 6.16.000 (MedCalc Software, Belgium), and Prisma version 3.0 (GraphPad Software, USA), were used in the statistical calculations. Statistical significance was established as p,0.05.
RESULTS The clinical response to neoadjuvant chemotherapy was analysed and correlated with the induction of apoptosis and the expression of Bcl-2 and mutant p53 proteins.
Fig. 2 Imunoexpression of p53 protein in invasive ductal carcinoma of the breast. Note the dark brown nuclear staining in more than 10% of tumour cells (6400).
Bcl-2 and p53 protein expression The expression of Bcl-2 and p53 protein was evaluated in biopsy samples taken from 43 patients with locally advanced breast cancer prior to neoadjuvant chemotherapy. We identified 19 cases with positive expression of Bcl-2 protein (44.2%) and 16 cases of p53 protein over-expression (37.2%).
24
Pathology (2006), 38(1), February
TIEZZI et al.
p53 protein expression and apoptotic index p53 protein expression and the AI were evaluated before and after chemotherapy to investigate a possible correlation. No correlation was found prior to chemotherapy (rho520.03, p50.81, n543; Spearman’s rank correlation coefficient); however, a statistically significant inverse relationship was seen between AI and p53 protein expression following chemotherapy (rho520.36, p50.01, n542).
Fig. 3 Morphology of apoptosis in an invasive ductal carcinoma of the breast after induction chemotherapy. The slide was obtained from a patient who had an objective response. There is post-chemotherapy fibrosis, and invasive carcinoma cells are present in loose collagenous tissue. The arrows show apoptotic cells. Note the nuclear condensation and fragmentation (upper left) and cell shrinkage. The cytoplasm exhibits intense eosinophilia (6400).
Expression of Bcl-2 protein and clinical response Bcl-2 expression and its association with the clinical response to induction chemotherapy were evaluated in the biopsy specimens of 43 patients before and after treatment in the two groups: the objective response group (OR), and the no response group (NR). In patients with positive expression of Bcl-2 protein, the OR rate was 89.4% (17/19), and in patients with negative expression, the OR rate was 79.1% (19/24). This difference was not statistically significant (x2 test; p50.43). In the OR group, Bcl-2 was positive in 47.2% (17/36) before and in 41.1% (14/34) after chemotherapy (McNemar test; p50.75). Bcl-2 positive expression was seen in 28.5% of tumours (2/7) both before and after chemotherapy in the NR group (McNemar test; p51.0).
Induction of apoptosis and clinical response Morphologically, apoptotic cells were infrequent in pretreatment tissue. Median AI was 0.82 (0.12–2.49) before chemotherapy and 1.61 (0.32–4.31) following chemotherapy. Neoadjuvant chemotherapy was able to induce apoptosis in breast cancer cells (paired t-test; p,0.0001).
Expression of p53 protein and clinical response p53 protein expression was evaluated before and after treatment in the biopsy samples of 43 patients in the two different groups, the objective response group (OR) and the no response group (NR). The OR rate in patients with positive p53 protein expression was 93% (15/16), while in patients with negative p53 protein expression it was 77.7%. There was no statistically significant difference between OR rates (x2 test; p50.22). In the OR group, p53 was positive in 41.6% (15/36) before and 22.2% (8/36) after chemotherapy (McNemar test; p50.03; difference516.6%; 95% confidence interval [95%CI]51.3–16.6%). p53-positive expression was seen in 14.25% (1/7) before and in 42.8% (3/7) of tumours after chemotherapy in the NR group (McNemar test; p50.5). Table 2 summarises the relationship between biological characteristics and the clinical response to neoadjuvant chemotherapy.
Apoptotic index and Bcl-2 protein expression We evaluated the rank correlation between AI and Bcl-2 protein expression before and after chemotherapy. A statistically significant inverse relationship was seen between AI and Bcl-2 before chemotherapy (Kendall’s tau520.33, p50.02, n543). No such correlation was observed after chemotherapy (Kendall’s tau520.125, p50.24, n541).
Apoptotic index and clinical response AI was evaluated before and after neoadjuvant chemotherapy in 35 OR patients and seven NR patients. In the OR group, the median AI before chemotherapy was 0.88 (0.12–2.49) and after chemotherapy the median was 1.84 (0.32–4.31). An increase in apoptotic rate was seen after chemotherapy in patients with an objective response (paired t-test; p,0.0001). In the NR group, median AI
Following chemotherapy, Bcl-2 and p53 protein expression was evaluated in 42 patients. The remaining patient presented complete pathological response. Bcl-2 protein expression was positive in 16 patients (39%), while p53 protein expression was positive in 12 patients (27.9%). Neoadjuvant chemotherapy had no influence on Bcl-2 expression (McNemar test; p50.77) or p53 protein expression (McNemar test; p50.28).
TABLE 2 Relationship between Bcl-2 and p53 protein expression and clinical response in 43 breast cancer patients treated with neoadjuvant chemotherapy
OR NR
p53+ pre
p53+ post
p (dif)
Bcl-2+ pre
Bcl-2+ post
p (dif)
41.6% (15/36) 14.2% (1/7)
22.2% (8/36) 42.8% (3/7)
0.03 (16.6%) NS
47.2% (17/36) 28.5% (2/7)
41.1% (14/34) 28.5% (2/7)
NS NS
OR, objective response; NR, no response; pre, positive protein expression before chemotherapy; post, positive protein expression after chemotherapy; dif, difference (McNemar test); NS, not significant.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
25
TABLE 3 Relationship between AI and tumour response in 42 breast cancer patients treated with neoadjuvant chemotherapy
Initial AI (median) Post-chemotherapy AI (median) AI change (median)
OR (range)
NR (range)
p (rho)
0.8 (0.12–2.5) 1.8 (0.3–4.3) 0.84 (20.19–2.6)
1.2 (0.65–2.4) 1.3 (0.64–2.2) 20.01 (20.31–0.68)
NS (–0.24) 0.03 (0.33) 0.006 (0.4)
OR, objective response; NR, no response; rho, rank correlation; AI, apoptotic index; NS, not significant.
before chemotherapy was 1.21 (0.65–2.43) and 1.28 (0.64– 2.2) after chemotherapy. There was no change in apoptotic rate after chemotherapy in patients with no clinical response (paired t-test; p50.64). The relationship between AI before and after chemotherapy and clinical response was evaluated. There was no correlation between AI before chemotherapy and the prediction of clinical response (rho520.24; p50.1); however, a statistically significant relationship was seen between AI and clinical response after chemotherapy (rho50.33; p50.03). The AI change (AIC), defined as the difference between AI before and after chemotherapy, was calculated. The median AIC observed was 0.84 in the OR group and 0.01 in the NR group (rho50.40; p50.006), as summarised in Table 3. As a predictor of response, a change in AI.0.03 was related with positive and negative predictive values of 94.4% and 71.4%, respectively, and the area under the ordinary ROC curve was 0.90, indicating that AI increases over 0.03 have an accuracy of 90% in predicting response (Fig. 4).
Fig. 4 Receiver operating characteristic (ROC) curve for predicting response in 43 breast cancer patients treated with neoadjuvant chemotherapy.
DISCUSSION Neoadjuvant chemotherapy plays an important role in the management of patients with locally advanced breast cancer. The aim of neoadjuvant therapy is to reduce the tumour volume in patients before surgical resection, thus increasing the likelihood of breast conservation. The clinical response to this treatment is reported as an independent prognostic factor and patients with complete pathological response have a longer disease-free interval and overall survival.15 More recently, the expression of biological markers has been evaluated as a predictive factor of response. As previously reported by Ellis et al.,16 chemotherapy can induce apoptosis in breast carcinomas in vivo. Failure to undergo apoptosis is considered a major mechanism of chemoresistance. We investigated the apoptotic rate in locally advanced breast cancer patients treated with neoadjuvant chemotherapy. Apoptosis has distinctive biochemical and morphological features that distinguish it from necrosis. These include endonuclease activation, chromatin condensation and margination, cell shrinkage, and fragmentation.9,10 Although there are many other sensitive methods for detecting apoptotic cells, we counted apoptotic cells directly under light microscopy based on these morphological features. This technique reliably measures the apoptotic index in biopsy specimens of tumours with a heterogeneous cell population. The range of prechemotherapy AI observed in this study is consistent with previously published data.17,18 In the present study, an increase in apoptotic rate was observed following chemotherapy. There was a significant change in AI in patients with an objective response to induction chemotherapy. In patients with no clinical response, no such change was observed, suggesting that chemoresistant tumours were able to block apoptosis usually induced by cytotoxic drugs. Over-expression of bcl-2, an antiapoptotic protein, has been reported in up to 50% of all cases of breast cancer.19 In a previous study, high levels of Bcl-2 protein were able to block apoptosis and prevent tumour cells from dying in response to the severe DNA damage caused by chemotherapy. 20 We expected to find a reduction in Bcl-2 protein expression following chemotherapy in patients who had an objective response to chemotherapy; however, no such change was observed. As demonstrated in some studies, chemotherapy can reduce the level of Bcl-2 protein in MCF-7 breast cancer cells.21,22 We were unable to observe any reduction in the expression of this protein following chemotherapy, and Bcl-2 expression was found to be a stable phenotype. The change reported by other investigators, however, occurred at an early stage in the study period.22 We evaluated Bcl-2 expression 21 days after the last cycle of chemotherapy, and this delayed biopsy may explain the expression pattern observed in our study.
26
Pathology (2006), 38(1), February
TIEZZI et al.
Repeat biopsies would be necessary to confirm this finding in vivo. An inverse correlation between Bcl-2 expression and AI in pre-chemotherapy specimens was demonstrated in this study. This finding in pre-chemotherapy specimens is compatible with Bcl-2 molecular function. Tumours with an over-expression of Bcl-2 protein show a low rate of apoptosis. Over-expression of Bcl-2 protein has been postulated to promote chemotherapy resistance in solid tumours. 23–25 Objective response rates of 89.4% and 79.1% in patients with and without Bcl-2 protein over-expression, respectively, were observed in the present study. Because taxanes can result in Bcl-2 hyperphosphorylation and consequent functional inactivation,26 tumours with Bcl-2 protein over-expression may be more susceptible to taxaneinduced apoptosis. In this study, 22 patients were treated with taxane-based chemotherapy. p53 protein is a transcription factor that regulates normal cell growth by controlling genes that promote progression through the cycle and by controlling those that cause arrest in G1 when the genome is damaged. Its activation can induce apoptosis and is an important mechanism of cell execution in response to DNA damage. The loss of p53 function as a result, for example, of a mutation in the p53 gene has been reported to enhance cell resistance to a number of chemotherapeutic agents; however, controversial data have been published with respect to chemotherapy resistance and p53 protein expression in breast cancer tumour specimens.27,28 This lack of concordance may be attributed to differences in immunohistochemical techniques, study design or sample population. This mutation leads to p53 protein overexpression and can be identified by immunoperoxidase assay.29 In this study, there was a tendency towards a higher OR in patients with mutant p53 protein overexpression. The OR rate was 93% in that group of patients compared with an OR rate of 77.7% in patients with no over-expression. Some studies suggest that the DNA damage induced by cytotoxic agents may potentially favour the mutation of the p53 gene.30,31 A change in p53 protein expression after treatment in originally positive p53 tumours was observed in the present study. The positive p53 protein expression rate was 41.6% prior to and 22.2% following chemotherapy in the OR group. There was a significant decrease in the expression rate in this group of patients. In patients who had no response to neoadjuvant chemotherapy, there was no significant change in p53 protein expression rates. Only two breast cancer specimens that were negative for p53 prior to chemotherapy became positive afterwards. The observation of a reduction in p53 protein expression rate after chemotherapy in patients who had a clinically objective response may reflect a clonal selection process. This change is part of a multifactorial mechanism involved in tumour sensitivity to cytotoxic drugs that leads to an increase in the cell death/cell mitosis ratio. The finding of the inverse correlation between mutant p53 protein expression and AI following chemotherapy supports the hypothesis that a high level of mutant p53 protein can protect chemotherapy-induced cell death. AI was shown to correlate with an objective clinical response in primary chemotherapy-treated breast cancer. These data support the hypothesis that the increase in tumour
apoptotic rate correlates with a higher OR rate, and we believe that the change in AI may be representative of the level of tumour protection against chemotherapy-induced DNA damage; in other words, the greater the change in AI, the lower the resistance of the tumour cell. ACKNOWLEDGEMENTS This research was supported by a grant from Fundac¸a˜o de Amparo ao Ensino, Pesquisa e Assisteˆncia (FAEPA), Brasil. Address for correspondence: Dr D. G. Tiezzi, Hospital das Clı´nicas da Faculdade de Medicina de Ribeira˜o Preto – USP, Departamento de Ginecologia e Obstetricia, Av Bandeirantes, nu 3900, 8u andar, Ribeira˜o Preto, Sa˜o Paulo, Brazil, CEP 14049-900. E-mail: daniel_tiezzi@yahoo. com.br
References 1. Fisher ER, Wang J, Bryant J, et al. Pathobiology of preoperative chemotherapy: findings from the National Surgical Adjuvant Breast and Bowel (NSABP) protocol B-18. Cancer 2002; 95: 681–95. 2. Dieras V, Fumoleau P, Romieu G, et al. Randomized parallel study of doxorubicin plus paclitaxel and doxorubicin plus cyclophosphamide as neoadjuvant treatment of patients with breast cancer. J Clin Oncol 2004; 22: 4958–65. 3. Smith IE, Mainwaring PN. Primary chemotherapy. In: Bonadonna G, Hortobagyi GN, Gianni AM, editors. Textbook of Breast Cancer: A Clinical Guide to Therapy. London: Martin Dunitz. 1998; 104–5. 4. Faneyte IF, Schrama JG, Peterse JL, et al. Breast cancer response to neoadjuvant chemotherapy: predictive markers and relation with outcome. Br J Cancer 2003; 88: 406–12. 5. Symmans WF, Volm MD, Shapiro RL, et al. Paclitaxel-induced apoptosis and mitotic arrest assessed by serial fine-needle aspiration: implications for early prediction of breast cancer response to neoadjuvant treatment. Clin Cancer Res 2000; 6: 4610–7. 6. Hayward JL, Carbone PP, Heusen JC, et al. Assessment of response to therapy in advanced breast cancer. Br J Cancer 1977; 114: 160–3. 7. Soini Y, Puhakka A, Kahlos K, et al. Endothelial nitric oxide synthase is strongly expressed in malignant mesothelioma but does not associate with vascular density or the expression of VEGF, FLK1 or FLT1. Histopathology 2001; 39: 179–86. 8. Nagle RB, Fitzgibbons PL. Consensual interpretive guidelines for diagnostic immunohistochemistry. Am J Surg Pathol 2002; 26: 816–7. 9. Ha¨cker G. The morphology of apoptosis. Cell Tissue Res 2000; 301: 5–7. 10. Cohen JJ. Apoptosis. Immunol Today 1993; 14: 126–30. 11. Meyn RE, Stephens LC, Ang KK, et al. Heterogeneity in the development of apoptosis in irradiated murine tumours of different histologies. Int J Radiat Biol 1993; 64: 583–91. 12. Wheeler JA, Stephens LC, Tornos C, et al. ASTRO Research Fellowship. Apoptosis as a predictor of tumor response to radiation in stage IB cervical carcinoma. American Society for Therapeutic Radiology and Oncology. Int J Radiat Oncol Biol Phys 1995; 32: 1487–93. 13. Cameron DA, Keen JC, Dixon JM, et al. Effective tamoxifen therapy of breast cancer involves both antiproliferative and pro-apoptotic changes. Eur J Cancer 2000; 36: 845–51. 14. Aihara M, Scardino PT, Truong LD, et al. The frequency of apoptosis correlates with the prognosis of Gleason grade 3 adenocarcinoma of the prostate. Cancer 1995; 75: 522–9. 15. Carey LA, Metzger R, Dees EC, et al. American Joint Committee on Cancer Tumor-Node-Metastasis Stage After Neoadjuvant Chemotherapy and Breast Cancer Outcome. J Natl Cancer Inst 2005; 97: 1337–42. 16. Ellis PA, Smith IE, Mccarthy K, et al. Preoperative chemotherapy induces apoptosis in early breast cancer. Lancet 1997; 349: 849. 17. Staunton MJ, Gaffney EF. Tumor type is a determinant of susceptibility to apoptosis. Am J Clin Pathol 1995; 103: 300–7. 18. Paradiso A, Mangia A, Chiriatti A, et al. Biomarkers predictive for clinical efficacy of taxol-based chemotherapy in advanced breast cancer. Ann Oncol 2005; 16: 14–9. 19. Silvestrini R, Veneroni S, Daidone MG, Benini E, Borscchi P. The bcl2 protein: a prognostic indicator strongly related top53 protein in lymph node-negative breast cancer in patients. J Natl Cancer Inst 1994; 86: 499–504. 20. Srivastava RK, Mi QS, Hardwick JM, et al. Deletion of the loop region of Bcl-2 completely blocks paclitaxel-induced apoptosis. Proc Natl Acad Sci USA 1999; 96: 3775–80.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
21. Leung LK, Wang TT. Differential effects of chemotherapeutic agents on the Bcl-2/Bax apoptosis pathway in human breast cancer cell line MCF-7. Breast Cancer Res Treat 1999; 55: 73–83. 22. Wang Q, Maloof P, Wang H, et al. Basic fibroblast growth factor downregulates Bcl-2 and promotes apoptosis in MCF-7 human breast cancer cells. Exp Cell Res 1998; 238: 177–87. 23. Ohmori T, Podack ER, Nishio K, et al. Apoptosis of lung cancer cells caused by some anti-cancer agents (MMC, CPT-11, ADM) is inhibited by bcl-2. Biochem Biophys Res Commun 1993; 193: 30–6. 24. Bold RJ, Chandra J, McConkey DJ. Gencitabine-induced programmed cell death (apoptosis) of human pancreatic carcinoma is determined by bcl-2 content. Ann Surg Oncol 1999; 6: 279–85. 25. Eliopoulos AG, Kerr DJ, Herod J, et al. The control of apoptosis and drug resistance in ovarian cancer: influence of p53 and bcl-2. Oncogene 1995; 11: 1217–28. 26. Scatena CD, Stewart ZA, Mays D, et al. Mitotic phosphorylation of bcl-2 during normal cell cycle progression and taxol-induced growth arrest. J Biol Chem 1998; 273: 30777–84.
27
27. Erdem O, Dursun A, Coskun U, et al. The prognostic value of p53 and c-erbB-2 expression, proliferative activity and angiogenesis in nodenegative breast carcinoma. Tumori 2005; 91: 46–52. 28. Tsutsui S, Yasuda K, Higashi H, et al. Prognostic implication of p53 protein expression in relation to nuclear pleomorphism and the MIB-1 counts in breast cancer. Breast Cancer 2004; 11: 160–8. 29. Harris CC. Structure and function of the p53 tumor suppressor gene: clues for rational cancer therapeutic strategies. J Natl Cancer Inst 1996; 88: 1442–55. 30. Bottini A, Berruti A, Bersiga A, et al. p53 but not bcl-2 immunostaining is predictive of poor clinical complete response to primary chemotherapy in breast cancer patients. Clin Cancer Res 2000; 6: 2751–8. 31. Brambilla E, Gazzeri S, Lantuejoul S, et al. p53 mutant immunophenotype and deregulation of p53 transcription pathway (bcl-2, bax and waf-1) in precursor bronchial lesions of lung cancer. Clin Cancer Res 1998; 4: 1609–18.
ANATOMICAL PATHOLOGY
Apoptosis induced by neoadjuvant chemotherapy in breast cancer DANIEL GUIMARA˜ES TIEZZI*{, JURANDYR MOREIRA DE ANDRADE*, FRANCISCO JOSE´ CAˆNDIDO DOS REIS*, HEITOR RICARDO COSISKI MARANA*, ALFREDO RIBEIRO-SILVA{, MARCELO GUIMARA˜ES TIEZZI§ AND ANTOˆNIO PLA´CIDO PEREIRA Departments of *Gynecology and Obstetrics and {Pathology, School of Medicine of Ribeira˜o Preto, University of Sa˜o Paulo, {Department of Gynecology and Obstetrics, School of Medicine of Presidente Prudente – UNOESTE, and §Presidente Prudente Laboratory of Pathology and Cytopathology, S/A, Brazil
Summary Aim: To evaluate the relationship between apoptosis induced by chemotherapy and clinical response in breast cancer. Methods: Apoptosis index (AI), mutant p53 and Bcl-2 protein expression were evaluated in 44 breast tumour samples from patients submitted to neoadjuvant chemotherapy. Objective response (OR) to primary chemotherapy was observed in 37 patients (84%) and no response (NR) in seven. AI was measured by the rate of apoptotic cells identified using morphological criteria. p53 and Bcl-2 protein expression were evaluated using an immunoperoxidase staining technique. Results: The median AI change observed between prechemotherapy AI and post-chemotherapy AI was 0.84 in the OR group and 0.01 in the NR group, (rho50.4; p50.006). There was no change in Bcl-2 protein expression following chemotherapy. In the OR group, p53 protein expression was positive in 41.6% of patients before and in 22.2% after chemotherapy (difference516.6%; p50.03). No change was detected in the NR group. Conclusion: A positive correlation was found between the increase in AI and clinical response to neoadjuvant chemotherapy in locally advanced breast cancer.
apparent that the response to neoadjuvant chemotherapy is dependent on the molecular biology of the tumour, this correlation still remains to be defined.4 Repeat biopsy has enabled investigators to assess the biological effects of neoadjuvant chemotherapy on breast cancer tissue and this methodology was used as an investigational model of tumour sensitivity that facilitated the biological study of cancer. Several authors have studied apoptosis induction; however, when fine-needle aspiration is the methodology used to obtain samples of tumour cells, this may lead to errors, such as the sampling of necrotic areas.5,6 The aim of this study was to obtain information on intratumoural apoptosis and biological markers, to study their relationship with the response to induction chemotherapy, and to test the hypothesis that inhibition of chemotherapy-induced apoptosis by Bcl-2 and p53 mutation is a mechanism of resistance in locally advanced breast cancer.
Key words: Apoptosis, breast cancer, immunohistochemical, neoadjuvant chemotherapy, p53 protein.
Forty-four breast cancer patients who had been submitted to neoadjuvant chemotherapy were admitted consecutively to one of the two separate phases of the study. From January 1994 to December 1996, 22 patients received treatment with FEC (5-fluorouracil 500 mg/m2, epirubicin 50 mg/ m2 and cyclophosphamide 500 mg/m2, IV infusion, for 1 day every 3 weeks; FEC group). The other 22 patients who were submitted to treatment from January 2000 to December 2001 received anthracycline and taxane (docetaxel 75 mg/m2 and epirubicin 50 mg/m2, IV infusion, for 1 day every 3 weeks; DE group). None of these patients had distant metastasis nor had they received prior therapy for breast 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. Chemotherapy was continued until tumours became operable in locally advanced disease or until breast conserving surgery could be performed in primary operable tumours. The patients who had no clinical response, defined after at least two cycles of chemotherapy, were submitted to a new incisional biopsy. This procedure was carried out with the prior informed consent of the patient. This study was approved by the local Ethics Committee in accordance with the ethical guidelines of the 1975 Declaration of Helsinki, revised 1983. Patients gave their written informed consent prior to enrollment.
Received 23 July, revised 27 September, accepted 4 October 2005
INTRODUCTION Neoadjuvant or induction chemotherapy has been used in the management of patients with large tumours (T.3 cm) and locally advanced breast cancer. Such treatment is administered to try to reduce the size of the primary tumour, increase the likelihood of breast conservation and abolish occult systemic metastasis, thereby improving survival rates.1 The response (both clinical and pathological) of the breast cancer to neoadjuvant chemotherapy is correlated with survival, and predictive factors need to be established in order to identify those groups of patients that would benefit from this kind of treatment.1 The assessment of clinical factors such as age, menopausal status and tumour size, has proved to be an inadequate predictor of the patient’s response to chemotherapy.2,3 Although it is
PATIENTS AND METHODS Characteristics of patients and description of therapy
Assessment of response to neoadjuvant chemotherapy Clinical staging and size of the primary tumour were recorded before treatment. The measurement of the primary tumour consisted of the
ISSN 0031-3025 printed/ISSN 1465-3931 # 2006 Royal College of Pathologists of Australasia DOI: 10.1080/00313020500465315
22
Pathology (2006), 38(1), February
TIEZZI et al.
product of its greatest diameter 6 its perpendicular diameter. The clinical response of the tumour was evaluated at each cycle of chemotherapy and prior to definitive surgery on day 21 of the last cycle of chemotherapy. The clinical response to treatment, measured and evaluated two-dimensionally, was classified as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD) according to World Health Organization (WHO) criteria. CR was defined as the disappearance of all clinical evidence of the tumour; PR was defined as a reduction of 50% or more in the sum of the products of measured lesions or an estimated decrease in tumour size of at least 50%, without the appearance of new lesions; 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. 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. Clinical response Thirty-seven patients (84%) were classified as OR. In the FEC group, four patients presented stable disease and one presented progressive disease. In the DE group, only two patients presented stable disease (p50.41). The patients’ characteristics are summarised in Table 1. Tissue processing Tissue was collected during incisional biopsy and definitive locoregional surgery following neoadjuvant treatment. Tissue samples were fixed in 10% formaldehyde for 24 h and then embedded in paraffin blocks. Histological diagnosis was made using light microscopy examination of tissue sections stained with H&E. The histological diagnosis of breast carcinoma was confirmed by the Department of Pathology of the same institution in which the study was being carried out. Immunohistochemistry for Bcl-2 protein All tissue samples were routinely fixed in 4% neutral formalin and embedded in paraffin. Briefly, 3-mm-thick sections were cut from paraffin blocks containing representative tumour 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 vapour lock for 40 min. After heating, the slides were allowed to cool to room temperature and were briefly washed with Tris-buffered saline. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol for 5 min. Normal serum (Novostain Super ABC kit; Novocastra, UK) was used for 30 min in order to block non-specific immunostaining. Immunohistochemical staining was performed using an avidin–biotin peroxidase system (Novostain Super ABC kit; Novocastra). Sections were then incubated overnight at room temperature with a Bcl-2 primary antibody (clone Bcl2/100/D5; 1:50; Novocastra). Following washes in PBS, biotinylated universal secondary antibody (Novostain Super ABC kit; Novocastra) was applied for 30 min. TABLE 1 Patients’ characteristics according to tumour response
No. cases Age, median (range) Type of chemotherapy FEC DE Cycles, median (range) Tumour size T2 T3 T4
OR
NR
37 49 (32–63)
7 44 (29–60)
17 20 3 (2–5)
5 2 3 (2–4)
8 20 9
1 2 4
p
0.12 0.41 0.73 0.22
OR, objective response; NR, no response; FEC, fluorouracil/epirubicin/ cyclophosphamide; DE, docetaxel/epirubicin.
The sections were incubated with the avidin–biotin complex reagent (Novostain Super ABC kit; Novocastra) for 30 min and developed with 3,39-diaminobenzidine tetrahydrochloride (DAB) in PBS, pH 7.5, containing 0.036% hydrogen peroxide for 5 min. Light Mayer’s haematoxylin was applied as a counterstain. The slides were then dehydrated in a series of ethanols and mounted with Permount (Fischer, USA). A normal tonsil was used as the positive control for Bcl-2. Negative controls for immunostaining were prepared by omission of the primary antibody. The percentage of positive cells rather than staining intensity was used for analysis of data. To assess the distribution of Bcl-2 staining in tumours, four immunohistochemical (IHC) scoring categories were used for classification according to the percentage of Bcl-2-positive cells: 0, none; 1, ,10%; 2, 11–50%; 3, 51–90%; and 4, .90%. Staining intensity was divided into four score categories: 0, no cytoplasmic staining; 1, weak cytoplasmic staining intensity; 2, moderate cytoplasmic staining intensity; 3, strong cytoplasmic staining intensity. The categories of intensity and proportion of staining were then added together to obtain a total score (range 0–7). The total score (percentage of staining + intensity of staining) was then divided into three groups: 0–2, no immunostaining; 3–5, weak immunostaining (a score of 5 indicated a cytoplasmic staining intensity of 1 and 4 percentage of staining); and 5–7, strong immunostaining (a score of 5 indicated a cytoplasmic staining intensity of 2 and 3 percentage of staining). For the purpose of statistical analysis, tumours were considered to be positive for Bcl-2 if the total score indicated strong immunostaining.7 Figure 1 shows different Bcl-2 protein staining patterns. Immunohistochemistry for p53 protein Immunoassay was carried out using the avidin–biotin peroxidase complex system (ABC). In brief, 4-mm sections were cut, mounted on poly-L-lysinecoated slides, deparaffinised in xylol and cleaned in absolute ethanol. After treatment with 0.3% hydrogen peroxide in methanol to diminish endogenous peroxidase activity, the slides were washed several times in PBS. Immunostaining was preceded by incubation in 10 mM citrate buffer (200 mL/20 slides) for 7 min at the highest potency of a domestic microwave. The primary monoclonal antibody used was mouse anti-human p53 (clone DO-7; 1:100; Dako, Denmark). Subsequently, the slides were washed in PBS overlaid with biotinylated anti-mouse secondary antibody for 30 min at room temperature. After three additional washes, the slides were incubated with ABC complex for 60 min at room temperature. The counterstain used was Harris’ haematoxylin. Positive control was a breast cancer tissue sample with known p53 mutation and immunoreactivity. Negative control was carried out by omission of the primary antibody. Positivity was indicated by the presence of dark brown nuclear staining in more than 10% of neoplastic cells.8 Figure 2 shows an invasive ductal carcinoma of the breast with positive p53 protein expression. Assessment of the apoptotic index (AI) Assessment of the AI was made by light microscopy in slides stained with H&E using standard morphological criteria: condensation of nuclear chromatin, nuclear fragmentation, separation of the surrounding cells and intense cytoplasmic eosinophilia.9,10 One thousand tumour cells were counted in each slide. The AI was expressed as the ratio of the number of apoptotic cells and the total number of tumour cells examined, as previously described.11–14 Figure 3 shows the morphology of apoptosis in H&E stained slides. Statistical analysis Pretreatment characteristics (age, type of chemotherapy, number of chemotherapy cycles and tumour size) were analysed as possible predictors of response using Fisher’s exact test and the Mann–Whitney U test. McNemar’s test was used to evaluate paired correlation between Bcl-2 and p53 protein expression before and after chemotherapy. The change in AI (AI before and after chemotherapy) was analysed using the paired t-test. Spearman’s coefficient of rank correlation (rho) and Kendall’s tau correlation were used to establish the relationship between AI and Bcl-2 and p53 protein expression and clinical response. The change in AI (AIC) was defined as the difference between AI before and after chemotherapy.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
23
Fig. 1 Immunoexpression of Bcl-2 protein in invasive ductal carcinoma of the breast. The slides show different IHC scores based both on the intensity of staining and the percentage of cells stained. (a) Weak cytoplasmic staining is observed in 100% of the tumour cells. The scoring category is 4, weak immunostaining (6100). (b) Moderate cytoplasmic staining. Note that the staining is observed in less than 50% of tumour cells. The scoring category is 4, weak immunostaining (6100). (c) Moderate cytoplasmic staining is observed in 100% of tumour cells. The scoring category is 6, strong immunostaining (6100). (d) Strong cytoplasmic staining intensity in 100% of tumour cells. The scoring category is 7, strong immunostaining (6400). When a score of 5 or 6 was obtained, this was considered Bcl-2 protein over-expression.
The correlation between AIC and clinical response was analysed using Spearman’s coefficient of rank correlation. The capacity of the change in AI to predict clinical response was assessed using the receiver operating characteristic (ROC) curve. The MedCalc software program version 6.16.000 (MedCalc Software, Belgium), and Prisma version 3.0 (GraphPad Software, USA), were used in the statistical calculations. Statistical significance was established as p,0.05.
RESULTS The clinical response to neoadjuvant chemotherapy was analysed and correlated with the induction of apoptosis and the expression of Bcl-2 and mutant p53 proteins.
Fig. 2 Imunoexpression of p53 protein in invasive ductal carcinoma of the breast. Note the dark brown nuclear staining in more than 10% of tumour cells (6400).
Bcl-2 and p53 protein expression The expression of Bcl-2 and p53 protein was evaluated in biopsy samples taken from 43 patients with locally advanced breast cancer prior to neoadjuvant chemotherapy. We identified 19 cases with positive expression of Bcl-2 protein (44.2%) and 16 cases of p53 protein over-expression (37.2%).
24
Pathology (2006), 38(1), February
TIEZZI et al.
p53 protein expression and apoptotic index p53 protein expression and the AI were evaluated before and after chemotherapy to investigate a possible correlation. No correlation was found prior to chemotherapy (rho520.03, p50.81, n543; Spearman’s rank correlation coefficient); however, a statistically significant inverse relationship was seen between AI and p53 protein expression following chemotherapy (rho520.36, p50.01, n542).
Fig. 3 Morphology of apoptosis in an invasive ductal carcinoma of the breast after induction chemotherapy. The slide was obtained from a patient who had an objective response. There is post-chemotherapy fibrosis, and invasive carcinoma cells are present in loose collagenous tissue. The arrows show apoptotic cells. Note the nuclear condensation and fragmentation (upper left) and cell shrinkage. The cytoplasm exhibits intense eosinophilia (6400).
Expression of Bcl-2 protein and clinical response Bcl-2 expression and its association with the clinical response to induction chemotherapy were evaluated in the biopsy specimens of 43 patients before and after treatment in the two groups: the objective response group (OR), and the no response group (NR). In patients with positive expression of Bcl-2 protein, the OR rate was 89.4% (17/19), and in patients with negative expression, the OR rate was 79.1% (19/24). This difference was not statistically significant (x2 test; p50.43). In the OR group, Bcl-2 was positive in 47.2% (17/36) before and in 41.1% (14/34) after chemotherapy (McNemar test; p50.75). Bcl-2 positive expression was seen in 28.5% of tumours (2/7) both before and after chemotherapy in the NR group (McNemar test; p51.0).
Induction of apoptosis and clinical response Morphologically, apoptotic cells were infrequent in pretreatment tissue. Median AI was 0.82 (0.12–2.49) before chemotherapy and 1.61 (0.32–4.31) following chemotherapy. Neoadjuvant chemotherapy was able to induce apoptosis in breast cancer cells (paired t-test; p,0.0001).
Expression of p53 protein and clinical response p53 protein expression was evaluated before and after treatment in the biopsy samples of 43 patients in the two different groups, the objective response group (OR) and the no response group (NR). The OR rate in patients with positive p53 protein expression was 93% (15/16), while in patients with negative p53 protein expression it was 77.7%. There was no statistically significant difference between OR rates (x2 test; p50.22). In the OR group, p53 was positive in 41.6% (15/36) before and 22.2% (8/36) after chemotherapy (McNemar test; p50.03; difference516.6%; 95% confidence interval [95%CI]51.3–16.6%). p53-positive expression was seen in 14.25% (1/7) before and in 42.8% (3/7) of tumours after chemotherapy in the NR group (McNemar test; p50.5). Table 2 summarises the relationship between biological characteristics and the clinical response to neoadjuvant chemotherapy.
Apoptotic index and Bcl-2 protein expression We evaluated the rank correlation between AI and Bcl-2 protein expression before and after chemotherapy. A statistically significant inverse relationship was seen between AI and Bcl-2 before chemotherapy (Kendall’s tau520.33, p50.02, n543). No such correlation was observed after chemotherapy (Kendall’s tau520.125, p50.24, n541).
Apoptotic index and clinical response AI was evaluated before and after neoadjuvant chemotherapy in 35 OR patients and seven NR patients. In the OR group, the median AI before chemotherapy was 0.88 (0.12–2.49) and after chemotherapy the median was 1.84 (0.32–4.31). An increase in apoptotic rate was seen after chemotherapy in patients with an objective response (paired t-test; p,0.0001). In the NR group, median AI
Following chemotherapy, Bcl-2 and p53 protein expression was evaluated in 42 patients. The remaining patient presented complete pathological response. Bcl-2 protein expression was positive in 16 patients (39%), while p53 protein expression was positive in 12 patients (27.9%). Neoadjuvant chemotherapy had no influence on Bcl-2 expression (McNemar test; p50.77) or p53 protein expression (McNemar test; p50.28).
TABLE 2 Relationship between Bcl-2 and p53 protein expression and clinical response in 43 breast cancer patients treated with neoadjuvant chemotherapy
OR NR
p53+ pre
p53+ post
p (dif)
Bcl-2+ pre
Bcl-2+ post
p (dif)
41.6% (15/36) 14.2% (1/7)
22.2% (8/36) 42.8% (3/7)
0.03 (16.6%) NS
47.2% (17/36) 28.5% (2/7)
41.1% (14/34) 28.5% (2/7)
NS NS
OR, objective response; NR, no response; pre, positive protein expression before chemotherapy; post, positive protein expression after chemotherapy; dif, difference (McNemar test); NS, not significant.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
25
TABLE 3 Relationship between AI and tumour response in 42 breast cancer patients treated with neoadjuvant chemotherapy
Initial AI (median) Post-chemotherapy AI (median) AI change (median)
OR (range)
NR (range)
p (rho)
0.8 (0.12–2.5) 1.8 (0.3–4.3) 0.84 (20.19–2.6)
1.2 (0.65–2.4) 1.3 (0.64–2.2) 20.01 (20.31–0.68)
NS (–0.24) 0.03 (0.33) 0.006 (0.4)
OR, objective response; NR, no response; rho, rank correlation; AI, apoptotic index; NS, not significant.
before chemotherapy was 1.21 (0.65–2.43) and 1.28 (0.64– 2.2) after chemotherapy. There was no change in apoptotic rate after chemotherapy in patients with no clinical response (paired t-test; p50.64). The relationship between AI before and after chemotherapy and clinical response was evaluated. There was no correlation between AI before chemotherapy and the prediction of clinical response (rho520.24; p50.1); however, a statistically significant relationship was seen between AI and clinical response after chemotherapy (rho50.33; p50.03). The AI change (AIC), defined as the difference between AI before and after chemotherapy, was calculated. The median AIC observed was 0.84 in the OR group and 0.01 in the NR group (rho50.40; p50.006), as summarised in Table 3. As a predictor of response, a change in AI.0.03 was related with positive and negative predictive values of 94.4% and 71.4%, respectively, and the area under the ordinary ROC curve was 0.90, indicating that AI increases over 0.03 have an accuracy of 90% in predicting response (Fig. 4).
Fig. 4 Receiver operating characteristic (ROC) curve for predicting response in 43 breast cancer patients treated with neoadjuvant chemotherapy.
DISCUSSION Neoadjuvant chemotherapy plays an important role in the management of patients with locally advanced breast cancer. The aim of neoadjuvant therapy is to reduce the tumour volume in patients before surgical resection, thus increasing the likelihood of breast conservation. The clinical response to this treatment is reported as an independent prognostic factor and patients with complete pathological response have a longer disease-free interval and overall survival.15 More recently, the expression of biological markers has been evaluated as a predictive factor of response. As previously reported by Ellis et al.,16 chemotherapy can induce apoptosis in breast carcinomas in vivo. Failure to undergo apoptosis is considered a major mechanism of chemoresistance. We investigated the apoptotic rate in locally advanced breast cancer patients treated with neoadjuvant chemotherapy. Apoptosis has distinctive biochemical and morphological features that distinguish it from necrosis. These include endonuclease activation, chromatin condensation and margination, cell shrinkage, and fragmentation.9,10 Although there are many other sensitive methods for detecting apoptotic cells, we counted apoptotic cells directly under light microscopy based on these morphological features. This technique reliably measures the apoptotic index in biopsy specimens of tumours with a heterogeneous cell population. The range of prechemotherapy AI observed in this study is consistent with previously published data.17,18 In the present study, an increase in apoptotic rate was observed following chemotherapy. There was a significant change in AI in patients with an objective response to induction chemotherapy. In patients with no clinical response, no such change was observed, suggesting that chemoresistant tumours were able to block apoptosis usually induced by cytotoxic drugs. Over-expression of bcl-2, an antiapoptotic protein, has been reported in up to 50% of all cases of breast cancer.19 In a previous study, high levels of Bcl-2 protein were able to block apoptosis and prevent tumour cells from dying in response to the severe DNA damage caused by chemotherapy. 20 We expected to find a reduction in Bcl-2 protein expression following chemotherapy in patients who had an objective response to chemotherapy; however, no such change was observed. As demonstrated in some studies, chemotherapy can reduce the level of Bcl-2 protein in MCF-7 breast cancer cells.21,22 We were unable to observe any reduction in the expression of this protein following chemotherapy, and Bcl-2 expression was found to be a stable phenotype. The change reported by other investigators, however, occurred at an early stage in the study period.22 We evaluated Bcl-2 expression 21 days after the last cycle of chemotherapy, and this delayed biopsy may explain the expression pattern observed in our study.
26
Pathology (2006), 38(1), February
TIEZZI et al.
Repeat biopsies would be necessary to confirm this finding in vivo. An inverse correlation between Bcl-2 expression and AI in pre-chemotherapy specimens was demonstrated in this study. This finding in pre-chemotherapy specimens is compatible with Bcl-2 molecular function. Tumours with an over-expression of Bcl-2 protein show a low rate of apoptosis. Over-expression of Bcl-2 protein has been postulated to promote chemotherapy resistance in solid tumours. 23–25 Objective response rates of 89.4% and 79.1% in patients with and without Bcl-2 protein over-expression, respectively, were observed in the present study. Because taxanes can result in Bcl-2 hyperphosphorylation and consequent functional inactivation,26 tumours with Bcl-2 protein over-expression may be more susceptible to taxaneinduced apoptosis. In this study, 22 patients were treated with taxane-based chemotherapy. p53 protein is a transcription factor that regulates normal cell growth by controlling genes that promote progression through the cycle and by controlling those that cause arrest in G1 when the genome is damaged. Its activation can induce apoptosis and is an important mechanism of cell execution in response to DNA damage. The loss of p53 function as a result, for example, of a mutation in the p53 gene has been reported to enhance cell resistance to a number of chemotherapeutic agents; however, controversial data have been published with respect to chemotherapy resistance and p53 protein expression in breast cancer tumour specimens.27,28 This lack of concordance may be attributed to differences in immunohistochemical techniques, study design or sample population. This mutation leads to p53 protein overexpression and can be identified by immunoperoxidase assay.29 In this study, there was a tendency towards a higher OR in patients with mutant p53 protein overexpression. The OR rate was 93% in that group of patients compared with an OR rate of 77.7% in patients with no over-expression. Some studies suggest that the DNA damage induced by cytotoxic agents may potentially favour the mutation of the p53 gene.30,31 A change in p53 protein expression after treatment in originally positive p53 tumours was observed in the present study. The positive p53 protein expression rate was 41.6% prior to and 22.2% following chemotherapy in the OR group. There was a significant decrease in the expression rate in this group of patients. In patients who had no response to neoadjuvant chemotherapy, there was no significant change in p53 protein expression rates. Only two breast cancer specimens that were negative for p53 prior to chemotherapy became positive afterwards. The observation of a reduction in p53 protein expression rate after chemotherapy in patients who had a clinically objective response may reflect a clonal selection process. This change is part of a multifactorial mechanism involved in tumour sensitivity to cytotoxic drugs that leads to an increase in the cell death/cell mitosis ratio. The finding of the inverse correlation between mutant p53 protein expression and AI following chemotherapy supports the hypothesis that a high level of mutant p53 protein can protect chemotherapy-induced cell death. AI was shown to correlate with an objective clinical response in primary chemotherapy-treated breast cancer. These data support the hypothesis that the increase in tumour
apoptotic rate correlates with a higher OR rate, and we believe that the change in AI may be representative of the level of tumour protection against chemotherapy-induced DNA damage; in other words, the greater the change in AI, the lower the resistance of the tumour cell. ACKNOWLEDGEMENTS This research was supported by a grant from Fundac¸a˜o de Amparo ao Ensino, Pesquisa e Assisteˆncia (FAEPA), Brasil. Address for correspondence: Dr D. G. Tiezzi, Hospital das Clı´nicas da Faculdade de Medicina de Ribeira˜o Preto – USP, Departamento de Ginecologia e Obstetricia, Av Bandeirantes, nu 3900, 8u andar, Ribeira˜o Preto, Sa˜o Paulo, Brazil, CEP 14049-900. E-mail: daniel_tiezzi@yahoo. com.br
References 1. Fisher ER, Wang J, Bryant J, et al. Pathobiology of preoperative chemotherapy: findings from the National Surgical Adjuvant Breast and Bowel (NSABP) protocol B-18. Cancer 2002; 95: 681–95. 2. Dieras V, Fumoleau P, Romieu G, et al. Randomized parallel study of doxorubicin plus paclitaxel and doxorubicin plus cyclophosphamide as neoadjuvant treatment of patients with breast cancer. J Clin Oncol 2004; 22: 4958–65. 3. Smith IE, Mainwaring PN. Primary chemotherapy. In: Bonadonna G, Hortobagyi GN, Gianni AM, editors. Textbook of Breast Cancer: A Clinical Guide to Therapy. London: Martin Dunitz. 1998; 104–5. 4. Faneyte IF, Schrama JG, Peterse JL, et al. Breast cancer response to neoadjuvant chemotherapy: predictive markers and relation with outcome. Br J Cancer 2003; 88: 406–12. 5. Symmans WF, Volm MD, Shapiro RL, et al. Paclitaxel-induced apoptosis and mitotic arrest assessed by serial fine-needle aspiration: implications for early prediction of breast cancer response to neoadjuvant treatment. Clin Cancer Res 2000; 6: 4610–7. 6. Hayward JL, Carbone PP, Heusen JC, et al. Assessment of response to therapy in advanced breast cancer. Br J Cancer 1977; 114: 160–3. 7. Soini Y, Puhakka A, Kahlos K, et al. Endothelial nitric oxide synthase is strongly expressed in malignant mesothelioma but does not associate with vascular density or the expression of VEGF, FLK1 or FLT1. Histopathology 2001; 39: 179–86. 8. Nagle RB, Fitzgibbons PL. Consensual interpretive guidelines for diagnostic immunohistochemistry. Am J Surg Pathol 2002; 26: 816–7. 9. Ha¨cker G. The morphology of apoptosis. Cell Tissue Res 2000; 301: 5–7. 10. Cohen JJ. Apoptosis. Immunol Today 1993; 14: 126–30. 11. Meyn RE, Stephens LC, Ang KK, et al. Heterogeneity in the development of apoptosis in irradiated murine tumours of different histologies. Int J Radiat Biol 1993; 64: 583–91. 12. Wheeler JA, Stephens LC, Tornos C, et al. ASTRO Research Fellowship. Apoptosis as a predictor of tumor response to radiation in stage IB cervical carcinoma. American Society for Therapeutic Radiology and Oncology. Int J Radiat Oncol Biol Phys 1995; 32: 1487–93. 13. Cameron DA, Keen JC, Dixon JM, et al. Effective tamoxifen therapy of breast cancer involves both antiproliferative and pro-apoptotic changes. Eur J Cancer 2000; 36: 845–51. 14. Aihara M, Scardino PT, Truong LD, et al. The frequency of apoptosis correlates with the prognosis of Gleason grade 3 adenocarcinoma of the prostate. Cancer 1995; 75: 522–9. 15. Carey LA, Metzger R, Dees EC, et al. American Joint Committee on Cancer Tumor-Node-Metastasis Stage After Neoadjuvant Chemotherapy and Breast Cancer Outcome. J Natl Cancer Inst 2005; 97: 1337–42. 16. Ellis PA, Smith IE, Mccarthy K, et al. Preoperative chemotherapy induces apoptosis in early breast cancer. Lancet 1997; 349: 849. 17. Staunton MJ, Gaffney EF. Tumor type is a determinant of susceptibility to apoptosis. Am J Clin Pathol 1995; 103: 300–7. 18. Paradiso A, Mangia A, Chiriatti A, et al. Biomarkers predictive for clinical efficacy of taxol-based chemotherapy in advanced breast cancer. Ann Oncol 2005; 16: 14–9. 19. Silvestrini R, Veneroni S, Daidone MG, Benini E, Borscchi P. The bcl2 protein: a prognostic indicator strongly related top53 protein in lymph node-negative breast cancer in patients. J Natl Cancer Inst 1994; 86: 499–504. 20. Srivastava RK, Mi QS, Hardwick JM, et al. Deletion of the loop region of Bcl-2 completely blocks paclitaxel-induced apoptosis. Proc Natl Acad Sci USA 1999; 96: 3775–80.
APOPTOSIS AND NEOADJUVANT CHEMOTHERAPY
21. Leung LK, Wang TT. Differential effects of chemotherapeutic agents on the Bcl-2/Bax apoptosis pathway in human breast cancer cell line MCF-7. Breast Cancer Res Treat 1999; 55: 73–83. 22. Wang Q, Maloof P, Wang H, et al. Basic fibroblast growth factor downregulates Bcl-2 and promotes apoptosis in MCF-7 human breast cancer cells. Exp Cell Res 1998; 238: 177–87. 23. Ohmori T, Podack ER, Nishio K, et al. Apoptosis of lung cancer cells caused by some anti-cancer agents (MMC, CPT-11, ADM) is inhibited by bcl-2. Biochem Biophys Res Commun 1993; 193: 30–6. 24. Bold RJ, Chandra J, McConkey DJ. Gencitabine-induced programmed cell death (apoptosis) of human pancreatic carcinoma is determined by bcl-2 content. Ann Surg Oncol 1999; 6: 279–85. 25. Eliopoulos AG, Kerr DJ, Herod J, et al. The control of apoptosis and drug resistance in ovarian cancer: influence of p53 and bcl-2. Oncogene 1995; 11: 1217–28. 26. Scatena CD, Stewart ZA, Mays D, et al. Mitotic phosphorylation of bcl-2 during normal cell cycle progression and taxol-induced growth arrest. J Biol Chem 1998; 273: 30777–84.
27
27. Erdem O, Dursun A, Coskun U, et al. The prognostic value of p53 and c-erbB-2 expression, proliferative activity and angiogenesis in nodenegative breast carcinoma. Tumori 2005; 91: 46–52. 28. Tsutsui S, Yasuda K, Higashi H, et al. Prognostic implication of p53 protein expression in relation to nuclear pleomorphism and the MIB-1 counts in breast cancer. Breast Cancer 2004; 11: 160–8. 29. Harris CC. Structure and function of the p53 tumor suppressor gene: clues for rational cancer therapeutic strategies. J Natl Cancer Inst 1996; 88: 1442–55. 30. Bottini A, Berruti A, Bersiga A, et al. p53 but not bcl-2 immunostaining is predictive of poor clinical complete response to primary chemotherapy in breast cancer patients. Clin Cancer Res 2000; 6: 2751–8. 31. Brambilla E, Gazzeri S, Lantuejoul S, et al. p53 mutant immunophenotype and deregulation of p53 transcription pathway (bcl-2, bax and waf-1) in precursor bronchial lesions of lung cancer. Clin Cancer Res 1998; 4: 1609–18.