Aberrant inhibition of programmed cell death (apoptosis) prevents normal homeostasis and promotes tissue tumorigenesis, but whether it also influences the outcome of common cancers has remained arguable. The expression of a novel IAP apoptosis inhibitor, survivin, in breast cancer and its association with tumor cell apoptosis and overall prognosis were examined in this study. Immunohistochemical analysis showed that survivin expression was positive in 118 of 167 cases(70.7%) of breast carcinomas of histological stages I to III. In contrast, no expression of survivin in adjacent normal tissue was detected. Although survivin expression was not correlated with p53 mutations, survivin-positive cases were strongly associated with bcl-2 expression (78.0% versus 47.5%; P = 0.0005) and reduced apoptotic index (0.62% ± 0.51%versus 1.27% ± 1.37%; P <0.0001). In addition, patients with low apoptotic index (<0.52%) had worse survival rates than the group with high apoptotic index(≥0.52%; P = 0.028), and multivariate Cox proportional hazard model analysis identified apoptotic index as an independent prognostic factor (P = 0.024). The results suggest that apoptosis inhibition by survivin,alone or in cooperation with bcl-2, is a significant prognostic parameter of worse outcome in breast carcinoma.

Abnormalities in the control of programmed cell death (apoptosis)play an important role in tumorigenesis (1, 2). This process involves an evolutionarily preserved multistep cascade and is regulated by proteins that promote or counteract apoptotic cell death (3). bcl-2 was the first protein shown to lead to prolonged survival of cells by preventing apoptosis (4). Several apoptosis inhibitors related to the baculovirus iapgene have been identified in humans, mice, and Drosophila (5, 6). Highly evolutionarily conserved,IAP2 proteins contain two to three Cys/His BIRs and a COOH-terminal RING finger (4). Recombinant expression of IAP proteins counteracted various forms of apoptosis in vivo(7) and in vitro(6). These molecules are thought to block an evolutionarily conserved step in apoptosis. At least in the case of XIAP, this may involve direct inhibition of the terminal effectors caspase-3 and caspase-7 through a BIR-dependent recognition (8). Recently, a novel gene encoding a structurally unique IAP apoptosis inhibitor, designated as survivin, has been identified. Survivin is an Mr∼16,500 cytoplasmic protein with a single BIR and no RING finger. Recombinant expression of survivin prevents apoptosis induced by growth factor interleukin 3 withdrawal in a pre-B cell line (9). Unlike bcl-2 (10) or other IAP proteins (5, 6, 7), survivin is undetectable in terminally differentiated adult tissues but becomes notably expressed in the most common human cancers, including stomach (11), colorectal (12), lung, breast, pancreatic, and prostate cancers and high-grade non-Hodgkin’s lymphomas in vivo(9, 13). Our previous studies demonstrated that the expression of survivin was significantly associated with bcl-2 expression and reduced apoptotic indices, which were strongly correlated with poor prognosis after surgery in both gastric and colorectal cancers (11, 12). In this study, we assessed the expression of survivin in breast cancer and its potential effect on tumor cell apoptosis and overall survival.

Patients and Samples.

We studied a series of 167 patients with invasive breast carcinomas who did not receive any form of treatment prior to surgery. The surgically resected specimens used for this study were obtained from consecutive patients with breast carcinoma who underwent potentially curative resection at the Department of General and Gastroenterological Surgery,Osaka Medical College Hospital, during the period from 1988 to 1994. Clinicopathological factors, tumor histologies, and disease stage were assigned according to the General Rules for Clinical and Pathological Recording on Breast Cancer (14). The specimens consisted of 32 cases of stage I, 104 cases of stage II, and 31 cases of stage III tumors. All of the patients were female, and the mean age of the patients was 53.1 years (SD, 11.84 years; range, 29–89 years). With regard to postoperative adjuvant therapy, the patients with four or more axillary lymph node metastases had six cycles of CMF (100 mg/m2 cyclophosphamide p.o. on days 1–14, 40 mg/m2 i.v. methotrexate on days 1 and 8, and 500 mg/m2 i.v. 5-fluorouracil on days 1 and 8, every 4 weeks) or CAF (100 mg/m2 cyclophosphamide p.o. on days 1–14, 30 mg/m2 Adriamycin i.v. on days 1 and 8, and 500 mg/m2 5-fluorouracil i.v. on days 1 and 8, every 4 weeks). The patients with three or less axillary lymph node metastases were treated with tegaful[(2-tetra-hydrofuryl)-5-fluorouracil] at an oral dose of 600 mg/day for 2 years after surgery. Node-negative premenopausal patients were followed without chemotherapy. Patients with ER-positive tumors and postmenopausal patients received adjuvant endocrine treatment with 20 mg/day oral dose of Tamoxifen. The median follow-up time in this study group was 96 months (range, 11–104 months). Routinely processed formalin-fixed, paraffin-embedded blocks containing the main tumor were prepared. Serial sections of 2–4 μm were cut from the blocks at the maximum cross-section of the tumor.

Immunohistochemical Staining for survivin and the Scoring Method for Its Expression.

A pilot study using the anti-survivin antibody was conducted on various neoplasms, including high-grade non-Hodgkin’s lymphoma, pancreatic cancer, colorectal cancer, and gastric cancer to confirm the specificity in staining tumor cells and to determine an appropriate dilution for mAb 8E2 (IgG1 kindly provided by Dr. D. C. Altieri,Department of Pathology, Yale University, School of Medicine, New Haven, CT) for staining of breast carcinomas. One case of stage III gastric cancer was stained intensively and reproducibly for survivin expression in >30% of tumor cells, and this was used as a positive control throughout the present study. Negative control slides processed without primary antibody were included for each staining. For immunohistochemical detection of survivin protein, the standard avidin-biotin-peroxidase complex technique was carried out by using an LSAB kit (Dako A/S, Carpinteria, CA). Before using the LSAB kit,antigen retrieval was done by the pressure cooking method as described previously (11, 12). In brief, deparaffinized and rehydrated sections were bathed in a 10−3m sodium citrate buffer (pH 6.0) after bringing the solution to a boil in a pressure cooker and boiled for 20 min while maintaining the pressure. After quenching in 3% hydrogen peroxide and blocking for 5 min, the sections were incubated overnight at 4°C with a 1:5 dilution of primary monoclonal antibody 8E2 raised against purified recombinant survivin. Biotinylated antimouse immunoglobulin and streptavidin conjugated to horseradish peroxidase were then added. Finally, 3,3′-diaminobenzidine was used for color development, and hematoxylin was used for counterstaining. Negative control slides processed without primary antibody were included for each staining. The scoring method for survivin was modified from that for bcl-2 expression described by Sinicrope et al.(15). The mean percentage of positive tumor cells was determined in at least five areas at 400-fold magnification and assigned one of the following five categories: 0, <5%; 1, 5–25%; 2, 25–50%; 3, 50–75%; and 4,>75%. The intensity of survivin immunostaining was scored as follows:1+, weak; 2+, moderate; and 3+, intense. Because tumors showed heterogeneous staining, the dominant pattern was used for scoring. The staining intensity of infiltrated lymphocytes for bcl-2 was judged as 3+ relative to the intensity of survivin staining. The scores indicating percentage of positive tumor cells and staining intensity were multiplied to produce a weighted score for each case. Cases with weighted scores <1 were defined as negative, and cases were otherwise defined as positive.

Immunohistochemical Staining for bcl-2 and p53.

For detection of bcl-2 and p53 protein, deparaffinized and rehydrated sections were immunostained using the same techniques as for survivin antigen staining. Monoclonal mouse antibody against bcl-2 (clone 124,diluted at 1:20; Dako, Copenhagen, Denmark) and a mouse antihuman p53 antibody (DO7, diluted 1:50; DAKO, Copenhagen, Denmark) were used as primary antibodies for bcl-2 and p53 immunostaining. Before addition of the primary antibodies, sections were heated in a microwave oven three times at 900 W for a total of 15 min in 10−3m sodium citrate buffer (pH 6.0). The other staining procedures were the same as those for survivin. The scoring criteria for bcl-2 were the same as those for survivin, and cases with weighted scores of <1 were judged as negative. For p53 expression, cases with<5% positively stained tumor cells were defined as negative;otherwise, they were defined as positive.

Histochemical Detection of Apoptosis and Determination of the AI.

Apoptotic cells and apoptotic bodies were detected by in situ labeling using an ApopTag in situ Detection kit(S7101-KIT; Oncor, Gaithersburg, MD). In brief, deparaffinized and rehydrated sections were digested with proteinase K (20 μg/ml in PBS;Wako, Osaka, Japan) for 20 min at room temperature and washed. After quenching in 3% hydrogen peroxide for 5 min, washing with PBS, and adding the equilibration buffer for 10 min, terminal deoxynucleotidyl transferase enzyme was pipetted onto the sections, which were then incubated at 37°C for 1 h. After stopping the reaction by putting sections in stop/wash buffer and washing,anti-digoxigenin-peroxidase was added to the slides. Finally, slides were washed with PBS, stained with diaminobenzine (DAKO A/S, Glostrup,Denmark) substrate, and counterstained with methyl green. A positive control was prepared by nicking DNA with DNase I (0.7 μg/ml;Stratagene Co., La Jolla, CA) for the first staining procedure. A specimen known to be positive for apoptotic cells was used as positive control for subsequent staining. Substitution of terminal deoxynucleotidyl transferase with distilled water was used as negative control. The AI was expressed as the ratio of positively stained tumor cells and bodies to all tumor cells, given as a percentage for each case, and determined according to the criteria described previously (11, 12, 16). In brief, a minimum of 3000 cells was counted at 400-fold magnification. Positively staining tumor cells with the morphological characteristics of apoptosis were identified using standard criteria, including chromatin condensation, nuclear disintegration, and formation of crescent caps of condensed chromatin at the nuclear periphery.

Statistical Analysis.

All statistical analysis was performed using the SPSS 6.1 J software package for Macintosh (SPSS, Inc., Chicago, IL). Variables associated with survivin expression as well as the correlation between survivin and p53 or bcl-2 expression were analyzed by theχ 2 test. Differences in tumor cell AI for groups differing according to survivin expression were checked by the independent Wilcoxon method. The survival curves were plotted according to Kaplan-Meier method and checked by the log-rank test. A value of P < 0.05 was considered statistically significant.

State of Expression of Survivin in Breast Carcinomas.

Immunohistochemical staining revealed that anti-survivin mAb 8E2 specifically reacted with breast carcinoma cells, with positive staining of the cytoplasm of cancer cells (Fig. 1 a), whereas no expression of survivin was observed in adjacent normal tissues. Unlike bcl-2, no reactivity of the anti-survivin mAb 8E2 with infiltrating lymphocytes and other stroma cells was detected. The intensity of survivin staining was usually homogenous within a given case, but the number of positive tumor cells stained by the anti-survivin mAb varied between 20 and 100%, depending on the case investigated. After multiplying the weighted survivin score, 118 cases of breast carcinoma in this series were defined as positive (70.7%), with a weighted survivin score from 1 to 12. The distributions of weighted survivin scores were 0 for 49 cases, 1 for 17 cases, 2 for 26 cases, 3 for 30 cases, 4 for 14 cases,6 for 13 cases, 8 for 8 cases, 9 for 7 cases, and 12 for 3 cases.

Relationship Between Expression of Survivin and Clinicopathological Factors.

Table 1 shows the correlation between expression of survivin and clinicopathological factors. It shows that none of the prognostic parameters analyzed, including menopausal status, different tumor size, clinical stage, lymph node status, ER status, lymphatic or venous invasion, tumor histology, or recurrence,was significantly correlated with survivin expression.

Relationship Between Expression of survivin and bcl-2 or p53.

Among the 167 breast carcinomas studied, 127 cases (76.0%) showed positive cytoplasmic immunoreactivity for bcl-2. Expression of survivin showed a strong positive correlation with expression of bcl-2, and it was very significantly segregated with bcl-2-negative cases (99 of 127,78.0% versus 19 of 40, 47.5%; P = 0.0005). In contrast, nuclear accumulation of p53 was demonstrated in 39 of 167 cases (23.6%). There was no correlation between survivin and p53 expression (P = 0.705; Table 2).

Association of Expression of survivin and bcl-2 with Tumor Cell Apoptosis.

Every case of breast carcinoma examined showed apoptotic cells and apoptotic bodies that were detected by in situ labeling(Fig. 1,b). The relationships between AI and expression of survivin and bcl-2 were examined (Table 3). The mean AI of the 167 cases was 0.81% (SD, 0.90%; range, 0.10–8.61%), with a median value of 0.54%. No significant correlation was observed between the AI and the tumor stage. The mean AI in survivin-positive tumors (n = 118) was 0.62% (SD, 0.51%), which was significantly lower than the mean AI of 1.27% (SD, 1.37%) observed in survivin-negative tumors(n = 49; P < 0.0001). However, when the cases with survivin weighted scores <2 were defined as negative and otherwise as positive, the mean AI for survivin-positive tumors(n = 101) was 0.63%, which was also significantly lower than the mean AI of 1.08% for survivin-negative tumors(n = 66; P = 0.0016). The mean AI in bcl-2-positive tumors was significantly lower than that in bcl-2-negative tumors (P = 0.0013). In survivin-positive tumors (n = 118), 99 cases were also positive for bcl-2 expression, and the mean AI in these cases was 0.56% (SD, 0.44%). In bcl-2-negative tumors (n = 40),the mean AI of 19 survivin-positive tumors was 0.94% (SD, 0.72%),which was lower than the value of 1.57% (SD, 1.86%) observed in survivin-negative tumors (P = 0.233).

Prognostic Analysis in Patients with Breast Carcinoma.

Fig. 2,a shows the Kaplan-Meier curves for patients with breast carcinoma categorized according to survivin expression. The overall 5-year survival rate for patients with survivin-positive breast carcinomas (87.3%; 103 of 118) was less than that for patients with survivin-negative tumors (95.9%; 47 of 49), but the difference was not significant. To study a possible correlation between AI and prognosis, the patients were dichotomized by the cutoff of 0.52% for AI, which provided the most sensitive parameter of survival difference in the present patient series. However, AI dichotomized by the median value (0.54%) was also proved to be prognostically significant in the multivariate analysis(P = 0.044). The overall 5-year survival rate for the group with high AI (≥0.52) was 94.4% (84 of 89), whereas that for patients with low AI (<0.52) was 84.6% (66 of 78; Fig. 2,b;P = 0.028). In addition, multivariate Cox proportional hazard model analysis using the variables lymph node metastasis, ER status, and AI, which were all identified as prognostically significant by univariate analysis (Table 4),demonstrated that each of these three variables was an independent predictor for the overall survival (P = 0.024; Table 5).

The results of the present study suggest that decreased apoptosis integrated partly by survivin expression (9, 13) is a new predictive indicator of poorer prognosis in breast carcinoma. Among the recently described IAP family, survivinis characterized by a unique structure with a single BIR and no zinc-binding domain known as the RING finger, and by a selective distribution in common human cancers but not in normal adjacent tissue in vivo(9). In this study, specific staining for survivin was detected in 118 cases (70.7%), with a variable proportion of positive tumor cells (20–100%). In contrast, the adjacent normal tissue or the infiltrating lymphocytes did not express survivin, in accordance with our prior studies (11, 12). One of the intriguing findings in this study was that the expression of survivin in breast carcinoma was associated with significantly reduced apoptosis, as compared with that in survivin-negative tumors. Apoptosis is a feature commonly seen in tumors, and in fact the ability to resist apoptosis may seem to offer an advantage to a rapidly growing tumor, by slowing down the cell loss rate, but potential effector molecules responsible for these aberrations have not been identified. In this study, we demonstrated that low AI, which is correlated with survivin expression, functions as a predictive/prognostic indicator of poor overall survival in breast carcinoma (P = 0.028),although this will have to be validated by additional clinical studies. It was independently confirmed by multivariate Cox proportional hazard model analysis, which identified the AI as a significant independent predictor of overall survival (P = 0.024), although the decrease in survival rate of survivin-positive patients did not reach statistical significance. However, in contrast to our findings, several studies, using AI assessment with a microscopy at 200–400-fold magnification, have demonstrated that apoptotic cell death is increased in tumors with high-grade malignancy in comparison with tumors with low-grade malignancy in various human tumors (17, 18),including breast carcinoma (19). Although morphological studies using light microscopy are useful for the detection of some apoptotic bodies, they may not reveal the early apoptotic stages (20), i.e., “normal-looking” apoptotic cells, which can be detected by the DNA nick end labeling technique used in this study. In addition, the morphological identification of apoptotic nuclei within large numbers of polymorphic nuclei may be extremely difficult (21). Bedi et al.(22) have shown by in situ end-labeling technique that the transformation of colorectal epithelium to carcinomas is associated with a progressive inhibition of apoptosis. Consequently, although it has been shown that biological processes other than apoptosis may lead to positive in situend-labeling (23, 24), this technique of simplifying the identification of apoptotic nuclei in routinely processed tissue sections seems more reliable than morphological assessment alone. Nevertheless, identification of prognostic relevance of tumor cell apoptosis seems problematic in various aspects of this study, including insufficient numbers of patients studied and the heterogeneous character in terms of adjuvant therapy, second-line and third-line treatments for recurrence and/or distant metastasis, who responded variously to combinations of chemotherapy and/or endocrine therapy.

We found compelling evidence that the presence of survivin in breast carcinoma was strongly associated with expression of bcl-2 and with reduced AI (Tables 2 and 3). Our results agree with the findings of previous investigations, which showed a similar association between survivin and bcl-2 expression in neuroblastoma (25),gastric cancer (11), colorectal cancer (12),and high-grade non-Hodgkin’s lymphoma.3 The percentage of bcl-2-positive cases in our series of breast cancer(76.0%) was comparable with those reported by other investigators (26). The accumulation of mutated p53 protein was detected in 23.6% of the present series of patients. The rate of p53 accumulation was within the range of values reported by others (27, 28, 29). The survivin gene is encoded at chromosome 17q25 (30), whereas the bcl-2 gene is located at chromosome 18q21 and may be involved in the tumorigenic t(14;18) translocation (10). These data imply that other transcriptional factors may contribute to the coregulation of both gene products in the progression of cancer. In this context, both survivin and bcl-2 genes are regulated by TATA-less, GC-rich promoter sequence in similar manners, and both are markedly transcribed in actively proliferating cell types (9), suggesting common mechanism(s) of transcriptional activation. However, regardless of the pathway of simultaneous coexpression, it appears that survivin and bcl-2 proteins may mediate nonoverlapping, antiapoptosis mechanisms. Although bcl-2 is an integral inner mitochondrial membrane protein implicated in counteracting cytochrome c release from the mitochondria, IAP molecules,potentially including survivin, prevent apoptosis by targeting the terminal effectors caspase-3 and caspase-7 (8, 31, 32). Survivin is expressed in the G2-M phase of the cell cycle in a cell cycle-regulated manner and associates with microtubules of the mitotic spindle. Disruption of survivin-microtubule interactions results in loss of survivin’s antiapoptosis function and increased caspase-3 activity during mitosis. The overexpression of survivin in cancer may obliterate this apoptotic check point and allow aberrant progression of transformed cells through mitosis (32). In breast carcinoma and in many of the most common human cancers, inhibition of apoptosis may be a general feature, and expression of survivin alone or survivin plus other antiapoptosis genes like bcl-2 may cause more pronounced antiapoptotic effects, as reflected in the significantly reduced apoptotic index observed in our series.

Although decreased AI associated with survivin expression was shown to be an indicator of poor prognosis in breast carcinoma, the predictive value was not as strong as was observed in colorectal carcinoma(P = 0.024 versusP =0.0001; Ref. 12). One of the factors relevant to this issue may be the limited number of patients studied in this trial and the various responses to chemotherapy and/or endocrine therapy (case by case) that were given postoperatively to the present series of patients. Clark et al.(33, 34) has recently proved that either mitosin expression or telomerase activity, a regulator of S-phase fraction, is an independent prognostic factor in node-negative and node-positive breast cancers. Therefore,identification of prognostic significance of apoptotic index should be clarified with additional use of these prognostic markers. bcl-2 has been reported to be frequently expressed in breast cancer and to be associated with positivity for ERs in both node-negative and node-positive breast cancers (35, 36). These findings are in agreement with our finding of a significant positive relationship between bcl-2 immunoreactivity and ER status (P =0.024). However, unlike survivin proteins, the association of bcl-2 with a favorable clinical prognosis in breast cancer is reported (26, 36). One explanation for this paradox is that bcl-2 may have still unrecognized and other nonapoptotic functions. In clinical studies, bcl-2 expression is inversely correlated with S-phase fraction and tumor size in breast cancer (37, 38). Furthermore, bcl-2 has proved to have potential to lead to prolongation of the cell cycle as well as a decrease in vitro breast cancer growth (39).

It is well known that there are multiple genetic pathways that control apoptosis, a part of which is probably regulated by survivin and/or bcl-2. Therefore, changing the level of expression of these proteins may not necessarily have an effect on outcome of therapy. A recent in vitro study demonstrated, however, that antisense survivin RNA down-regulated expression of endogenous survivin in transformed cells and resulted in increased apoptotic cell death (32). In this context, in addition to chemotherapy and endocrine therapy, targeted antagonists of survivin may be beneficial as apoptosis-based therapy for breast carcinoma.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

                
2

The abbreviations used are: IAP, inhibitor of apoptosis; BIR, baculovirus IAP repeat; AI, apoptotic index; ER,estrogen receptor; mAb, monoclonal antibody.

        
3

Unpublished observations.

Fig. 1.

Histochemical staining for survivin and apoptotic tumor cells. a, survivin expression was restricted to the cytoplasm of tumor cells (arrows;×400). b, in situ labeling for apoptosis in a similar area detected only one apoptotic tumor cell(arrow; ×400).

Fig. 1.

Histochemical staining for survivin and apoptotic tumor cells. a, survivin expression was restricted to the cytoplasm of tumor cells (arrows;×400). b, in situ labeling for apoptosis in a similar area detected only one apoptotic tumor cell(arrow; ×400).

Close modal
Fig. 2.

a, Kaplan-Meier curves for overall survival rates of patients with breast carcinoma categorized according to survivin expression. No significant difference was found between the groups (P = 0.108; log-rank test). b, Kaplan-Meier curves for overall survival rates of patients with breast carcinoma categorized by high versus low AI (P = 0.028; log-rank test).

Fig. 2.

a, Kaplan-Meier curves for overall survival rates of patients with breast carcinoma categorized according to survivin expression. No significant difference was found between the groups (P = 0.108; log-rank test). b, Kaplan-Meier curves for overall survival rates of patients with breast carcinoma categorized by high versus low AI (P = 0.028; log-rank test).

Close modal
Table 1

Correlation between clinicopathological factors and expression of survivin in breast carcinomas

Characteristicansurvivin expression (%)P
Age (yr)    
<53 86 58 (67.4) 0.441 
≥53 81 60 (74.1)  
Menopausal status    
Pre 75 51 (68.0) 0.610 
Post 92 67 (72.8)  
Tumor size (cm)    
<2 41 27 (65.9) 0.562 
≥2 126 91 (72.2)  
Clinical stage    
I, II 136 94 (69.1) 0.487 
III 31 24 (77.4)  
Lymph node metastasis    
Negative 56 40 (71.4) 0.968 
Positive 89 62 (69.7)  
Lymphatic invasion    
Negative 75 55 (73.3) 0.607 
Positive 92 63 (68.5)  
Venous invasion    
Negative 20 16 (80.0) 0.474 
Positive 147 102 (69.4)  
ER status    
Negative 49 37 (75.5) 0.483 
Positive 118 81 (68.6)  
Histological typeb    
Pap, solid 109 76 (69.7) 0.805 
Scirrhous 38 25 (65.8)  
Recurrence    
Negative 133 94 (70.7) 1.000 
Positive 34 24 (70.6)  
Characteristicansurvivin expression (%)P
Age (yr)    
<53 86 58 (67.4) 0.441 
≥53 81 60 (74.1)  
Menopausal status    
Pre 75 51 (68.0) 0.610 
Post 92 67 (72.8)  
Tumor size (cm)    
<2 41 27 (65.9) 0.562 
≥2 126 91 (72.2)  
Clinical stage    
I, II 136 94 (69.1) 0.487 
III 31 24 (77.4)  
Lymph node metastasis    
Negative 56 40 (71.4) 0.968 
Positive 89 62 (69.7)  
Lymphatic invasion    
Negative 75 55 (73.3) 0.607 
Positive 92 63 (68.5)  
Venous invasion    
Negative 20 16 (80.0) 0.474 
Positive 147 102 (69.4)  
ER status    
Negative 49 37 (75.5) 0.483 
Positive 118 81 (68.6)  
Histological typeb    
Pap, solid 109 76 (69.7) 0.805 
Scirrhous 38 25 (65.8)  
Recurrence    
Negative 133 94 (70.7) 1.000 
Positive 34 24 (70.6)  
a

Pathological variables were defined in accordance with the General Rules for Clinical and Pathological Recording of Breast Cancer (14).

b

pap, papillotubular carcinoma; solid,solid-tubular carcinoma; scirrhous, scirrhous carcinoma.

Table 2

Expression of bcl-2 and p53 in survivin-positive breast carcinoma

nsurvivin expression (%)P
bcl-2 expression    
Negative 40 19 (47.5) 0.0005 
Positive 127 99 (78.0)  
p53 expression    
Negative 128 89 (69.5) 0.705 
Positive 39 29 (68.9)  
nsurvivin expression (%)P
bcl-2 expression    
Negative 40 19 (47.5) 0.0005 
Positive 127 99 (78.0)  
p53 expression    
Negative 128 89 (69.5) 0.705 
Positive 39 29 (68.9)  
Table 3

Correlation between tumor cell AI and expression of survivin and bcl-2

nAI (%)P
survivin expression    
Negative 49 1.27 ± 1.37 <0.0001 
Positive 118 0.62 ± 0.51  
bcl-2 expression    
Negative 40 1.27 ± 1.46 0.0013 
Positive 127 0.66 ± 0.57  
nAI (%)P
survivin expression    
Negative 49 1.27 ± 1.37 <0.0001 
Positive 118 0.62 ± 0.51  
bcl-2 expression    
Negative 40 1.27 ± 1.46 0.0013 
Positive 127 0.66 ± 0.57  
Table 4

Univariate Cox proportional hazard model

CharacteristicRisk rate95% CIaP
Tumor size (cm) (≥2 vs. <2) 1.292 0.37–4.50 0.688 
Lymph node metastasis (positive vs. negative) 18.073 2.39–136.87 0.005 
Lymphatic invasion (positive vs. negative) 2.842 0.93–8.72 0.067 
Venous invasion (positive vs. negative) 1.824 0.52–6.36 0.346 
ER status (positive vs. negative) 0.332 0.13–0.86 0.023 
p53 expression (positive vs. negative) 2.363 0.90–6.22 0.081 
bcl-2 expression (positive vs. negative) 0.631 0.22–1.80 0.389 
survivin expression (positive vs. negative) 3.093 0.71–13.53 0.134 
AI (≥0.52 vs. <0.52) 0.328 0.12–0.93 0.037 
CharacteristicRisk rate95% CIaP
Tumor size (cm) (≥2 vs. <2) 1.292 0.37–4.50 0.688 
Lymph node metastasis (positive vs. negative) 18.073 2.39–136.87 0.005 
Lymphatic invasion (positive vs. negative) 2.842 0.93–8.72 0.067 
Venous invasion (positive vs. negative) 1.824 0.52–6.36 0.346 
ER status (positive vs. negative) 0.332 0.13–0.86 0.023 
p53 expression (positive vs. negative) 2.363 0.90–6.22 0.081 
bcl-2 expression (positive vs. negative) 0.631 0.22–1.80 0.389 
survivin expression (positive vs. negative) 3.093 0.71–13.53 0.134 
AI (≥0.52 vs. <0.52) 0.328 0.12–0.93 0.037 
a

CI, confidence interval.

Table 5

Multivariate Cox proportional hazard model

CharacteristicRisk rate95% CIaP
Lymph node metastasis (positive vs. negative) 22.518 2.86–177.54 0.003 
ER status (positive vs. negative) 0.224 0.08–0.62 0.004 
AI (≥0.52 vs. <0.52) 0.300 0.10–0.86 0.024 
CharacteristicRisk rate95% CIaP
Lymph node metastasis (positive vs. negative) 22.518 2.86–177.54 0.003 
ER status (positive vs. negative) 0.224 0.08–0.62 0.004 
AI (≥0.52 vs. <0.52) 0.300 0.10–0.86 0.024 
a

CI, confidence interval.

We are grateful to Dr. Dario C. Altieri (Boyer Center for Molecular Medicine, Department of Pathology, Yale University School of Medicine, New Haven, CT) for providing the anti-survivin antibody(8E2).

1
Strasser A., Harris A. W., Cory S. bcl-2 transgene inhibits T cell death and perturbs thymic self-censorship.
Cell
,
67
:
889
-899,  
1991
.
2
Cross M., Dexter T. M. Growth factors in development, transformation, and tumorigenesis.
Cell
,
64
:
271
-280,  
1991
.
3
Thompson C. B. Apoptosis in the pathogenesis and treatment of disease.
Science (Washington DC)
,
267
:
1456
-1462,  
1995
.
4
Garcia I., Martinou I., Tsujimoto Y., Martinou J. C. Prevention of programmed cell death of sympathetic neurons by the bcl-2 proto-oncogene.
Science (Washington DC)
,
258
:
302
-304,  
1992
.
5
Clem R. J., Miller L. K. , Control of programmed cell death by the baculovirus genes p35 and iap.
Mol. Biol. Cell
,
14
:
5212
-5222,  
1994
.
6
Hay B. A., Wassarman D. A., Rubin G. M. Drosophila homologs of baculovirus inhibitor of apoptosis proteins function to block cell death.
Cell
,
83
:
1253
-1262,  
1995
.
7
Duckett C. S., Nava V. E., Gedrich R. W., Clem R. J., Van-Dongen J. L., Gilfillan M. C., Shiels H., Hardwick J. M., Thompson C. B. A conserved family of cellular genes related to the baculovirus iap gene and encoding apoptosis inhibitors.
EMBO J.
,
15
:
2685
-2694,  
1996
.
8
Takahashi R., Deveraux Q., Tamm I., Welsh K., Assa-Munt N., Salvesen G. S., Reed J. C. A single BIR domain of XIAP sufficient for inhibiting caspases.
J. Biol. Chem.
,
273
:
7787
-7790,  
1998
.
9
Ambrosini G., Adida C., Altieri D. C. A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma.
Nat. Med.
,
3
:
917
-921,  
1997
.
10
Yang E., Korsmeyer S. J. Molecular thanatopsis: a discourse on the bcl-2 family and cell death.
Blood
,
88
:
1456
-1462,  
1996
.
11
Lu C-D., Altieri D. C., Tanigawa N. Expression of a novel antiapoptosis gene, survivin, correlated with tumor cell apoptosis and p53 accumulation in gastric carcinomas.
Cancer Res.
,
58
:
1808
-1812,  
1998
.
12
Kawasaki H., Altieri D. C., Lu C-D., Toyoda M., Tenjo T., Tanigawa N. Inhibition of apoptosis by survivin predict shorter survival rates in colorectal cancer.
Cancer Res.
,
58
:
5071
-5074,  
1998
.
13
Adida C., Crotty P. L., McGrath J., Berrebi D., Diebold J., Altieri D. C. Developmentally regulated expression of the novel cancer anti-apoptosis gene survivin in human and mouse differentiation.
Am. J. Pathol.
,
152
:
43
-49,  
1998
.
14
The Japanese Breast Cancer Society. General Rules for Clinical and Pathological Recording of Breast Cancer, Ed. 12. Tokyo: Kanehara Public Co., 1996.
15
Sinicrope F. A., Ruan S. B., Cleary K. R., Stephens L. C., Lee J. J., Levin B. bcl-2 and p53 oncoprotein expression during colorectal tumorigenesis.
Cancer Res.
,
55
:
237
-241,  
1995
.
16
Lu C-D., Tanigawa N. Spontaneous apoptosis is inversely related to intratumoral microvessel density in gastric carcinoma.
Cancer Res.
,
57
:
221
-224,  
1997
.
17
Leoncini L., Del Vecchio M. T., Megha T., Barbini P., Galieni P., Pileri S., Sabattini E., Gherlinzoni F., Tosi P., Kraft R., Cottier H. Correlations between apoptotic and proliferative indices in malignant non-Hodgkin’s lymphomas.
Am. J. Pathol.
,
142
:
755
-763,  
1993
.
18
Lipponen P. K., Aaltomaa S. Apoptosis in bladder cancer as related to standard prognostic factors and prognosis.
J. Pathol.
,
173
:
333
-339,  
1994
.
19
Lipponen P. K., Aaltomaa S., Kosma V. M., Syrjanen K. Apoptosis in breast cancer as related to histopathological characteristics and prognosis.
Eur. J. Cancer
,
30
:
2068
-2073,  
1994
.
20
Frankfurt O. S., Robb J. A., Sugarbaker E. V., Villa L. Apoptosis in human breast and gastrointestinal carcinomas. Detection in histological sections with monoclonal antibody to single-stranded DNA.
Anticancer Res.
,
16
:
1979
-1988,  
1996
.
21
Wyllie A. H., Kerr J. F., Currie A. R. Cell death: the significance of apoptosis.
Int. Rev. Cytol.
,
68
:
251
-306,  
1980
.
22
Bedi A., Pasricha P. J., Akhtar A. J., Barber J. P., Bedi G. C., Giardiello F. M., Zehnbauer B. A., Hamilton S. R., Jones R. J. Inhibition of apoptosis during development of colorectal cancer.
Cancer Res.
,
55
:
1811
-1816,  
1995
.
23
Ansari B., Coates P. J., Grenstein B. D., Hall P. A. In situ end-labeling detects DNA breaks in apoptosis and other physiological states.
J. Pathol.
,
170
:
1
-8,  
1993
.
24
Coulton G. R., Rogers B., Strutt P., Skynner M. J., Watt D. J. In situ localisation of single-stranded DNA breaks in nuclei of a subpopulation of cells within regenerating skeletal muscle of the dystrophic mdx mouse.
J. Cell Sci.
,
102
:
653
-662,  
1992
.
25
Adida C., Berrebi D., Peuchmaur M., Reyes-Mugica M., Altieri D. C. Anti-apoptosis gene, survivin, and prognosis of neuroblastoma.
Lancet
,
351
:
882
-883,  
1998
.
26
Hellemans P., van-Dam P. A., Weyler J., van-Oosterom A. T., Buytaert P., Van-Marck E. Prognostic value of bcl-2 expression in invasive breast cancer.
Br. J. Cancer
,
72
:
354
-360,  
1995
.
27
Levesque M. A., Katsaros D., Yu H., Giai M., Genta F., Roagna R., Ponzone R., Massobrio M., Sismondi P., Diamandis E. P. Immunofluorometrically determined p53 accumulation as a prognostic indicator in Italian breast cancer patients.
Int. J. Cancer
,
79
:
147
-152,  
1998
.
28
Thor A. D., Moore D. H., II, Edgerton S. M., Kawasaki E. S., Reihsaus E., Lynch H. T., Marcus J. N., Schwartz L., Chen L. C., Mayall B. H., Smith H. S. Accumulation of p53 tumor suppressor gene protein: an independent marker of prognosis in breast cancers.
J. Natl. Cancer Inst.
,
84
:
845
-855,  
1992
.
29
Rosen P. P., Lesser M. L., Arroyo C. D., Cranor M., Borgen P., Norton L. p53 in node-negative breast carcinoma: an immunohistochemical study of epidemiologic risk factors, histologic features, and prognosis.
J. Clin. Oncol.
,
13
:
821
-830,  
1995
.
30
Ambrosini G., Adida C., Sirugo G., Altieri D. C. Induction of apoptosis and inhibition of cell proliferation by survivin gene targeting.
J. Biol. Chem.
,
273
:
11177
-11182,  
1998
.
31
Deveraux Q., Takahashi R., Salvesen G. S., Reed J. C. X-linked IAP is a direct inhibitor of cell-death proteases.
Nature (Lond.)
,
388
:
300
-304,  
1997
.
32
Li F., Ambrosini G., Chu E. Y., Plescia J., Tognin S., Marchisio P. C., Altieri D. C. Control of apoptosis and mitotic spindle checkpoint by survivin.
Nature (Lond.)
,
396
:
580
-584,  
1998
.
33
Clark G. M., Allred D. C., Hilsenbeck S. G., Chamness G. C., Osborne C. K., Jones D., Lee W. H. Mitosin (a new proliferation marker) correlates with clinical outcome in node-negative breast cancer.
Cancer Res.
,
57
:
5505
-5508,  
1997
.
34
Clark G. M., Osborne C. K., Levitt D., Wu F., Kim N. W. Telomerase activity and survival of patients with node-positive breast cancer.
J. Natl. Cancer Inst.
,
89
:
1874
-1881,  
1997
.
35
Leek R. D., Kaklamanis L., Pezzella F., Gatter K. C., Harris A. L. bcl-2 in normal human breast and carcinoma, association with oestrogen receptor-positive, epidermal growth factor receptor-negative tumours and in situ cancer.
Br. J. Cancer
,
69
:
135
-139,  
1994
.
36
Silvestrini R., Benini E., Veneroni S., Daidone M. G., Tomasic G., Squicciarini P., Salvadori B. p53 and bcl-2 expression correlates with clinical outcome in a series of node-positive breast cancer patients.
J. Clin. Oncol.
,
14
:
1604
-1610,  
1996
.
37
Lipponen P., Pietilainen T., Kosma V. M., Aaltomaa S., Eskelinen M., Syrjanen K. Apoptosis suppressing protein bcl-2 is expressed in well-differentiated breast carcinomas with favourable prognosis.
J. Pathol.
,
177
:
49
-55,  
1995
.
38
Joensuu H., Pylkkanen L., Toikkanen S. Bcl-2 protein expression and long-term survival in breast cancer.
Am. J. Pathol.
,
145
:
1191
-1198,  
1994
.
39
Knowlton K., Mancini M., Creason S., Morales C., Hockenbery D., Anderson B. O. bcl-2 slows in vitro breast cancer growth despite its antiapoptotic effect.
J. Surg. Res.
,
76
:
22
-26,  
1998
.