SEARCH

SEARCH BY CITATION

Keywords:

  • breast carcinoma;
  • flow cytometry;
  • grade;
  • ploidy;
  • proliferative activity;
  • S phase

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND

The goal of the current study was to investigate the prognostic impact of proliferative activity, together with the other classic clinicopathologic prognostic factors (tumor size, tumor grade, receptor status, ploidy, and lymph node status), in breast carcinoma by counting mitoses and evaluating S phase fraction (SPF) in fresh and frozen tumor samples.

METHODS

From March 1, 1990, to July 1, 1999, a total of 1984 previously untreated invasive breast carcinoma samples were snap-frozen for flow cytometry.

RESULTS

After multivariate analysis incorporating all classic prognostic factors, SPF combined with mitotic activity (i.e., proliferative activity) remained the sole prognostic factor in the lymph node–negative group; proliferative activity was accompanied by tumor size as a prognostic factor in patients with lymph node–positive disease and by lymph node status, lymphatic invasion, and receptor status in the overall population. The predictive value of proliferative activity was superior to that of the reference standards (classic prognostic predictors according to the guidelines of our institution [common oncology practice] and the St. Gallen classification). A review of the literature, focusing on series in which fresh material was used, allowed us to demonstrate that there is widespread agreement regarding the correlation between SPF and prognosis, even after multivariate analysis.

CONCLUSIONS

S phase fraction is a valuable predictor of survival and can confidently be assessed in approximately 80% of cases. In conjunction with mitotic activity, SPF should become a prognostic factor that is used in daily practice by oncologists for the management of breast carcinoma. Cancer 2004. © 2003 American Cancer Society.

The literature on prognostic factors in breast carcinoma is rich, describing numerous predictive factors. Among the most useful factors are tumor size (T status), lymph node status (N status), and histologic grade, which, along with age and hormone receptor status, are used on a daily basis by oncologists in treatment management.1

Nearly half of all patients with breast carcinoma have disease that is limited to the breast, with no axillary lymph node invasion (N0). Although patients with N0 disease have a relatively good prognosis, approximately one-third eventually will die of their disease2–4; thus, there is a need for new prognostic factors that can identify a low-risk subgroup of patients who do not require adjuvant chemotherapy.

Despite the extensive literature on flow cytometry (FCM) in breast carcinoma, this technique is not yet used in therapeutic management, because results appear to be discrepant across the body of relevant literature. Neither American Society of Clinical Oncology (ASCO) conference5, 6 recommended its routine use, due to the lack of standardization and the use of retrospectively chosen cutoff points in specific definitions of high and low S phase fraction (SPF).

As has been demonstrated previously,7, 8 standardized methods that are based on the findings of the American Consensus Conference (ACC) on DNA cytometry in breast carcinoma9 and that employ tertiles with prospectively defined cutoffs respond to the criticisms raised by ASCO. The literature indicates that SPF is one of the most significant prognostic factors for patients with breast carcinoma; SPF was found to be significant in 21 of 24 studies examined, even after multivariate analysis in 15.

In the current study, which involved patients with previously untreated breast carcinoma, we searched for correlations between FCM parameters and classic clinicopathologic factors (T status, tumor grade, receptor status, etc.) and between FCM parameters and survival (disease free, overall, recurrence free, or metastasis free) in the overall population and by lymph node status. The current analysis paid particular attention to patients with low proliferative activity (i.e., SPF and mitotic activity), with a focus on the effect of treatment on outcome.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

From March 1, 1990, to July 1, 1999, as part of a prospective study, 1984 previously untreated invasive breast carcinomas without systemic metastases or any other synchronously or metachronously occurring carcinomas of significant prognostic incidence were sampled for FCM analysis. Clinicopathologic characteristics are summarized in Table 1. All patients were treated at the Centre François Baclesse (Caen, France). Complete information on follow-up, secondary events, and treatment (radiotherapy, hormonotherapy, and/or chemotherapy) was available for all patients. During the study period, 5565 patients were treated for breast carcinoma, with 2904 undergoing a surgical procedure (biopsy, lumpectomy, or mastectomy).

Table 1. Clinicopathologic Characteristics of the Overall Study Population
VariableNo. of patients (%)
  1. CCI:; CLI:.

Tumor size (cm)1772 (100)
 ≤ 2 914 (52)
 > 2 858 (48)
Type of surgery1853 (100)
 Tumorectomy1129 (61)
 Mastectomy 661 (35.75)
 Biopsy  63 (3.25)
Histologic type1853 (100)
 CCI1459 (78.7)
 CLI 313 (17)
 Low-grade  48 (2.5)
 High-grade  33 (1.8)
Scarff–Bloom grade1787 (100)
 I 243 (13.6)
 II 844 (47.2)
 III 700 (39.2)
Modified Scarff—Bloom classification1786 (100)
 Low (1–3) 784 (43.9)
 High (4–5)1002 (56.1)
Mitotic index1833 (100)
 < 10 925 (50.5)
 ≥ 10 908 (49.5)
Maximum no. of mitoses per field1842 (100)
 < 3 897 (48.7)
 ≥ 3 945 (51.3)
Lymphatic invasion1853 (100)
 No1401 (75.6)
 Yes 452 (24.4)
Concentration of at least one receptor (fmol/mg protein)1824 (100)
 < 10 358 (19.6)
 ≥ 101466 (80.4)

Flow Cytometry

Sampling, preparation, and staining

Samples were obtained from the most representative area of the tumor by a pathologist during a frozen section examination and preserved in liquid nitrogen until analysis. As a control, one adjacent section was excised for routine pathologic examination. Tumors that had unrepresentative adjacent sections compared with the whole tumor (lower-grade tumors, most often, because of their decreased mitotic activity) and/or exhibited tissue remnants without tumor cells along with a diploid DNA histogram were excluded from further analysis. When a small number of tumor cells were observed in tissue remnants or the DNA histogram indicated an aneuploid stemline, the case was considered to be representative. Details on the implementation of FCM according to ACC guidelines can be found in our previous study.7

Data analysis

Flow cytometric data analysis was performed using MultiCycle software (McycleAV Version 3.01; Phoenix Flow Systems, San Diego, CA). More extensive details on ploidy classification can be found in our previous study.7 Conditions for data acquisition and histogram analysis essentially were defined according to the guidelines of the ACC.10 SPF was calculated in DNA multiploid cases by considering the two most represented peaks.

SPF results were divided into 3 classes (low, intermediate, and high), according to tertiles and DNA ploidy, as recommended by the ACC.9 We adhered to the criteria set forth by the consensus conference, retaining only histograms with a coefficient of variation < 8 and > 2000 cells and/or 15% of events in the aneuploid cycle studied; also excluded were cases with too much debris (cases with > 20% or with a ‘bad’ slope) and cases exhibiting abnormalities in the software calculation.

Pathology

Tumor size was measured using fresh material. Histologic type was divided into 4 categories: 1) ductal carcinoma; 2) lobular carcinoma; 3) low-grade carcinoma (papillary, cribriform, medullary, mucinous, or tubular); and 4) high-grade carcinoma (undifferentiated and metaplastic or spindle cell). Histologic grade was evaluated according to the Scarff–Bloom (SB) system11 and the modified SB system,12 for which we considered two groups: low modified SB grade (Grade 1–3) and high modified SB grade (Grade 4–5). Quality of surgical margins was assessed whenever possible.

Mitoses were counted in the most proliferative area using the mitotic index (indmit: maximum number of mitoses per 1.7 mm2; and maximit: maximum number of mitoses per high-power field), which is called for in SB grading. Hormone receptor status was determined using the charcoal method, with a clinical cutoff level of 25 fmol/mg.

Statistical Methods

Clinical data and comprehensive data regarding treatment (radiotherapy, hormonotherapy, and/or chemotherapy), follow-up, and pathologic and biologic tumor characteristics were stored in a database. The primary endpoints were the recurrence-free, disease-free, metastasis-free, and overall survival rates (RFS, DFS, MFS, and OS, respectively). DFS was defined as the interval between diagnosis and the first event (recurrence or metastasis). Patients who died without documented disease recurrence or metastasis were censored at the date of death. DFS was calculated according to the Kaplan–Meier method; the differences between curves were assessed with the log-rank test for censored survival data. The Cox regression model was used for multivariate analysis, which included all factors that were found to be significant on univariate analysis and excluded all factors for which P > 0.0001, after stratification according to chemotherapy use. (Stratification according to treatment demonstrated that only chemotherapy was linked to prognosis [P < 0.0001].)

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Population

One thousand eight hundred sixty-one of 1984 cases (93.8%) were retained for analysis after histologic quality controls. Patient age ranged from 23 to 93 years (mean, 58 years; median, 58 years), and follow-up ranged from 1 to 140 months (median, 55 months). Fifty percent of all tumors occurred in the right breast, 48% occurred in the left breast, and 2% were bilateral synchronous tumors. Half of all patients had N0 disease. Tumor size ranged from 4 to 70 mm (mean, 27 mm; median, 20 mm)

Patients were treated according to the running protocols and received the following types of therapy alone or in combination with each other: chemotherapy (42%), radiotherapy (73%), and hormonotherapy (70%). At the end of follow-up, 1392 of 1861 patients (75%) were alive with no evidence of disease, 148 (8%) were alive with disease, 65 had died of intercurrent disease, and 256 (14%) had died of breast carcinoma. DFS rates at 5 and 10 years were 76% and 64%, respectively (Fig. 1). All pathologic results are summarized in Table 1. Regarding mitotic activity in the overall study population, the median indmit was 9 (mean, 15), and the median maximit was 3 (mean, 3.3); we selected 10 and 3, respectively, as the cutoff values for indmit and maximit.

thumbnail image

Figure 1. Disease-free survival among the entire study population.

Download figure to PowerPoint

A comparison of patients with FCM data and those without FCM data revealed no significant difference with respect to date of treatment or patient age. A small but statistically significant difference with respect to clinically evolutive tumors (inflammatory cancer) was found between patients with FCM data (13%) and those without (14%). In the FCM group, tumor grade was higher, and tumors were larger (26 mm vs. 20 mm) and more frequently receptor-positive (80% vs. 60% with a positive status for at least 1 receptor). In addition, in the FCM group, more lymph nodes were found in the axillary clearance (10 vs. 7), and a greater proportion of these lymph nodes were positive (50% vs. 32%; mean number of positive lymph nodes, 1.8 vs. 1.4). Patients in the FCM group received more mastectomies, more hormonotherapy, less neoadjuvant chemotherapy, and more adjuvant chemotherapy; no difference was noted with respect to radiotherapy. The risk of local recurrence or secondary disease (breast or other) was similar in the FCM and non-FCM groups, but more lymph node recurrences and metastases were observed in the FCM group.

Flow Cytometric Results

All but 8 samples (1853 of 1861) were evaluable for DNA ploidy: there were 725 (39%) diploid tumors, 929 (50%) aneuploid tumors, and 201 (10.8%) multiploid tumors. One thousand five hundred fourteen samples (81.7%) were assessed for SPF according to the guidelines of the ACC.10

High SPF and aneuploidy were strongly correlated (P < 0.0001) with large tumor size, high histologic grade, positive lymph nodes, hormone receptor negativity, lymphatic vessel invasion, histologic type, and mitotic activity. Only SPF was correlated with young age.

Prognostic Factors

Univariate analysis of the entire study population indicated that the following clinicopathologic factors were strongly correlated with DFS (P < 0.0001): T status, tumor grade, mitotic activity (indmit and maximit), lymphatic and/or vascular invasion, and lymph node and/or receptor status together with ploidy (diploid vs. nondiploid) and SPF. Age (threshold at 50 years) and postmenopausal status were not correlated with DFS (Table 2). On multivariate analysis, the following factors were found to be correlated with DFS (P < 0.0001): lymph node status, proliferative activity, receptor status, and lymphatic or vascular invasion (Table 3). A graph of DFS versus number of risk factors demonstrates that prognosis becomes significantly worse with each additional unfavorable prognostic factor (Fig. 2). Graphs of DFS according to proliferative activity show 3 groups with significantly different outcomes (P < 0.0001), with tumors with higher proliferative activity associated with poorer outcome. Low-risk patients who did not receive chemotherapy had a better outcome than those who did (P < 0.0001) (Fig. 3), with a similar but nonsignificant tendency in the intermediate-risk arm (P = 0.35); in the high-risk group, chemotherapy use had no effect on outcome.

Table 2. Univariate Analysis of the Impact of Prognostic Factors on Disease-Free Survival
VariableP valueHR (95% CI)
  • HR: hazard ratio; CI: confidence interval; SPF: S phase fraction.

  • a

    Threshold at age 50 years.

  • b

    Positivity for at least one hormone receptor vs. no hormone receptors.

  • c

    Diploidy vs. aneuploidy.

Agea0.540.94 (0.76–1.15)
Tumor size< 0.00012.74 (2.2–3.42)
Tumor grade< 0.00012.77 (2.27–3.37)
Lymphatic/vascular invasion< 0.00013.09 (2.56–3.78)
Lymph node status< 0.00014.37 (3.43–5.57)
Hormone receptor statusb< 0.00010.45 (0.36–0.56)
DNA ploidyc< 0.00010.54 (0.44–0.67)
Mitotic activity< 0.00013.66 (2.92–4.60)
SPF< 0.00012.51 (2.01–3.14)
Table 3. Multivariate Analysis of the Impact of Prognostic Factors on Disease-Free Survival
VariableP valueHR (95% CI)
  1. HR: hazard ratio; CI: confidence interval.

Hormone receptor status< 0.00010.63 (0.49–0.82)
Lymph node status< 0.00013.48 (2.62–4.62)
Proliferative activity< 0.00013.28 (2.38–4.54)
Lymphatic permeation< 0.00011.58 (1.24–2.3)
thumbnail image

Figure 2. Disease-free survival according to number of unfavorable prognostic factors (after Cox analysis). Pf: number of unfavorable prognostic factors.

Download figure to PowerPoint

thumbnail image

Figure 3. Disease-free survival according to adjuvant chemotherapy use for (A) patients with tumors that had low proliferative activity (P < 0.0001), (B) patients with positive lymph nodes and tumors that had low proliferative activity (P = 0.027), and (C) patients with negative lymph nodes and tumors that had high proliferative activity (P = 0.14). 0: without chemotherapy; 1: with chemotherapy.

Download figure to PowerPoint

When only patients with N0 disease (n = 747) were considered, the results of univariate analysis of DFS did not change significantly; the only differences were weaker correlations with ploidy (P = 0.012) and T status (P = 0.0004). After multivariate analysis, two significant factors remained: mitotic activity (indmit) and SPF. The combination of these two factors allowed us to divide patients with N0 disease into three groups (Table 4).

Table 4. Events by Proliferative Activity and Prognostic Group
Prognostic groupLow SPF and low mitotic indexHigh SPF or high mitotic indexHigh SPF and high mitotic indexTotal
  1. LR: local recurrences; M: metastases; DOD: died of disease; DFS: disease-free survival; CT: number of patients who underwent chemotherapy.

Low-risk    
 No. of patients212629283
 Events58215
 LR1124
 M48214
 DOD3216
 5 yr DFS0.970.850.70 
 CT (%)17 (8) 7 (11) 4 (44) 
 % of events   5.3
High-risk    
 No. of patients151168145464
 Events6182953
 LR451423
 M3172444
 DOD3111529
 5 yr DFS0.950.860.81 
 CT (%)17 (12)56 (33)81 (56) 
 % of events   11.4
All patients    
 No. of patients363230154 
 Events112631 
 LR5616 
 M72526 
 DOD61316 
 % of events311.320 

Three hundred sixty-three patients with both low proliferative activity and low SPF exhibited a low rate of carcinologic events (3%); of these 363 patients, 9.3% received chemotherapy, 69% received hormonotherapy, and 73.5% received local radiotherapy. No difference in DFS was observed between patients who received chemotherapy and those who did not in this subgroup.

In the group of 154 patients with both high mitotic index and high SPF, 30 carcinologic events (20%) were observed. Fifty-five percent of these 154 patients were treated with chemotherapy, 60% were treated with hormonotherapy, and 69% were treated with local radiotherapy. In this subgroup, patients who received chemotherapy tended to fare better than those who did not (P = 0.14) (Fig. 3)

Among the 230 patients with either high SPF or high mitotic index (but not both), 23 carcinologic events (11.3%) were observed. Twenty-seven percent of these 230 patients received chemotherapy, 70% received hormonotherapy, and 65% received local radiotherapy. There was no significant difference in DFS between patients who received chemotherapy and those who did not.

To assess the predictive value of proliferative activity, we compared it with reference standards—the factors put forth by the St. Gallen Conference (CSG) and the set of prognostic factors used in daily oncology practice. The comparison of our 3 prognostic groups with those proposed by the CSG (Table 5) shows that according to the CSG, 2% of patients were at low risk (no events), approximately one-third were at intermediate risk (5.4% of events), and two-thirds were at high risk (10.4% of events); according to our criteria, 49% of patients were at low risk (3% of events), 30% were at intermediate risk (10.6% of events), and 20% were at high risk (20% of events). There was no significant difference in the frequency of events according to proliferative activity when patients were stratified using the CSG classification, the converse being wrong.

Table 5. Comparison of the St. Gallen and Centre François Baclesse Classification Systems for Patients with Lymph Node–Negative Disease
St. Gallen risk groupCFB risk group% of events
LowMediumHigh
  1. CFB: Centre François Baclesse (Caen, France); DOD: died of disease; M: metastases; LR: local recurrences.

Low   0
 No. of patients1110 
 No. of events000 
Medium   5.4
 No. of patients170494 
 No. of events651 
 DOD300 
 M351 
 LR401 
High   10.4
 No. of patients177175146 
 No. of events51829 
 DOD41215 
 M41724 
 LR1514 
% of events310.220 

At our institution, patients with N0 disease are eligible for chemotherapy if they are age < 35 years, if their tumors measure > 2 cm, or if they have high-grade disease and/or negative hormone receptor status. Fifteen carcinologic events (5.3%) were observed in the low–clinical risk group (n = 283), compared with 53 events (11.4%) in the high-risk group (n = 464) (Table 4).

Univariate analysis of patients with lymph node–positive disease revealed the following significant prognostic factors for DFS (all with P < 0.0001): T status, hormone receptor status, tumor grade, mitotic activity (maximit and indmit), SPF, ploidy, and lymphatic or vascular invasion. (Only age was not found to be correlated with DFS [P = 0.7].) After multivariate analysis of these significant factors using a Cox model, two prognostic factors remained: tumor size and proliferative activity. Using these factors, we created 3 subgroups of lymph node–positive patients—patients with either 0, 1, or 2 risk factors, who accounted for 5%, 28%, and 43% of carcinologic events, respectively. Based on proliferative activity alone, 3 groups with significantly different outcomes (P < 0.0001) were identified. Only among patients whose tumors exhibited weak proliferation did those who received chemotherapy fare better than those who did not (P = 0.027); in the other two arms, prognosis was not affected by chemotherapy use (Fig. 3).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Discrepancies regarding the relevance of FCM have been widely reported by oncology panels and standard books. These discrepancies can be explained by the type of material used but are partially ascribable to the lack of standardization that prevailed before the publication of the ACC guidelines in 1993. (One should note that only a limited number of studies published since 1993 have specified that they followed these guidelines.)

Implementation of the ACC guidelines on breast carcinoma9 resolves the majority of standardization problems. The use of SPF tertiles, which also was strongly recommended by the ACC, optimizes the delineation of prognostic risk groups and allows the comparison of results among different teams.

In the current study, there were strong correlations between DNA aneuploidy and unfavorable prognostic factors; such as high tumor grade, negative hormone receptor status, unfavorable histologic type, large tumor size, and lymphatic invasion; but no correlation was found between DNA ploidy and age. Analysis of the overall study population demonstrated that the classic first-category prognostic factors (T status, N status, tumor grade, hormone receptor status, and mitotic activity) were correlated with survival. Lymphatic and/or vascular invasion also was a significant prognostic indicator, although it was considered to be only a second-category factor.1

FCM is recognized as one prognostic factor among numerous others, but until the current study, it was considered to be less predictive than T status, tumor grade, hormone receptor status, and N status. Throughout the literature on FCM involving fresh or frozen material, despite the lack of a standardized method (like the one yielded by the ACC guidelines), there is agreement regarding the prognostic impact of DNA ploidy; 17 of 23 studies13–29 (including 813, 15–17, 19, 20, 22, 24 after multivariate analysis) revealed significant prognostic impact. There was even better agreement regarding the prognostic impact of SPF; 21 of 24 studies13–16, 19–23, 25–27, 30–38 (including 1714–16, 19–23, 25–27, 30, 31, 34, 35, 37, 38 after multivariate analysis) revealed significant prognostic impact. The same conclusions were drawn by Wenger et al.,39 who also reviewed the relevant literature (series involving > 100 patients) on SPF and breast carcinoma in the last decade.

Another method for assessing proliferative activity is the evaluation of mitotic activity. We are aware of seven studies that address this topic; all seven found a correlation between mitotic activity and SPF.18, 40–45 Nonetheless, mitotic activity was found to have independent prognostic value in studies46–48 as in the current one. In the current study, we identified the most proliferative area on the periphery of the tumor at low magnification and then counted mitoses in 10 consecutive high-power fields (area, 1.7 mm2; objective, ×40; numeric aperture, 0.75; field diameter, 450 μm). Janninck et al.48 demonstrated that this simple method of counting is at least as effective as more complicated methods that take into account the proportions of tumor and stroma or involve counting in randomly chosen fields.

We are aware of seven studies that focus on N0 disease. In five of these studies,14, 19, 23, 30, 36 including four after multivariate analysis14, 23, 30 with one involving only DNA diploid tumors,19 SPF was found to be correlated with outcome. In 2 other studies, involving 421 and 340 patients, respectively, no correlation between SPF and outcome was found.17, 24

To our knowledge, only four studies have focused on lymph node–positive breast carcinoma and FCM (involving fresh or frozen material). One study49 involved image analysis of a population of 47 patients with short survival duration (< 2 years) compared with 47 patients with long survival duration (> 7.5 years); SPF was found to be correlated with prognosis, but only in univariate analysis. Another study, which involved 1283 patients with lymph node–positive disease,35 found a correlation between SPF and prognosis even after multivariate analysis. Like Simpson et al.,50 we found that a low level of proliferative activity in patients with lymph node–positive disease was correlated with better prognosis, irrespective of chemotherapy use. Chang et al.51 investigated 346 patients with lymph node–positive disease who developed distant or recurrent disease; they also found that SPF was correlated with outcome after multivariate analysis.

We have demonstrated that the objections raised by ASCO regarding FCM in breast carcinoma,5 specifically with respect to previously characterized methods and prospectively defined high-risk and low-risk groups, can be disregarded if the ACC guidelines are adhered to strictly and if tertiles are used. Proliferative activity, whether assessed using SPF or mitotic activity (the former involves S phase, which has significant intertumor variability, and the latter involves M phase, which has essentially fixed length), allowed us to identify a group, representing 48.5% of patients with N0 breast carcinoma, with a low risk of carcinologic events (3% of events in the study population after a mean follow-up period of 55 months). In our opinion, this finding would justify a less aggressive course of treatment for this group of patients.

We compared our risk groups of patients with N0 disease with the risk groups used in the CSG classification.52 The CSG considers patients with Grade 1 tumors measuring < 1 cm and positive receptor status to be at low risk, patients with Grade 1 or 2 tumors measuring 1–2 cm and positive receptor status to be at intermediate risk, and all other patients to be at high risk. We noted that 76% and 35%, respectively, of intermediate-risk and high-risk patients according to the CSG classification were at low risk according to our criteria.

The CSG proposed the use of tamoxifen with or without chemotherapy for all intermediate-risk patients except for older ones, the same therapy for postmenopausal patients with positive receptor status in the high-risk group, and chemotherapy for all other high-risk patients. We believe that these recommendations probably are excessive, particularly for patients in the intermediate-risk group, 76% of whom are at low risk according to our criteria (< 3% of carcinologic events). The same low rate of events was observed among 35% of patients in the CSG high-risk group; for these patients, the treatment recommendations of the CSG, which state that chemotherapy is mandatory for all young patients and/or patients with negative receptor status, are even more excessive.

In the current study, proliferative activity is a salient prognostic factor, with predictive ability that is at least similar to that of classic factors such as age, receptor status, T status, and tumor grade. In fact, fewer events occurred in the low-risk group identified using proliferative factors (3%) than in the low-risk group identified using classic prognostic indicators (5.3%). Opposite results were obtained, with 20% and 11.3% of carcinologic events, respectively, occurring in the high-risk and intermediate-risk groups according to proliferative activity, compared with 11.4% of events in the high-risk group according to classic prognostic indicators.

The results of the current study, combined with the near-consensus in the relevant literature, argue in favor of everyday use of proliferative activity (and thus, SPF) for determining prognosis in breast oncology. The identification of a large group of patients with very low risk of recurrence or metastasis provides a strong argument against the quasisystematic use of adjuvant therapy that has been recommended by several oncology panels and could benefit patients and significantly alleviate the financial burden on the healthcare system.

In conclusion, the current study should encourage other investigators to attempt to reproduce our results with a longer follow-up period. A trial comparing no chemotherapy with chemotherapy for patients considered clinically at risk in the low–proliferative activity group (10% of patients with N0 disease in the current study) certainly would be of great interest but would require a sizable study population. With adjuvant treatment focusing on cell proliferation and cell cycle dynamics, it appears necessary to explore these areas as they relate to therapeutic management. At present, the most studied and accessible method for evaluating proliferative activity remains the determination of SPF using FCM, which was shown to be reliable and reproducible by Bergers et al.53 Other methods, such as immunohistochemical quantification of proteins involved in the cell cycle (e.g., p21, cyclins, p53, Ki-67, etc.), are under development. To the best of our knowledge, these methods have been investigated only in small series, without standardization, and complementary studies involving significant numbers of patients are still required. In a previous study54 involving 104 patients drawn from the current study, Ki-67 exhibited no prognostic impact, a finding that definitely was ascribable to the small study population. At present, evaluation of Ki-67 expression is, at best, a rough estimation made by pathologists. We currently are developing a relatively inexpensive device that would allow semiautomatic assessment of the percentage of stained nuclei on slides, but this device probably will not be commercially available for several years. In the literature, there is a near-consensus regarding potential interest in investigating Ki-67, despite all of the limitations of the method described above.

The FCM technique has been criticized for its lack of reproducible results; however, when this technique is applied correctly, following published guidelines, its results are, in fact, useful. The drawbacks of this technique are related to the need for dedicated equipment, the difficulties associated with analyzing certain histograms, and the inability to evaluate all cases. Such drawbacks are even more pronounced for proteomics and DNA analysis, two novel techniques. These new techniques pose significant problems with respect to the sheer numbers of parameters involved, require significant data processing, and, at present, do not provide reliable prognostic data.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank Françoise Duigou for her technical expertise and Dr. M. Henry-Amar for his pertinent advice regarding the statistical analysis.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    Fitzgibbons PL, Page DL, Weaver D, et al. Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med. 2000; 124: 966978.
  • 2
    Bonadonna G, Valagussa P. Systemic therapy in resectable breast cancer. Hematol Oncol Clin North Am. 1989; 3: 727742.
  • 3
    Pritchard KI. Systemic adjuvant therapy for node-negative breast cancer: proven or premature? Ann Intern Med. 1989; 111: 14.
  • 4
    Harris J, Morrow M, Norton L. Malignant tumors of the breast. In: De VitaVT HellmanS, RosenbergSA, editors. Cancer: principles and practice of oncology. ( 5th edition) Philadelphia: Lippincott-Raven, 1997: 15911592.
  • 5
    American Society of Clinical Oncology. Clinical practice guidelines for the use of tumor markers in breast and colorectal cancer. Adopted on May 17, 1996 by the American Society of Clinical Oncology. J Clin Oncol. 1996; 14: 28432877.
  • 6
    American Society of Clinical Oncology. 1997 update of recommendations for the use of tumor markers in breast and colorectal cancer. Adopted on November 7, 1997 by the American Society of Clinical Oncology. J Clin Oncol. 1998; 16: 793795.
  • 7
    Duigou F, Herlin P, Marnay J, Michels JJ. Variation of flow cytometric DNA measurement in 1,485 primary breast carcinomas according to guidelines for DNA histogram interpretation. Cytometry. 2000; 42: 3542.
  • 8
    Michels JJ, Duigou F, Marnay J. Flow cytometry in primary breast carcinomas. Prognostic impact of proliferative activity. Breast Cancer Res Treat. 2000; 62: 117126.
  • 9
    Hedley DW, Clark GM, Cornelisse CJ, Killander D, Kute T, Merkel D. Consensus review of the clinical utility of DNA cytometry in carcinoma of the breast. Report of the DNA Cytometry Consensus Conference. Cytometry. 1993; 14: 482485.
  • 10
    Shankey TV, Rabinovitch PS, Bagwell B, et al. Guidelines for implementation of clinical DNA cytometry. International Society for Analytical Cytology. Cytometry. 1993; 14: 472477.
  • 11
    Bloom HC, Richardson WW. Histological grading and prognosis in breast cancer. A study of 1409 cases of which 359 have been followed for 15 years. Br J Cancer. 1957; 11: 359376.
  • 12
    Le Doussal V, Tubiana-Hulin M, Friedman S, Hacene K, Spyratos F, Brunet M. Prognostic value of histologic grade nuclear components of Scarff-Bloom-Richardson (SBR). An improved score modification based on a multivariate analysis of 1262 invasive ductal breast carcinomas. Cancer. 1989; 64: 19141921.
  • 13
    Bracko M, Us-Krasovec M, Cufer T, Lamovec J, Zidar A, Goehde W. Prognostic significance of DNA ploidy determined by high-resolution flow cytometry in breast carcinoma. Anal Quant Cytol Histol. 2001; 23: 5666.
  • 14
    Wingren S, Stal O, Sullivan S, Brisfors A, Nordenskjold B. S-phase fraction after gating on epithelial cells predicts recurrence in node-negative breast cancer. Int J Cancer. 1994; 59: 710.
  • 15
    Pinto AE, Andre S, Soares J. Short-term significance of DNA ploidy and cell proliferation in breast carcinoma: a multivariate analysis of prognostic markers in a series of 308 patients. J Clin Pathol. 1999; 52: 604611.
  • 16
    Baldetorp B, Ferno M, Fallenius A, et al. Image cytometric DNA analysis in human breast cancer analysis may add prognostic information in diploid cases with low S-phase fraction by flow cytometry. Cytometry. 1992; 13: 577585.
  • 17
    Balslev I, Christensen IJ, Rasmussen BB, et al. Flow cytometric DNA ploidy defines patients with poor prognosis in node-negative breast cancer. Int J Cancer. 1994; 56: 1625.
  • 18
    Bergers E, van Diest PJ, Baak JP. Cell cycle analysis of 932 flow cytometric DNA histograms of fresh frozen breast carcinoma material. Correlations between flow cytometric, clinical, and pathologic variables. MMMCP Collaborative Group. Multicenter Morphometric Mammary Carcinoma Project Collaborative Group Cancer. 1996; 77: 22582266.
  • 19
    Clark GM, Dressler LG, Owens MA, Pounds G, Oldaker T, McGuire WL. Prediction of relapse or survival in patients with node-negative breast cancer by DNA flow cytometry. N Engl J Med. 1989; 320: 627633.
  • 20
    Ewers SB, Attewell R, Baldetorp B, Borg A, Langstrom E, Killander D. Prognostic potential of flow cytometric S-phase and ploidy prospectively determined in primary breast carcinomas. Breast Cancer Res Treat. 1992; 20: 93108.
  • 21
    Ferno M, Baldetorp B, Borg A, Olsson H, Sigurdsson H, Killander D. Flow cytometric DNA index and S-phase fraction in breast cancer in relation to other prognostic variables and to clinical outcome. Acta Oncol. 1992; 31: 157165.
  • 22
    Lawry J, Rogers K, Duncan JL, Potter CW. The identification of informative parameters in the flow cytometric analysis of breast carcinoma. Eur J Cancer. 1993; 29A: 719723.
  • 23
    Sigurdsson H, Baldetorp B, Borg A, et al. Indicators of prognosis in node-negative breast cancer. N Engl J Med. 1990; 322: 10451053.
  • 24
    Silvestrini R, Daidone MG, Del Bino G,et al. Prognostic significance of proliferative activity and ploidy in node-negative breast cancers. Ann Oncol. 1993; 4: 213219.
  • 25
    Stal O, Wingren S, Carstensen J, et al. Prognostic value of DNA ploidy and S-phase fraction in relation to estrogen receptor content and clinicopathological variables in primary breast cancer. Eur J Cancer Clin Oncol. 1989; 25: 301309.
  • 26
    Stal O, Carstensen J, Hatschek T, Nordenskjold B. Significance of S-phase fraction and hormone receptor content in the management of young breast cancer patients. Br J Cancer. 1992; 66: 706711.
  • 27
    Wenger CR, Beardslee S, Owens MA, et al. DNA ploidy, S-phase, and steroid receptors in more than 127,000 breast cancer patients. Breast Cancer Res Treat. 1993; 28: 920.
  • 28
    Spyratos F, Briffod M, Gentile A, Brunet M, Brault C, Desplaces A. Flow cytometric study of DNA distribution in cytopunctures of benign and malignant breast lesions. Anal Quant Cytol Histol. 1987; 9: 485494.
  • 29
    Meyer JS, Province MA. S-phase fraction and nuclear size in long term prognosis of patients with breast cancer. Cancer. 1994; 74: 22872299.
  • 30
    Stal O, Dufmats M, Hatschek T, et al. S-phase fraction is a prognostic factor in Stage I breast carcinoma. J Clin Oncol. 1993; 11: 17171722.
  • 31
    Chassevent A, Jourdan ML, Romain S, et al. S-phase fraction and DNA ploidy in 633 T1T2 breast cancers: a standardized flow cytometric study. Clin Cancer Res. 2001; 7: 909917.
  • 32
    Pinto AE, Andre S, Pereira T, Nobrega S, Soares J. Prognostic comparative study of S-phase fraction and Ki-67 index in breast carcinoma. J Clin Pathol. 2001; 54: 543549.
  • 33
    Klintenberg C, Stal O, Nordenskjold B, Wallgren A, Arvidsson S, Skoog L. Proliferative index, cytosol estrogen receptor and axillary node status as prognostic predictors in human mammary carcinoma. Breast Cancer Res Treat. 1986; 7 Suppl: S99S106.
  • 34
    Chassevent A, Geslin J, Bertrand G, et al. Place de l'analyse de DNA par cytométrie en flux dans l'évaluation du pronostic des cancers du sein [Place of flow cytometry DNA analysis in the prognosis of breast carcinoma]. Bull Cancer. 1990; 77( Suppl 1): 149S154S.
  • 35
    Clark GM, Wenger CR, Beardslee S, et al. How to integrate steroid hormone receptor, flow cytometric, and other prognostic information in regard to primary breast cancer. Cancer. 1993; 71: 21572162.
  • 36
    Muss HB, Kute TE, Case LD, et al. The relation of flow cytometry to clinical and biologic characteristics in women with node negative primary breast cancer. Cancer. 1989; 64: 18941900.
  • 37
    Ottestad L, Pettersen EO, Nesland JM, Hannisdal E, Fossa SD, Tveit KM. Flow cytometric DNA analysis as prognostic factor in human breast carcinoma. Pathol Res Pract. 1993; 189: 405410.
  • 38
    Sigurdsson H, Baldetorp B, Borg A, et al. Flow cytometry in primary breast cancer: improving the prognostic value of the fraction of cells in the S-phase by optimal categorisation of cut-off levels. Br J Cancer. 1990; 62: 786790.
  • 39
    Wenger CR, Clark GM. S-phase fraction and breast cancer—a decade of experience. Breast Cancer Res Treat. 1998; 51: 255265.
  • 40
    Toikkanen S, Joensuu H, Klemi P. The prognostic significance of nuclear DNA content in invasive breast cancer—a study with long-term follow-up. Br J Cancer. 1989; 60: 693700.
  • 41
    Meyer JS, Friedman E, McCrate MM, Bauer WC. Prediction of early course of breast carcinoma by thymidine labeling. Cancer. 1983; 51: 18791886.
  • 42
    Rondez R, Yoshizaki C, Pirozynski W. Determination of nuclear DNA content and hormone receptors in breast cancer by the CAS 100 cell analysis system as related to morphologic grade and biochemical results. Anal Quant Cytol Histol. 1991; 13: 233245.
  • 43
    Hatschek T, Grontoft O, Fagerberg G, et al. Cytometric and histopathologic features of tumors detected in a randomized mammography screening program: correlation and relative prognostic influence. Breast Cancer Res Treat. 1990; 15: 149160.
  • 44
    Bosari S, Lee AK, Tahan SR, et al. DNA flow cytometric analysis and prognosis of axillary lymph node-negative breast carcinoma. Cancer. 1992; 70: 19431950.
  • 45
    McDivitt RW, Stone KR, Craig RB, Palmer JO, Meyer JS, Bauer WC. A proposed classification of breast cancer based on kinetic information: derived from a comparison of risk factors in 168 primary operable breast cancers. Cancer. 1986; 57: 269276.
  • 46
    Clayton F. Pathologic correlates of survival in 378 lymph node-negative infiltrating ductal breast carcinomas. Mitotic count is the best single predictor. Cancer. 1991; 68: 13091317.
  • 47
    Lynch J, Pattekar R, Barnes DM, et al. Mitotic counts provide additional prognostic information in Grade II mammary carcinoma. J Pathol. 2002; 196: 275279.
  • 48
    Jannink I, van Diest PJ, Baak JP. Comparison of the prognostic value of four methods to assess mitotic activity in 186 invasive breast cancer patients: classical and random mitotic activity assessments with correction for volume percentage of epithelium. Hum Pathol. 1995; 26: 10861092.
  • 49
    Gilchrist KW, Gray R, Driel-Kulker AM, et al. High DNA content and prognosis in lymph node positive breast cancer. A case control study by the University of Leiden and ECOG. ( Eastern Cooperative Oncology Group) Breast Cancer Res Treat. 1993; 28: 18.
  • 50
    Simpson JF, Gray R, Dressler LG, et al. Prognostic value of histologic grade and proliferative activity in axillary node-positive breast cancer: results from the Eastern Cooperative Oncology Group Companion Study, EST 4189. J Clin Oncol. 2000; 18: 20592069.
  • 51
    Chang J, Clark GM, Allred DC, Mohsin S, Chamness G, Elledge RM. Survival of patients with metastatic breast carcinoma: importance of prognostic markers of the primary tumor. Cancer. 2003; 97: 545553.
  • 52
    [No authors listed]. Adjuvant Therapy of Primary Breast Cancer 6th International Conference. Olma Messen St. Gallen. February 25-28, 1998 Abstracts. Eur J Cancer. 1998; 34 Suppl 1: S3S45.
  • 53
    Bergers E, Montironi R, van Diest PJ, Prete E, Baak JP. Interlaboratory reproducibility of semiautomated cell cycle analysis of flow cytometry DNA-histograms obtained from fresh material of 1,295 breast cancer cases. Hum Pathol. 1996; 27: 553560.
  • 54
    Michels JJ, Duigou F, Marnay J, et al. Flow cytometry and quantitative immunohistochemical study of cell cycle regulation proteins in invasive breast carcinoma. Cancer. 2003; 97: 13761386.