Independent factors FOR poor prognosis in young patients with stage I-III breast cancer.

Breast cancer is the most common malignancy in the population of women under 40 years of age. Young age is an independent factor for poor prognosis. In this research, we tried to establish other factors for poor prognosis in stage I-III breast cancer. The following parameters were observed: tumor size, lymph node status, histologic grade, hormonal receptor status, Ki-67 prognostic index, Her2 neu status, histologic type of the tumor, local recurrence and metastases. Logistic regression was used to evaluate the effect of specific factors on the probability of lethal outcome and development of distant metastases. Our patients showed a predominance of T1 tumor (49.4%), had positive lymph nodes (62%) and most of them were pN1 (61.2%). Up to one-third of patients had triple negative status. Ki-67 proliferation index was high (25%). Multicentric tumor was detected in 23% of patients. There was no difference in overall survival between the two types of surgical procedures. Patients with pN0 status had better overall survival. Breast cancer in the population of young women has a more aggressive nature. Study results indicated positive lymph node status as an independent factor for poor prognosis of stage I-III breast cancer.

Introduction

Breast cancer represents a global public health issue. It is the most common malignancy and mortality factor in the population of women. Due to the expansion of novel therapeutic options, from surgical and oncologic perspective, and with greater potential for recovery, breast cancer is one of the most researched malignancies in the past twenty years. Young women with breast cancer are considered to be those under 40 years of age (, ). In this population, breast carcinoma represents the most common malignant disease with highest mortality, although breast cancer accounts for 6% of the total number of cancer cases (). Numerous clinical studies have confirmed that breast cancer in young women has a higher histologic grade, unfavorable hormonal status, and overall higher death rate compared to the older population of women (-). Young age is an adverse prognostic factor in primary breast cancer. Various studies found young women to have worse outcomes than older patients (, ). Young age is also an independent risk factor for disease-free survival (DFS) and overall survival (OS) in women with operable breast cancer (, ). In the group of young women, there are limited data on prognostic factors. Only a few studies recognize nodal status, estrogen receptor status and molecular subtype as independent factors for poor survival (, ).

We recognized young age as an independent factor for poor prognosis but also wanted to establish other possible independent factors for DFS and OS in the group of young women with breast cancer from our population.

Patients and Methods

This was a retrospective study. Data were collected from medical charts of young female patients having undergone surgical treatment at the Department of Surgery, Osijek University Hospital Centre, Osijek, Croatia, during the 2002-2010 period. Study patients were under 40 years of age at the time of surgery. Data on the following parameters were collected: tumor size, lymph node status, histologic grade, hormonal receptor status, Ki-67 prognostic index, Her2 neu status, histologic type of tumor, local recurrence, metastases, and outcome (alive or death). All patients included in the study had complete medical documentation. Those with incomplete documentation and data were excluded from the study. All study patients had stage I-III tumor. Patients who had positive distant metastasis (M) at the time of diagnosis were not included in the study. All patients were monitored during routine and urgent checkups. All patients included in this study were monitored for 5-13 (median 8.7) years. We collected data on 79 young female patients, median age 37 (range 33-39) years.

Breast tissue was fixed, embedded in paraffin and cut into 5-µm sections. After that, it was stained and observed under a microscope. Hormonal status, Her2 neu status and Ki 67 were determined by immunochemical staining. Among patients with Her2 neu 2+, the FISH method was performed to determine whether the patient was Her2 neu positive or negative. Histologic grading of breast carcinoma was performed by Elston and Ellis method. We used the TNM (tumor-nodes-metastases) classification of the American Joint Committee on Cancer Staging System for Breast Cancer, 2010 (, ).

Local recurrence is defined as recurrence in the field of mastectomy or in the original tumor location for breast conserving resection. Local recurrence was diagnosed pathologically after surgical biopsy. Metastases are defined as recurrence of the tumor in distant organ during regular or urgent follow up examinations using abdominal ultrasound, lung radiography, bone scintigraphy, surgical biopsy, computed tomography (CT), magnetic resonance imaging (MRI) and positron emission/computed tomography (PET/CT) scans. Outcome was defined as the patient being dead or alive at the end of the follow up period.

Statistical analysis was performed with SPSS 13.0 (Chicago, IL, USA) software. Logistic regression (univariate and multivariate) was used to evaluate the impact of several factors on the probability of negative outcome in study subjects.

Results

Data were collected on 79 patients, median age 37 (range 33-39) years (Table 1). There were no significant differences according to the side operated on. Invasive ductal carcinoma was the most common histologic type of tumor (70.9%). Invasive lobular carcinoma was the second most common type (13.9%).

Table 1

Study parameters in young women with breast cancer

Affected side, n (%)
right	40 (50.6)
left	39 (49.4)
Histologic type, n (%)
ductal invasive	56 (70.9)
lobular invasive	11 (13.9)
other	12 (15.2)
Tumor size (T), n (%)
T1 ≤2 cm	39 (49.4)
T2 >2 cm, ≤5cm	30 (38)
T3 >5 cm	7 (8.9)
T4 any size with skin or chest wall spread	3 (3.8)
Positive lymph nodes, n (%)	49 (62)
Lymph nodes status, n (%)
1-3 lymph nodes	30 (61.2)
4-9 lymph nodes	13 (26.5)
≥10 lymph nodes	6 (12.2)
Estrogen status, n (%)
negative	36 (45.6)
positive	43 (54.4)
Progesterone status, n (%)
negative	35 (44.3)
positive	44 (55.7)
HER2 neu status, n (%)
negative	60 (75.9)
positive	19 (24.1)
Triple negative, n (%)	25 (32.1)
Ki67, median (25%-75%)	25 (11-48)
Type of surgical procedure, n (%)
modified radical mastectomy	52 (65.8)
breast conserving resection	27 (34.2)
Multicentric tumor position, n (%)	18 (23.1)
Histologic grade, n (%)
I	13 (16.5)
II	43 (54.4)
III	23 (29.1)
Local recurrence, n (%)	MRM 0 (0)BCR 3 (3.8)
Metastases, n (%)	21 (26.6)
Outcome, n (%)
alive	64 (81)
dead	15 (19)

Regarding the size of the tumor, most patients had T1 and T2 (87.4%) tumor size, with a predominance of tumor size ≤2cm (49.4%). The majority of patients had positive lymph nodes (62%), most of them with pN1 status (1-3 positive nodes; 61.2%). There were no significant differences in estrogen or progesterone status. The majority of patients were her2 neu negative (75.9%) and up to one-third of patients had triple negative immunohistochemical status. Ki-67 proliferation index was rather high (25%; interquartile range 11% to up to 46%). We found a high incidence of multicentric tumors (n=18, 23.15%) in our group of young women. More than one-half of patients were histology grade II (54.4%) (Table 1).

Modified radical mastectomy was the preferred type of surgical procedure, performed in 65.8% of patients. Only three patients developed local recurrence. They all were primarily treated with breast conserving resection.

At the end of follow up, metastases were detected in 26.6% of patients and 15 (19%) of them died (Table 1).

Effect of parameters on predicting lethal outcome

Logistic regression was used to estimate the potential impact of specific factors on the possible lethal outcome in the study group of patients. The model contains fourteen parameters: age, histologic type, tumor size, lymph node status, estrogen status, progesterone status, Her2 neu, Ki-67, type of surgical procedure, multicentric position, triple negative status, histology grade, local recurrence, and metastases (Table 2).

Table 2

Impact of parameters on prediction of lethal outcome (univariate regression analysis)

Variable	β	p	Odds ratio(OR)	95% confidence interval
Age	-0.025	0.75	0.975	0.84 -0.94
Histology type  ductal invasive  lobular invasive  other	-1.003 -0.087	0.840.360.91	0.3670.917	0.04-3.150.17-4.89
Tumor size T (cm)	0.079	0.54	1.083	0.837-1.40
Lymph nodes status  pN0  pN1  pN2  pN3	2.1782.5563.367	0.040.030.030.01	8.812.8929	1.012-76.91.27-130.52.3-373.7
Estrogen  negative  positive	-1.073	0.04	0.342	0.105-0.812
Progesterone  negative  positive	-0.785	0.18	0.456	0.145-1.44
HER2 neunegativepositive	-0.288	0.68	0.750	0.188-2.9
Ki-67 (%)	0.019	0.03	1.019	0.797-0.97
Type of surgery  BCR  MRM	0.875	0.21	2.4	0.614-9.4
Multicentric position  no  yes	1.041	0.09	2.833	0.845-9.49
Triple negative  no  yes	0.783	0.18	2.187	0.691-6.92
Histology grade  I  II  III	0.2290.424	0.890.790.64	1.2571.528	0.232-6.80.252-9.3
Local recurrence  no  yes	2.271	0.07	9.692	0.82-114.9
Metastases  no  yes	22.119	0.997	4.0x109	0

BCR = breast conserving resection; MRM = modified radical mastectomy

From the predictor variables that showed statistical significance, and exclusion of variables that did not change the probability (P) model by 20%, the following five variables were selected for the model on which to apply multivariate logistic regression: estrogen and progesterone positive status, lymph node status, Ki-67, and local recurrence. The model was statistically significant, χ2=20.4 (df=7, p=0.005), and explained between 22.7% (after Cox & Snell) and 36.6% (after Negelkerke) of variance for death outcome, and accurately classified 86.1% of cases. The strongest independent factor for death outcome was lymph node status (Table 3).

Table 3

Impact of parameters on prediction of lethal outcome (multivariate regression analysis)

Variable	β	p	Odds ratio(OR)	95% confidence interval
Lymph nodes status  pN0  pN1  pN2  pN3	2.0163.4333.579	0.040.090.010.01	7.50930.9735.83	0.707-79.7702.044-468.92.339-548.8
Estrogen positive	-1.856	0.14	0.156	0.013-1.871
Progesterone positive	0.677	0.57	1.968	0.191-20.296
Ki-67 (%)	0.019	0.17	1.019	0.992-1.048
Local recurrence  yes	2.240	0.12	9.398	0.553-159.78
Constant	-3.843	0.005	0.021

Effect of parameters on predicting development of metastases

The probability of developing distant metastases in the study sample was evaluated by screening for the specific factors applying the logistic regression methods. The model contains thirteen parameters: age, histologic type, tumor size, lymph node status, estrogen status, progesterone status, Her2 neu, Ki-67, type of surgical procedure, multicentric position, triple negative status, histology grade, and local recurrence (Table 4).

Table 4

Impact of parameters on prediction of metastasis development (univariate regression analysis)

Variable	β	p	Odds ratio(OR)	95% confidence interval
Age	-0.157	0.03	0.855	0.745-0.981
Histology type  ductal invasive  lobular invasive  other	-0.674-0.556	0.810.420.51	0.5100.574	0.099-2.6140.110-2.989
Tumor size T (cm)	0.048	0.70	1.049	0.823-1.337
Lymph nodes status  pN0  pN1  pN2  pN3	1.3501.7272.890	0.040.060.040.006	3.8575.62518	0.927-16.0481.097-28.832.26-143.3
Estrogen  negative  positive	-0.905	0.08	0.404	0.145-1.128
Progesterone  negative  positive	-0.708	0.17	0.493	0.179-1.356
HER2 neu  negative  positive	0.651	0.25	1.917	0.633-5.802
Ki-67 (%)	0.007	0.52	1.007	0.987-1.027
Type of surgery  BCR  MRM	-0.671	0.25	0.511	0.164-1.592
Multicentric position  no  yes	1.386	0.015	4	1.305-12.256
Triple negative  no  yes	0.370	0.49	1.448	0.508-4.128
Histology grade  I  II  III	0.7560.878	0.600.370.33	2.1292.406	0.410-11.0570.419-13.832
Local recurrence  no  yes	22.37	0.99	5.2 x109	0

BCR = breast conserving resection; MRM = modified radical mastectomy

Four predictor variables that showed statistical significance were chosen for the model on which multivariate logistic regression was applied. These variables were found to be predictors for metastasis development: age, lymph node status, estrogen status and multicentric position. The model was statistically significant, χ2=20.1 (df=5, p=0.001), and explained between 22.5% (after Cox & Snell) and 32.7% (after Negelkerke) of variance for metastasis development, and accurately classified 78.5% of cases. The strongest independent factor for metastases was lymph node status (Table 5).

Table 5

Impact of parameters on prediction of metastasis development (multivariate regression analysis)

Variable	β	p	Odds ratio(OR)	95% confidence interval
Lymph nodes status  pN0  pN1  pN2  pN3	1.6261.9953.523	0.070.040.060.02	5.0867.35433.874	1.020-25.3490.953-56.7291.718-667.8
Estrogen positive	-1.52	0.03	0.219	0.054-0.885
Age	-0.160	0.05	0.852	0.723-1.003
Multicentric position, yes	0.719	0.39	2.052	0.054-0.885
Constant	3.854	0.18	47.1

Kaplan-Meier’s analysis of survival

Kaplan-Meier’s analysis of survival was applied in patients with distant metastases. Patients without positive lymph node (n=8) had OS at 7.8 years (95% CI 7.3-8,2). Patients with N1 (n=27) had OS at 6.5 years (95% CI 5,6-7,4). Lowest OS was recorded in patients with N2 (n=13), at 5.2 years (95% CI 3.8-6.7). Patients with N3 (n=3) had OS at 6.7 years (95% CI 3.8-9.5). Patients with positive N had poor OS, but without differences according to the number of positive nodes (Mantel-Cox test, p=0.08) (Fig. 1).

Fig. 1

Kaplan-Meier’s analysis of survival with lymph node involvement.

The mean survival period in our group of young patients was 7.3 years (95% CI 6.6 to 8.1). Survival time of patients operated on with modified radical mastectomy with axillary dissection was 6.8 years (95% CI 5.7 to 7.8), and of patients operated on with breast conserving resection 7.3 years (95% CI 6.6-8), with no significant difference (Mantel-Cox test, p=0.13) (Fig. 2).

Fig. 2

Kaplan-Meier’s analysis of survival with type of surgical procedure. BCR = breast conserving resection; MRM = modified radical mastectomy

Discussion

Breast cancer in the population of young women accounted for only 5%-7% of all breast cancer patients (). In Far East countries, the incidence of breast cancer in young patients was found to be twofold that in developed countries (). Breast cancer among young women is considered to have a more aggressive nature and higher tumor grade with more vascular invasion than breast cancer in older women (). When compared to older women, breast cancer in younger women tends to have a worse prognosis in terms of OS and DFS (, ). Young age has been identified as an independent factor for poor prognosis (, ).

In our sample, we found invasive ductal carcinoma as the most frequent histologic type of tumor. The second most common type was invasive lobular carcinoma. Among other different studies, Rosenberg et al. in their case-control study confirmed invasive ductal carcinoma as the most frequent one, and invasive lobular carcinoma as the second most common (, ). We found a predominance of tumors smaller than 5 cm, especially T1, which is consistent with the findings reported by Colleoni et al. (). Prognostic value of tumor size is independent of the state of lymph nodes. Patients with small tumors have better prognosis (, ). The prognosis of multicentric tumor is worse than for solid tumors of similar size (often axillary metastases) (). Multicentric cancer has worse biologic behavior with frequently present multiple foci and therefore should be considered in planning adjuvant treatments (). We found a rather high proportion of multicentric breast cancer.

Positive axillary lymph nodes have been considered as the most important prognostic factor for poor prognosis of breast cancer patients (). We found the majority of our patients to have positive lymph nodes. Our positive lymph node patients had worse OS and DFS (positive distant metastases) than lymph node negative patients, however, without difference between pN1-pN3 lymph node status. This is somewhat different from the findings reported from one study, which established difference in OS between pN1 and a greater number of lymph nodes involved (). Tai et al. also report on difference in prognosis within pN1 group between 1-2 and 3 lymph nodes involved (). Our findings might be somewhat imprecise due to the small sample size.

Triple-negative breast cancer is more aggressive, more likely to metastasize, and has more frequent recurrences after treatment. This is visible in the first 5 years after treatment. Generally, triple negative breast cancer has poor prognosis and higher mortality rate compared to other molecular subtypes (). In our study sample, up to one-third of patients were triple negative. Studies emphasize that the Ki-67 index ≥20 is a poor prognostic sign. Breast cancer patients with high Ki-67 had a significantly worse prognosis due to local recurrence, distant metastasis and poor OS (, ). In our study, Ki-67 was found to be 25%, which is a high value pointing to poor prognosis of the disease in our group of young women.

When compared with modified radical mastectomy, breast conserving surgery is similar in OS and the chance for distant metastasis development. Of course, breast conserving resection must be combined with radiotherapy. Breast conserving resection has more local recurrences; all local recurrences in our sample were in the breast conserving resection group. We proved that there was no difference between these two surgical procedures in OS, which is consistent with recent literature (, ).

On statistical analysis, we identified local recurrence, hormone receptor status, lymph node status and Ki-67 index as predictors of poor OS. Lymph node status, hormone receptor status and multicentric position were predictive of developing distant metastases. However, lymph node status was an independent factor for both OS and DFS. Other authors recognize hormone receptor status and lymph node status as predictors of poor prognosis (, ).

Conclusion

Breast cancer in the population of young women has a more aggressive nature, positive lymph nodes, often triple negative molecular subtype, high Ki-67 index, multicentric tumor position and distant metastases. These are predictors of poor prognosis, but positive lymph node status is an independent prognostic factor for OS and DFS.