A comparison of the clinical outcomes of patients with invasive lobular carcinoma and invasive ductal carcinoma of the breast according to molecular subtype in a Korean population

Background To investigate the clinicopathological characteristics and the survival outcomes of invasive lobular carcinoma (ILC) patients compared to invasive ductal carcinoma (IDC) patients according to their molecular subtype. Methods We compared the clinicopathological characteristics, breast cancer-specific survival (BCSS) and overall survival (OS) between patients with IDC (n = 14,547) and ILC (n = 528). Results The ILC presented with a larger tumor size, more advanced cancer stage, increased rate of hormonal receptor positivity, human epidermal growth factor 2 (HER2) negativity and mastectomy than the IDC. The ILC patients more frequently presented with the luminal A subtype, whereas the IDC patients more frequently presented with the luminal B, HER2-overexpression, or triple negative subtype. The BCSS and OS were not significantly different between the IDC and ILC for each molecular subtype. Conclusions Similar to IDC patients, molecular subtype should be considered when determining the prognosis and treatment regimen for ILC patients.


Background
Invasive lobular carcinoma (ILC), also known as infiltrating lobular carcinoma, is the second most frequent histological subtype of breast cancer. It was first described by Foote and Stewart in 1941 and it is found in approximately 5 to 15% of patients in western countries [1][2][3][4]. The incidence rate of ILC has steadily increased over the last 20 years [5].
In the past, the prognosis of ILC compared with invasive ductal carcinoma (IDC) has been controversial [6][7][8][9][10]. Although ILC has been reported as more multifocal and bilateral than IDC [11][12][13][14], IDC and ILC present with similar clinical manifestations. Moreover, the treatment strategies for IDC and ILC are similarly based on TNM staging.
The importance of molecular subtype in the treatment and prognosis of IDC has been increasingly emphasized in the recent literature. However, relatively little is known aboutILC despite its increasing incidence. Moreover, reports on the prognostic significance of the molecular subtypes of ILC in Asia are limited because there is a lower incidence of ILC in Asia compared to western countries [15,16].
Therefore, in this study, we aimed to compare the association between the molecular subtype and the clinical outcomes of IDC and ILC in Korea using patient information from the nationwide Korean Breast Cancer Registry (KBCR) database. The primary objective of this investigation was to compare the survival outcomes of IDC and ILC according to molecular subtype. The second objective was to determine the association between various clinicopathological factors and survival outcomes.

Korean breast cancer registry
The KBCR database is a nationwide database that includes 41 university hospitals and 61 surgical training hospitals [17]. This database provides information pertaining to patient survival, sex, age, the surgical method used, the stage of cancer based on the seventh American Joint Committee on Cancer (AJCC) classification, the pathological characteristics of the patient's tumor, and any adjuvant treatment received.

Study population
All female breast cancer patients who were listed in the KBCR and were diagnosed between January 1995 and December 2006 were selected for this study. Clinicopathological data, including age, date of surgery, method of surgery, tumor size at presentation, axillary lymph node status, TNM stage andlymphovascular invasion were collected. Immunohistochemical results evaluating the expression of the estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2(HER2) were also collected. A patients was considered ER and PR positive if 10% or more of their tumor was positively stained. For HER2, an immunohistochemical staining score of 3+ was considered positive. Because fluorescence in situ hybridization (FISH) was unavailable during most of the study period, a HER2 score of 2+ was considered negative.
All of the patients at risk for relapse received adjuvant chemotherapy followed by local radiotherapy and/or hormonal therapy according to the recommended therapeutic regimen at the time of surgery as determined by international guidelines, such as the national comprehensive cancer network (NCCN).
Patients were excluded if they had metastatic disease at the time of presentation, bilateral breast cancer, a history of previous malignancy, or had received neoadjuvant chemotherapy. We also excluded patients who were not treated with a curative intent (no surgery, no axillary staging, or had tumor tissue remaining after their final surgery), patients without follow-up data, and patients whose ER, PR, HER2, and lymphovascular invasion status was unknown.

Statistical analysis
Molecular subtype was categorized as follows: luminal A (LA; ER + and/or PR+, and HER2-), luminal B(LB; ER + and/or PR+, and HER2+), HER2+ (HER2; ER-and PR-, and HER2+), and triple negative (TN; ER-and PR-, and HER2-). Breast cancer-specific survival (BCSS) was defined as the time from the date of breast cancer diagnosis until the date of breast cancer-related death or the date of the last follow-up. Overall survival (OS) was defined as the time from the date of breast cancer diagnosis until the date of death (from any cause) or the date of the last follow-up.
Characteristic differences between the IDC and ILC groups were compared using independent t-test and chi-square analyses, as appropriate. Survival curves were obtained using the Kaplan-Meier method, and the survival curves were compared using the log rank test. Multivariate Cox proportional hazard regression analysis was used to assess the independent prognostic significance of various clinical and histopathological characteristics of the tumors. All of the statistical analyses were performed using SPSS version 11.0 (SPSS, Chicago, IL,USA).

Patients' characteristics and distribution based on molecular subtype
A total of 41,813 patients diagnosed with breast cancer between 1995 and 2006, whose information was available in the KBCR database, were selected for this study. After exclusion, we identified 15,075 invasive breast cancer patients. Of the 15,075 patients in the study population, 14,547 (96.5%) presented with IDC and 528 (3.5%) presented with ILC. The clinical, demographic, and treatment features of the patients in the study population are summarized in Table 1.
The ILC patients presented with larger (P < 0.001) and more advanced stage (P = 0.001) tumors. The rate of hormone receptor positivity (P < 0.001) and HER2 negativity (P < 0.001) was increased in the ILC patients, as was the rate of mastectomy (P < 0.001). Statistically significant differences in the percentage of patients receiving postoperative external radiotherapy might be explained by the reduced percentage of ILC patients receiving breastconserving operations (Table 1). Table 1 also shows the distribution of the IDC and ILC patients based on their molecular subtypes. Whereas the ILC patients more frequently presented with the luminal A subtype, the IDC patients more frequently presented with either the luminal B, HER2-overexpression, or triple negative subtype. The differences in the molecular subtypes between the IDC and ILC patients were statistically significant (P < 0.001).

Breast cancer-specific survival and overall survival of the IDC and ILC patients
The median follow-up period was 81.91 months, 82.37 months, and 69.41 months for the total patient populations, the patients with IDC, and the patients with ILC, respectively. Figure 1 shows the survival curves for the IDC and ILC cohorts. The BCSS (Figure 1a, P = 0.500) and OS (Figure 1b, P = 0.503) were not significantly different between these two groups.  Effect of molecular subtype on the breast cancer-specific survival and overall survival of IDC and ILC patients Figure 2 shows the impact of molecular subtype on breast cancer-specific and overall survival in the IDC and ILC patients.

Discussion
This study compared the clinicopathological characteristics and survival outcomes of patients with ILC and IDC according to their molecular subtype in a Korean population. Several studies have reported the relatively lower incidence rate of ILC cases in Asia compared to western countries [15,16]. In a study by Ko et al. [15], the incidence rate of ILC among the whole breast cancer population was reported to be 2 to 4% in Korea. Consistent with previous reports, the incidence rate of ILC was 3.5% in our study [15,16].
Our study show that ILC is associated with increased tumor size, advanced TNM stage, hormone receptor positivity, HER2 negativity, and increased mastectomy rate compared to patients with IDC. Among these characteristics, increased tumor size and advanced staging are generally accepted as poor prognostic factors. Therefore, discrepancies between OS and BCSS for the IDC and ILC cases are expected. Although it is not possible to determine precisely why the IDC and ILC cases exhibited similar patterns of OS and BCSS, despite not sharing any known prognostic markers, a reduction in HER2 positivity and an increase in the percentage of patients with the luminal type A subtype may contribute to the relatively favorable prognosis of ILC cases. However, stage III cancer status and the TN molecular subtype were found to be correlated with a decreased OS and BCSS in the ILC cases. Thus, attending physicians should discuss these matters with their patients when deciding on adjuvant therapy.
Some studies have shown that compared to IDC patients, ILC patients present with larger tumors at the time of diagnosis [11,[18][19][20]. Consistent with these reports, the median tumor size in our study was higher in the ILC group (2.5 cm) compared to the IDC group (2.0 cm). One possible explanation for these results is  the indistinct growth pattern of these tumors, which renders ILC unclear and sometimes invisible in clinical and mammographic investigations [21][22][23]. This unique characteristic of ILC might contribute to its late diagnosis and, consequently, to the increased size and TNM staging of the tumor at the time of diagnosis. However, despite conflicting reports regarding the degree of lymph node positivity in ILC patients compared to patients with IDC [6,23,24], we found no significant differences in lymph node positivity between the analyzed IDC and ILC cases. In this study, the incidence of hormone receptor positivity was significantly increased in the patients with ILC compared to the patients with IDC, which is consistent with previous reports [24][25][26]. We also confirmed an increased incidence of the luminal subtype in the ILC patients compared to the IDC patients. In addition, HER2 positivity was significantly lower in the ILC patients compared to the IDC patients, which is consistent with previous reports showing an increased incidence of HER2-tumors in patients with ILC compared to patients with IDC [11,26,27]. However, because we considered a HER2 score of 2+ to be HER2-due to a lack of FISH amplification information for the study period, we cannot exclude the possibility that the HER2 positivity was underestimated, which would influence the incidence of both the luminal B and HER2-overexpression subtypes. However, it has been shown in various studies that approximately 25 to 50% of HER2 2+ tumors are positive by FISH amplification [28,29], and this value may provide a better estimation of the molecular subtype distribution in this study population.
Mastectomy was performed more often for the ILC patients than the IDC patients. This trend is likely due to the larger size of ILC tumors as well as their multifocality compared to IDC tumors. In addition, because the rate of breast conservation surgery is reduced in ILC patients, postoperative radiation therapy was performed less frequently in these patients, which is consistent with previous reports [20,24]. Additionally, some reports have suggested a relatively higher rate of multifocality and multicentricity for ILC tumors compared to IDC tumors, which may influence the increased rate of mastectomy in these patients [21,26]. However, we could not investigate the association between these factors because the relevant information was not available.
There are many conflicting reports regarding the prognostic differences between ILC and IDC. Pestalozzi et al. [1] reported a poorer survival outcome for ILC than for IDC, while many reports have suggested a similar or more favorable survival outcome for ILC patients compared with IDC patients [30][31][32][33][34]. In this study, we demonstrated that the survival outcome for the ILC patients was similar to that of the IDC patients. Furthermore, although the triple negative cohort of ILC patients tended to exhibit a worse survival outcome than the same cohort of IDC patients, we also found similar survival outcomes between the IDC and ILC patients of each molecular subtype. Accordingly, similar to IDC, we conclude that the molecular subtype classification should be considered as an important prognostic indicator for ILC patients. Moreover, a recent subgroup analysis of the HERA trial revealed the beneficial effects of trastuzumab therapy on survival in HER2 + ILC patients [35], which reflects the importance of incorporating molecular subtype classification into the therapeutic treatment of ILC to produce a better clinical outcome.
To our knowledge, this is the first report comparing the differences in survival outcome between IDC and ILC according to molecular subtype within an Asian population. However, our study has several limitations. First, the study design is retrospective; therefore, a risk of selection bias is present. Second, due to the lack of information and the need of a consistent protocol for determining positivity in immunohistochemical staining, we could not apply Ki-67 values for the classification of the luminal subtypes. Finally, the sample sizes of certain ILC subgroups are relatively small, and a larger study is warranted for this type of analysis.

Conclusion
In conclusion, despite some characteristic differences, our study demonstrated a similar survival outcome for ILC patients among all molecular subtypes compared to IDC patients. Although studies with a larger sample size and a longer follow-up period should be performed to

Competing interests
There is no conflict of interest regarding this study