Survival analysis in patients with invasive lobular cancer and invasive ductal cancer according to hormone receptor expression status in the Korean population

Background We compared the clinicopathological characteristics and survival outcomes of invasive lobular carcinoma (ILC) cases with those of invasive ductal carcinoma (IDC) cases in various hormone receptor expression subgroups. Methods We compared clinicopathological characteristics, overall survival (OS), and breast cancer-specific survival (BCSS) between patients with IDC (n = 95,486) and ILC (n = 3,023). In addition, we analyzed the effects of different hormone receptor expression subgroups on survival. Results The ILC group had more instances of advanced stage and hormonal receptor positivity than did the IDC group (p < 0.001), but the IDC group had higher histological grade and nuclear grade, as well as higher frequency of human epidermal growth factor receptor 2 and Ki67 expression than did the ILC group (p < 0.001). The OS and BCSS were not significantly different between the IDC and ILC groups. The 5-year OS of the IDC group was 88.8%, while that of the ILC group was 90.6% (p = 0.113). The 5-year BCSS of the IDC group was 94.8%, while that of the ILC group was 95.0% (p = 0.552). When analyzing each hormone receptor expression subgroup, there were no significant differences in survival between the IDC and ILC groups. However, the estrogen receptor (ER) negative/progesterone receptor (PR) negative subgroup showed differences in survival between the IDC and ILC groups. Moreover, the hazard ratio of ILC in the ER negative/PR negative subgroup was 1.345 (95% confidence interval: 1.012–1.788; p = 0.041). Conclusions Hormone receptor expression should be considered when determining prognosis and treatment regimen for IDC and ILC. Researchers should further study the ER negative/PR negative population to identify treatment and prognostic models that will facilitate the development of individualized therapy for these patients, which is needed for good outcomes.

As individualized therapy has become important, studies on hormone receptor expression subtypes have been conducted, mainly in the West. According to the Surveillance, Epidemiology, and End Results (SEER) Program database, compared to IDC, ILC is associated with larger tumor size, older diagnosis age, advanced stage, lower histological grade, higher estrogen receptor (ER)/progesterone receptor (PR) expression, and lower human epidermal growth factor receptor 2 (HER-2) expression. Higher percentages of lymph node positivity and distant metastasis are also found in ILC cases than in IDC cases. In an analysis of hormone receptor expression status that excluded the ER negative/PR negative subgroup, the ER positive/PR positive subgroup showed the best survival, while the ER positive/PR negative subgroup had the worst outcomes [6].
As in the West, studies are being conducted in Asia, including Korea [13][14][15]. However, few have compared invasive breast cancer survival outcomes among different hormone receptor expression subgroups. Therefore, in the present work, we conducted a study on invasive breast cancer in Korea using data from the Korean Breast Cancer Registry (KBCR) to compare and analyze survival among various hormone receptor expression subgroups.

Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments.

Patient selection
In the KBCR database, we identified 98,509 patients with invasive breast cancer diagnosed between 2001 and 2013 who aged more than 18 years old. The KBCR database is a nationwide, Korean, multi-institutional online database. The Korean Breast Cancer Society (KBCS) prospectively keeps the information of patients diagnosed with breast cancer in 102 hospitals. The following information is included: patient identification number, age at operation, sex, tumor stage based on the American Joint Committee on Cancer classification, pathophysiology, and type of surgery. Expression of ER and PR was considered positive if more than 10% of the tumor stained positive, HER-2 status was evaluated using HER-2 overexpression analysis with any grade over 2+ being considered positive. Fluorescence in situ hybridization was used when HER-2 status was graded as 2+, and considered positive if graded 3+ for its result. We excluded patients with metastatic breast cancer at the time of diagnosis, as well as those with carcinoma in situ or poorly evaluated axillary lymph nodes, and those without biological subtype information [S1 Table]. This study was approved by the Institutional Review Board of Asan Medical Center, Seoul, South Korea (20171341). Given that the study was based on retrospective clinical data, the need for informed consent was waived.

Statistical analysis
The clinicopathological features of invasive breast cancer cases were analyzed using a Pearson's chi-square test. We used the Kaplan-Meier method and log-rank test to analyze and compare survival outcomes. Overall survival (OS) was defined as the time from the date of breast cancer diagnosis until the date of death (from any cause) or last follow-up. 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 last follow-up. A Cox proportional hazard analysis was used to obtain hazard ratios (HRs) with 95% confidence intervals (CIs) in uni-and multivariable analyses. All p-values less than 0.05 were considered statistically significant. We used SPSS statistical software, version 26.0 (SPSS Inc., Chicago, USA) for all statistical analysis.

Clinicopathological characteristics of patients with invasive breast cancer
In total, 98,509 patients diagnosed with invasive breast cancer between 2001 and 2013 were selected from the KBCR database, and their data were analyzed. Among them, 95,486 (96.9%) patients had IDC and 3,023 (3.1%) had ILC. The clinicopathological characteristics of the study population are summarized in Table 1. Patients with ILC were older at the time of surgery than those with IDC (�41 years of age at operation: ILC group, 89.3% vs. IDC group, 81.5%; p < 0.001). Compared to the IDC group, the ILC group more frequently presented with advanced stage and positive ER and PR expression (p < 0.001). The IDC group had higher histological grade and nuclear grade as well as higher frequency of HER-2 and Ki67 expression than did the ILC group (p < 0.001).
In Table 2, we have compared the clinicopathological characteristics of the study population according to the hormonal receptor expression subgroups. The ER negative (−)/PR− group presented much higher histological grade and nuclear grade, as well as higher frequency of HER-2 and Ki67 expression than did other groups (p < 0.001). Regarding TNM stage, the ER positive (+)/PR+ subgroup was the least advanced, while the ER−/PR+ subgroup was the most advanced (p < 0.001). Patients in the ER−/PR+ were the oldest and showed the highest frequency of lymphovascular invasion (p < 0.001).

Comparison of survival between IDC and ILC in each hormone receptor expression subgroup
We compared survival between the IDC and ILC populations in each hormone expression subgroup. There were no differences in survival between the IDC and ILC populations in the

Discussion
In the present study, we compared the clinicopathological characteristics and survival outcomes of invasive breast cancer cases in Korea. In Asia, fewer studies than in the West have compared clinicopathological characteristics and survival between ILC and IDC cases of different molecular subtypes. According to a study using the SEER database, compared to patients with IDC, patients with ILC are older at diagnosis, have larger tumor size, show more advanced stage, have lower histological grade, and display more hormone expression positivity [6]. We found similar clinicopathological tendencies in the KBCR database. Several studies have asserted that ILC cases show larger tumor size and more advanced stage than IDC cases because the indistinct tumor growth pattern leads to delays in diagnosis and detection failure [16][17][18][19]. In the present study, total mastectomy was more common in patients with ILC than in those with IDC, as in other studies, probably because of the larger tumor size, more advanced stage, and multifocal tendency.  [6]. In other reports, ILC cases had similar or better survival outcomes compared to those of IDC cases [8][9][10][11][12][13][14]. In the present study, there were no meaningful differences in OS or BCSS between patients with ILC and IDC in the KBCR database.
Breast cancer is a heterogeneous disease with varying hormone receptor status, and each subtype has different clinical features, treatment options, outcomes, and prognoses. For this reason, the hormone receptor subtypes are being studied worldwide. The ER+/PR+ subgroup presented the best survival in the present study, while the ER−/PR− subgroup presented the worst when the total study population was compared according to hormone expression status. The HR of ER+/PR− expression on OS was 1.419 (95% CI: 1.331-1.513; p < 0.001), while that of ER−/PR+ expression was 1.344 (95% CI: 1.237-1.459; p < 0.001) and that of ER −/PR − expression was 1.620 (95% CI: 1.528-1.718; p < 0.001) when ER+/PR+ expression was used as a reference. The HR of ER+/PR− expression on BCSS was 1.516 (95% CI: 1.364-1.685; p < 0.001), while that of ER−/PR+ expression was 1.625 (95% CI: 1.435-1.841; p < 0.001) and that of ER−/PR− expression was 1.915 (95% CI: 1.747-2.098; p < 0.001) when ER+/PR + expression was used as a reference.
In the IDC group, the order of survival HR was similar to that in the total population. In contrast, the ILC group showed a slightly different order of survival HR. When analyzing each hormone receptor expression subgroup, as shown in Fig 5, there were no significant differences in survival between the IDC and ILC populations. However, the ER−/PR− subgroup showed differences in survival between the IDC and ILC populations (5-year OS in IDC group: 87.8% vs. ILC group: 87.5%; p = 0.040; 5-year BCSS in IDC group: 92.0% vs. ILC group: 87.1%; p = 0.049). In some studies, hormone receptor negativity was shown to reduce survival rates in the ILC group. In a study by Francesca et al., triple negative ILC showed the worst survival outcomes (79.7% at 5-years and 73.8% at 10-years) among all histological types. The same study reported that triple negative ILC showed a higher metastatic lymph node ratio (> 0.65) and lower response to chemotherapy than those of other triple negative breast cancer histological types [20]. It is hard to predict outcomes based on current classification and treatment regimens because the ER−/PR− population shows heterogeneity. Therefore, researchers must identify new molecular targets and subtypes. Understanding the heterogeneous molecular subtypes will allow targeted treatments in the future.
There were some limitations to this study. First, it was retrospective, so selection bias may have been present. However, the incidence of ILC is too small to study prospectively. Second, some data were clearly missing during the follow-up period, but we reasoned that the data were valuable and reliable because they were sourced from a large-scale database of one country with long-term follow-up and they corroborated findings of previous studies. Moreover, hormone receptor expression of 1-10% of tumor nuclei positivity was considered as negative, because it was initially registered as negative when the database started to be built. Another limitation may be that the neoadjuvant chemotherapy is not isolated, but in Korea, since it started in the 2010s, the number of patients who received neoadjuvant chemotherapy is small, so it will not have a significant effect.
In conclusion, we reviewed the clinicopathological characteristics and survival outcomes of invasive breast cancer cases in Korea. There was no difference in survival outcomes between ILC and IDC cases in the present study. However, in the ER−/PR− subgroup, the survival outcomes of ILC cases were worse than those of IDC cases. Given that the ER−/PR− group was heterogeneous and the incidence of ILC was low, further large studies are needed to allow comprehensive classification and identification of treatment regimens.