IJC International Journal of Cancer
Heterogeneity of epidemiological factors by breast tumor subtypes in Korean women: A case–case study Nan Song1, Ji-Yeob Choi1,2,3, Hyuna Sung2,4, Seokang Chung2, Minkyo Song2,3, Sue K. Park1,2,3, Wonshik Han1,5, Jong Won Lee6, Mi Kyung Kim7, Keun-Young Yoo3, Sei-Hyun Ahn6, Dong-Young Noh1,5 and Daehee Kang1,2,3 1
Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea Department of Biomedical Sciences, Seoul National University College of Medicine, Seoul, Korea 3 Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Korea 4 Genetic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, Maryland, USA 5 Department of Surgery, Seoul National University College of Medicine, Seoul, Korea 6 Department of Surgery, University of Ulsan College of Medicine and ASAN Medical Center, Seoul, Korea 7 Division of Cancer Epidemiology and Management, National Cancer Center, Goyang-Si, Gyeonggi-Do, Korea 2
Key words: breast cancer, heterogeneity, tumor subtypes, epidemiological factors, case-case study Abbreviations: AMC: ASAN Medical Center; BMI: body mass index; CI: confidence interval; ER: estrogen receptor; FFTP: first full-term pregnancy; FISH: fluorescence in situ hybridization; HR: hormone receptor; HER2: human epidermal growth factor receptor 2; IHC: immunohistochemistry; OR: odds ratio; PR: progesterone receptor; SEBCS: Seoul Breast Cancer Study; SNUH: Seoul National University Hospital Additional Supporting Information may be found in the online version of this article. Grant sponsor: Ministry of Education, Science and Technology (Basic Research Laboratory Program through the National Research Foundation of Korea); Grant number: 2012-0000347 DOI: 10.1002/ijc.28685 History: Received 23 Aug 2013; Accepted 5 Dec 2013; Online 19 Dec 2013 Correspondence to: Ji-Yeob Choi, Department of Biomedical Sciences, Seoul National University College of Medicine, 103 Daehak-Ro, Jongno-Gu, Seoul 110-799, Korea, Tel.: 182-2-740-8922, Fax: 182-2-747-4830, E-mail:
[email protected]
C 2013 UICC Int. J. Cancer: 135, 669–681 (2014) V
Breast cancer is a heterogeneous disease in terms of clinical behavior. Breast tumors have been classified into intrinsic subtypes using gene expression patterns measured by microarrays.1–3 The intrinsic subtypes were explained as the clinicopathological definition based on the expression of estrogen and progesterone receptors (ER and PR), human epidermal growth factor receptor 2 (HER2) and Ki-67.4 For the basallike subtype of breast cancer, epidermal growth factor receptor (EGFR) and cytokeratin 5/6 (CK 5/6) were also used as surrogates.5 These tumor subtypes have shown significant differences in their clinical features.2,3,6 Notwithstanding the clinical heterogeneity of the tumor subtype, the etiological heterogeneity has not been established.7 Through previous epidemiological studies, several reproductive factors, such as early menarche, late menopause and late age at first full-term pregnancy (FFTP), and family history have been established as risk factors of breast cancer, and obesity has been probable as protective or a risk factor of breast cancer depending on the menopausal status.8–10 Furthermore, well-known risk factors of breast cancer have been associated with the ER-positive subtype rather than the ER-negative subtype.8
Epidemiology
Breast cancer is heterogeneous in clinical behavior by subtypes; however, it is unclear how this heterogeneity is related to epidemiological factors. To evaluate the differences in epidemiological factors by breast tumor subtypes, we investigated the associations of epidemiological factors between tumor subtypes in Korean women. From the Seoul Breast Cancer Study, a total of 3,058 patients with breast cancer were included in the analyses. Tumor subtypes were classified based on hormone receptor (HR) and human epidermal growth factor receptor 2 (HER2) statuses. The epidemiological factors of each subtype were compared through case–case analyses using multivariate a polytomous logistic regression model adjusted for age and recruiting centers. The distribution of the subtypes was as follows: 1,714 HR1 HER22 (56.0%), 414 HR1 HER21 (13.5%), 423 HR2 HER21 (13.9%) and 507 HR2 HER22 (16.6%) patients with breast cancer. There were significant differences in age, menopausal status, age at menarche, number of children, age at first full-term pregnancy (FFTP), duration of breastfeeding and duration of endogenous estrogen exposure between tumor subtypes (p < 0.05). Compared to HR1 HER22 patients, the other subtype patients showed more frequency in having more numbers of children and less frequency in having earlier menarche, later FFTP and longer endogenous estrogen exposure. Although HR2 HER21 patients were less obese, HR2 HER22 patients were younger and more obese. In conclusion, age, body mass index and reproductive factors were differentially associated with breast tumor subtypes suggesting a possible distinct etiology for tumor progression.
670
Epidemiological heterogeneity by breast tumor subtypes
What’s new? Breast cancer behaves differently depending on whether tumor cells express certain hormone or growth-factor receptors (ER, PR, HER2, etc.) However, it isn’t clear how these subtypes are related to epidemiological factors. In addition, only a few epidemiological studies have been conducted in Asian populations. In this case-case study of Korean women, the authors found that age, body-mass index, and reproductive factors were differentially associated with the various receptor-status subtypes. These findings provide epidemiological evidence regarding the etiological heterogeneity of breast tumor subtypes.
Although several case–control studies have shown that these risk factors are differentially associated with breast cancer subtypes, those risk factors were classified based on ER and PR statuses only.8,11 Furthermore, there have been many studies that compared epidemiological factors between tumor subtypes based on hormone receptor (HR) and HER2 status in the Western population12–19; however, there has been only a few studies in the Asian population.20,21 In this study, we describe the characteristics of breast cancer classified into tumor subtypes based on ER, PR and HER2 statuses and assess the heterogeneity in association with the epidemiological factors within tumor subtypes as evidence for the etiological differences between breast cancer subtypes through case– case analyses.
Material and Methods
staining. The IHC test is described in previous studies.24,25 Furthermore, the ER and PR statuses were combined and denoted as HR status. For HER2 status, 0 and 11 were considered as negative and 31 as positive by IHC scores. When an IHC score of 21 was recorded, the HER2 status was confirmed either as positive or negative according to the gene amplification ratio of the fluorescence in situ hybridization (FISH). Based on the results of IHC and FISH, breast tumors were classified into the following subtypes such as ER and/or PR positive and HER2 negative (HR1 HER22), ER and/or PR positive and HER2 positive (HR1 HER21), ER and PR negative and HER2 positive (HR2 HER21) and ER and PR negative and HER2 negative (HR2 HER22). When at least either the HR or HER2 status was unknown through the patients’ medical records, it was considered as an unclassified subtype and excluded from the analyses.
Epidemiology
Study population
The Seoul Breast Cancer study (SEBCS) is a multicenterbased case–control study of female breast cancer conducted in Seoul, Korea, as previously described.22,23 A total of 4,040 patients with breast cancer were registered from the Seoul National University Hospital (SNUH) and the ASAN Medical Center (AMC) from 2001 to 2007. After excluding patients with a previous cancer history, 3,689 patients with incident breast cancer were eligible for the study. In case that the tumor subtype of the patients was unknown, the patients were excluded from the analyses. Finally, a total of 3,058 patients with in situ and invasive breast cancer were included in the study. After obtaining written informed consent, information on epidemiological factors including demographic factors, anthropometry, personal and family history of diseases, reproductive factors and lifestyle factors were obtained with face-to-face interviews by trained interviewers using structured questionnaires. Through a review of the computerized medical records of patients, the clinicopathological information including ER, PR and HER2 statuses was abstracted. The study design was approved by the Committee on Human Research of SNUH (IRB No. H-0503-144-004).
Literature review
Tumor subtypes of breast cancer
Statistical analyses
The ER, PR and HER2 statuses were classified as positive or negative based on the laboratory results and a pathologist’s interpretation in the patients’ medical records. The ER and PR statuses were considered as positive when staining of 10% or more was assessed based on immunohistochemistry (IHC)
Breast tumor subtypes (HR1 HER22, HR1 HER21, HR2 HER22 and HR2 HER22) were considered as the outcome variables. The explanatory variables consisted of epidemiological factors including established or probable breast cancer risk factors. The categorical variables were specified as age at
Previous relevant studies on the distribution of breast tumor subtypes were searched by using PubMed with the search restrictions that full-text is available and they were human studies (until October 18, 2013) using the following search term in the title: “breast cancer” and “subtypes.” A total of 409 articles were evaluated according to the inclusion and exclusion criteria (shown in Supporting Information Fig. S1). By reviewing the title and abstract of articles, 196 relevant observational studies were selected after excluding review, animal and in vitro/in vivo studies, clinical trials, meta-analyses and irrelevant studies. The inclusion criteria were as follows: the study population was female patients with incident breast cancer, and the tumor subtypes were classified into at least four groups based on ER, PR and HER2 statuses with available data on the proportion. By reviewing the references of relevant articles, 11 additional studies were included in the literature review. When there was more than one publication for each study population, the article with the larger study population was included in this review. Finally, a total of 65 previous studies on the distribution of breast tumor subtypes are summarized in Supporting Information Table S1.
C 2013 UICC Int. J. Cancer: 135, 669–681 (2014) V
671
Song et al.
C 2013 UICC Int. J. Cancer: 135, 669–681 (2014) V
determine statistical significance. All statistical analyses were performed with SAS 9.3 and STATA 12.
Results Among the 3,058 patients with breast cancer, the mean age was 47.9 6 9.9 years, and 61.8% of the patients were premenopausal women (Table 1). When divided into tumor subtypes, 1,714 (56.0%, 95% CI 5 54.3–57.8%), 414 (13.5%, 95% CI 5 12.4–14.8%), 423 (13.9%, 95% CI 5 12.6–15.1%) and 507 (16.6%, 95% CI 5 15.3–18.0%) patients were classified as the HR1 HER22, HR1 HER21, HR2 HER21 and HR2 HER22 subtype, respectively. There were significant differences for age, menopausal status, age at menarche, number of children, age at FFTP, duration of breastfeeding and duration of endogenous estrogen exposure between the breast tumor subtypes (p < 0.05; Table 1). The associations of the epidemiological factors by tumor subtypes in the case–case analyses are given in Table 2. Overall, the patients with an older age more frequently had HR2 HER21 tumors; however, patients aged below 40 years also frequently had HR2 HER21 tumors (OR5-year increase 5 1.07, 95% CI 5 1.02–1.13), whereas younger patients more frequently had HR2 HER22 tumors (OR5-year increase 5 0.93, 95% CI 5 0.89–0.98) compared to patients with HR1 HER22 tumors. For the HR2 HER21 subtype, women with a BMI of or more than 25.0 kg/m2 were less frequent in this group (OR25 vs.
