Fracture in Asian Women with Breast Cancer Occurs at Younger Age

Background Western breast cancer survivors have an increased risk of osteoporosis and bone fracture. Breast cancer occurs 10 to 20 years earlier in Asian women than in Western women. We investigated if younger Asian women with breast cancer also have increased risk of fracture. Methods We used the universal insurance claims data from 2000 to 2003 to identify 22,076 patients with breast cancer and 88,304 women without cancer, frequency matched with age and index date (the date for a health care visit). The incidence of fracture in both cohorts and the hazard ratios (HRs) of fracture in the cancer cohort were estimated by the end of 2009. Results The incidence of all types of fracture was higher in the breast cancer cohort than in the comparison cohort (46.72 vs. 42.52 per 10,000 person-years), with adjusted HRs (aHRs) of 1.18 (95% confidence intervals [CI], 1.03–1.35) for hip fractures, 1.12 (95% CI, 0.98–1.28) for forearm fractures and 1.24 (95% CI, 1.04–1.48) for vertebral fractures. The aHRs were significant in both non-traumatic fractures (1.29; 95% CI, 1.11–1.51) and traumatic fractures (1.12; 95% CI, 1.01–1.23). The age-specific aHR was higher for younger breast cancer patients, and was significant for <50 years old patients in both traumatic (aHR 1.35; 95% CI 1.08–1.68) and non-traumatic (aHR, 1.72; 95% CI, 1.21–2.44) fractures. Conclusion This study suggests that Asian women with breast cancer might have an increased risk of fracture.


Introduction
Both breast cancer and osteoporosis are disorders primarily associated with aging in women, and have been a medical challenge worldwide. Osteoporosis and the associated fractures have become important global public health issues. Almost 56 million people were diagnosed with various types of fracture in 2000, with approximately 9 million new osteoporotic fractures occur annually [1]. The incidence of breast cancer has increased globally over the past few decades [2,3], with greater increase observed in Asian populations [4]. However, no apparent biological difference in the disease has been found between Asian and Western women [4]. Previous studies have noted that breast cancer survivors are at an increased risk of osteoporosis [5] and fracture [6]. The elevated risk of fracture in patients with breast cancer has been attributed to the effects of chemotherapy, ovarian failure, early menopause, and the use of aromatase inhibitors (AI) [7,8,9]. However, most clinical trials or cohort studies on fractures associated with breast cancer have been performed on Caucasian postmenopausal patients [6]. The association in other ethnic groups may be significantly different. For instance, basal bone mineral density (BMD) and the incidence of bone fracture differ among ethnic groups [10]. Even without significant biological difference in breast cancer, the incidence of breast cancer in Asian women peaks in the age of 40-50 years, whereas in Western women it peaks in the age of 60 to 70 [4]. Whether Asian women with breast cancer are also at elevated risk of fracture and if fractures occur in the younger age groups should be investigated. We, therefore, used Taiwan's National Health Insurance (TNHI) claims data to assess the relationship using a retrospective cohort study.

Data Source
TNHI is a universal health insurance system established in 1995 by the Department of Health of Taiwan

Study Subjects
From the catastrophic illness dataset, we identified 22,812 women with newly diagnosed breast cancer (ICD-9-CM 174), who are free from other cancers and are aged 20 years and above in 2000-2003. The diagnosis date of breast cancer was used as the index date. Women with history of hip, distal forearm, and vertebral fracture at the baseline or those who have these types of fracture within one month after the index date were excluded from the study. A total of 21,952 women were included in the breast cancer cohort. Among the women without any cancer, we randomly selected 87,808 women as non-cancer comparison cohort, and frequency matched with age and index date (the date for a health care visit). Both cohorts were followed up until the end of 2009. The subsequent fractures including the hip (ICD-9-CM 820), vertebrae (ICD-9-CM 806.20-806.9, forearm (ICD-9-CM 813) and the other type of fractures (ICD-9-CM 800-806.5, 807-812,814-819, and 821-829) were investigated.

Statistical Analyses
Data analysis first measured the annual incidence of osteoporosis-related fracture by the type of fracture in women with breast cancer. We compared the distributions of age, location of subject's residential area and the history of non-osteoporosis fracture, between the breast and comparison cohorts. Each study subject was followed up from the index date to the event when the fracture was diagnosed, or the date censored for loss to follow-up, death, termination of insurance, or the end of 2009. The person-years of follow-up were measured for all subjects and the incidence density was estimated by per 10,000 person-years during the follow-up period. We estimated the incidence rates of fractures for both cohorts. Hazard ratio (HR) and 95% confidence interval (CI) associated with fracture were estimated using Cox proportional hazards regression analysis. The multivariable analysis model estimated adjusted HR (aHR) controlling for age at the index date (,50, 50-64, and $65 years), the residential area, and nonosteoporosis fracture history, etc. We also assessed the HRs of fracture among age groups to evaluate if the fracture risk was higher for a specific age group. All analyses were performed used SAS statistical package (SAS institute Inc., Cary, NC. Version 9.1), and the significance level was set at 0.05.

Results
Annual Osteoporosis-related Fracture Incidence in Breast Cancer Cohort Figure 1 shows that the annual incident cases of fracture in the breast cancer cohort increased from 89 in 2000 to 319 in 2009. The overall annual incidence decreased in 2000 and then increased to a plateau from 2002. The average overall incidence of the 3 types of fractures studied was 35.4 per 10,000 personyears, and it was higher for hip fractures than for distal forearm and vertebral fractures (14.80, 13.28, and 7.96 per 10,000 personyears, respectively). The overall age-specific incidence of fracture in the 10 years peaked during 60-69 years of age in both breast cancer and comparison cohorts ( Figure 2).

Characteristics of Study Cohorts
The baseline mean age was slightly higher in the breast cancer cohort than in the comparison cohort (51.4 (SD 12.0) vs. 51.2 (SD 12.3) years) (p = 0.02) ( Table 1). Higher portion of study subjects resided in northern Taiwan, which is more urbanized than other areas. Non-osteoporosis-related fracture history (included hip, vertebral and distal forearm) was less prevalent in the breast cancer cohort than in the comparison cohort (1.62 vs. 1.94%, p = 0.002). Hazard Ratio and Incidence of Osteoporosis-related Fracture in the Patients with Breast Cancer The incidence of subsequent fractures was 1.10-fold higher in the breast cancer cohort than in the comparison cohort (46.72 vs. 42.52 per 10,000 person-years), with an aHR of 1.16 (95% CI 1.07-1.27) ( Table 2). The site-specific data showed significant differences for hip (aHR, 1.18; 95% CI, 1.03-1.35) and vertebral fractures (aHR, 1.24; 95% CI, 1.04-1.48) after controlling for age, area of residence, and other fracture history. The stratified analysis further showed a higher incidence of traumatic fracture than nontraumatic fracture in both cohorts (Table 2). When compared to subjects without the breast cancer, however, the breast cancer patients had significant adjusted hazard ratios for non-traumatic hip and forearm vertebral fractures, not for the traumatic fractures, particularly for hip and forearm vertebral fractures.

Age-specific Fracture
The age-specific incidence of fracture increased with age in both cohorts, with the peak appeared in the 60-69 ages group in both the breast cancer cohort and comparison cohort ( Figure 2). The average age at which a fracture occurred in the breast cancer group was approximately 2 years younger than that in the comparison cohort (66.0612.7 years vs. 67.9612.6 years, p = 0.0006; data not shown).

Discussion
Our study on breast cancer in relation to fractures yielded an aHR of 1.16 for the cancer patients versus non-cancer comparison women after controlling for covariates. The incidence of traumatic fractures was greater than that of non-traumatic fractures. But, the breast cancer cohort to comparison cohort aHR estimate was significant for non-traumatic fractures not for traumatic fractures. In a case-control study, Newcomb et al. reported women with breast cancer are 20% less likely to have the history of fracture [11]. Our study also showed that women with breast cancer are less likely to have previous fracture history at the baseline compared with those without breast cancer. We also found women in Taiwan have their breast cancer and fracture occurred at younger ages than Western women. The aHRs were stronger for younger breast cancer women, particularly for the nontraumatic hip fracture with an aHR of 5.32. These data indicate a greater relative impact on non-traumatic events for women with the cancer.
Postmenopausal women with high bone density have an elevated risk of breast cancer [12][13][14][15][16], but have a lower risk of bone fracture [11]. Studies have shown that postmenopausal women with longer or higher estrogen exposure are associated with increased BMD, and increased breast cancer as well [17][18][19]. In addition, factors regulating the ossification process, such as insulin, insulin-like growth factor type 1, insulin-like growth factor  type 2, and insulin-like growth factor binding protein 3, may also have association with the breast cancer risk [20][21][22][23][24]. However, breast cancer accelerates bone loss in patients. During natural menopause, women may suffer from bone loss for 3% per year in the earlier two years, and slows down to approximately 1% annually thereafter [25]. In women with breast cancer, the osteoclastic activity is increased by releasing transforming growth factors [6], even in the absence of bone metastases. Breast cancer treatment may thus enhance bone loss in women undergoing a natural menopause. Furthermore, breast cancer women with postmenopausal estrogen deficiency are at an elevated risk of bone loss with age. The use of estrogen-depleting therapies, such as third-generation aromatase inhibitors (AI), accelerates age-and menopause-related BMD loss [26][27][28][29][30][31]. Chemotherapy may induce ovarian failure or ovarian function suppression and cause low estrogen levels caused, leading to bone loss [5].
Chen et al. [6] reported a higher risk of fracture for breast cancer survivors in Women's Health Initiative Observational  Study (WHI). However, their study population and evaluation method are different from those in our study. The breast cancer patients in our study were mainly premenopausal breast cancer women and who were younger. Moreover, most breast cancer patients were screened after fifty years old for postmenopausal women diagnosed in the WHI study [6]; by contrast, more than 50% of breast cancer patients were diagnosed under the age of 50 in premenopausal women in this study. Oriental women have an elevated risk of premenopausal breast cancer and are at the risk of bone loss, which is associated with cancer treatment [32]. For example, ovarian suppression with goserelin in premenopausal women decreases BMD by 6% to 10% during the first two years of treatment [33]. Moreover, an Austrian study found that women with the goserelin-induced ovarian suppression on AIs medication may have 17.3% enhanced BMD loss within 3 years [34]. Ovarian ablation (either medical or surgical) leads to increased bone loss in premenopausal women. Women with premenopausal breast cancer may thus have lower BMD later in their life and are at an increased risk of osteoporosis, compared with those without premenopausal breast cancer [35,36]. Since the early premenopausal age, these women suffer not only losing trabecular connectivity in cancellous bone structural, and cortical thinning and porosity, but also experiencing reduced toughness of bone and the resistant to crack propagation. Pores coalesce and the low bone mass cannot absorb energy radiating from a fall. Hip and distal forearm fractures occurred more frequently in the present study, and hip fracture was the worst non-traumatic osteoporosis-related fractures. Therefore, the higher risk of hip fracture in women of younger age in our study groups may be related to breast cancer.
This study has several limitations. First, a few minor subclinical vertebral or wrist fractures do not systematically lead to medical management or hospitalization. The claims data also included few self-reported fractures. The incidence of vertebral fracture is low in our study and the findings are likely underestimated because of subclinical status. This inference can be verified in further analysis that compared traumatic and non-traumatic events (Table 3).
Women with breast cancer have a higher risk of vertebral fracture among non-traumatic fracture probably because they visit clinics more often and probably have these subclinical events identified. However, the difference is no more significant in the analysis by age, probably because of the small sample size. Furthermore, studies have shown that self-reported fracture is generally reliable [37][38][39]. Women with breast cancer are at higher risk than the general population. Second, insurance claims files do not provide information on cancer stages. We were unable to determine if patients with advanced stages of the cancer are at considerably greater risk of fracture. Third, information on lifestyle, such as alcohol consumption and smoking, is also unavailable in the claims file; thus, we were unable to assess the association between fracture and lifestyle factors. However, lifestyle is probably not an important factor in this study because only approximately 4% of the women in Taiwan smoke [40].
The 10-year overall survival of patients with breast cancer is 75% in Taiwan [4]. As the fracture is an important factor that affects the quality of life, a multidisciplinary treatment team of breast cancer should encompass the issue of bone protection and fracture prevention, such as regular examination of BMD [41], early use of antiresorptive agents [36,41,42], or fall prevention facility, to improve the quality of life of young breast cancer survivors.
To the best of our knowledge, this study is the first national population-based report on fracture risk among Asian breast cancer survivors. This study suggests that women with breast cancer, particularly those diagnosed at a relatively early age, below 50 years old, should undergo prophylactic treatment to counter the increased risk of fractures.