Abstract
BACKGROUND:
Breast cancer is the leading cause of cancer death in Hispanic/Latina women nationwide. Limited cancer research has been conducted in this population. El Paso, Texas is a large border city with a population of around 900,000, of which 85% are Latinos and would provide a suitable setting for this study. The aim of this study is to evaluate ethnic differences and cancer characteristics in Hispanic/latina women with breast cancer.
METHODS:
After IRB approval, we retrospectively analyzed the variables of patients with breast cancer treated consecutively at a large tertiary medical center in El Paso, TX between 2005–2015. Descriptive statistics, bivariate, and multivariable analyses were conducted.
RESULTS:
1,252 patients were identified. Mean age at diagnosis was 57 years. 1074 were Hispanics/Latinas (86%). When comparing Hispanics versus non-Hispanics, 31% of Hispanics compared to 24% Non-Hispanics were diagnosed at age <50 (P = 0.043). More Hispanics are uninsured (34%) compared to Non-Hispanics (25%) (p = 0.008). Hispanics presenting with advanced stages were more likely to be uninsured (P = 0.02).
CONCLUSIONS:
This analysis confirms that Hispanics/Latinas are diagnosed with breast cancer at a younger age and are more commonly uninsured than Non-Hispanics. We did not observe significant differences in the prevalence of
Background
Breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death among women in the United States (U.S) [1,2]. In Hispanic American women, breast cancer is the leading cause of cancer death [3]. Ethnic -related health disparities in breast cancer screening rates, stages at presentation and mortality have been previously reported [3]. However, despite these ethnic differences and being the fastest growing major demographic group currently in the U.S [4], Hispanics in the US, the majority of whom are of Mexican descent, remain understudied with limited representation in research overall. El Paso, TX is a large American–Mexican border city with a majority of Hispanic residents. Therefore, it provides a desirable setting to study breast cancer among Hispanics especially that most are of Mexican descent which limits the variability in evaluating race and ethnicity, a common challenge encountered in Hispanic related research overall. A previous study by our group has suggested that breast cancer in Hispanic women is usually diagnosed at younger age, with more advanced tumor stage, higher tumor grades, estrogen receptor negative tumors and a higher prevalence of BRCA1 and BRCA2 mutations than in non-Hispanic whites [5]. Other studies have reported that breast cancer in Hispanics is likely to be associated with increased tumor size at diagnosis and not having received breast conservative surgery [6,7].
In this study, we sought to determine the clinical and pathological characteristics of breast cancer more precisely in a large predominantly Hispanic breast cancer population treated at a tertiary care teaching medical center in El Paso, TX. We also evaluated potential disparities in characteristics and cancer management between Hispanics and Non-Hispanics. A better understanding of cancer characteristics in this understudied patient population and identifying potential disparities in cancer management and outcome can lead to more focused efforts toward standardizing treatment options and addressing unmet needs in Hispanics and other minorities.
Methods
After obtaining IRB approval, we conducted a retrospective study of consecutive patients diagnosed with breast cancer treated at a tertiary academic medical center in El Paso, TX between January 1, 2005 and December 31, 2015. This facility diagnoses and treats over one third of the patients diagnosed with breast cancer in the region. Patients’ ethnicity was self- identified according to their records. For simplicity, we categorized patients into the following major categories: Hispanics (regardless of race), White Non-Hispanics, African–Americans, Native-Americans, and others (including Asians, Middle Eastern etc...). We used the tumor registry at the teaching hospital (University Medical Center) as well as the Texas Tech Breast Care Center database to identify cases. This database includes clinical, pathological and demographic characteristics. The clinical and pathological data were extracted from electronic medical records (EMR) for those patients seen since 2011 (date of implementation of EMR). The rest of the data was obtained manually from archived medical charts, included in the database by the co-authors and verified for accuracy by the senior author. In this manuscript, Hispanic and Latino/Latina are used interchangeably. We sought to (1) assess the characteristics of the predominantly Hispanic breast cancer patients including mean age at diagnosis, race, Body mass index (BMI), stage, histology of the tumor, cancer treatment, type of surgery, and prevalent comorbidities; (2) compare characteristics of breast cancer and cancer treatment between Hispanics and Non-Hispanics at this facility; (3) compare clinical and pathological characteristics of patients who have insurance coverage (any financial insurance including 3
Statistical considerations
Continuous data was described using mean and standard deviation (SD) while categorical data was described using frequencies and percentages. An unpaired t -test was used to compare all continuous data while a Fisher’s exact or Chi square test was used to test the associations between the categorical variables. Multiple variable logistic regression analysis was carried out to determine factors associated with insured status compared to uninsured status. Similarly, logistic regression analysis was also used to determine factors associated with advanced stages at presentation (stage III and IV) compared to early stages (0, I, and II) for the entire cohort and separately for Hispanics and non-Hispanics cases. Significant variables from univariate analysis were included in multivariable analysis. The results of logistic regression analysis were presented using odds ratio (OR) along with 95% confidence interval (CI) and p-value. P-values less than 5% were considered as significant results. SAS 9.3 was used for data analysis.
Results
A total of 1252 cases were identified and their characteristics analyzed. 817 patients had available electronic medical records (EMR) and the rest was obtained from archived medical charts. Mean age at diagnosis was 57 years (SD 12.2, range 25–95). Table 1 summarizes patient and tumor characteristics as well as treatment distribution (surgical type, chemotherapy, radiation therapy, anti-estrogen therapy), and compares the clinical, pathological and treatment characteristics according to ethnic groups. The majority of patients were Hispanics (86%), 384 patients (31%) were diagnosed at younger than 50 years and 8% were diagnosed at ≤ than 40 years. 51% of patients had a medical insurance coverage, 49% had no or unknown insurance coverage. Stage distribution at initial diagnosis was as follows: Stage 0 (11%); Stage I (28%); Stage II (33%); Stage III (20%); and Stage IV (8%). Pathology of invasive tumors are listed in Table 1, with the most common type being invasive ductal carcinoma (86%), estrogen Receptor (ER) positive tumors 68% compared to ER negative 30% and unknown 2%, and Human Epidermal Growth Factor Receptor 2 neu (HER2) positive tumors 19%. Of note is that 20% of patients had Triple Negative i.e. ER and progesterone receptor (PR) negative and HER2 negative tumors. Cancer treatment received was as follows: Mastectomy was performed in 464 patients (37%) while breast conserving surgery/lumpectomy was performed in 649 (52%); 737 (67%) received hormonal therapy and 517 patients (58%) received chemotherapy. The major identified comorbidities were as follows: Hypertension (34%), Diabetes Mellitus (26%); hyperlipidemia (20%); coronary artery disease (3%). Body mass index (BMI) was noted to be in the obesity range ≥30 in 46% of patients. Mean (BMI) was 30.86 (SD 6.38, range 14–58).
When comparing Hispanic versus Non-Hispanics, Hispanic individuals tend to have a slightly lower age at diagnosis (56 years versus 58 years, P = 0.045), but 32% were younger than 50 years at diagnosis versus 24% non-Hispanics, with a higher prevalence of obesity (47% versus 39%, p = 0.062), and significantly more patients without insurance coverage (34% versus 25%, p = 0.008). There was no significant difference between the 2 groups in regards to histological type, frequency of ER positive or HER2 positive, nor stage at presentation. In addition, no statistically significant differences were noted regarding the systemic treatment received and the type of surgery (mastectomy versus lumpectomy). More Hispanic patients appear to have comorbidities but that did not reach statistical significance (60% versus 56%, p = 0.284).
Table 1 also shows patient’s characteristics in association with insurance status (insured versus non-insured). Those with unknown insurance status were included with the non-insured patients since in the institution where the study was conducted, patients who are indigent or self-paying are included under the category of “unknown insurance”. Insured patients included all those with any type of private of governmental medical insurance coverage. Several of the studied tumor and baseline characteristics were found to be different between insured and uninsured cases.
Compared to insured patients, those without insurance coverage present at a younger age (59 years versus 55 years, p < 0.0001), are diagnosed at more advanced stages (stage IV 11% versus 5%), have a higher prevalence of Triple Negative breast cancer (23% versus 17%, p = 0.023), and receive less chemotherapy, hormonal therapy, breast-conserving surgery and radiation therapy. After adjusting for factors associated with insured patients compared to uninsured patients, insured patients were likely to be older, presenting with early stage, less likely to receive chemotherapy, more likely to receive radiation therapy and surgery, and less likely to have triple negative tumors. Adjusted analyses showed that uninsured patients are more likely to be diagnosed at younger age, advanced stage, and receive chemotherapy, and less likely to have surgery compared to insured patients.
Table 2 summarizes the factors associated with advanced stages (Stage III and IV) at diagnosis. Adjusted association of considered cofactors for the entire cohort shows that uninsured status, younger age (<50 years), ER negative, and HER positive were associated with increased odds of advanced stage at presentation. Hispanic patients presenting with advanced stages were more associated with uninsured status, ER negative and HER2 positive tumors while non-Hispanics were more likely to be younger (p = 0.04), and have coronary artery disease (CAD) as a comorbidity (p = 0.02). In addition, both ethnic groups presenting with advanced stages were more likely to have ER or PR negative disease,
Discussion
This large study confirms that Hispanic/Latina breast cancer patients are diagnosed at a younger age compared to Non-Hispanics and historical controls nationwide [8]. Also patients evaluated in our Hispanic cohort tend to present at more advanced stages of breast cancer with 8% presenting at stage IV compared to 5% presenting at this stage nationwide [9]. Hispanics appear to have a higher prevalence of Triple Negative breast cancer (20%) compared to reported 13%–15% in all race/ethnicities nationwide, and have more ER/PR negative/HER2 positive breast cancer (10%) compared to 5% nationwide [2]. This study is representative of Hispanics particularly those of Mexican descent since the city of El Paso, TX is a large border city with 85% of its population consisting of Mexican Hispanics [5]. We did not observe significant differences between stages at presentation and breast cancer types between Hispanics and Non-Hispanics breast cancer patients in this studied cohort, but significant differences were found in this study related to younger age at diagnosis, more prevalence of obesity and lack of medical insurance in Hispanics compared to non-Hispanics. The lack of statistically significant differences in other variables which is inconsistent with published literature could be attributed to a relatively small sample size in the non-Hispanic group (14%) that might not be representative of non-Hispanics overall. Also, in this border city with possibility of mixed heritage, true ethnicity might not be accurately represented and many of the self-labeled “white, non-Hispanics” might have Mexican genetic lineage. Therefore, and short of genetic testing and accurate lineage assessment, we cannot say confidently that this comparison is generalizable and another multi-institutional comparative study is being conducted to clarify these findings. We favor considering the totality of the data that did indeed show significant differences between Hispanics and non-Hispanics with breast cancer nationwide as noted above based on reported literature in the US population.
Our study noted that Hispanic women are diagnosed with breast cancer at a younger age than non-Hispanic women. This is consistent with what has been reported in the literature [4,10,11] and confirms that Hispanic women especially those of Mexican origin tend to have high-risk for early onset, premenopausal breast cancer. Potential explanations for this observation may be related to the higher prevalence of BRCA mutations in Mexican women with European (Spanish, and particularly of Ashkenazi Jewish descent) ancestry. Future studies examining the migratory patterns from Europe and the region of Mexico women migrate from, and their possible role in higher rates of breast cancer in this population might be helpful. Also future studies should investigate change in behaviors and social environments over time to further understand the underlying mechanisms for these increased risks in young Hispanic women with breast cancer.
Ethnic disparities related to breast cancer in Hispanics might be related to several factors including inferior participation of Hispanic women in cancer screening programs [12], prevalence of obesity and comorbidities, less access, adherence and inferior treatment response as well as inadequate insurance coverage [1,6,13,14].
This study showed significant differences between patients with no health insurance coverage and those with insurance. Non-insured patients presented at younger ages and with more advanced stages at diagnosis and received less loco -regional therapy and systemic therapy (chemotherapy, hormonal therapy and anti-Her2 therapy). This is consistent with other studies suggesting that individuals without health insurance are less likely to have a usual source of health care and less likely to receive preventive services including cancer screening and are, therefore, more likely to be diagnosed with late stages of cancer [15,16]. A study using data from the National Cancer Database examined the potential impact of health insurance status and stage at diagnosis for 8 common cancers including breast cancer [15] among 843,177 patients. The study was co-sponsored by the American Cancer Society and the American College of Surgeons and showed that for each cancer site, uninsured and Medicaid-insured patients had the highest proportion of stages III and IV cancers at diagnosis, while those with private insurance and Medicare plus supplemental insurance had the lowest.
In this study, we did not see significant difference related to treatment received between Hispanics and non-Hispanics. This is inconsistent with other studies and could be attributed to the fact that this is a single institution study and all patients treated followed similar treatment guidelines. Also, this data could be unique to the single facility that treats a large percentage of uninsured and underinsured patients but with no treatment discrimination made possible by a special discount service provided for uninsured patients by the affiliated safety net hospital. This is inconsistent with other studies showing that not only race/ethnicity, but insurance as well as socioeconomic status (SES) influence breast cancer treatment. Friedman et al., [17] used multivariate logistic regression to assess the probability of definitive loco- regional therapy, hormone receptor testing, and adjuvant systemic therapy among 662,117 white, black, and Hispanic women diagnosed with invasive breast cancer during 1998–2005 at National Cancer Data Base hospitals. After adjustment, black (versus white) women had less definitive loco- regional therapy (odds ratio [OR], 0.91; 95% confidence interval [CI], 0.88–0.94), hormonal therapy (OR, 0.90; 95% CI, 0.87–0.93), and chemotherapy (OR, 0.87; 95% CI, 0.84–0.91). Hispanic (versus white) women were also less likely to receive hormonal therapy while hormone receptor testing did not differ by race/ethnicity. Racial disparities persisted despite adjusting for insurance and SES.
The disparity in treatment and stages at presentation between insured and non-insured is quite alarming and is likely related to delay in presentation, lack of access to care and underuse of screening mammograms.
Strengths of the study included reliance on a robust cancer registry data that is representative of the large underserved breast cancer population diagnosed in a large border geographic area, using well-documented and standardized collection methods as part of a nationally accredited breast care facility by National Accreditation Program for Breast Centers (NAPBC), affiliation with a tertiary care teaching medical center accredited by Commission on Cancer (COC), and verification by several co-authors including the PI (Z.N) with standardized categorization of data. In addition, this study provided the opportunity to focus on a relatively homogenous population of Hispanic Mexican ethnicity.
In this study, we did not capture the difference between USA born and foreign born patients, as this information is not routinely documented and is a limitation in this study. The risk for developing breast cancer is higher in USA born Hispanics than who are born outside [4,18]. It would be desirable in future studies to study the characteristics of breast cancer between these 2 groups. Also, the data did not capture survival data by stage which would have also been valuable.
In conclusion, ethnic disparity related to breast cancer and patients’ characteristics were observed in this study. Hispanics are diagnosed with breast cancer at a younger age, more advanced stages and tend to be uninsured and have prevalence of obesity. These findings suggest important ethnic disparities linked to cancer presentation but also to socio-economic status and access to care and health care services.
The data described in this work may be utilized to identify and better characterize areas of unmet needs for Hispanics with breast cancer. These patients will likely benefit from future interventions aimed at reducing the cancer burden by improving access to health care coverage and screening services. Future studies are needed to clarify whether breast cancers among Hispanic women consist of a distinct biologic spectrum of tumors causing an increased prevalence of young age at presentation, more TNBC and Her2 positive tumors. The implications of these findings can be significant given that a large number of Hispanic patients are uninsured/underinsured and might receive less systemic therapy including chemotherapy and anti-Her2 therapy. Also, increasing awareness particularly among young Hispanic women and improving access to genetic testing and counselling and preventive services would be important. Finally, more focused implementation of healthy dietary choices and lifestyle habits are recommended in all patients and specifically in this patient population in view of the prevalence of obesity and related co-morbidities. A follow up larger multi-institutional study is ongoing.
Footnotes
Conflict of interest
All authors declare no conflict.
Funding
Cancer Prevention and Research Institute of Texas (CPRIT)-RP120528. (PI Z.Nahleh).