Estimation of healthcare‐related charges in women with BRCA mutations and breast cancer

Background Breast cancer costs were estimated at $16.5 billion in 2010 and were higher than other cancer costs. There are limited studies on breast cancer charges and costs by BRCA mutations and receptor status. We examined overall health care and breast cancer-related charges by BRCA status (BRCAm vs. BRCAwt), receptor status (HER2+ vs. HER2-), and treatment setting (neoadjuvant vs. adjuvant). Methods Retrospective cohort study of charge data from 1995-2014 in an academic medical center. Facilities, physician, pharmacy, and diagnosis-related charges were presented as mean and median charges with standard deviation (SD) and interquartile ranges (25%-75%). Wilcoxon rank-sum test was used to assess statistically significant differences in charges between comparators. Results Total median breast-cancer related charges were $65,414 for BRCAm and $54,635 for BRCAwt (p=0.19); however all-cause charges were higher for BRCAm patients ($145,066 vs. $119,119, p<0.001). HER2+ status was associated with higher median breast cancer charges ($152,159 vs. $44,087, p<0.0001) that was driven by the charges for biological agents. Patients initially seen in the neoadjuvant setting had higher mean breast cancer charges than in the adjuvant setting ($117,922 vs. $80,061, p<0.0001). Conclusion BRCA mutation status was not associated with higher breast cancer charges but HER2+ status had significantly higher charges, due to charges for biological agents. Patients who initially received neoadjuvant treatment had significantly higher overall treatment charges than adjuvant therapy patients. With the advent of novel therapies for BRCAm, the economic impact of these treatments will be important to consider relative to their survival benefits.


Background
Breast cancer is the most common type of cancer observed in women in the US and is associated with significant clinical and economic burden [1]. In 2015, there were nearly 3.5 million women living in the US with a breast cancer diagnosis and it is estimated that approximately 12.4% of women in the US will receive a breast cancer diagnosis at some point in their lives [1].
Breast cancer is associated with significant costs. In 2010, breast cancer treatment costs in the US were approximately $16.5 billion, which was higher than those for any other cancer and these costs are expected to rise to $20.5 billion by 2050 [2]. Similar results have also been observed in Europe wherein breast cancer costs accounted for the highest healthcare costs of all cancers [3]. The published economic studies in breast cancer vary widely by perspective, methodology, time horizon, and patient populations [4]. While breast cancer costs have been estimated by tumor stage, time after diagnosis, and treatment options, there are limited studies on such costs categorized by BRCA mutations and receptor status as well as neoadjuvant and adjuvant treatment settings [4][5][6].
The assessment of costs within a healthcare system is challenging given the complex interplay between costs, charges, and reimbursed amounts for the service provided [7]. In many instances, it is difficult to access costs due to their proprietary nature, which is dependent on the contractual agreements between the healthcare systems and payer organizations. Studies have used costto-charge ratios, the ratio of charges over CMS allowable costs, to describe the costs assigned by the healthcare system. However, there are significant variations in these cost estimates depending on the type of institution and geographical location [8,9]. Thus, healthcare studies typically use the more readily available charge data to make economic assessments.
The purpose of this study was to explore cumulative health care charges in women diagnosed with breast cancer and tested for BRCA mutations with charges stratified by BRCA mutation status, receptor status, and treatment settings.

Study design, population, and data source
This was a retrospective cohort study utilizing charge data from 1995 to 2014 in the University of Utah Clinical Enterprise Data Warehouse (EDW), a repository containing records for more than 3.3 million patients dating back to 1995. The EDW contains data from electronic medical records, laboratory and radiology findings, administrative claims, and patient encounter data. Patients included in the study were adult women with breast cancer who were tested for BRCA mutations (BRCA positive [BRCAm] or BRCA wildtype [BRCAwt]) and received treatment at the Huntsman Cancer Institute (HCI) in Salt Lake City, Utah. HCI is the major cancer center in the Intermountain West region as well as a National Cancer Institute-Designated Cancer Center and member of the National Comprehensive Cancer Network. Patients were identified from the Huntsman Cancer Institute Tumor Registry (HCI-TR) using site (ICD-O C50.x) and histology codes for the diagnosis of breast cancer between January 1, 1995 and December 31, 2014. To our knowledge, healthcare providers from HCI had unrestricted availability to request BRCA testing during the study period. The date of breast cancer diagnosis was defined as the index date. Electronic medical records of eligible patients were manually reviewed to identify and obtain information regarding BRCA mutation, HER-2, and ER/PR status, treatment patterns, and disease progression. Chart abstraction was conducted by an oncology clinical pharmacist using a chart abstraction form and reviewed by another oncology clinical pharmacist. Demographic and clinical characteristics of these patients were obtained through a EDW query.
Patients were excluded from the study if they were < 18 years of age at the time of breast cancer diagnosis, had < 2 encounters (clinic visits) separated by ≥ 30 days with relevant ICD-9 codes for breast cancer in the EDW, were male patients, were not tested for a BRCA mutation or had unknown BRCA mutation status, or had no charge data.
The charge amount was the actual amount charged by the facility based on the International Classification of Diseases, 9th Clinical Modification (ICD-9-CM) diagnosis and Current Procedural Terminology (CPT) codes. The estimated yearly charges were adjusted for inflation to 2014 values using the Bureau of Labor Statistics Consumer Price Index for Medical Care [10].

Study outcomes
The patients identified in this study were linked to the administrative data in the EDW to obtain all charge data related to the healthcare utilization from the index date until they were lost to follow-up (no additional encounters in the database or death occurred). Charges (the monetary amount billed for services provided to a patient) were initially categorized based on paid claims linked to the corresponding ICD-9 and CPT codes as breast cancer-related, other (non-breast) cancers related, non-cancer related, and all-cause charges. Each categorical charge was stratified as inpatient-and outpatientrelated charges based on site of care. Inpatient charges refer to charges incurred whn a patient has been admitted to a hospital for care. Outpatient charges refer to charges incurred when a patient is treated in an ambulatory clinic setting (not admitted to a hospital for care). All-cause charges were defined as the sum of all charges (breast cancer-related, other cancers related, and non-cancer related) incurred from index date to death or the end of study whichever occurred earlier.
Charges were also categorized as facilities/technical charges, physician/professional charges, and pharmacy charges. The pharmacy charges were sub-divided by the type of treatment provided (chemotherapy, hormonal therapy, biological therapy, targeted small molecule therapy, investigational drug, chemotherapy administration charges, and other medications).
Breast cancer-related charges were further stratified by BRCA mutation status, HER2 status, and initial treatment setting at cancer diagnosis. Initial treatment setting was defined as the first treatment for the breast cancer patient at diagnosis and included non-metastatic adjuvant therapy, non-metastatic neoadjuvant therapy, surgical resection, and treatment for metastatic disease. BRCA mutation status was categorized as BCRA mutated gene (BRCAm) and BCRA wild-type gene BCRAwt. HER2 status was categorized as HER2+ (HER-2 amplified tumors) that included ER+/PR+/HER2 + and ER-/PR-/HER2 + patients; and HER2-(HER2 non-amplified tumors) that included ER-/PR-/HER2-(also called triple negative breast cancer [TNBC]) and ER+/PR+/HER2-patients. If a patient progressed to metastatic disease after receiving adjuvant, neoadjuvant or surgical resection, their charges were assigned to the initial treatment setting until progression to metastatic disease at which point subsequent charges were assigned to the metastatic group. Hence, metastatic patients could contribute charges to more than one treatment setting; however, their charges were not double counted.

Statistical analysis
Charges were presented as mean and median charges with standard deviation (SD) and interquartile ranges (25%-75%). Wilcoxon rank-sum test was used to assess statistically significant difference in health care charges across comparator groups. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary, NC).  (Fig. 1a).

Results
With respect to initial treatment setting after breast cancer diagnosis, 553 (67.8%) patients received adjuvant therapy, 148 (18.1%) patients received neoadjuvant therapy, and 73 (8.9%) patients had surgical resection alone. There were 42 (5.1%) patients who presented with metastatic breast cancer at the time of diagnosis. Of the patients who received adjuvant, neoadjuvant and surgical resection at diagnosis, 58 (10.5%), 25 (16.9%), and 15 (20.5%) progressed to metastatic disease during the study period respectively. The median time to progression for these 98 patients was 3.5 years (± IQR 2.2-9.7 years). Thus, of the total 140 patients who had metastatic disease during the study period, 133 (16.3% of 816 patients) had valid charges (Fig. 1b).  Fig. 3). In addition, HER2 + patients incurred significantly higher charges across nearly all breast cancer related charge categories (  Table 3).

Comparison of health care cumulative charges by initial treatment setting
Patients seen initially in the neoadjuvant setting had higher mean breast cancer related charges vs. those seen in the adjuvant treatment setting ($117,922 vs. $80,061, p < 0.0001; Fig. 4). Patients receiving neoadjuvant treatment had significantly higher charges across most of the charge categories vs. patients receiving adjuvant treatment (Table 4). Patients with metastatic disease had a mean breast cancer related total charge of $101,969 (Fig. 4).

Discussion
The purpose of this study was to identify the important drivers of charges in women with breast cancer who were tested for BRCA mutation status and received treatment at a National Comprehensive Cancer Network/National Cancer Institute (NCCN/NCI)-accredited comprehensive cancer treatment center. In addition to BRCA status, the study also categorized patients by the HER2 status and type of treatment received at the time of breast cancer diagnosis. The study outcomes included comparison of breast cancer related charges between BRCAm and BRCAwt patients; HER2 + and HER2-patients; patients receiving adjuvant and neoadjuvant treatments overall, and by charge type. The study also compared the breast cancer related charges vs. charges associated with other cancer types and non-cancer related charges in these patients.      [14].
With respect to treatment costs by type, Campbell and Ramsey, in a synthesis of published cost studies for breast cancer, noted that patients who received adjuvant therapy had significantly higher costs ($23,000-31,000) vs. those who did not. [4] Though not exactly comparable with our study due to differences in methodology (BRCA tested population), we observed significantly higher charges for neoadjuvant vs. adjuvant therapy ($117,922 vs. $80,061, p < 0.0001).
This study has a number of limitations. First, the results may not be generalizable since the results were for a single site (HCI) in Utah with a younger, Caucasian population. Some of the high treatment costs seen in this study may be a result of the younger age of the breast cancer cohort. Also, the community practice patterns may be different from what was observed at HCI, which is the only NCCN/NCI designated comprehensive cancer center in the Intermountain West region. Secondly, the study results were limited by the relatively small number of BRCAm patients (n = 134) vs. the BRCAwt (n = 682) group. Thirdly, the charges were aggregated across breast cancer disease stage, hence were not comparable to the more commonly published breast cancer cost studies that assessed costs by stage.

Conclusions
Our assessment of the treatment charges for breast cancer at a single NCCN/NCI-designated comprehensive cancer center in the Intermountain West region demonstrates that BRCA mutation status was not associated with higher breast cancer charges, but were associated with significantly higher all-cause charges. However, HER2 + status was associated with higher breast cancer charges versus HER2-tumors and this was driven by the cost of biological agents. Breast cancer patients who initially received neoadjuvant treatment after diagnosis had significantly higher breast cancer related treatment charges than those who received adjuvant therapy. These economic differences by mutation, receptor and treatment setting are important to consider when evaluating the impact of targeted therapy on overall survival benefit.