Outcomes research examining treatments, quality of life and costs in HER2-negative and triple-negative metastatic breast cancer: a systematic literature review

We searched PubMed and four online conference databases: ASCO Annual Meeting, ASCO Breast Cancer Symposium, ASCO Quality Care Symposium and San Antonio Breast Cancer (SABC) Symposium. To supplement this search, we also reviewed the references of articles included for data abstraction.

PubMed was searched to identify studies in HER2-negative and TN mBC patients, accounting for a variety of ways the terms of interest may be referenced in the literature using keywords and Medical Subject Headings terms: ‘metastatic breast cancer’, ‘breast neoplasms’, ‘triple negative’, ‘estrogen receptor negative’, ‘progesterone receptor negative’, ‘human epidermal growth factor receptor 2 negative’, ‘costs and cost analysis’, ‘economics’, ‘cost of illness’, ‘quality of life’, ‘health services’, ‘morbidity’, ‘survival’, ‘mortality’, ‘epidemiologic studies’, ‘therapeutics’, ‘comparative’, ‘effectiveness’, ‘guidelines as topic’, ‘pain’, ‘self report’, ‘evaluation studies as a topic’ and ‘disease management’ (Tables 1 & 2).

The search strings, shown in Tables 1 & 2, were developed based on review of keywords in relevant published literature and the US National Library of Medicine Health Services Research PubMed Queries [19]. The PubMed search (conducted in January 2016) was limited to literature published within the past 5 years and written in the English language. The 5-year search limit was imposed to capture the most current literature, while also considering the recent availability of everolimus for women with advanced hormone receptor (HR)-positive/HER2-negative mBC [20]. The four conferences (ASCO Annual Meeting, ASCO Breast Cancer Symposium, ASCO Quality Care Symposium and SABC Symposium) were searched online for abstracts presented in 2013 through 2015, using generic keywords – breast, metastatic and negative.

Identified articles were screened in three phases. In the first phase, the article titles and abstracts were screened to ensure the reported results were in human subjects, in patients with HER2-negative or TN mBC and in the US market. In the second Phase, the titles and abstracts of the articles that passed the first Phase were screened for acceptable study types: for example, not systematic reviews/meta-analyses reporting on one or fewer treatments, Phase I, II or III clinical trials and clinical studies on biomarkers in vivo/vitro. In the final screening phase, the full-texts of the remaining articles were screened to ensure the studies were in treated US human subjects with HER2-negative or TN mBC, reporting real-world research of interest. Finally, the references of the accepted articles were mined for key articles. This was done by flagging any reference for a study published in 2010 or later that mentioned ‘human epidermal growth factor receptor 2’, ‘HER2’ or ‘negative’ in the title. All flagged articles were then screened using the approach described above. Conference abstracts were screened in one phase, and all studies (full-text articles and conference abstracts) that passed the screenings were abstracted. The entire screening and abstraction process of this review was conducted using specialized software [21].

The search of PubMed, online ASCO and SABC conference databases, conducted on 12 January 2016, yielded 1070 studies identified in PubMed and 712 identified in conference searchers (Figure 1). After screening the literature, 22 full text articles and 11 conference abstracts (seven from ASCO; four from SABC) were accepted for abstraction; among these, four studies were included after mining references of accepted articles. Of the 34 accepted records, seven were on CER (six full-length articles; one abstract), 18 were on treatment patterns (ten full-length articles; eight abstracts), one was on treatment navigation and categorized as a treatment appropriateness study in this review (one full-length article; zero abstracts), two were on QoL (two full-length articles; zero abstracts) and five were on costs (three full-length articles; two abstracts).

There were six full-length CER studies identified: Dawood et al., Dranitsaris et al., Li et al., Li et al., Lin et al., Xie et al. (Table 3) [22,23,24,25,26,27]. One CER study was reported in a conference abstract: Dranitsaris et al. [28].

Dawood et al. [22] conducted a retrospective chart review, using data from the Breast Medical Oncology Department of the University of Texas MD Anderson Cancer Center, to compare the prognosis of women with metastatic HER2-positive breast cancer with (n = 191) and without (n = 118) the addition of trastuzumab with the prognosis of women with HER2-negative disease (n = 1782). The main outcomes examined were 1-year survival rates among patients with HER2-negative disease, HER2-positive disease and trastuzumab treatment and HER2/neu-positive disease and no trastuzumab treatment, which were 75.1, 86.6 and 70.2%, respectively. After adjustment, women with HER2-positive disease who received trastuzumab had a 44% reduction in the risk of death compared with women with HER2-negative disease. Dranitsaris et al. [28] carried out a retrospective chart review to examine the time to treatment failure, safety and efficacy of eribulin in 90 patients with mBC, treated at eight US community oncology clinics. About 17 of 90 patients (18.9%) had TN mBC and 57 of 90 (63.3%) had prior anthracycline treatment. The differences in time to failure by TN status (p = 0.49) and prior exposure to anthracyclines (hazard ratio: 0.78; p = 0.15) were not significant. The authors concluded that time to treatment failure with eribulin was comparable regardless of TN status and prior exposure to anthracyclines. Dranitsaris et al. [23] conducted a retrospective comparative effectiveness analysis of four different monotherapies with cytotoxic agents (eribulin, capecitabine, gemcitabine or vinorelbine) in 225 TN mBC treated at 19 community oncology clinics across the USA. The four treatment groups were comparable with respect to age, performance status, duration of disease-free survival, presence of co-morbidities and hemoglobin level before the start of chemotherapy. The median durations of treatment were roughly 2 months with eribulin, capecitabine and gemcitabine compared with 1.6 months with vinorelbine. Overall, eribulin was used as a later line of therapy relative to capecitabine, gemcitabine and vinorelbine, although eribulin was reported to have had at least comparable drug activity and tolerability, even when used in more heavily pretreated patients. Li et al. [24] conducted a retrospective chart review study to compare the real-world effectiveness of everolimus-based therapy and chemotherapy in a national sample of 371 postmenopausal women with HR-positive/HER2-negative mBC. Data came from patient charts at community oncology/hematology practices and outcomes assessed included overall survival (OS), progression-free survival (PFS) and time on treatment (TOT). Results were compared between patients that received everolimus-based therapy or chemotherapy, after failure of a nonsteroidal aromatase inhibitor (NSAI). The authors concluded patients treated with everolimus-based therapy tended to have less aggressive mBC than patients treated with chemotherapy. After controlling for the observed baseline characteristics, everolimus-based therapy was associated with significantly longer OS, PFS and TOT than chemotherapy. Li et al. [25] conducted a secondary data analysis of commercial health insurance claims to compare the TOT among 3298 patients treated with everolimus and chemotherapy among postmenopausal women with HR-positive/HER2-negative mBC. Data were from MarketScan and PharMetrics databases (2002Q1–2014Q2). HR-positive/HER2-negative mBC patients receiving everolimus experienced significantly longer TOT than those on chemotherapy overall and those on capecitabine monotherapy. Lin et al. [26] conducted a retrospective chart review at community oncology clinics to examine the comparative effectiveness of everolimus-based therapy versus endocrine monotherapy and chemotherapy in the treatment of 202 patients with HR-positive/HER2-negative breast cancer with liver metastasis in the USA. Patients treated with everolimus-based therapy had significantly longer treatment duration than those treated with endocrine monotherapy or chemotherapy, and longer PFS than those treated with endocrine monotherapy. These findings were similar in the first- and second-line settings. Xie et al. [27] conducted a retrospective chart review study in several community oncology practices across the USA to examine the comparative effectiveness of everolimus-based therapy versus endocrine monotherapy among 513 postmenopausal women with HR-positive/HER2-negative mBC. The review included postmenopausal HR-positive/HER2-negative mBC patients who received everolimus-based therapy or endocrine monotherapy as any line of therapy for mBC after NSAI failure. The study evaluated TOT, PFS and time to chemotherapy. In this nationwide sample of postmenopausal HR-positive/HER2-negative mBC patients treated in community oncology settings, treatment with everolimus-based therapy was associated with significantly longer TOT and PFS compared with endocrine monotherapy.

Overall, CER studies most commonly reported treatment duration (e.g., TOT) and various survival outcomes (e.g., OS, PFS, mortality); each reported in at least 71% of these studies. The most common study design was a retrospective chart review, with data acquired most commonly from community-oncology centers.

Of all the topics we searched, we found the greatest number of studies on treatment patterns in HER2-negative and TN mBC: ten articles [29–38] and eight conference abstracts (Table 4) [39,40,41,42,43,44,45,46]. Ten of these were retrospective chart reviews [29,30,32–34,37,38,41,44,45], two were physician surveys [31,46] and six were analyses of secondary databases (e.g., Surveillance, Epidemiology and End Results [SEER]-Medicare database, commercial claims) [35,36,39,40,42,43]. These studies examined a variety of therapeutic options, including endocrine therapy, chemotherapy and surgery. Most of the studies examined HR-positive/HER2-negative metastatic (or ‘advanced’) breast cancer patients, except for five studies that examined patients with TNBC [29,33,39,41,45].

Seah et al. [34] examined the use and duration of chemotherapy in 318 patients with mBC according to line of therapy and tumor subtype (HR-positive, TN or HER2-negative amplified breast cancer). Data were from medical records of patients with mBC at Dana-Farber Cancer Institute. Median OS was 32 months: 54 months for HER2-positive disease, 36 months for HR-positive and 17 months for TNBC. Overall, tumor subtypes influenced the number of lines, duration of chemotherapy and survival. Morris et al. [33] retrospectively assessed OS, patterns of recurrence, treatment and incidence of brain metastases in 1323 patients with TNBC, using data from patient medical records at a single institution. A total of 298 (23%) patients developed metastases. Of these patients, 240 had died, which resulted in a crude death rate of 81%. The median OS from the time of metastatic disease diagnosis for the 298 patients with metastatic disease was 14 months (95% CI: 13–16 months). Of 1323 patients, 99 (7.5%) patients developed brain metastases. Following brain metastasis, observed treatments were radiotherapy (88%), resection (26%) and systemic chemotherapy (71%). The median survival from diagnosis of brain metastasis in mTNBC was 5 months. Li et al. [30] conducted a retrospective chart review of 699 postmenopausal HR-positive/HER2-negative mBC women who received everolimus, endocrine therapy or chemotherapy after a NSAI failure, with data from community-based oncology practices across the USA. The study aim was to examine the factors associated with everolimus use in these patients in the real-world setting. The top reason for prescribing everolimus was efficacy (69–85%). About 15 and 29% of everolimus users in second-line and third-line or above received prior chemotherapy. Exemestane was the most common concomitant therapy with everolimus (56–87%). The majority of patients used everolimus according to the labeled combination and dose. This study indicated that in HR-positive/HER2-negative mBC, everolimus is used in more severe (e.g., metastasis, tumor volume and performance status) patients than endocrine therapy but in less severe patients than chemotherapy. Lin et al. [31] assessed physicians’ preferences and self-reported prescribing patterns for endocrine therapy and chemotherapy in the treatment of HR-positive/HER2-negative mBC at community practices. Data were from a survey of 213 physicians from small/intermediate practices; of which, 58% referred to the National Comprehensive Cancer Network (NCCN) guidelines when treating mBC. Survey results suggested that after first-line endocrine therapy, mono- or combination-endocrine therapy was commonly used for second- and third-line treatments and chemotherapy monotherapy for third- or later-line treatments. Chemotherapies were used in early lines for patients with visceral symptoms. Efficacy was the most important factor for treatment choice. Karve et al. [43] reported the treatment patterns and survival of 13,170 women aged ≥65 years diagnosed with HR-positive/HER-negative mBC based on the SEER-Medicare linked database (2000–2009). 54% of patients had surgery, followed by radiation (50%), hormone therapy (49%) and chemotherapy (40%). Of patients treated with hormonal therapy, 20% received megestrol, 16% received anastrazole and 11% had fulvestrant. Most commonly used chemotherapy agents were cyclophosphamide (21%), doxorubicin (17%) and paclitaxel (14%). Patients in this study had a median OS of 50 months; the 5-year survival rate was 45%. The authors found that survival remains lower in older and ER-/PR+ mBC despite advances in treatment. Krook et al. [44] described the patterns of use of anthracyclines in a community setting among 247 patients with HER2+ and HER2-negative stage IV mBC. Of 420 unique chemotherapy requests in these patients, 54 were anthracycline requests (50 patients); 357 of these requests were for treatment of 205 HER2-negative mBC patients. Approximately 83% of anthracycline requests were for conventional doxorubicin, while the rest were for liposomal doxorubicin. This analysis shows that anthracyclines are primarily used in HER2-negative mBC in the southeast USA, with use occurring in about one out of four patients. Patt et al. [46] analyzed data from a US oncology survey of 104 network physicians to examine the treatment experience of nab-paclitaxel (nab-P) in HER2-negative mBC patients. The 100 mg/m² every week was the most common first line (51%) and second line (58%) dosage, followed by 125 mg/m² every week (22 and 14%), 150 mg/m² every week (4 and 2%) and 260 mg/m² every three weeks (14 and 7%). Common factors affecting choice of nab-P monotherapy for more and less experienced users included co-morbidities/performance status, quality of prior taxane response, steroid intolerance/contraindications/hypersensitivity, predominant visceral metastases and age. Overall, the study showed that 100 mg/m² every week nab-P is the most commonly used starting dosage for HER2-negative mBC patients, and peripheral neuropathy and neutropenia were common dose-limiting toxicities. Hao et al. [42] reported patterns of chemotherapy and endocrine therapy use in patients with HR-positive/HER2-negative advanced breast cancer (stage IV or metastases) based on a retrospective analysis of physician-reported clinical data linked to medical and pharmacy claims (2008–2012) from a large national US health plan. Of 317 metastatic patients, 50% started chemotherapy without prior endocrine treatment (CH), 30% received only endocrine therapy (OT), 17% had endocrine therapy prior to chemotherapy and 3% had neither endocrine therapy nor chemotherapy. Of CH patients, 55% of patients later initiated treatment with aromatase inhibitors, compared with 64% of OT patients (p = 0.104). Overall, this study showed that a large proportion of patients with HR-positive/HER2-negative advanced breast cancer initiated chemotherapy without prior endocrine therapy. Engel-Nitz et al. [40] described the mortality outcomes of patients with HR-positive/HER2-negative advanced and mBC based on a retrospective analysis of clinical data linked to medical and pharmacy claims (2008–2012) from a large national US health plan. There were 263 stage III, 71 stage IV and 317 patients with metastases. Of these, 3.4% of stage III, 22.5% of stage IV and 10.7% of metastatic patients died during follow-up (p ≤ 0.010). 1-year mortality also varied by disease stage: 1.1% (stage III), 12.7% (stage IV) and 5.1% (metastatic patients; p < 0.001). Metastatic patients had significantly shorter mean survival if they did not have endocrine treatment prior to chemotherapy versus other metastatic patients: 35.7 versus 76.1 months; a similar pattern was observed for stage IV patients: 10.0 versus 14.8 months. This study showed that mortality rates varied by stage of HR-positive/HER2-negative disease, while length of survival varied by pretreatment with endocrine therapy prior to chemotherapy. Koshy et al. [29] conducted a single institution retrospective chart review study of 36 patients at the Louisiana State University Health Sciences Center to determine if TNBC (n = 17) is preferentially sensitive to platinum therapy as compared with non-TNBC by reviewing the outcomes of patients treated with cisplatin–gemcitabine chemotherapy. The median PFS for TNBC was 5.3 months compared with 1.7 months in non-TNBC. The risk of progression was reduced by 47% for metastatic TNBC (mTNBC) compared with non-TNBC patients treated with the cisplatin–gemcitabine. The median OS after the start of cisplatin–gemcitabine was 10.8 months in TNBC patients compared with 4.3 months in non-TNBC patients. OS from date of breast cancer diagnosis for the TNBC group was 47.8 months compared with 66.8 months in non-TNBC. The authors concluded that cisplatin–gemcitabine combination chemotherapy is an effective regimen for patients with mTNBC. Xie et al. [37] conducted a retrospective chart review study to compare clinical outcomes, PFS, TOT and OS between 699 HR-positive/HER2-negative mBC patients with multiple metastases versus those with a single metastasis. The authors concluded that patients with multiple metastases had significantly shorter PFS, TOT and OS than single metastasis patients, with similar outcomes in each line of therapy, highlighting the substantial clinical burden and unmet need for more efficacious treatments for the former group of patients. Zeichner et al. [38] examined the survival of patients with de novo mBC by retrospectively analyzing breast cancer data from patient records from a large breast cancer oncology private practice (n = 62) and from the University of Miami/Sylvester Comprehensive Cancer Center tumor database (n = 91). The authors also conducted a review of literature. The goal was to assess how advances in treatment have impacted OS. The authors concluded that the median survival for patients with de novo metastatic disease is in excess of 3 years compared with studies reported from prior to 1996. In the HER2-negative subset, the OS was 39 months among the patients in the private practice group and 36 months in the University of Miami/Sylvester Comprehensive Cancer Center group. They were unable to confirm whether survival for de novo metastatic disease is better than for patients with recurrent disease, however, a subset of these patients can have a prolonged survival with good QoL in excess of 10 years. Macalalad et al. [32] collected data from patient charts across a network of community-based oncology practices to describe the treatment patterns and treatment duration by lines of therapy for endocrine therapy and chemotherapy in 144 postmenopausal women with HR-positive/HER2-negative mBC. Considering that clinical guidelines recommend endocrine therapy as initial treatment for postmenopausal HR-positive/HER2-negative mBC women, with chemotherapy reserved for patients that develop symptomatic visceral disease or have no clinical benefit from three lines of endocrine therapy, the authors examined whether real-world clinical practice reflects these recommendations. Of 144 patients in this study, less than 10% had three or more lines of endocrine therapy before chemotherapy was initiated; a median of two lines of endocrine therapy was observed in these patients. The authors observed that few patients received three lines of endocrine therapy as suggested by NCCN guidelines, and concluded that single-agent endocrine therapy may not be as effective beyond first line, given most patients received less than three lines of endocrine therapy before receiving chemotherapy and the median duration of endocrine therapy decreased from first to second line. Swallow et al. [35] evaluated the real-world patterns of endocrine therapy for HR-positive/HER2-negative mBC in 19,120 postmenopausal mBC patients in the USA from 2002 to 2012. The authors examined mean numbers of lines of therapy and median duration of each line using the MarketScan commercial health insurance claims database (2002Q3–2012Q2). Although endocrine therapy is the preferred first-line treatment option in HER-positive/HER2-negative mBC according to guidelines, only 11,545 (60%) patients initiated endocrine therapy as the first treatment following metastatic diagnosis in this study. Most of these patients received only one line of endocrine therapy (n = 8524; 74%) before discontinuation or transition to chemotherapy, and only 821 (7%) patients were treated with three or more lines of endocrine therapy as recommended by the NCCN treatment guidelines. Overall, patients were treated with a mean of 1.36 lines of endocrine therapy during the study period. In patients that were treated with more than one line of endocrine therapy, the duration of this therapy decreased with each subsequent line. Vaz-Luis et al. [36] conducted a secondary database analysis of Medicare data to examine the impact of race on survival of 4364 women ≥66-year old with de novo mBC by disease subtype. The investigators used trastuzumab as a proxy for HER2-positive status. Patients with HER2-positive disease experienced longer median OS compared with HER2-negative/unknown/HR-positive and HER2-negative/unknown/HR-negative disease. Racial differences in OS were only observed among patients with HER2-positive tumors. Results for breast cancer-specific survival were similar. Overall, racial differences in survival were only apparent for those with inferred HER2-positive tumors. In a chart review of 173 patients, Faria et al. [41] compared time to treatment progression and time to treatment failure in mBC patients with TN disease versus those with ER-positive/HER2-negative mBC receiving eribulin therapy. The authors concluded that in these heavily pretreated HER2-negative mBC patients, treated in a community oncology setting, patients with TNBC achieved similar treatment benefit from eribulin therapy compared with the ER-positive/HER2-negative cohort. Clarke et al. [39] examined 6268 de novo mBC patients by HER2 subtypes in California. The authors found that a high proportion of patients with mBC are not treated surgically and that survival remains poor, with worse survival strongly associated with tumor biology (TN and HR-negative/HER2-positive molecular subtypes). Pahuja et al. [45] retrospectively examined medical charts of 117 TNBC patients to assess responses to subsequent therapies after failure to achieve pathologic complete response after neoadjuvant chemotherapy in patients with TNBC. The authors determined that response to subsequent chemotherapeutic regimens in mTNBC patients who do not achieve pathologic complete response to anthracycline and taxane-based neoadjuvant chemotherapy is limited.

Overall, treatment patterns studies varied both by methodology and objectives. The most common end points examined were treatment patterns and survival. Most common data sources used included physician reports and medical records from cancer centers and university hospitals. We found that most, but not all, patients with HR-positive/HER2-negative disease are treated with at least one line of endocrine therapy. Beyond first line, multiple regimens may be used. Studies often reported that most important factor for treatment choice was efficacy. Unlike for HR-positive/HER2-negative mBC, real-world treatment patterns data in TN mBC are limited.

Treatment appropriateness was addressed by one study, in terms of patient treatment navigation (Table 5) [47]. Ko et al. [47] examined whether patient navigation improved receipt of recommended breast cancer care. In this study, women with breast cancer who participated in the national Patient Navigation Research Program were examined to determine whether the care they received included: initiation of antiestrogen therapy in patients with HR-positive breast cancer; initiation of postlumpectomy radiation therapy; and initiation of chemotherapy in women less than 70-year-old with TN tumors greater than 1 cm. Due to small sample size and limited variation in the receipt of recommended chemotherapy treatment, the authors were unable to assess the relationship between patient navigation and chemotherapy, and concluded that the value of the addition of a navigator is inconclusive from the available data. Overall, studies on treatment appropriateness were limited for TNBC, considering the single study identified was inconclusive for TNBC patients due to a small sample size [47].

Only two full-length articles were identified that reported QoL in mBC patients (Table 6) [48,49]. Both studies examined HER2-negative mBC but QoL measures reported varied widely, including overall QoL (using Patient Care Monitor [PMC]), symptom burden, treatment satisfaction and the use of the Functional Assessment of Cancer Therapy-Breast.

Walker et al. [48] conducted a retrospective chart review study to evaluate the impact of disease progression and of specific sites of metastasis on patient reported outcomes in 102 women with HER2-negative mBC from seven community oncology practices. The study used PMC assessments, a self-report measure of symptom burden and QoL, which was administered as part of routine care. PMC assessments of fatigue, physical pain and trouble sleeping were sensitive to effects of disease progression. Progression was also associated with worsening of physical symptoms, treatment side effects, acute distress and impaired performance scores. Gupta et al. [49] examined the association of chemotherapy versus hormonal therapy on health outcomes among patients with HR-positive, HER2-negative mBC. Data came from a cross-sectional survey of postmenopausal women (200 from the USA; 160 from EU) aged ≥50 years with HR-positive, HER2-negative mBC, undergoing treatment with hormonal therapy or chemotherapy for ≤1 year. Measures included cancer therapy satisfaction questionnaire, Functional Assessment of Cancer Therapy-Breast and work and productivity loss impairment items. Hormonal therapy was associated with better patient-reported outcomes, including better health-related QoL, greater satisfaction with treatment, better feelings about side effects, less bother with treatment side effects and activity impairment and lower family burden, than chemotherapy in first line mBC.

Overall, studies that reported QoL research were based on validated questionnaires and instruments, and included assessments of patient reported outcomes to be able to understand QoL from the patient's perspective. The two QOL studies not only varied in study designs (one was a patient survey and the other a retrospective chart review) but in patient populations, where one study examined ≥50-year-old patients and the other ≥18-year-old patients with HER2-negative mBC.

Findings on costs were reported by three articles [50–52] and two conference abstracts (Table 7) [53,54]. These studies varied in aims, designs, data sources and therapies: for example, three were cost–effectiveness analyses [50,51,54], one study was a budget impact model [52] and one study was a retrospective secondary database analysis [53]. All patients examined were women with HER-negative mBC.

Pawaskar et al. [53] assessed the cost and healthcare utilization burden in HER2-negative, HR-positive, ≥65-year-old women with mBC using the SEER-Medicare Database (2000–2009). The study focused on patients who were not part of a health maintenance organization but who had continuous Medicare (Parts A and B) enrollment for ≥12 months before the date of mBC diagnosis or progression or 2009. The main end points examined were incremental cost–effectiveness ratio, all-cause and mBC-related annual healthcare utilization and costs, and budget impacts. Among 13,170 study patients, the mean number of hospitalizations was 2.5 and emergency department visits were three. Mean (standard deviation) annual all-cause costs per patient were US$34,264 (US$33,429), of which US$16,397 (US$18,284) were attributable to mBC. The authors found hospitalizations, hospital outpatient visits and physician office visits accounted for over three-quarters of total all-cause costs and 75% of mBC-related costs. Of mBC-related costs, mean (standard deviation) annual costs were US$1442 (US$4095) for chemotherapy, US$1872 (US$4086) for radiation and US$579 (US$1197) for hormone therapy. Xie et al. [52] conducted an analysis to estimate the budget impact of everolimus as the first and second treatment option after letrozole or anastrozole failure for postmenopausal women with HR-positive/HER2-negative advanced breast cancer. The study found that introduction of everolimus in combination with exemestane is projected to increase the pharmacy budget while reducing medical service expenditures, yielding a modest net increase of US$522,336 in total budget for the first year from a US payer perspective. Diaby et al. [50] conducted a cost–effectiveness analysis (CEA) of everolimus plus exemestane versus exemestane alone for treatment of HR-positive/HER2-negative mBC. Using the BOLERO-2 data and costs obtained from the Center for Medicare Services drug payment table and physician fee schedule, the benefits were expressed as quality-adjusted PFS weeks and quality-adjusted progression-free years, with utilities/disutilities derived from the literature. Considering an estimated incremental cost–effectiveness ratio (ICER) of US$265,498.5/quality-adjusted progression-free years, the authors concluded that everolimus plus exemestane is not cost-effective compared with exemestane alone. Zeichner et al. [54] conducted a CEA of everolimus in combination with exemestane versus exemestane alone in postmenopausal women with HR-positive/HER2-negative mBC, using the BOLERO-2 trial data, utilities derived from literature and costs obtained from the Center for Medicare Services drug payment table and physician fee schedule (US$2012). The main end points and corresponding results were overall cost of US$62,751.54 per year of PFS gained and an ICER of US$79,376/quality-adjusted life year, and by local assessment an overall cost of US$83,222 per year of PFS gained with an ICER of US$108,131/quality-adjusted life year. The authors concluded that everolimus plus exemestane versus exemestane alone is cost-effective in the treatment of mBC. Refaat et al. [51] conducted a CEA of bevacizumab plus paclitaxel versus paclitaxel alone as first-line chemotherapy in HER2-negative mBC using efficacy and toxicity data from the E2100 study. This study demonstrated that, despite a significant PFS advantage, the addition of bevacizumab to paclitaxel is not cost-effective (marginal cost–effectiveness was US$232,721) in comparison to paclitaxel given alone for the cohort of patients with HER2-negative mBC.

Overall, the cost studies reported a variety of results in HER2-negative mBC patients. Most cost studies were economic models, specifically cost–effectiveness models and the most common end points were ICERs. Data were derived from a variety of sources including the published literature, clinical trials and Center for Medicare Services.