Cost of adverse events during treatment with everolimus plus exemestane or single-agent chemotherapy in patients with advanced breast cancer in Western Europe

Abstract

Objective:

Treatment options for recurrent or progressive hormone receptor-positive (HR+) advanced breast cancer include chemotherapy and everolimus plus exemestane (EVE + EXE). This study estimates the costs of managing adverse events (AEs) during EVE + EXE therapy and single-agent chemotherapy in Western Europe.

Methods:

An economic model was developed to estimate the per patient cost of managing grade 3/4 AEs for patients who were treated with EVE + EXE or chemotherapies. AE rates for patients receiving EVE + EXE were collected from the phase III BOLERO-2 trial. AE rates for single-agent chemotherapy, capecitabine, docetaxel, or doxorubicin were collected from published clinical trial data. AEs with at least 2% prevalence for any of the treatments were included in the model. A literature search was conducted to obtain costs of managing each AE, which were then averaged across Western European countries (when available). Per patient costs for managing AEs among patients receiving different therapies were reported in 2012 euros (€).

Results:

The EVE + EXE combination had the lowest average per patient cost of managing AEs (€730) compared to all chemotherapies during the first year of treatment (doxorubicin: €1230; capecitabine: €1721; docetaxel: €2390). The most costly adverse event among all patients treated with EVE + EXE was anemia (on average €152 per patient). The most costly adverse event among all patients treated with capecitabine, docetaxel, or doxorubicin was lymphocytopenia (€861 per patient), neutropenia (€821 per patient), and leukopenia (€382 per patient), respectively.

Conclusions:

The current model estimates that AE management during the treatment of HR+ advanced breast cancer will cost one-half to one-third less for EVE + EXE patients than for chemotherapy patients. The consideration of AE costs could have important implications in the context of healthcare spending for advanced breast cancer treatment.

Keywords::

• Adverse events
• Everolimus
• Exemestane
• Metastatic breast cancer
• Chemotherapy
• Costs and cost analysis

Introduction

Breast cancer is the most common form of cancer diagnosed in women worldwide, with 1.67 million new cases diagnosed in 2012 (25% of all cancers) and a yearly incidence of 27–96 cases per 100,000 people, depending on geographic region1,2. More than half of patients3–5 with breast cancer (estimates range from 39–69%5–7) are classified as having hormone receptor-positive/human epidermal growth factor receptor-2 negative (HR+/HER2−) breast cancer, also known as endocrine receptor (estrogen or progesterone receptor) positive/HER2− (ER+/HER2−) breast cancer8. Endocrine/hormone therapy is recommended to treat HR+/HER2− advanced breast cancer as first-line treatment9. However, not all patients respond to hormonal therapy, and some patients who initially respond will later develop resistance to treatment10,11. For these patients, treatment options include alternative hormone monotherapy, chemotherapy, and the recently approved everolimus in combination with hormone therapy, specifically exemestane (EVE + EXE)9,12.

The efficacy of EVE + EXE in breast cancer was demonstrated in the multi-center phase III BOLERO-2 (NCT00863655) clinical trial. In the trial, EVE + EXE more than doubled the median progression-free survival (PFS) period compared to exemestane alone, 7.8 vs 3.2 months13,14. EVE + EXE was approved to be used following failure of hormone therapy with non-steroid aromatase inhibitors (letrozole or anastrozole) among women with HR+/HER2− advanced breast cancer by the US Food and Drug Administration (FDA) in July 201215 and by Health Canada in January 201316, and among HR+ post-menopausal women with advanced breast cancer by the European Medicines Agency (EMA) in July 201217,18.

Chemotherapy is another treatment option among patients who failed or developed resistance to hormone therapy. It is generally recommended to be used after hormone therapy is no longer effective8. The most commonly administered first-line single-agent chemotherapies for the treatment of advanced breast cancer include anthracyclines and taxanes9,19. Doxorubicin is the most commonly used anthracycline. For patients who are ineligible for anthracycline, single-agent docetaxel is recommended as the initial choice; single-agent capecitabine or vinorelbine are equally recommended as the next line of therapy12, although capecitabine has been shown to be more effective than vinorelbine20. In summary, doxorubicin, docetaxel, and capecitabine are all treatment options for advanced breast cancer patients who failed prior hormone therapy. They are effective in delaying progression and reducing cancer-related symptoms21,22.

In addition to clinical efficacy in delaying progression and reducing cancer-related symptoms, tolerability and its associated costs is another important factor to consider when evaluating treatment options. Prior research indicates that the costs of treating adverse events (AEs) associated with chemotherapy are an important economic consideration alongside the cost of therapy itself among women with advanced breast cancer23,24. Patients who experience serious chemotherapy-related AEs incur hospitalization costs over 4-times greater and emergency room costs nearly 5-times greater than patients who do not experience serious chemotherapy-related AEs24. Common AEs experienced among chemotherapy-treated patients can be severe and include neutropenia, hand–foot syndrome, diarrhea, cardiotoxicity, and neuropathy, depending on the therapy25,26. On the contrary, use of EVE + EXE is more often associated with AEs such as stomatitis, anemia, and pneumonitis13,27. Differences in the AE profiles of EVE + EXE− vs chemotherapy-treated patients may translate into substantial cost differences. Accordingly, the current study aims to evaluate the expected costs of managing and treating AEs during EVE + EXE therapy and single-agent chemotherapy for HR+ advanced breast cancer patients living in Western Europe.

Method

An AE cost model was developed to estimate the per patient cost of managing adverse events during the first year of treatment among patients receiving EVE + EXE and three different single-agent chemotherapies: capecitabine, docetaxel, and doxorubicin. The model included two main inputs: (1) frequencies of grade 3 and 4 AEs for breast cancer patients; and (2) costs for each included AE (averaged across western European countries). These inputs to the model were used to calculate the average cost of treating each AE (per patient) and the total average cost of treating all AEs (per patient) for each therapy.

For the first model input, the AE rates of EVE + EXE were based on the reported rates among all patients enrolled in the BOLERO-2 trial; a targeted literature review of clinical trial publications and drug labels was conducted to obtain the AE rates for chemotherapy. Frequency of occurrence of each grade 3/4 AE was extracted from these publications, and a weighted average of the rate of each AE was calculated by weighting the rates reported in each publication by the number of patients included in the study. The AE rate estimates did not take into consideration multiple episodes of the same event, and the reported rates were used as a 1-year rate with the assumption that the majority of AEs will occur early in the treatment course (the follow-up time for each trial was longer than 1 year). Only grade 3/4 AEs with a prevalence of at least 2% during any of these treatments were included in the study.

For the second input to the model, a targeted literature search was performed for relevant articles, using the following criteria: (1) articles were required to contain information on the per event costs associated with each of the previously selected grade 3/4 AEs; and (2) these per event costs were reported for patients living in Western Europe, including the following countries: the UK, Germany, France, Italy, Belgium, Spain, Sweden, and Switzerland. When a single study reported costs for a specific AE for multiple Western European countries, an average was taken. If the costs of a given AE were not reported for Europe, US data was used instead. If the costs of a selected AE were not available in the literature, costs were estimated based on inpatient and outpatient resource use and unit costs. Data on the proportion of patients with a specific AE resulting in hospitalization and an associated length of stay were obtained from the BOLERO-2 clinical trial. For patients whose AEs did not lead to an inpatient stay, two physician visits per AE event were assumed. Unit costs for the assumed visits were obtained from the literature. All costs were converted and/or inflated to 2012 euros (€). Costs were first inflated to 2012 euros using the Harmonized Index of Consumer Prices (HICP) provided by the European Central Bank (ECB)28 when the original costs were reported in euros and using the Consumer Price Index for UK for costs expressed in pounds sterling29; US costs were inflated to 2012 USD using the US Bureau of Labor Statistics’ Consumer Price Index data30. After inflation, costs were converted, when needed, to 2012 euros using the conversion rate published on January 31, 201331.

The model outputs, the per patient costs of treating each AE, and all AEs were calculated by multiplying the previously estimated cost of managing a particular AE in Western Europe by the rate of occurrence of that AE for each treatment. The costs for treating each AE per patient were summed to compute the total average cost per patient for each drug of interest. No statistical comparisons were performed in this study.

Results

Literature search and data extraction: AE rates

The AE rates for EVE + EXE, reported in Table 1, were collected from the phase III BOLERO-2 trial, with a data cut-off of December 15, 2011 (based on patient-level data analyzed by the sponsor). The most common grade 3 or 4 AEs for this drug treatment were stomatitis (occurring in 7.9% of patients), anemia (7.5%), infection (6.6%), gamma-glutamyltransferase increase (6.2%), and hyperglycemia (5.6%).

Table 1. Cost and rates of adverse events.

Grade 3/4 adverse event	Cost per event (2012 EUR)^a	Adverse event rates
		EVE + EXE^b	Capecitabine^c	Docetaxel^d	Doxorubicin^e
Abdominal pain38	€157	0.2%	3.4%	1.4%	0.4%
Alanine aminotransferase increased45	€271	3.3%	0.0%	0.0%	0.0%
Alopecia39	€194	0.0%	0.0%	4.2%	0.0%
Anemia40	€2024	7.5%	3.4%	4.0%	6.8%
Anorexia40	€701	0.0%	2.5%	0.5%	0.1%
Aspartate aminotransferase increased45	€297	3.5%	0.0%	0.0%	0.0%
Asthenia40	€689	2.1%	0.0%	8.6%	3.4%
Dehydration41	€2417	0.4%	4.2%	0.3%	0.0%
Diarrhea40	€1735	2.5%	14.7%	6.0%	1.2%
Dyspnea40	€495	5.2%	1.0%	1.4%	0.0%
Fatigue40	€689	4.6%	7.8%	4.2%	0.6%
Gamma-glutamyltransferase increased45	€241	6.2%	0.0%	0.0%	0.0%
Hand-and-foot syndrome40	€1401	0.0%	9.3%	0.4%	0.0%
Hyperbilirubinemia45	€241	0.2%	9.3%	1.6%	0.0%
Hyperglycemia42	€287	5.6%	0.0%	0.0%	0.0%
Hypokalemia45	€241	3.3%	3.4%	0.0%	0.0%
Infection43	€1851	6.6%	0.0%	1.8%	2.6%
Leukopenia40	€1730	0.6%	0.0%	31.2%	22.1%
Lymphocytopenia40	€1730	1.5%	49.8%	33.7%	0.0%
Neutropenia40	€1385	2.3%	3.4%	59.3%	24.4%
Pneumonitis44	€2102	3.3%	0.0%	0.0%	0.0%
Stomatitis40	€763	7.9%	5.9%	3.4%	6.5%
Thrombocytopenia40	€1911	2.9%	0.0%	1.5%	4.0%
Vomiting and nausea40	€1202	1.6%	8.3%	4.0%	12.3%

^aCosts were inflated to 2012 EUR using the Harmonized Index of Consumer Prices (HICP) provided by the European Central Bank (ECB)28 for most European countries, the Consumer Price Index for UK29, and the Consumer Price Index for the US30. Western Europe costs were averaged across countries based on data availability. Countries considered include: Belgium, France, Germany, the UK, Italy, Spain, Sweden, and Switzerland.

^bRates of adverse events for Everolimus + Exemestane are from Novartis’ data-on-file based on the BOLERO-2 trial (data cut-off at December 15, 2011).

^cCapecitabine AE rates come from a weighted average of rates reported in the Phase II stage IV breast cancer Capecitabine monotherapy trial (see US label32) and the open label sequential dosing trial33.

^dRates of adverse events for Docetaxel are a weighted average of rates reported in the (i) Phase III Clinical Trial Capecitabine + Docetaxel combination therapy vs Docetaxel monotherapy (see US label for Capecitabine32), (ii) Phase III Clinical Trial Docetaxel vs Doxorubicin34, and (iii) Phase III Clinical AVADO35.

^eRates of adverse events for Doxorubicin are a weighted average of rates reported in the (i) Phase III Clinical Trial Doxorubicin vs Caelyx/Doxil36, (ii) Phase III Clinical Trial Docetaxel vs Doxorubicin34, and (iii) Phase III Intergroup Trial E119337.

The AE rates for single-agent capecitabine were extracted from the US drug label (reporting Phase II stage IV breast cancer capecitabine monotherapy trial data32) and from the published information on an open-label sequential dosing trial33. Based on the weighted average of rates for each AE, the most common grade 3 or 4 AEs for capecitabine patients were lymphocytopenia (occurring in 49.8% of patients), diarrhea (14.7%), hyperbilirubinemia (9.3%), hand-and-foot syndrome (9.3%), and vomiting and nausea (8.3%).

The rates of AEs for single-agent docetaxel were extracted from the US label of capecitabine (reporting data on the Phase III clinical trial capecitabine + docetaxel combination therapy vs docetaxel monotherapy), the publication of the results of the Phase III clinical trial docetaxel vs doxorubicin34, and the publication of the outcomes of the Phase III AVADO clinical trial35. From the weighted average of rates of these reported AEs, the most common grade 3 or 4 AEs for docetaxel patients were neutropenia (occurring in 59.3% of patients), lymphocytopenia (33.7%), leukopenia (31.2%), asthenia (8.6%), and diarrhea (6.0%).

The rates of AEs for single-agent doxorubicin were extracted from publications reporting the results of the Phase III clinical trial doxorubicin vs Caelyx/Doxil36, the Phase III clinical trial of docetaxel vs doxorubicin34, and the Phase III Intergroup trial E119337. Based on the weighted average of these rates of AEs, the most common grade 3 or 4 AEs for doxorubicin patients were neutropenia (occurring in 24.4% of patients), leukopenia (22.1%), vomiting and nausea (12.3%), anemia (6.8%), and stomatitis (6.5%).

Literature search and data extraction: costs of each AE

Information on the costs associated with the management of each AE listed in Table 1 was extracted from publications38–45. For the majority of AEs, the studies reported average per episode costs of treating each grade 3/4 AE in euros, accounting for both treatment and hospitalization costs, when relevant. The average cost of an AE was calculated when costs were reported across multiple European countries40,41,43. For AEs for which only US costs were obtained from the literature, (e.g., alopecia, hyperglycemia)39,42, the final cost was converted to euros.

Cost information was lacking in the literature for some AEs. These AEs were mostly toxicities associated with metabolic dysfunction: increased serum levels of alanine aminotransferase, elevated levels of aspartate aminotransferase, increased levels of gamma-glutamyltransferase, hyperbilirubinemia, and hypokalemia. Among these AEs, a small proportion of patients with alanine aminotransferase increase and aspartate aminotransferase had associated hospitalization. To estimate the costs for the two AEs, the cost of hospitalization per day/the cost per outpatient visit was derived from the listed cost of an oncologist visit, and this value was inflated and converted from UK pounds to the final amount of €120.2529,45,46. This cost per day/visit was then multiplied by the number of days/visits (as reported in the BOLERO-2 trial, when available) to generate an estimate of cost per AE. To estimate costs for the rest of the AEs, patients with elevated levels of gamma-glutamyltransferase, hyperbilirubinemia, or hypokalemia were assumed to have visited a physician twice (e.g., one initial visit and one follow-up visit).

Costs of treating AEs

Based on AE rates reported for each treatment and unit costs per AE, the per patient costs of managing AEs were estimated for patients receiving each treatment. The per patient costs of managing AEs for the first year of treatment with EVE + EXE were €730. Per patient costs of managing AEs during treatment with docetaxel, capecitabine, and doxorubicin were greater than the costs for AEs occurring with EVE + EXE treatment by €1660, €991, and €500, respectively. Per patient costs for all treatments are shown in Figure 1.

Figure 1. Cost of managing adverse events—per patient in euros.

The top two most costly AEs on a per patient basis by treatment were: anemia (€152) and infection (€122) for patients receiving EVE + EXE, lymphocytopenia (€861) and diarrhea (€256) for the use of capecitabine, neutropenia (€821), and lymphocytopenia (€583) for the use of docetaxel, and leukopenia (€382) and neutropenia (€337) for the use of doxorubicin (Table 2). Neutropenia and lymphocytopenia were the most prevalent AE for docetaxel and capecitabine, respectively, and they were also the most costly AE for each drug on a per patient basis. Costs per patient for anemia and leukopenia were the highest for EVE + EXE and doxorubicin, respectively, despite being the second most prevalent AEs for each treatment.

Table 2. Most costly adverse event on a per patient basis, by treatment.

No.	EVE + EXE		Capecitabine		Docetaxel		Doxorubicin
	AE	Costs	AE	Costs	AE	Costs	AE	Costs
1	Anemia	€152	Lymphocytopenia	€861	Neutropenia	€821	Leukopenia	€382
2	Infection	€122	Diarrhea	€256	Lymphocytopenia	€583	Neutropenia	€337
3	Pneumonitis	€69	Hand-and-food syndrome	€130	Leukopenia	€540	Vomiting and nausea	€148
4	Stomatitis	€60	Dehydration	€102	Diarrhea	€104	Anemia	€139
5	Thrombocytopenia	€55	Vomiting and nausea	€100	Anemia	€81	Thrombocytopenia	€77
6	Diarrhea	€43	Anemia	€68	Asthenia	€59	Stomatitis	€50
7	Neutropenia	€32	Fatigue	€54	Vomiting and Nausea	€49	Infection	€48
8	Fatigue	€32	Neutropenia	€47	Infection	€33	Asthenia	€24
9	Lymphocytopenia	€26	Stomatitis	€45	Fatigue	€29	Diarrhea	€21
10	Dyspnea	€26	Hyperbilirubinemia	€22	Thrombocytopenia	€29	Fatigue	€4
11	Vomiting and nausea	€19	Anorexia	€18	Stomatitis	€26	Anorexia	€1
12	Hyperglycemia	€16	Hypokalemia	€8	Alopecia	€8	Abdominal pain	€1
13	Gamma-glutamyltransferase increased	€15	Abdominal pain	€5	Dehydration	€7	Alanine aminotransferase increased	€0
14	Asthenia	€14	Dyspnea	€5	Dyspnea	€7	Alopecia	€0
15	Aspartate aminotransferase increased	€10	Alanine aminotransferase increased	€0	Hand-and-foot syndrome	€6	Aspartate aminotransferase increased	€0
16	Leukopenia	€10	Alopecia	€0	Hyperbilirubinemia	€4	Dehydration	€0
17	Dehydration	€10	Aspartate aminotransferase increased	€0	Anorexia	€3	Dyspnea	€0
18	Alanine aminotransferase increased	€9	Asthenia	€0	Abdominal pain	€2	Gamma-glutamyltransferase increased	€0
19	Hypokalemia	€8	Gamma-glutamyltransferase increased	€0	Alanine aminotransferase increased	€0	Hand-and-foot syndrome	€0
20	Hyperbilirubinemia	€0	Hyperglycemia	€0	Aspartate aminotransferase increased	€0	Hyperbilirubinemia	€0
21	Abdominal pain	€0	Infection	€0	Gamma-glutamyltransferase increased	€0	Hyperglycemia	€0
22	Alopecia	€0	Leukopenia	€0	Hyperglycemia	€0	Hypokalemia	€0
23	Anorexia	€0	Pneumonitis	€0	Hypokalemia	€0	Lymphocytopenia	€0
24	Hand-and-foot syndrome	€0	Thrombocytopenia	€0	Pneumonitis	€0	Pneumonitis	€0

Discussion

Historically, guidelines for treating HR+/HER2− advanced breast cancer patients who failed hormone therapy have recommended chemotherapy (National Institute for Health and Care Excellence [NICE]12 and National Comprehensive Cancer Network [NCCN]9 guidelines). Recently, the EVE + EXE combination was recommended in the US9 and was approved for marketing in Europe for treatment of advanced breast cancer patients who failed hormone therapy with non-steroid aromatase inhibitors17. Although the effectiveness of chemotherapy and EVE + EXE is established, there is no available information on the comparative effectiveness and tolerability of EVE + EXE treatment vs chemotherapy. While the current study is not designed to evaluate the comparative effectiveness and safety of different treatment options, knowledge of the type of AEs associated with each choice of treatment, as well as the costs incurred for treating these AEs, can provide valuable information to patients and physicians.

AEs could affect patients’ incurred costs, quality-of-life (QoL), and treatment choices. The management of AEs has been shown to increase economic burden to patients and society47. Prior research has shown that chemotherapy-associated AEs can be very costly: incremental monthly costs associated with managing AEs in chemotherapy patients can range from $3000 to nearly $7000 (USD 2011)23. A prior study has also estimated that the costs of AEs can constitute approximately one fifth of the overall costs among docetaxel-treated metastatic breast cancer patients48. AEs associated with chemotherapy not only increase costs, but also decrease QoL49; in contrast, the combination of EVE + EXE presents a higher rate of AEs compared to EXE alone50, but QoL remains unaffected51. AEs could also impact patients’ treatment choices. Prior studies have shown that willingness to receive potentially toxic chemotherapy treatment may vary widely by patient, and that some cancer patients may avoid aggressive treatments with high risk of severe side-effects52–55.

In the current study, we have investigated the rates of grade 3/4 AEs for four advanced breast cancer treatment choices and evaluated the costs associated with each event, as well as the total annual costs incurred by patients for AE management. We found that the per patient costs for managing AEs in advanced breast cancer patients varied by treatment, and that EVE + EXE had the lowest total cost of all therapies, amounting to only 31–59% of the cost of treating AEs associated with chemotherapeutic treatments. In addition, the costliest AEs per patient for chemotherapy (neutropenia, lymphocytopenia, and leukopenia) were also the most common or the second most common AEs for those therapies. The costs per patient for these AEs were 2.5–5.7-times higher than the cost of treating anemia, the costliest AE on a per patient basis for the EVE + EXE patients. The most commonly reported AEs for advanced breast cancer treatment with EVE + EXE were stomatitis, anemia, and infection.

For payers and policy-makers, the substantial cost of AE management should be a factor considered alongside efficacy and therapy costs. Since no prior study has compared the direct economic impact of AEs arising from EVE + EXE treatment vs chemotherapies, the information presented in this current study aims to fill this knowledge gap. Further studies are warranted to examine the overall net cost of using a therapy—including the cost of therapy, costs for treating AEs, and other costs associated with disease progression and management.

Limitations

The current study is subject to several limitations. First, the current model does not take into account AE recurrences for the same patient; the model assumes that patients will only have one occurrence of the same event, which may lead to under-estimation of the AE costs. Second, the generalizability of this study may be affected by the fact that only grade 3/4 AEs with a prevalence of at least 2% during any of the evaluated treatments were included in the study. Third, rates of AEs reported in clinical trials may be different from those observed in the real world24. The current study has used the AE rates reported in clinical trials (with progression-free survival ranging in length from 4 to 8 months)14,32–37 and assumed that all the AEs occurred within 1 year. However, patients in the real-world could be exposed to drugs for different durations than those observed in clinical trials, which could impact the rate of AEs. This study estimates the average per patient cost of treating AEs; however, the cost could vary between individuals. In addition, cross-trial differences in patient populations could influence the AE rates, and AEs might be measured and reported differently across trials. These heterogeneities among the trials could impact our cost calculations. Finally, the AE rates used in the current model were calculated based on the reported rates from clinical trials for each drug. An indirect comparison was not attempted given the lack of a common comparator across trials. Some of the included trials had only a single arm. Future studies measuring AE rates for different treatments in the same setting are needed to verify the findings from the current study. The BOLERO-6 trial, with an expected completion date in 2015, will provide a comparison of efficacy and safety for EVE + EXE vs capecitabine56.

Conclusion

The current study shows that the expected costs of managing AEs in patients with advanced breast cancer receiving an EVE + EXE are lower than the costs incurred by patients receiving chemotherapies with capecitabine, docetaxel, or doxorubicin. Economic considerations of AE management in patients with HR+/HER2− breast cancer can have important implications for assessing the cost and cost-effectiveness of treatments for advanced breast cancer. The development and use of AE management guidelines and early identification of AEs may reduce the overall costs of managing such occurrences.

Transparency

Declaration of funding

This study was funded by Novartis Pharmaceuticals Corporation.

Declaration of financial/other relationships

J.Z. and H.G. are employees of Novartis and own stock/stock options. H.Y., E.F., A.K., and J.S. are employees of Analysis Group Inc., which has received consultancy fees from Novartis. M.C. acted in a consulting capacity for Novartis for this project. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgments

We thank Ana Bozas, PhD, an employee of Analysis Group Inc., who contributed to the preparation and editing of the manuscript.