The real-world cost-effectiveness of adjuvant trastuzumab in HER-2/neu-positive early breast cancer in Taiwan

Abstract

Background: Trastuzumab was considered a cost-effective adjuvant treatment for HER 2-positive early breast cancer. Since 2010, the Taiwanese National Health Insurance (NHI) has started to reimburse for 1-year adjuvant treatment. This study aims to provide an updated cost-effectiveness analysis from the NHI perspective, which explores assumptions about long-term cardiac toxicity and treatment benefit of 1-year adjuvant treatment sequentially after chemotherapy.

Methods: A Markov model was used to evaluate the cost-effectiveness of 1-year adjuvant trastuzumab for HER-2/neu positive early breast cancer over a 20-year life-time horizon. A probability sensitivity analysis using Monte Carlo simulation was performed to characterize uncertainties in the expected outcomes, which are expressed as an incremental costs effectiveness ratio (ICER, cost/QALY). A willingness-to-pay threshold of 3-times the per capita gross domestic product was adopted according to the WHO definition. The Taiwan per capita gross domestic product in 2015 was US$22,355; thus, a threshold was considered as NT$2,011,950 (US$67 065, 1USD =30 NTD in 2015).

Results: The model showed that adjuvant trastuzumab treatment in HER-2/neu positive early breast cancer yielded 1.631 quality-adjusted life-years (QALY) compared with no trastuzumab treatment. The ICER was US $51,863 per QALY gained in the base-case scenario. The Monte Carlo simulation by varying all variables simultaneously demonstrated that the probability of cost-effectiveness at the willingness-to-pay threshold of US$67,065 was 50% for 1-year adjuvant trastuzumab.

Conclusions: From this real-world study, 1-year adjuvant trastuzumab treatment is likely to be a cost-effective therapy for patients with HER-2 positive breast cancer at the willingness-to-pay threshold of 3-times GDP per capita in Taiwan.

Keywords:

• Trastuzumab
• HER2-positive
• Early breast cancer

Introduction

Breast cancer is the fourth leading cause of cancer death in Taiwan¹. According to the registry data available in 2012, the standardized incidence rate of female breast cancer was 10,056 and 10,525 per 100,000 females in 2011 and 2012, respectively². The Taiwanese mortality rate was ranked at 31 (10.5%) compared to the 35 OECD countries, and there has been an uptrend in morbidity and mortality over the past several years. Interestingly, the median age at first diagnosis of female breast cancer has remained at 45–49 years in Asian countries such as Taiwan, China, South Korea, and Japan during the past 2 decades—at least 10 years younger than women in Western countries^3–8. Therefore, preventing and reducing breast cancer deaths is one of the most important issues for the Health Promotion Administration in Taiwan.

Trastuzumab use as adjuvant treatment for early breast cancer that over-expresses Human Epidermal growth factor Receptor 2 (HER2) is well established^9–11; however, trastuzumab-related cardiotoxicity has been raising oncologist and cardiologist concerns, although it may reverse after a withdrawal from treatment¹². Patients with cardiotoxicity may have a worse prognosis and an important impact on average life expectancy and consequent cost-effectiveness, but only a few studies reported data regarding the association of cardiotoxicity with trastuzumab outside clinical-trial settings. Until recently, there have been studies confirming that trastuzumab may associate with cardiotoxicity in treated women, particularly in the elderly^13,14. Therefore, an updated cost-effectiveness analysis (CEA) including cardiotoxicity-incurred treatment cost is needed. There is no CEA study based on actual claim data since the approval of reimbursement for 1-year adjuvant trastuzumab therapy for early breast cancer in Taiwan. We intended to perform this study assessing whether long-term reimbursement of adjuvant trastuzumab for 1 year is cost-effective from the perspective of the Taiwan Health Insurance Bureau.

Methods

Markov model

A Markov model was used to assess the cost-effectiveness of 1-year adjuvant trastuzumab treatment for women with HER2-positive early breast cancer. The model incorporated five key health states identified in the progression of breast cancer (1) disease-free state (DFS), defined as the health state after surgery and chemotherapy; (2) local recurrence; (3) cardiac events; (4) metastasis; and (5) death¹⁵. The model was simulated from the National Health Insurance (NHI) perspective. The starting age of the patient cohort was 50 years, matching the majority of clinical trials for HER2-positive early breast cancer and actual data^9–11. A patient in the model was considered to be in one of five health states at any time. All patients entered the model at the initiation of trastuzumab therapy (or at the disease-free state in the non-trastuzumab group) and transitioned from one state to another on the basis of the transition probabilities. The cycle length was 1 year and the time horizon was the 20-year lifetime. We used the definition of willingness-to-pay (WTP) threshold suggested by the WHO: 3-times the per capita gross domestic product (GDP)^15,16. The Taiwan per capita gross domestic product in 2015 was US$22,355; thus a WTP threshold was considered as US$67 065 (NTD 2,011,950; 1 USD = NT$30 in 2015). The transition probabilities and quality-of-life using utility values assigned to each state were derived from available published literature^17–23. Results are presented as incremental cost-effectiveness ratios (ICERs) in cost per quality-adjusted life year (QALY) gained. All costs and health outcomes were discounted at 3% to adjust for the relative value of the Taiwan dollar at present. The schematic of the Markov model is shown in Figure 1. Model parameters are described in Table 1.

Figure 1. Markov model. Arrows indicate transitions. Arrows that curve back to the same state represent remains in the same state. DFS: Disease free survival. CHF: congestive heart failure.

Figure 2. Toranado analysis. EV, expect value of ICER for Trastuzmab

Table 1. Parameters data in sensitivity analysis.

Variable	Base-case value	Minimum value	Maximum value	Assumed distribution	Mean	SD	Sources
Transition probability
Trastuzumab gr.
DF to LR	0.02	0.016	0.037	Beta	0.026	0.014	[14,22]
DF to M	0.05	0.041	0.078	Beta	0.060	0.026	[15,21]
LR to M	0.03	0.031	0.078	Beta	0.055	0.034	[15,21]
M to D	0.22	0.220	0.295	Beta	0.258	0.053	[14,22]
Non-Trastuzumab
DF to LR	0.036	0.022	0.039	Beta	0.031	0.012	[15,21]
DF to M	0.079	0.079	0.091	Beta	0.085	0.0085	[15,21]
LR to M	0.078	0.078	0.091	Beta	0.085	0.0085	[15,21]
M to D	0.33	0.295	0.33	Beta	0.313	0.0245	[14,15]
Utility
Trastuzumab gr.
DF	0.749	0.696	0.97	Beta	0.833	0.194	[17,18]
LR to M	0.810	0.700	0.935	Beta	0.813	0.048	[19,20]
M to D	0.685	0.484	0.822	Beta	0.653	0.239	[15,20,22]
Non-Trastuzumab
DF	0.779	0.749	0.97	Beta	0.859	0.156	[16,17,22]
LR to M	0.810	0.700	0.935	Beta	0.818	0.166	[19,20]
M to D	0.685	0.484	0.822	Beta	0.653	0.239	[15,20,22]
Cost
Trastuzumab*
DF 1st year	27,256	19,736	35,724	Constant	27,256	2 550	Claim database
DF subsequence year	16,185	5634	31,104	Constant	16,185	10,842	Claim database
LR	40,115	40,115	40,115	Constant	40,115	–	Claim database
M	38 403	10,609	57,327	Constant	38,403	14,435	Claim database
Non-Trastuzumab
DF 1st year	16,241	13,287	19,378	Constant	16,241	2 495	Claim database
DF subsequence year	5 171	1533	8087	Constant	5 171	2 599	Claim database
LR	14,415	14,415	14,415	Constant	14,415	–	Claim database
M	18,856	7689	28,106	Constant	18,856	7 568	Claim database

All costs are in US dollars, exchange rate: 1.00 US$= NT$30.

DF, Disease free; LR, local recurrence; M, metastases; D, death; NT, non Trastuzumab.

* Cardiotoxicity treatment cost included.

Treatment regimen

Two treatment scenarios were examined from the data retrieved from the NHI database. The first, a trastuzumab group, defined by the reimbursement criteria that trastuzumab is used as an adjuvant treatment only reimbursed for 1 year pre- or post-operation or after standard chemotherapy, and a second non-trastuzumab group defined as the commonly used chemotherapy regimen for HER-2+ early breast cancer women without trastuzumab (such as combination of decetaxel or paclitaxel, doxorubicin and cyclophosphamide, which were identified by ATC code and ICD-9 code 174.9). The initiate dose of trastuzumab sequentially administered within 3 weeks after chemotherapy was 8 mg/kg IV infusion over 90 min. The dose of subsequent therapy was 6 mg/kg IV infusion over 30–90 min every 3 weeks for a total of 17 doses (52 weeks of therapy).

Costs

All direct medical costs for each patient, including physician visits, pharmacist-dispensing fees, laboratory and diagnostic tests as well as additional treatments (surgery, hormonal therapy, chemotherapy, radiotherapy and special interventions) were retrieved directly from the NHI database. The average direct medical cost for each health state was then calculated and used to estimate the annual medical expenditures under different health states. Costs associated with detection and treatment of cardiac events were also retrieved from the database.

Sensitivity analysis

A one-way sensitivity analysis was conducted to determine the potential impact of different parameters on this analysis; the result was presented as a tornado diagram. We hypothesized that the parameters varied over a range of minimum and maximum values (Table 1), which was retrieved from the published literature and approved by the experts. A probability sensitivity analysis (PSA) using a Monte Carlo simulation was performed to assess the impact of uncertainty around the key parameters of the model on the incremental cost-effectiveness ratio (ICER). That is, distributions for each parameter with the probabilistic sensitivity analysis were modeled. Log-normal distributions were adopted for all costs and beta distributions were adopted for probabilities, utilities, and toxicity. The probabilistic sensitivity analysis was based on 10,000 samples, and the results are presented as a cost-effectiveness acceptability curve.

Results

Base-case analysis

The direct medical costs in different health states were calculated by the total direct medical cost retrieved from the database as the criteria for patients with HER-2 positive early breast cancer who received treatment regimen with trastuzumab or without trastuzumab. The first year of medical costs, including cost of surgery, chemotherapy, toxicity, and trastuzumab, in the DFS was higher than that of the subsequent year. Adjuvant trastuzumab treatment yielded an ICER of USD $51 863 (NTD 1,711,485) per QALY gained in the base-case scenario (Table 2). Based on the WTP threshold as defined in Taiwan, 1-year adjuvant trastuzumab for breast cancer was a cost-effective therapy.

Table 2. Cost-effectiveness analysis for lifetime model.

	Regimen
	Trastuzumab		Non-trastuzumab		Incremental cost		Incremental effectiveness		ICER (cost/QALY)
	Base	PSA	Base	PSA	Base	PSA	Base	PSA	Base	PSA
Direct medical cost (US$)	151,080	136,112	58,052	56,069	93,028	80,042
QALYs gained	4.984	4.431	3.353	3.375			1.631	1.056	57,037	75,798

PSA, Probability sensitivity analysis; 1US$= NT$30.

Sensitivity analysis

One-way sensitivity analyses showed important parameters driving the model. The tornado analysis showed that the key parameters of the model were transition probabilities of the DFS for both the trastuzumab and non-trastuzumab group; health utility for non-trastuzumab group in disease free state; utility for trastuzumab in metastasis state (Figure 2).

Figure 3. (a) Incremental cost-effectiveness plane. A sample of 10,000 results of the Markov Chain Monte Carlo simulation are plotted. The diagonal line represents a willingness-to-pay threshold of this study, where estimates below this line are cost-effective. (b) Cost-effectiveness acceptability curve. These show that probability that a specific treatment is cost-effective at a given willingness-to-pay threshold of US$67,060 (NT$2,213,145).

The Monte Carlo simulation by varying all variables simultaneously demonstrated that the probability of cost-effectiveness at the WTP threshold of US$67 065 per QALY gained was 50% chance for 1-year adjuvant trastuzumab (Figure 3). The ICER range is shown in Table 2.

Discussion

Our analysis indicates that 1 year of adjuvant trastuzumab after surgery or chemotherapy is a cost-effective treatment option compared with chemotherapy alone. The incremental cost per QALY gained was estimated to USD 51 863 (NTD 1 711 485) for women with HER2-positive early breast cancer. These estimates of the cost per QALY gained are 50% chance at the WTP threshold of US$67 065 used in Taiwan (no consensus WTP threshold has yet been determined so we, therefore, used the WHO cost-effectiveness threshold). To our knowledge, this is a first study in Taiwan to assess the cost-effectiveness of 1-year adjuvant trastuzumab after surgery or chemotherapy for HER +2 early breast cancer after the reimbursement by the national healthcare plan. Therefore, the result of this study could present valuable information for decision-markers to re-evaluate their payment policy on whether it is still of value to pay for the 1-year adjuvant treatment strategy continuously.

Adjuvant trastuzumab has previously been proven to be a cost-effective treatment for HER2-positive early breast cancer in many countries. Most of these studies were performed before 2009 and did not include the treatment cost of grade 3 or 4 cardiotoxicity^23–25. Therefore, the ICER may be under-estimated for the lifetime model; however, few studies performed beyond 2009 including treatment cost of cardiotoxicity reported that 1-year adjuvant trastuzumab treatment is not cost-effective using the definition of WHO cost-effectiveness threshold of 3-times GDP per capita^26,27. The reason may be due to the high price of drugs and lower per capita GDP in those countries.

The strength of our study was that our cost data was retrieved directly from actual claims data, eliminating the impact of bias from using questionnaire surveys of clinical experts. The limitation of our study was that some clinical oncologist omitted precise ICD-9 of cardiotoxicity in the patients’ medical record. Therefore, the cost of treating cardiotoxicity may be under-estimated because it may not be directly reflected in the claim database.

In conclusion, from this real-world study, 1-year adjuvant trastuzumab treatment is likely to be a cost-effective therapy for patients with HER-2 positive breast cancer at the WTP threshold of 3-times GDP per capita in Taiwan.

Transparency

Declaration of funding

The authors report no funding for this manuscript.

Declaration of financial/other relationships

The authors report no financial/other relationships to declare for this manuscript. JME peer reviewers on this manuscript also have no relevant financial or other relationships to disclose.