Cost effectiveness and affordability of trastuzumab in sub-Saharan Africa for early stage HER2-positive breast cancer

Non-communicable diseases, are responsible for about 70% of deaths worldwide. It was predicted that by 2050, 24 million people will be diagnosed with cancer annually. Among them, up to 70% will be from low-income and middle-income (LMICs) countries [1]. The increasing incidence in these countries is due to lifestyle changes, increased life expectancy, and the improvements in treating infectious diseases [1]. Health care systems in many African countries are struggling to deal with the increasing demand caused by the increasing number of cancer patients. Furthermore, using the same care guidelines from high-income countries in a region with less resources and fewer personnel is inappropriate. The high mortality of cancer patients in Africa is multifactorial—due to poor infrastructure, lack of skilled health-care workers, advanced stage at diagnosis, reliance on traditional therapy, few treatment options, and poor compliance [2]. According to a report published by the Institute of Medicine (IOM), the burden of cancer is growing in many poor countries [3]. Breast cancer is the second most common cancer worldwide, and the most common among women. Moreover, breast cancer is the fifth most common cause of death overall, the most frequent cause of death among women in less developed regions, and the second cause of death in more developed regions [4].

Trastuzumab is a humanized monoclonal antibody that downregulates the extracellular domain of the HER2 protein. Trastuzumab significantly increases the cure rate in patients with HER2 positive localized breast cancer, and has therefore become a standard adjuvant treatment for early stage breast cancer in many countries [5]. In a survey conducted on breast cancer management in Africa, trastuzumab was available in 10 out of 19 facilities. However, only 5% of the patients were able to afford it [6]. Due to minimal availability of data, there is very scant evidence regarding actual treatment patterns in sub-Saharan Africa. Furthermore, it was found that many breast cancer patients in sub-Saharan Africa are usually treated with tamoxifen regardless of their receptor status [2].

The aim of this study was to explore the cost-effectiveness and affordability of adjuvant trastuzumab treatment for HER2-positive breast cancer in 11 countries in sub-Saharan Africa, from a societal perspective. The decision whether an intervention is good value for money or not is determined by comparing the incremental cost-effectiveness ratio result (ICER) in each country to a specific willingness to pay (WTP) threshold [7]. In LMICs, the threshold suggested by the World Health Organization (WHO) is related to the annual gross domestic product (GDP) per capita of each country. If the ICER of the intervention is less than 1xGDP per capita, the intervention is considered as very cost-effective. If the ICER is between 1 and 3× GDP per capita, the intervention is considered cost-effective. Otherwise, it is considered not cost-effective [7]. In order to use WHO’s suggested threshold, the ICER should be in terms of cost per DALY (disability-adjusted life year) [7], however according to a study that reviewed the cost-effectiveness literature, it was found that cost-per quality adjusted life year (QALY) is used to address diseases in wealthier countries such as cancer, whereas cost-per DALY tend to address more prevalent diseases in low income countries such as HIV. Furthermore, QALYs tend to be used for interventions that evaluate pharmaceuticals, while DALYs are used to evaluate interventions that are more focused on immunizations [8]. Additionally, the use of a figure close to the GDP per capita as a cost-effectiveness threshold is widespread in many countries. For example, the United States cost-effectiveness ratio, as for 2016 was 50,000 USD per QALY, while their GDP per capita was 57,588 USD. The United Kingdom cost-effectiveness threshold as for 2016, was set to be 20,000–30,000 GBP/QALY (~ 25,245–37,867 USD per QALY), and the GDP per capita was 40,412 USD [9]. The Netherlands cost-effectiveness threshold was set to 40,000 EUR per QALY (~ 45,498 USD), and the GDP per capita was 45,637 USD [10]. Australia’s GDP per capita is 53,800 USD, however they do not have a specific threshold for funding a new medicine. However, it is more likely that a new drug that costs less than 50,000 USD per QALY will recommended for funding [11]. In Mexico, when deciding whether or not to include a certain technology in the public healthcare system, the GDP per capita is defined to be the cost-effectiveness threshold [12]. This threshold is also being used in Chile [13] and Colombia [14] to define a technology as cost-effective. Following that, we decided to use the GDP per capita threshold to determine whether trastuzumab is cost-effective or not in sub-Saharan Africa. Furthermore, with DALY being typically an intermediate between LY and QALY, since in our paper the LY and QALY obtained are not far apart (and thus, replacing either by DALY will not change much), the extra page-space and demand from the reader seem unwarranted. An additional objective of this paper was to estimate a value-based price (VBP) for each country based on each country’s GDP per capita.

We developed a Markov model with monthly cycles and life time horizon using the Rstudio platform. The model was used to estimate the costs and health outcomes (LYs, QALYs) associated with two treatment strategies (chemotherapy with and without trastuzumab) for treating early stage HER2 positive breast cancer. The chemotherapy regimen used in this model is anthracycline based chemotherapy as depicted in the HERA trial [15, 16]. However, there are multiple different approaches regarding timing and precise therapy such as the use of taxane based therapy, depending on the patient and stage of disease. However, as the precise treatment patterns in sub-Saharan Africa are unknown, it would only be speculation to suggest that one chemotherapy is used in preference of another. This limits the applicability of this model.

In the base case analysis, trastuzumab yielded a gain ranging from 0.92 LYs in Nigeria to 1.07 LYs in South Africa, and 0.9 QALYs in Nigeria to 1.02 QALYs in South Africa. The ICER ranged from 19,534 USD/QALY in South Africa to 21,697 USD/QALY in Nigeria. Results are presented in Table 4 and in Additional file 3: Figure S2.

Uncertain values of the model were varied in a one-way probabilistic sensitivity analysis. Results for all countries are presented in Fig. 2. The diagram presents ICER results from the 0.25 percentile to the 0.75 percentile. The probabilistic sensitivity analysis results were consistent through all 11 countries. The most sensitive variables were the discount rate, trastuzumab cost, and the hazard ratio. Furthermore, as can be seen in the figure, trastuzumab is not considered cost-effective even with the low ICER values in the sensitivity analysis. Even though the prices used through most countries are identical as well as the drug efficacy, the ICER of each country is different due to the variation in background mortality of each country.

The individual estimates of valued based prices for trastuzumab are presented in Figs. 3, 4 and in Table 5. Since it was concluded that trastuzumab is not cost-effective in the examined countries, we calculated the VBP in such a way that would make the treatment considered as cost-effective in each country, when the WTP threshold is one GDP per capita. In order for the treatment to be considered as cost-effective in African countries, significant price reductions would be required.

In this paper, the cost-effectiveness of trastuzumab for treating breast cancer was examined. Trastuzumab was compared to conventional chemotherapy treatment. In this analysis, trastuzumab treatment yields a gain ranging from 0.92 LYs in Nigeria to 1.07 LYs in South Africa, and 0.9 QALYs in Nigeria to 1.02 QALYs in South Africa, with an incremental cost ranging from 19,561 USD in Nigeria to 19,997 USD in Congo, and an incremental cost-effectiveness ratio ranging from 19,534 USD/QALY in South Africa to 21,697 USD/QALY in Nigeria per gained QALY.

The ICER results were comparable to previously published results [19, 43‐45]. Prior analyses in the UK, Australia, United States and Italy yielded similar results to those in our analysis. Other papers presented higher ICERs, which may be due to calculated costs and survival rates in Africa being significantly lower [18, 46, 47].

Although the ICER results are quite similar to other studies, their meaning is different when examined in Africa. While trastuzumab can be considered as cost-effective in HICs [5, 21, 42‐44, 48], this is not the case in LMICs. ICER results in HICs were close or below the suggested threshold by WHO, one GDP per capita, as can be seen in Table 6, while the calculated ICERs in Africa were significantly higher than one GDP per capita for each examined country.

This study has several limitations. Since no clinical trials were performed in Africa, there was a lack of information regarding the efficacy of the drug in the population in this area. Therefore, for this analysis it was assumed that the findings from the performed clinical trials are relevant in Africa as well. Additionally, although the median age of patients in the HERA trial was 49, we used a different age (45) in our model under the assumption that the reported hazard ratio and the transition probabilities depend on disease progression rather than the age. Furthermore, as indicated by previous literature [30], 45 years old is more representative of the general population of patients in sub-Saharan Africa. Another limitation concerns the omission of heart failure from the model. Although heart failure is reversible and is not associated with increased mortality it may have a mild effect on quality of life estimates. However, since the incidence is very low [27, 28], we excluded it as it would have only a very minor impact on the results of the model. Moreover, given that the gap between the current results and the cost-effectiveness threshold is very large, it can be assumed that the inclusion of heart failure will not change the results significantly. Moreover, as stated before, the transition probabilities were based on the HERA trial, which was conducted in HICs. These probabilities were used due to lack of trials performed in sub-Saharan Africa. Therefore, it can be assumed that in reality these probabilities would be a bit different due to differences between health care systems in HICs and sub-Saharan Africa. Furthermore, not all the countries in Africa provide trastuzumab due to its high price. There is a significant lack of price transparency globally. Hence, the prices for all examined countries were merely estimated. Lastly, since we concluded trastuzumab is not cost-effective in the current setting (12 months treatment), further research concerning a recently suggested treatment duration (6/9 months treatment/9 weeks) [24, 49] needs to be further investigated and maybe adopted in sub-Saharan Africa.

Africa is facing a major public health challenge from non-communicable diseases. Even though infectious diseases are still a major concern in Africa, the overall disease burden proportion in Africa associated with cancer is rising. It was predicted that by 2030, this region will experience an increase of more than 85% cancer burden [50]. One way to improve the current state in Africa is to make cancer drugs more accessible and affordable. Major improvements in the health system and cancer awareness are also crucial.

In this work we calculate the expected ICER gained by using trastuzumab in sub-Sahara Africa, and compare it to the threshold recommended by the WHO. Although the ICER results are quite similar to other studies, their meaning is different when examined in sub-Sahara Africa. While a trastuzumab can be considered as cost-effective in HICs, this is not the case in LMICs. ICER results in HICs were close or below the suggested threshold by WHO, one GDP per capita, while the calculated ICERs in Africa were significantly higher than one GDP per capita for each examined country.

Despite several limitations of our study, it seems that a significant reduction in trastuzumab’s price (compared to its price in HICs) would be required in order for it to be cost-effective in sub-Sahara Africa.