Abstract
Objective
The aim of the study was to demonstrate the clinical and economic impact of two PD-L1 IHC assays, SP142 versus 22C3, to identify the eligibility of the patients with advanced triple negative breast cancer (aTNBC) to the treatment with atezolizumab plus nab-paclitaxel in the Brazilian private healthcare system (BPHS).
Methods
The study performed a cost-effectiveness analysis based on a partitioned-survival model with three mutually exclusive health states: progression-free (PF), progression, and death. Data of progression-free survival and overall survival were extracted from a retrospective exploratory analysis of IMpassion130, an analytical harmonization of PD-L1 IHC assays. The analyses included only direct costs (drug acquisition and management of adverse events) that were based on CBHPM (Classificação Brasileira Hierarquizada de Procedimentos Médicos) and CMED PF18% (Câmara de Regulação do Mercado de Medicamentos) tables. A probabilistic sensitivity analysis was performed as a second-order Monte Carlo Simulation in order to evaluate the uncertainties of the model.
Results
The SP142 assay has the potential to improve PFS and generate savings to the BPHS. The incremental cost-effectiveness ratio (ICER) was –USD 4,119.43 per month of progression-free survival.
Conclusions
The SP142 assay demonstrated to be a dominant alternative compared to 22C3 to guide the treatment with atezolizumab plus nab-paclitaxel in patients with aTNBC.
Introduction
In Brazil, female breast cancer occupies the first most frequent cancer in all regions, without considering non-melanoma skin tumors. According to the National Cancer Institute (INCA), 66,280 new cases of breast cancer are expected in 2020Citation1.
Breast cancer can be classified by three tumor markers: estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor 2–neu (HER2)Citation2. Triple-negative breast cancer (TNBC) is defined by a lack of ER and PR expression and a lack of HER2 overexpressionCitation3,Citation4. Approximately 15% of all breast cancers are diagnosed with TNBCCitation3.
Atezolizumab is a monoclonal antibody that binds to programmed cell death ligand 1 (PD-L1) and enhances T-cell activity against tumor cellsCitation5. A phase III trial, IMpassion130, demonstrated the efficacy and safety of atezolizumab plus nab-paclitaxel compared to placebo plus nab-paclitaxel in patients with untreated advanced TNBC (aTNBC)Citation6.
The results of the study demonstrated that atezolizumab plus nab-paclitaxel improved progression-free survival (PFS) compared to placebo plus nab-paclitaxel (7.5 × 5.0 months; hazard ratio = 0.62; 95% CI = 0.49–0.78; p < 0.001) in PD-L1 positive patientsCitation6.
Based on the results of IMpassion130, atezolizumab plus nab-paclitaxel was FDA approved for patients with aTNBC if the tumor expresses PD-L1 (PD-L1 stained tumor-infiltrating immune cells [IC] of any intensity covering ≥ 1% of the tumor area) by the SP142 assayCitation7,Citation8.
In a post hoc analysis of IMpassion130, the 22C3 and SP263 PD-L1 IHC assays were assessed for analytical concordance with the SP142 assay and their association with clinical activity of atezolizumab plus nab-paclitaxel was also evaluated. The retrospective exploratory analysis showed that the SP142 assay can identify patients with aTNBC that are most likely to benefit from the treatment with atezolizumab plus nab-paclitaxelCitation9.
Research question
Is the SP142 (tumor-infiltrating immune cells 1% [IC ≥ 1%]) assay cost-effective in guiding the treatment with atezolizumab plus nab-paclitaxel in patients with aTNBC when compared to the 22C3 (positive combined score ≥ 10 [CPS ≥ 10]) assay?
Objective
The aim of this study is to estimate the clinical and economic impact of using the SP142 (IC ≥ 1%) and 22C3 (CPS ≥10) assays to identify the eligibility of the patients with aTNBC to the treatment with atezolizumab plus nab-paclitaxel through a cost-effectiveness analysis.
Methods
Patient population
The target population for the model were untreated aTNBC patients, which is consistent with the patient population from the IMpassion130 trialCitation6. Patients with aTNBC are eligible to atezolizumab plus nab-paclitaxel if they were positive for PD-L1 expression through the SP142 assay.
The characteristics of the patients considered in the model:
Average weight: 67.04 kgCitation10
Average height: 155.91 cmCitation10
Body surface area: 1.70 m2
Perspective
The perspective of the study was the Brazilian Private Healthcare system (BPHS). In Brazil, all patients with health insurance have access to outpatient cancer therapies, therefore they have access to atezolizumab and nab-paclitaxelCitation11.
Comparators
The SP142 is the only assay FDA approved as a companion diagnosis for the treatment of atezolizumab plus nab-paclitaxel in patients with aTNBCCitation7. The SP142 (IC ≥ 1%) assay was considered the intervention in our study. The comparator included was the 22C3 (CPS ≥ 10) assay based on the retrospective exploratory analysis of IMpassion130, an analytical harmonization of PD-L1 IHC assays, that evaluated the association of its result with clinical activity of atezolizumab plus nab-paclitaxelCitation12.
The SP263 assay was not included in our study because the harmonization exploratory sub-analysis of Impassion130 did not publish a Kaplan-Meier curve for SP263 resultsCitation12.
Patients with PD-L1 positive results with SP142 or 22C3 assay were treated with atezolizumab plus nab-paclitaxel and the patients with negative results were treated with nab-paclitaxel (). Although in Brazil nab-paclitaxel monotherapy is not considered on label for aTNBCCitation13, the study considered the treatment with nab-paclitaxel because it is currently the only chemotherapy compared to atezolizumab plus nab-paclitaxel. Besides, no network meta-analysis that compares the posology of nab-paclitaxel used in the IMpassion130 trial with other chemotherapy regimens for aTNBC has been published yet.
Figure 1. Strategies for PD-L1 testing.
The prevalence of PD-L1 expression by SP142 (IC ≥ 1%) and 22C3 (CPS ≥ 10) were 46% and 53%, respectivelyCitation12. These results were used to guide the patients to atezolizumab plus nab-paclitaxel in the case of PD-L1 positive results and to monotherapy of nab-paclitaxel in the case of PD-L1 negative results.
Type of economic analysis
The cost-effectiveness analysis was performed in order to evaluate the clinical outcomes and the cost associated with the first line of treatment for patients with aTNBC guided by both assays. The study was limited to the first line of treatment because there were many uncertainties of selecting the 2 L for aTNBC as the patients received different cancer therapies during the IMpassion130 trial follow-upCitation6.
Model structure
A partitioned-survival model with three mutually exclusive health states: progression-free (PF), progression, and death, was used to estimate the costs and clinical outcomes of the treatments guided by both assays. All patients started in the PF state and could either transit to progression or death, or remain in the PFS state after each monthly cycle. Patients who progressed could remain in this state or die per cycle.
Time horizon
As the aim of the study was to evaluate the PFS of PD-L1 assays in the first line of treatment, the time horizon considered in the model was 12 months. As the median PFS with atezolizumab plus nab-paclitaxel reported in the IMpassio130 trial was 7.5 months,Citation6 most of the patients would have progressed in 12 months. Therefore, 12 months was considered an appropriate time horizon since our study did not consider a second line of treatment.
Clinical input
All clinical inputs were obtained from a retrospective exploratory analysis of IMpassion130Citation12. The PFS and OS data were extracted from the Kaplan-Meier curve of the sub-populations defined by SP142 (IC ≥ 1%) and non-standard 22C3 (IC ≥ 1%): SP142 + 22C3+, SP142-22C3+, and SP142-22C3-.
The following parametric functions were used to model the data in order to determine the proportion of patients in each health state: Exponential, Weibull, Log-normal, Gompertz, and Log-logistic. The Akaike's information criteria (AIC) was estimated in order to identify the best predictive function.
In order to compare the SP142 (IC ≥ 1%) assay with the 22C3 (CPS ≥ 10) assay, the atezolizumab plus nab-paclitaxel curves with the 22C3 (CPS ≥ 10) assay had to be projected. It was assumed that the clinical outcomes of nab-paclitaxel are independent of the PD-L1 results with 22C3 (IC ≥ 1%) and 22C3 (CPS ≥ 10). Thus, the atezolizumab plus nab-paclitaxel curve for 22C3 (CPS ≥ 10) were projected by the premise of proportional risks through the application of the Hazard Ratio (HR) between atezolizumab plus nab-paclitaxel versus nab-paclitaxel (SP142 [IC ≥ 1%] and 22C3 [CPS ≥ 10]) to the nab-paclitaxel curve (SP142 [IC ≥ 1%] and 22C3 [IC ≥ 1%]).
The HR of PFS and OS defined by SP142 (IC ≥ 1%) and 22C3 (CPS ≥ 10) are represented in Citation12. The atezolizumab plus nab-paclitaxel curves for PFS SP142 + 22C3+ and SP142 + 22C3- were projected by applying their HR to the nab-paclitaxel curves. Since only these HRs were statistically significant, the atezolizumab plus nab-paclitaxel curve for PFS SP142-22C3+ was considered to be the same as the nab-paclitaxel PFS SP142-22C3+ (IC ≥ 1%) curve.
Table 1. Hazard ratio of atezolizumab plus nab-paclitaxel versus nab-paclitaxel SP142 (IC ≥ 1%) and 22C3 (CPS ≥ 10)Citation12.
Adverse events (AEs)
Only adverse events of Grade 3 or 4 whose incidence rate was higher than 2% were included into the model. The frequencies of occurrence of each event were obtained from the pivotal study IMpassion130Citation6.
Use of resources and costs
In this analysis, only direct medical costs (drug acquisition and management of adverse events) were considered (16/07/2020 1 USD = 5.356 BRL). The diagnostic procedure cost for both PD-L1 assays were not considered because both assays are reimbursed under the same CBHPM codeCitation14. Indirect costs, which is the cost of lost productivity as a result of the morbidity, were not included.
Drug treatment: The unit costs of medicines were obtained from the price list of the Medicines Market Regulation Chamber (CMED) of July 2020, considering the factory price (PF 18%)Citation15. The cost and posology used are described in .
Unit cost of adverse events: The micro-costing of adverse events was obtained through the standard use of resources defined by expert opinion. The micro-costing considered the unit cost of medical procedures and drugs to manage the adverse events based on the price list of CMED and the CBHPM table (Supplementary Table S1)Citation15,Citation16.
Table 2. Unit cost and dosages.
Sensitivity analysis
A probabilistic sensitivity analysis (PSA) was carried out in order to evaluate the uncertainties of the model. The PSA was performed as a second-order Monte Carlo Simulation. The parameters were varied simultaneously in 1,000 iterations in a stochastic way. Key model parameters and their distributions are presented in Supplementary Table S2.
Results
Parametric distribution
The results of AIC for each parametric distribution are shown in Supplementary Table S3 and Supplementary Table S4. Based on these results, we could select the most appropriate parametric distribution to fit the PFS and OS data ().
Table 3. Parametric distribution.
Table 4. Cost-effectiveness analysis (CEA) in 1 year.
SP142 (IC ≥ 1%) versus 22C3 (CPS ≥ 10)
The results of cost-effectiveness analysis of the SP142 (IC ≥ 1%) assay compared to the 22C3 (CPS ≥ 10) assay in 1 year are described in . The SP142 (IC ≥ 1%) assay was demonstrated to generate savings of USD 1,358.25 and also gains of 0.33 months in PFS compared to 22C3 (CPS ≥ 10) in 1 year per patient. Thus, the incremental cost-effectiveness ratio (ICER) was –USD 4,119.43 per month of progression-free survival.
Probabilistic sensitivity analysis
PSA produced a similar ICER to the deterministic base case result (‒USD 5,262.13 vs –USD 4,119.43 per month of PFS in 1 year). The PSA results demonstrated that 68% of the simulations were dominant (greater effectiveness and lower cost), whereas 32% of the simulations had greater effectiveness and higher cost ().
Figure 2. Probabilistic sensitivity analysis results.
Discussion
Cost-effectiveness analyses are useful studies to support the decision-making regarding the use of different technologies for the same pathology. Few studies have evaluated the clinical and economic impact of diagnostic tests on the optimization of high-cost therapies.
This is the first cost-effectiveness study that compares two different assays, SP142 and 22C3, in the diagnosis and treatment of aTNBC.
One of the limitations of the study is the absence of Kaplan-Meier curves for 22C3 (CPS ≥ 10). The survival curves of atezolizumab plus nab-paclitaxel for 22C3 (CPS ≥ 10) were projected based on the Hazard Ratio between atezolizumab plus nab-paclitaxel versus nab-paclitaxel (SP142 [IC ≥ 1%] and 22C3 [CPS ≥ 10]).
Another limitation of the study is the drug treatment for patients who were PD-L1 negative. These patients were treated with nab-paclitaxel monotherapy which does not have a label indication for TNBC in Brazil, but, in order to compare the overall impact of both assays, nab-paclitaxel was considered in our study. Therefore, no network meta-analysis that compared the posology of nab-paclitaxel used in the IMpassion130 trial with other chemotherapy regimens for aTNBC has been published yet.
Despite these limitations, the study confirmed the clinical and economic benefit of using an FDA approved PD-L1 assay, SP142, as a companion diagnostics for the treatment of atezolizumab plus nab-paclitaxel in patients with aTNBCCitation7.
The savings generated with the SP142 (IC ≥ 1%) assay are mainly due to a lower proportion of patients defined as eligible to the treatment with atezolizumab plus nab-paclitaxel, in comparison with the 22C3 (CPS ≥ 10) assay (46% and 53% of the patients, respectively). In the other hand, the clinical outcome improvement observed with the SP142 (IC ≥ 1%) vs 22C3 (CPS ≥ 10) indicates this assay selecting the patients who are more likely to benefit from atezolizumab plus nab-paclitaxel and, therefore, SP142 can avoid the use of immunotherapy in those patients who will not benefit from it.
The PSA confirmed the determinist results, demonstrating that SP142 (IC ≥ 1%) assay is a dominant alternative compared to 22C3 (CPS ≥ 10).
The results of the study demonstrated that the SP142 assay has the potential to identify patients who are most likely to benefit from atezolizumab plus nab-paclitaxel and avoid its waste with those who do not benefit from it. The study demonstrates the value of diagnostic tests in precision cancer medicine.
Conclusion
The present study demonstrated the importance of using a right companion diagnostic test as a tool to optimize the health resources and guarantee that the right patients receive the right treatment. The use of SP142 assay has the potential to improve the PFS and also generate savings compared to 22C3 assay.
Having that said, it is important that HCS (healthcare systems) differentiate companion diagnostic tests from other types of assays (in terms of regulatory approval and reimbursement), as its use has a major impact for the patient outcome and costs for the system.
Transparency
Declaration of funding
This study was conducted internally by Roche Diagnostic Brazil.
Declaration of financial/other relationships
RSH, MMS, and MN are employed at Roche Diagnostics Brazil. JPVC and TN are employed at Roche Diagnostics LATAM. No other potential conflicts of interest were reported by the authors. JME peer reviewers on this manuscript have received an honorarium from JME for their review work, but have no other relevant financial relationships to disclose.
Institution where the work was performed
Roche Diagnostic Brazil.
Supplemental Material - Appendix
Download MS Word (28.2 KB)Acknowledgements
None reported.
References
- Instituto Nacional de Câncer José Alencar Gomes da Silva. Estimativa/2020. Incidência de Câncer no Brasil; 2019. Available from: https://www.inca.gov.br/sites/ufu.sti.inca.local/files/media/document/estimativa-2020-incidencia-de-cancer-no-brasil.pdf
- Howlader N, Altekruse SF, Li CI, et al. US incidence of breast cancer subtypes defined by joint hormone receptor and HER2 status. 2014;106(5):dju055.
- Pal SK, Childs BH, Pegram M. Triple negative breast cancer: unmet medical needs. Breast Cancer Res Treat. 2011;125(3):627–636.
- Prado-Vazquez G, Gámez-Pozo A, Trilla-Fuertes L, et al. A novel approach to triple-negative breast cancer molecular classification reveals a luminal immune-positive subgroup with good prognoses. Sci. Rep. 2019;9:1538.
- Krishnamurthy A, Jimeno A. Atezolizumab: A novel PD-L1 inhibitor in cancer therapy with a focus in bladder and non-small cell lung cancers. Drugs Today. 2017;53(4):217–237.
- Schmid P, Adams S, Rugo HS, et al. Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer. N Engl J Med. 2018;379(22):2108–2121.
- U.S. Food and Drug Administration. List of cleared or approved companion diagnostic devices (in vitro and imaging tools); 2020. https://www.fda.gov/medical-devices/vitro-diagnostics/list-clearedor-approved-companion-diagnostic-devices-vitro-and-imaging-tools
- Genentech. TECENTRIQ (atezolizumab) - FDA. 1–53; 2020. https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/761034s027lbl.pdf
- Rugo HS, Loi S, Adams S, et al. Performance of PD-L1 immunohistochemistry assays in unresectable locally advanced or metastatic triple-negative breast cancer: post hoc analysis of IMpassion130. ESMO Congr. 2019;30(5):v858–v859.
- Instituto Brasileiro de Geografia e Estatística (IBGE). Medidas Antropométricas; 2013. Available from: https://www.ibge.gov.br/estatisticas/sociais/saude/9160-pesquisa-nacional-de-saude.html?=&t=resultados.
- Agência Nacional de Saúde Suplementar. RESOLUÇÃO NORMATIVA - RN No 428, DE 7 DE NOVEMBRO DE 2017; 2017. https://www.ans.gov.br/component/legislacao/?view=legislacao&task=TextoLei&format=raw&id=MzUwMg==
- Rugo H, Loi S, Adams S, et al. Abstract PD1-07: exploratory analytical harmonization of PD-L1 immunohistochemistry assays in advanced triple-negative breast cancer: a retrospective substudy of IMpassion130. Cancer Res. 2020;80:PD1-07.
- Celgene Brasil Produtos Farmacêuticos Ltda. Abraxane. 1–23; 2020. http://www.anvisa.gov.br/datavisa/fila_bula/index.asp
- Agência Nacional de Saúde Suplementar (ANS). Correlação TUSS x ROL 2018; 2018.
- Câmara de Regulação do Mercado de Medicamentos (CMED). Lista de preços de medicamentos - preços fábrica e máximos ao consumidor; 2020. http://portal.anvisa.gov.br/listas-de-precos
- Associação Médica Brasileira (AMB). Classificação Brasileira Hierarquizada de Procedimentos Médico (CBHPM); 2018.
- Produtos Roche Químicos e Farmacêuticos S.A. Tecentriq (bula); 2020. http://www.anvisa.gov.br/datavisa/fila_bula/index.asp