Cost-effectiveness analysis in the Spanish setting of the PEAK trial of panitumumab plus mFOLFOX6 compared with bevacizumab plus mFOLFOX6 for first-line treatment of patients with wild-type RAS metastatic colorectal cancer

Abstract

Objective: To assess the cost-effectiveness of panitumumab in combination with mFOLFOX6 (oxaliplatin, 5-fluorouracil, and leucovorin) vs bevacizumab in combination with mFOLFOX6 as first-line treatment of patients with wild-type RAS metastatic colorectal cancer (mCRC) in Spain.

Methods: A semi-Markov model was developed including the following health states: Progression free; Progressive disease: Treat with best supportive care; Progressive disease: Treat with subsequent active therapy; Attempted resection of metastases; Disease free after metastases resection; Progressive disease: after resection and relapse; and Death. Parametric survival analyses of patient-level progression free survival and overall survival data from the PEAK Phase II clinical trial were used to estimate health state transitions. Additional data from the PEAK trial were considered for the dose and duration of therapy, the use of subsequent therapy, the occurrence of adverse events, and the incidence and probability of time to metastasis resection. Utility weightings were calculated from patient-level data from panitumumab trials evaluating first-, second-, and third-line treatments. The study was performed from the Spanish National Health System (NHS) perspective including only direct costs. A life-time horizon was applied. Probabilistic sensitivity analyses and scenario sensitivity analyses were performed to assess the robustness of the model.

Results: Based on the PEAK trial, which demonstrated greater efficacy of panitumumab vs bevacizumab, both in combination with mFOLFOX6 first-line in wild-type RAS mCRC patients, the estimated incremental cost per life-year gained was €16,567 and the estimated incremental cost per quality-adjusted life year gained was €22,794. The sensitivity analyses showed the model was robust to alternative parameters and assumptions.

Limitations: The analysis was based on a simulation model and, therefore, the results should be interpreted cautiously.

Conclusions: Based on the PEAK Phase II clinical trial and taking into account Spanish costs, the results of the analysis showed that first-line treatment of mCRC with panitumumab + mFOLFOX6 could be considered a cost-effective option compared with bevacizumab + mFOLFOX6 for the Spanish NHS.

Keywords:

• Panitumumab
• metastatic colorectal cancer
• wild-type RAS
• cost-effectiveness analysis

Introduction

Worldwide, colorectal cancer (CRC) is the third most common diagnosed cancer in men and the second in women¹. In 2012, 1.4 million cases of CRC were estimated, where Australia/New Zealand, Europe, and North America had the highest regional incidence¹. In Spain, CRC is the cancer with the highest incidence (15% of the total incidence of cancer), presenting 32,240 new cases annually^2,3. In line with the global trend, Spanish men have a higher incidence than women, with an age-standardized rate of 43.9 and 24.2 cases per 100,000 persons, respectively^2,3.

In Spain, CRC causes 14,700 deaths each year and is the second most common cause of cancer deaths, with an age-standardized mortality rate of 17.1 and 8.4 cases per 100,000 persons in men and women, respectively^2,3. The high mortality of CRC is related to the large number of patients with metastatic CRC (mCRC) (25% at diagnosis)⁴. It is calculated that 50% of CRC patients at earlier stages will develop metastases⁴. Furthermore, it is estimated that patients with mCRC have a total unadjusted prevalence of RAS mutation of 55.9%⁵.

In 2015, Corral et al.⁶ published a study carried out in Catalonia (Spain) where hospital costs associated with CRC were assessed. In this study, a cohort of 699 patients diagnosed with CRC were retrospectively analysed, identifying mean costs per patient by stage at diagnosis according to the American Joint Committee on Cancer classification (in situ, I, II, III, IV, or unknown), cost type (hospitalization, outpatient, chemotherapy, etc.), and disease phase (initial, monitoring, or advanced). The mean direct cost per patient with CRC increased from €6,573 in patients with cancer in situ to €36,894 in patients with disease in stage III (€2005)⁶.

The clinical implementation of the most appropriate treatment for mCRC is complex due to the heterogeneity of the disease, since it is caused by the interaction of genetic and environmental factors and because a major part of the genes involved in CRC are perceived in a small proportion of cases⁷. When choosing treatment, the characteristics of the patient, the tumor, and the treatment to be administered must be considered⁸. Molecular biology has allowed targeting the treatment of mCRC and has increased survival in these patients⁷. The determination of the molecular biomarkers predictive of response to treatment enables the choice of the therapy that may be more effective. Thus, less suitable therapies that are associated with undesirable costs and effects can be avoided. In the management of mCRC, anti-epidermal growth factor receptor (EGFR) monoclonal antibodies are only indicated in patients with wild-type RAS^8,9.

Panitumumab and bevacizumab are monoclonal antibodies. Panitumumab acts by binding to the external domain of EGFR and competing with its natural ligands¹⁰ and bevacizumab is an inhibitor of vascular endothelial growth factor¹¹. Both drugs in combination with chemotherapy are indicated as first-line treatment of adult patients with wild-type RAS mCRC in clinical practice guidelines and the recommendations of Spanish (Spanish Society of Medical Oncology) and European [European Society for Medical Oncology (ESMO)] scientific societies and the current literature^4,12–15.

These recommendations have been updated as a result of the evidence observed in recent studies where not only KRAS exon 2 mutations, but also KRAS mutations in exons 3 and 4 and NRAS mutations in exons 2, 3, and 4, are considered negative predictors of treatment^4,9,15. Accordingly, current ESMO guidelines on mCRC indicate that RAS testing should be carried out in all patients at diagnosis and is mandatory before treatment with EGFR monoclonal antibodies. It should include at least KRAS exons 2, 3, and 4 (codons 12, 13, 59, 61, 117, and 146) and NRAS exons 2, 3, and 4 (codons 12, 13, 59, 61, and 117)^4,15.

The PEAK Phase II clinical trial showed that patients with wild-type RAS (KRAS and NRAS in exons 2, 3, and 4) treated with panitumumab plus mFOLFOX6 chemotherapy regimen (panitumumab + mFOLFOX6) for the first-line treatment of mCRC had greater progression-free survival (PFS) compared with patients treated with bevacizumab plus mFOLFOX6 (bevacizumab + mFOLFOX6) [13.0 months vs 9.5 months; hazard ratio (HR) = 0.65; 95% confidence intervals (CI) = 0.44–0.96; p = .029]¹⁶.

Cost-effectiveness analysis (CEA) is one of the main tools in health economic evaluation that allow the analysis of both the cost of a specific medical intervention and the therapeutic or social value a drug provides^17,18. At present, some of the available CEA between these two treatments are now outdated, as they analysed only patients with wild-type KRAS¹⁹. Consequently, they did not consider clinical outcomes in the current population for which panitumumab is indicated, namely patients with wild-type RAS¹⁹. Moreover, few published CEA have evaluated wild-type RAS population, and none of them considered the Spanish National Health System (NHS) perspective^20–23.

The aim of this study was to evaluate the efficiency of panitumumab + mFOLFOX6 vs bevacizumab + mFOLFOX6 as first-line treatment of patients with wild-type RAS mCRC in the Spanish setting.

Patients and methods

Population

The model analysed a subset of patients with wild-type RAS from the PEAK clinical trial, the only available trial that evaluated the first-line treatment of mCRC with panitumumab + mFOLFOX6 vs bevacizumab + mFOLFOX6. The subset included patients aged ≥18 years diagnosed with wild-type RAS mCRC who had not received prior chemotherapy, anti-EGFR therapy, or bevacizumab therapy¹⁶.

Comparators

The mCRC treatment comparators were:

• Panitumumab administered every 2 weeks plus mFOLFOX6 chemotherapy regimen.

• Bevacizumab administered every 2 weeks plus mFOLFOX6 chemotherapy regimen.

The mFOLFOX6 chemotherapy regimen for both treatment options was that used in the PEAK trial¹⁶: oxaliplatin administered on day 1 of each cycle as an intravenous infusion of 85 mg/m²; 5-fluorouracil administered on days 1 and 2 of each cycle as an intravenous bolus of 400 mg/m² on day 1, followed by a 2,400 mg/m² intravenous infusion administered over 46 h on day 2; and leucovorin 400 mg/m² administered on day 1 of each cycle. It was considered that patients were treated with this chemotherapy regimen every two weeks after receiving treatment with panitumumab or bevacizumab.

Type of analysis

The efficiency of panitumumab + mFOLFOX6 vs bevacizumab + mFOLFOX6 as first-line treatment was evaluated using a CEA of the data from the primary analysis of the PEAK trial¹⁶. A semi-Markov model programmed in Microsoft Excel and adapted to the Spanish context was used to compare the effects on health and the costs associated with the two therapeutic options. This model has been used previously in other published economic evaluations between these two treatments in the settings of the Czech Republic, Greece, France, and Mexico^20–23. Semi-Markov models allow us to consider time dependency for state transitions; whereas in standard Markov models the state transitions remain constant over time, since they are based on the premise that state transitions are independent of any knowledge from the past (such as time since entry into a certain health state)^24–29. The incremental cost-effectiveness ratio (ICER) was calculated using the following equation: where Cost_Pmab and Cost_Bmab represent the costs associated with treatment with panitumumab and bevacizumab in combination with mFOLFOX6, respectively, while Effectiveness_Pmab and Effectiveness_Bmab represent the clinical consequences in terms of quality-adjusted life years (QALY) and life-years gained (LYG). The indicators of cost effectiveness were the incremental cost per QALY gained with panitumumab + mFOLFOX6 vs bevacizumab + mFOLFOX6, and per LYG.

The perspective of the Spanish NHS was used, including only direct healthcare costs (indirect costs were not considered). The time horizon was the lifetime of patients with mCRC (assuming a maximum of 20 years of life after the start of treatment). Costs were expressed in 2015 euros and an annual discount rate of 3% for both costs and health outcomes was applied, in line with Spanish health economic evaluation recommendations¹⁷.

Model description

The model simulated a cohort of patients from the initiation of first-line treatment of mCRC until death. The model consists of mutually-exclusive health states, with patients transitioning from one state to another, simulating the course of the disease: Progression free; Progressive disease: treat with best supportive care (BSC); Progressive disease: treat with subsequent active therapy; Attempted resection of metastases; Disease free after metastases resection; Progressive disease: after resection and relapse; and Death (Figure 1). Patients entered the model in the progression free state and were randomly assigned to first-line treatment with panitumumab + mFOLFOX6 or bevacizumab + mFOLFOX6. When the disease progressed after first-line treatment, patients were treated with an active second-line treatment or BSC. It was assumed that as active second-line treatment patients would receive the monoclonal antibody not received in first-line, combined with FOLFIRI chemotherapy regimen (irinotecan, 5-fluorouracil, and leucovorin). The model also included resection of tumor metastases in eligible patients.

Figure 1. Semi-Markov model structure. BSC: best supportive care; mCRC: metastatic colorectal cancer. 1a: Progression Free to Death; 1b: Progression Free to Progressive Disease: Treat with Subsequent Active Therapy; 1c: Progression Free to Progressive Disease: Treat with Best Supportive Care; 1d: Progression Free to Attempted Resection of Metastases; 2a: Progressive Disease: Treat with Subsequent Active Therapy to Death; 2b: Progressive Disease: Treat with Subsequent Active Therapy to Progressive Disease: Treat with Best Supportive Care; 3: Progressive Disease: Treat with Best Supportive Care to Death; 4a: Attempted Resection of Metastases to Progression Free; 4b: Attempted Resection of Metastases to Death; 4c: Attempted Resection of Metastases to Disease Free After Metastases Resection; 5a: Disease Free After Metastases Resection to Death; 5b: Disease Free After Metastases Resection to Progressive Disease: After Resection and Relapse; 6: Progressive Disease: After Resection and Relapse to Death.

The duration of the Markov cycles (time in which patients remain in a specific health state) in the model was 2 weeks, which corresponds with the frequency of administration of the monoclonal antibody in wild-type RAS patients in the PEAK trial. At the end of each cycle patients transitioned to another health state or remained in the same state, depending on the transition probabilities described below.

Transition probabilities

Transition probabilities for the states of disease progression and death for the two study options were estimated according to the parametric extrapolation of data from the PEAK trial¹⁶ using a Weibull distribution. This distribution was selected as the best-fit to the Kaplan-Meier curves, based upon both the Akaike information criterion and the graphic representation (Figure 2). The parametric survival modeling was coded in SAS (version. 9.3; Cary, NC) using the LIFEREG procedure.

Figure 2. Survival curves. (a) Progression-free survival; (b) Overall survival. Bmab: bevacizumab; Pmab: panitumumab.

The probabilities of the PEAK trial related to resection were also incorporated; attempted resection (13.64% of patients with panitumumab and 10.98% of patients with bevacizumab) and successful resection (66.7% for panitumumab and 77.8% for bevacizumab)²⁰. In patients with successful resection, probabilities for PFS and overall survival (OS) were extrapolated from the data in the study by Adam et al.³⁰ using the log-logistic distribution that best adjusted to the parametric curves.

When disease progressed after first-line treatment, patients received active second-line treatment (61.4% in the panitumumab group, 69.5% in the bevacizumab group) or BSC (38.6% for panitumumab, 30.5% for bevacizumab)³¹. Based on the PEAK trial, this transition was calculated taking into account the likely progression in each cycle according to the percentage of patients receiving each type of treatment. In the context of this analysis, it was considered that the second-line therapies received did not affect OS, and only had an impact on costs and the quality-of-life. The duration of active second-line treatment was determined from the PFS shown in the literature^32,33. Upon disease progression, the model assumed that patients received BSC until death.

Costs

The direct costs evaluated were: drug costs, cost of drug administration, and other medical costs (medical visits, diagnostic tests, adverse events, resection, and end of life). The costs and resource use included in the model were validated by Spanish clinical experts in mCRC in order to be consistent with Spanish clinical practice. Thus, inputs provided included active subsequent lines, BSC, resection of metastasis, medical visits, management of adverse events, end of life costs and utilities.

Drug cost and cost of administration

Drug costs for panitumumab and bevacizumab were expressed as ex-factory price (wholesale price) including mandatory discounts according to Royal Decree Law 8/2010 (an urgent measure enacted by the Spanish Government to control public spending)³⁴ (Table 1). Both the total consumption of vials (mean number of vials used and treatment cycles per patient) and the doses used were from the PEAK study. After model adaptation to routine clinical practice in Spain, an analysis per milligram was carried as base case^35,36.

Table 1. Drug acquisition and administration costs.

	Cost per cycle	Source
Drug acquisition costs
Panitumumab	€1,554.37	[16,34]
Bevacizumab	€1,126.78	[16,34]
mFOLFOX6	€242.66	[34]
FOLFIRI	€118.62	[34]
Best supportive care	€120.56	[34]
Chemotherapy administration costs
Day hospital cost	€267.65	[37]

Drug costs associated with chemotherapy were established according to the use observed in the PEAK trial of mFOLFOX6 as frontline therapy and for subsequent lines with FOLFIRI other studies were used^32,33. The cost of chemotherapy administration was estimated according to Spanish day hospital costs³⁷.

Drug costs associated with BSC were defined considering Spanish clinical practice and were validated by the clinical experts consulted. BSC drug costs included were the cost of painkillers (morphine, metamizole, fentanyl, tramadol, and midazolam), anti-cachexia agents (megestrol), anti-emetics (ondansetron), steroids (dexamethasone and prednisone), non-steroidal anti-inflammatory drugs (ibuprofen and naproxen), and low molecular weight heparin (tinzaparin).

Other medical costs

Other medical costs included in the model were: cost of RAS testing, cost of outpatient visits, cost of diagnostic tests, cost of serious adverse events management, and cost of resection (Table 2).

Table 2. Incidence/costs of serious adverse events and other medical costs.

Serious adverse events incidence and cost	Panitumumab + mFOLFOX6	Bevacizumab + mFOLFOX6	% Requiring hospitalization	Treatment/management costs		Source
				Hospital setting	Outpatient setting
Skin disorders	34.00%	1.00%	10.00%	€3,860.41	€103.40	[16,37]; Considering the incidence of serious adverse events occurring in >2% of wild-type RAS subjects in either treatment arm
Hypomagnesemia	8.00%	0.00%	0.00%	€3,919.87	€101.33
Decreased appetite	6.00%	0.00%	0.00%	€96.68	€101.33
Fatigue	12.00%	10.00%	0.00%	€570.11	€101.33
Mucosal inflammation	7.00%	3.00%	0.00%	€2,835.57	€101.33
Stomatitis	7.00%	0.00%	0.00%	€1,382.81	€101.33
Hypokalemia	8.00%	8.00%	0.00%	€4,012.55	€101.33
Weight loss	2.00%	1.00%	0.00%	€96.68	€101.33
Paronychia	2.00%	0.00%	0.00%	€3,799.71	€101.33
Dehydration	6.00%	1.00%	0.00%	€4,367.51	€101.33
Dysesthesia	0.00%	3.00%	0.00%	€4,623.12	€101.33
Hypertension	0.00%	8.00%	0.00%	€4,845.66	€101.33
Other medical costs	Cost
RAS test	€225.00					Amgen data on file 2015
Wild-type RAS frequency	46%					[38]
General practitioner office visit	€27.49					[37]
Oncology specialist office visit	€101.33					[37]
Computed tomography scan	€243.13					[37]
Resection surgery and hospitalization cost	€14,353.76					[37]
Disease relapse following resection cost	€1,726.84					[34,37]
End-of-life costs	€3,726.80					[37]

Utilities

The impact of the disease development on the quality-of-life was assessed using utilities, through the different health states in the model. Utilities were expressed on a numerical scale from 0 (death) to 1 (perfect health) and based on the EuroQol-5 Dimension (EQ-5D) questionnaire. EQ-5D responses from patients with wild-type RAS from the PRIME study^20,38 were used for the progression-free state (since PEAK trial did not collect this data) and calculated according to Dolan’s³⁹ algorithm (Table 3). Given the lack of specific data for patients with wild-type RAS, for subsequent lines of treatment the utilities observed in trials of panitumumab as second- and third-line treatment in patients with KRAS were used^20,32,40. Like other previous studies²⁰, it was assumed that the utility of patients with wild-type RAS was the same to that reported by patients with KRAS. Thus, for active treatment, the utility observed in trials of panitumumab as second-line line treatment was used³², for BSC, the utility observed in trials of third-line treatment⁴⁰, and the mean of these two utilities was used for the utility associated with recurrence of the disease. The utilities used in the model were validated by local clinical experts in mCRC.

Table 3. Utility weight values.

Health state	Utility weight	Source
Progression-free	0.821	[20,38]
Disease progression
Subsequent active treatment	0.782	[20,32]
Best supportive care	0.681	[20,40]
Metastases resection
Disease free	0.821	Assumption [20,38]
Disease recurrence*	0.731	[20,32,40]

* Assumed to be the mean utility of subsequent active treatment and best supportive care in the disease progression health state.

Sensitivity analysis

Scenario analyses were made after variation of the following parameters: adjustment of the drug cost per vial (wastage of vial remainder) and tolerance (using a new vial only if the calculated dose exceeded 30 mg); the assumption that all patients received BSC after progression of the disease in subsequent lines of therapy; no consideration of the drug costs, BSC, and end of life costs after disease progression in later lines of treatment; reduction of the cost of bevacizumab by 7.5%, 10%, 12.5%, 15%, 17.5%, and 20%.

Additionally, a probabilistic sensitivity analysis (PSA) was run analyzing 10,000 iterations through a Monte Carlo simulation where all parameters evaluated in the model were modified simultaneously. A normal multivariate distribution was assigned for the PFS and OS curves, a normal distribution for the number of vials consumed and the mean number of treatments observed, a Dirichlet distribution for distribution of subsequent therapy, a beta distribution for the probabilities of toxicity and utilities, and a gamma distribution for costs and hospital days.

Results

The results of the CEA showed that first-line treatment of RAS wild type mCRC with panitumumab + mFOLFOX6, based on efficacy from the PEAK trial, was more effective than bevacizumab + mFOLFOX6, with a QALY gain of 2.753 and 2.107, respectively (Table 4). Treatment with panitumumab had a higher overall cost than with bevacizumab: €72,203 and €57,485, respectively. The ICER between treatments was €22,794/QALY gained.

Table 4. Base case results of the cost-effectiveness analysis.

Outcome	Panitumumab + mFOLFOX6	Bevacizumab + mFOLFOX6	Difference
Life years (a)	3.685	2.796	0.888
QALY (b)	2.753	2.107	0.646
Cost category
RAS test	€225	€225	€0
Biologic drug*	€36,652	€27,729	€8,922
Drug administration and chemotherapy	€12,129	€10,081	€2,048
Best supportive care	€5,520	€3,645	€1,874
Disease monitoring	€5,792	€4,788	€1,004
Toxicity treatment and management	€221	€39	€182
Resection related	€8,357	€7,570	€787
End-of-life costs	€3,306	€3,406	−€100
Total costs (c)	€72,203	€57,485	€14,718
Incremental cost per LYG (a/c)	€16,567
Incremental cost per QALY gained (b/c)	€22,794

* Estimated number of first-line treatment cycles (from PEAK trial): 19.82 for panitumumab and 14.10 for bevacizumab; Estimated number of second-line treatment cycles: 17.02 for panitumumab³² and 14.43 for bevacizumab³³.

LYG: life-year gained; QALY: quality-adjusted life year.

Panitumumab + mFOLFOX6 resulted in more LYG vs bevacizumab + mFOLFOX6: 3.685 and 2.796, respectively, resulting in an incremental cost per LYG of €16,567.

The majority of the assessed scenarios showed an ICER below €30,000/QALY gained, reaching a maximum of €31,842/QALY gained in the scenario that considered that all patients received BSC in subsequent lines of treatment, therefore demonstrating the robustness of the results of the base case (Table 5).

Table 5. Scenario analyses results.

	Incremental cost	Incremental QALY	Incremental LYG	€/QALY	€/LYG
Base-case	€14,718	0.646	0.888	€22,794	€16,567
Drug consumption
Wastage with tolerance	€15,015	0.646	0.888	€23,254	€16,901
Wastage	€15,545	0.646	0.888	€24,075	€17,498
5% Discount rate for costs and outcomes	€14,145	0.588	0.806	€24,058	€17,550
All patients receive BSC post-progression	€20,603	0.647	0.888	€31,842	€23,191
No drug, BSC, or end-of-life costs post-progression	€19,231	0.646	0.888	€29,784	€21,647
Reduction in bevacizumab drug acquisition costs (ex-factory price discounted according to RDL 8/2010)
Reduction of 7.5%	€15,384	0.646	0.888	€23,826	€17,317
Reduction of 10%	€15,607	0.646	0.888	€24,170	€17,567
Reduction of 12.5%	€15,829	0.646	0.888	€24,514	€17,817
Reduction of 15%	€16,051	0.646	0.888	€24,858	€18,067
Reduction of 17.5%	€16,273	0.646	0.888	€25,202	€18,317
Reduction of 20%	€16,495	0.646	0.888	€25,546	€18,567

BSC: best supportive care; LYG: life-year gained; QALY: quality-adjusted life year.

The PSA results showed that the ICER between treatments was below €30,000/QALY gained in 62% of the generated iterations (Figures 3 and 4).

Figure 3. Probabilistic sensitivity analysis results. QALY: quality-adjusted life years.

Figure 4. Cost-effectiveness acceptability curve. C/E: Cost-effectiveness.

Discussion

In recent years, the high levels of mortality of CRC have been reduced as a result of the incorporation of new treatment alternatives, the individualization of treatment with the use of molecular biomarkers, the introduction of screening programs, and reductions in the prevalence of risk factors^1,8. In order to improve the current management of the disease, it is essential to incorporate in the treatment paradigm the options that proved effectiveness, safety, and efficiency of treatment^1,8,41.

Based on these grounds, this CEA took into account recent information on the use of anti-EGFR therapies in patients with wild-type RAS, comparing first-line treatment with panitumumab vs bevacizumab using data from the PEAK trial, as best available evidence¹⁶. The results showed that panitumumab + mFOLFOX6 provided increased survival and more QALY gained than treatment with bevacizumab + mFOLFOX6 (increase of 0.646). Although panitumumab + mFOLFOX6 had a higher total cost than bevacizumab + mFOLFOX6 (difference of €14,718), this increase in cost was due to a greater PFS in the PEAK trial, hence greater drug costs¹⁶, together with a greater percentage of patients undergoing tumor resection¹⁶, which increased cost of treatment.

When health outcomes and costs were compared, first-line management of mCRC with panitumumab + mFOLFOX6 may be considered a cost-effective option vs bevacizumab + mFOLFOX6 in the Spanish setting, as the ICER of €22,794/QALY gained is below the commonly accepted threshold of efficiency in Spain (≤30,000€/QALY gained)^42,43. Furthermore, the incremental cost per LYG was €16,567. The robustness of these results was confirmed in the PSA, where panitumumab + mFOLFOX6 was cost-effective from the Spanish NHS perspective (ICER ≤€30,000/QALY) in 62% of cases. Besides, the sensitivity analysis of scenarios evaluating the uncertainty of the parameters included in the model (such as variation in the consumption of monoclonal antibodies, which may depend on the patient's tolerance of treatment or the optimization of the vials during administration^35,36) confirmed the robustness of the results. Even in the scenario that considered a 20% reduction of the cost of bevacizumab, the first-line treatment with panitumumab remained as a cost-effective option within the Spanish NHS.

The results obtained are in line with current literature analyzing the efficiency of the first-line management of mCRC with anti-EGFR compared with bevacizumab in patients with wild-type RAS^20–23,44. For example, the study by Graham et al.²⁰, made from the French NHS perspective, showed that treatment with panitumumab + mFOLFOX6 was cost-effective at a willingness-to-pay of between €40,000–€60,000/QALY gained. This was a result of a greater number of QALY gained vs bevacizumab + mFOLFOX6, 2.68 vs 2.05, respectively, together with a higher total cost due largely to increased survival. Similar conclusions highlighting the improved efficiency of panitumumab have been demonstrated in studies from the Czech Republic²¹, Greece²³, and Mexico²², which used the same model as that of the current analysis. All these studies were based on PEAK first-line results and included the same health states, disease evolution rationale, and cost-effectiveness outcome (cost per LYG and cost per QALY)^20,21,23. Furthermore, all the studies were adapted to the characteristics of each setting, including treatments received as second-line, costs, and temporal discounts^20–23. As a result, comparison of the results obtained should be made with caution.

Other publications have also evaluated the economic consequences of panitumumab for the first-line treatment of mCRC in combination with FOLFOX in patients with wild-type RAS vs cetuximab plus FOLFIRI⁴⁵. In a cost minimization analysis carried out in the US setting, panitumumab resulted in savings of $23,125 (11.4%) in comparison with cetuximab, due to the lower acquisition and management costs and a lower cost of adverse events⁴⁵. This is in line with another Spanish study in patients with wild-type RAS, with first-line treatment with panitumumab resulting in savings of €5,801 (18%) compared with treatment with cetuximab, both in combination with FOLFOX4⁴⁶. Although cetuximab was not the comparator in the current analysis, these results show the cost differences between these two EGFR monoclonal antibodies, and suggest that the results of the present analysis cannot be extrapolated to cetuximab.

Other cost-effectiveness studies have examined panitumumab vs bevacizumab in first-line mCRC, but only in wild-type KRAS populations, which are not the current populations for which panitumumab is indicated^19,47,48. As explained previously, RAS testing is mandatory before treatment initiation with EGFR targeted monoclonal antibodies, since only wild-type RAS patients can be treated with panitumumab^9,10. Therefore, it is critically important to have economic evaluations in this population. Thus, in light of the latest available evidence, these analyses in wild-type KRAS populations would be less relevant, given that the determination of RAS is required for panitumumab administration in its current indication^9,10. In addition, these studies are not comparable with the current analysis, as the efficacy evaluated did not come from directly-comparable clinical trials, and the studies also included different health states and time horizons. Similarly, they cannot be extrapolated to the Spanish context given the large variability in drug costs between countries⁴⁹ and the differences in the cost categories included and the therapeutic algorithms applied.

Currently, the indication of anti-EGFR is subject to the state of the RAS gene. This is of such importance that some major international drug agencies that include economic evaluation in their assessments are updating their clinical practice guidelines on this basis^50,51. Consequently, the present study represents an advance in the evaluation of the efficiency associated with the first-line management of mCRC with panitumumab, given the current need for studies in this area^52,53.

As noted before, an increase in long-term survival in patients with mCRC has been observed in recent evidence^1,7,8,54. In consequence, the current analysis considered a lifetime horizon of 20 years. This time horizon was considered as appropriate to collect the main clinical and economic outcomes related to the disease evolution, due to the fact that Markov modeling simulation showed that 99% of the patients in each arm would be deceased by this time.

The study has some limitations. First, the inherent limitations of a simulation model, which means the results should be interpreted with caution. Nonetheless, this model has been used in other mCRC studies²⁰, and, furthermore, the use of simulation models is one of the only alternatives available to represent the natural history of a disease and assess the long-term impact of interventions, and they have also been widely used to study the efficiency of therapies available for the management of mCRC^19,47,48,55. To overcome this limitation, the model was adapted to Spanish clinical practice based on the available literature and validated with clinical experts in the treatment of mCRC in Spain. Due to the Spanish cost perspective and considering the large variability in drug costs between countries, the results of this CEA cannot be directly extrapolated to other settings, such as the US⁴⁹.

Another limitation might be the use of a phase 2 clinical trial (PEAK). However, this was the best available evidence and the only available direct comparison between the study treatments for the studied indication. CEAs in other oncological diseases⁵⁶ have also been based on phase II trials, although the majority are not. Therefore, further studies including phase III results are desirable. Nevertheless, some recent indirect comparisons have shown results in line with those of the PEAK trial, where panitumumab had greater efficacy^14,57. Khattak et al.¹⁴ found that patients with wild-type RAS mCRC who received first-line with EGFR monoclonal antibodies vs VEGF had longer OS with a HR of 0.77 (95% CI = 0.63–0.95; p = .016), whereas the HR in the meta-analysis by Pietrantonio et al.⁵⁷ was 0.80 (95% CI = 0.69–0.92; p = .003).

This CEA did not model effectiveness in OS associated with subsequent therapy lines. Although it is expected that newer therapies would increase the total cost of treatment, it would not modify the conclusions of the present analysis given the first-line focus and the fact that it will affect both arms.

Conclusions

In conclusion, based on the PEAK Phase II clinical trial and taking into account Spanish costs, the results of this analysis showed that first-line treatment of mCRC with panitumumab + mFOLFOX6 could be considered a cost-effective treatment option vs bevacizumab + mFOLFOX6 for the Spanish NHS.

Transparency

Declaration of funding

The study was financed by AMGEN, S.A. AMGEN, S.A. carried out the analysis with the help of Oblikue Consulting, which received financing from AMGEN, S.A.

Declaration of financial/other relationships

FR and MV participated as independent assessors and have received honoraria from AMGEN for participation in local and international advisory boards. SG works for AMGEN, S.A. NLM works for Oblikue Consulting. The other authors report no conflicts of interest. Peer reviewers on this manuscript have received an honorarium from JME for their review work, but have no other relevant financial relationships to disclose.

Acknowledgments

The authors would like to acknowledge Miriam Fernández from AMGEN, S.A. and Ferran Pérez from Oblikue Consulting for their support and insights during all the stages of this project.