Cost-Minimization Analysis of the Treatment of Patients With Metastatic Colorectal Cancer in Greece

Clinical Therapeutics/Volume 34, Number 10, 2012

Cost-Minimization Analysis of the Treatment of Patients With Metastatic Colorectal Cancer in Greece Vassilis Fragoulakis, MSc1; Vasilia Papagiannopoulou, MSc2; Georgia Kourlaba, MSc, PhD1; Nikolaos Maniadakis, MSc, PhD1; and Georgios Fountzilas, MSc, PhD3 1

National School of Public Health, Department of Health Services Management, Athens, Greece; 2Amgen Hellas, Athens, Greece; and 3Department of Medical Oncology, “Papageorgiou,” Hospital, Aristotle University of Thessaloniki School of Medicine, Thessaloniki, Greece

ABSTRACT Background: In 2008, colorectal cancer was the fourth most common cause of cancer-related death worldwide. Monotherapy with monoclonal antibodies directed against the epidermal growth factor receptor, such as cetuximab and panitumumab, has recently been introduced in the management of metastatic colorectal cancer (mCRC) patients. Objective: The aim of this study was to conduct a cost-minimization analysis comparing panitumumab with cetuximab in the treatment of patients with epidermal growth factor receptor– expressing mCRC with nonmutated (wild-type) Kirsten rat sarcoma viral oncogene homolog in Greece. The perspective of analysis was that of payers (Social Security Sickness Fund) and the country’s National Health Service (NHS). Methods: The model was designed to contain probabilistic parameters to account for uncertainty and variation in these parameters. All resources consumed in local hospitals in the management of patients in each case were evaluated. Two analyses were performed: 1 evaluating cost per milligram and another evaluating cost per vial. Results: From a payer perspective, the mean 20week total cost per patient for panitumumab and cetuximab was: (1) per-milligram analysis: €16,349 and €18,242, respectively; and (2) per-vial analysis: €18,808 and €19,701. From the NHS perspective, the mean total costs per patient were slightly higher; however, the use of panitumumab was associated with a 17.7% and 12.4% cost reduction in per-milligram and per-vial analysis, respectively. The results of probabilistic models confirmed those of the deterministic analyses. Conclusion: In the Greek NHS and Social Security Sickness Fund setting, panitumumab monotherapy

2132

potentially constitutes a cost-saving option (versus cetuximab monotherapy) in the management of patients with mCRC and no mutation of Kirsten rat sarcoma viral oncogene homolog. (Clin Ther. 2012; 34:2132–2142) © 2012 Elsevier HS Journals, Inc. All rights reserved. Key words: antiepidermal growth factor receptors, colon and rectum patients, cost-minimization analysis, monoclonal antibodies.

INTRODUCTION In 2008, according to figures from the International Agency for Research on Cancer, colorectal cancer (CRC) was the third most commonly diagnosed cancer worldwide (9.7%) and the fourth leading cause of cancer-related mortality, accounting for 610,000 deaths (8.1% among all cancer deaths).1,2 In the same year in Europe, CRC was the most common type of malignancy (13.6% of the total) and the second leading cause of cancer-related deaths (12.4%). In Greece, incidence rates have been estimated at 15.4 and 11.15 per 100,000 men and women, respectively, and the age-standardized mortality rate was estimated, in 2008, at 9.1 and 6.1 per 100,000 men and women.2,3 The choice of treatment depends on the stage of the cancer. Surgery remains the definitive treatment for localized CRC, offering the only chance of cure, whereas chemotherapy and/or radiotherapy may be used as adjuvant interventions depending on the loca-

Accepted for publication September 10, 2012. http://dx.doi.org/10.1016/j.clinthera.2012.09.005 0149-2918/$ - see front matter © 2012 Elsevier HS Journals, Inc. All rights reserved.

Volume 34 Number 10

V. Fragoulakis et al. tion of the patient’s staging tumor and other medical factors. Nonetheless, the disease may be diagnosed at a late stage or patients originally treated with surgery may eventually develop metastatic disease, for which palliative systemic chemotherapy is recommended.4 There is evidence from many studies and systematic reviews that chemotherapy for metastatic CRC (mCRC) can alleviate symptoms and prolong survival.5,6 Conventional chemotherapy used to treat mCRC include intravenous drugs such as oxaliplatin, irinotecan, and 5-fluorouracil (5-FU; usually given with intravenous leucovorin) and the 5-FU prodrug capecitabine. These drugs work by interfering with the ability of the rapidly growing cancer cells to divide or reproduce. In addition to chemotherapy, 3 other agents are available for the treatment of mCRC. To date, 2 types of monoclonal antibodies have been approved by the US Food and Drug Administration and the European Medicines Agency for clinical use in the first- and second-line management of mCRC: the anti–vascular endothelial growth factor (VEGF) receptor antibody bevacizumab and the anti– epidermal growth factor receptor (EGFR) antibodies cetuximab and panitumumab.7 Clinical trials have shown that these regimens improve the overall survival and progression-free survival of patients with mCRC.7–12 These antibodies inhibit specific proteins that are important for the growth and/or survival of colon cancer cells. Because targeted chemotherapy does not directly interfere with rapidly dividing cells, they do not have the usual adverse effects of conventional chemotherapy. Bevacizumab binds to a protein called VEGF. Bevacizumab is generally given in combination with other drugs, such as oxaliplatin or irinotecan. Cetuximab and panitumumab target a different protein, the EGFR, which is found in ⬃80% of CRC. Cetuximab is a chimeric antibody whereas panitumumab is a fully human antibody designed to act against EGFR. Both cetuximab and panitumumab are indicated for use in patients in whom the tumor does not possess the Kirsten rat sarcoma viral oncogene homolog gene mutation; it has been shown that neither agent is effective in patients whose tumor has the Kirsten rat sarcoma viral oncogene homolog mutation. Administration of these 2 antibodies is dependent on each country’s authorities; however, in Greece, cetuximab can be given either as monotherapy or in combination with chemo-

October 2012

therapy, while panitumumab was available as a monotherapy until November 2011.* Several combination chemotherapy regimens may be considered for the initial treatment (first-line) of mCRC. Each of these regimens consists of 2 or 3 drugs, used together in a specific way: oxaliplatin plus 5-FU and leucovorin, irinotecan plus 5-FU and leucovorin, oxaliplatin plus capecitabine, or irinotecan plus capecitabine. In Greece, as elsewhere, adding bevacizumab to these 4 regimens is often an available therapy option.* When, despite initial treatment with chemotherapy, CRC tumor continues to grow or when it begins to progress after an initial response to chemotherapy, patients have to be managed with second- or third-line chemotherapy, as long as the patient is well enough to tolerate additional therapy.7,13 The choice of subsequent treatment typically depends on what was administered originally. Because bevacizumab, which targets VEGF, is often administered during first-line therapy, cetuximab and panitumumab, which target EGFR, are often considered in second- and third-line therapy.7,9,14,15 Several studies have established the role of panitumumab and cetuximab in the management of patients with nonmutated (wild-type) Kirsten rat sarcoma viral oncogene homolog mCRC, after failure of first-line therapies containing fluoropyrimidine, oxaliplatin, and irinotecan regimens.7–9,16,17 The related literature and published studies indicate that the 2 regimens may be comparable in terms of efficacy.8,9 However, differences in the unit drug price of the 2 alternatives, their model of administration, premedication, and monitoring requirements still exist, in which case their resource utilization and consequently their total treatment cost may differ substantially. For instance, panitumumab is more costly compared with cetuximab when their unit drug cost is considered. Conversely, the treatment schedule contains 2 visits per month for panitumumab but 4 visits per month for cetuximab; thus, total costs for these 2 drugs are different. In light of these factors, it may be important to examine which treatment is superior from an economic point of view for hospitals and payers. Therefore, the aim of the current study was to conduct an economic evaluation comparing, from varying perspectives, panitumumab with cetuximab in the *National Formulary, National Organization for Medicines, http://www.eof.gr.

2133

Clinical Therapeutics treatment of patients with EGFR-expressing mCRC with nonmutated (wild-type) Kirsten rat sarcoma viral oncogene homolog in Greece. In this study, because there were no clear data to indicate a clear superiority of either of the 2 products in terms of efficacy, a costminimization analysis was conducted.

METHODS Model Description The current study was based on a cost-minimization model that was developed in consultation with local physicians according to each product’s characteristics and use in clinical practice in Greece. It is well known that a cost-minimization analysis can only be used to compare 2 health technologies that have proven to be fully equivalent in survival, therapeutic effect, tolerability, safety, and compliance.18 In such a case, the focus of the analysis shifts only to their treatment cost to choose the least costly as the preferable therapeutic option. It must be noted, however, that in the context discussed here, the differences (or the equivalence) of safety and efficacy for the 2 treatment regimens, if any, have not yet been fully established by using head-tohead trials.19 For instance, in a Phase III mCRC trial comparing panitumumab monotherapy with best supportive care (BSC), the authors tested the effect of panitumumab on progression-free survival (PFS).8 Results of the analysis showed that the treatment effect (hazard ratio) on PFS in the wild-type group was 0.45 (95% CI, 0.34 – 0.59). Similarly, in another study that analyzed 572 patients who were randomly assigned to receive cetuximab plus BSC or BSC alone, the hazard ratio for progression or death was estimated at 0.40 (95% CI, 0.30 – 0.54).9 From these studies, it is obvious that there were no statistically significant differences and, thus, the treatments could be indirectly considered to be equivalent. In addition, a recent meta-analysis, which included the 2 aforementioned studies among others, confirmed that cetuximab and panitumumab increase the response rate and significantly reduce the risk of progression and death in wild-type populations who have advanced CRC.20 The relative risk for response and PFS was not statistically significant between comparators; the meta-analysis found that panitumumab significantly reduced the risk of death (by 13%) compared with cetuximab. Nonetheless, to avoid selection bias that may result from synthesized data, the manufacturer of panitumumab (Amgen Hellas, Athens, Greece) has been con-

2134

ducting an open-label, randomized trial (ASPECCT trial [A Study of Panitumumab Efficacy and Safety Compared to Cetuximab in Subjects With KRAS WildType Metastatic Colorectal Cancer21]) to directly compare the safety and efficacy of panitumumab versus cetuximab in subjects with wild-type MCC; it has been ongoing since 2009. This study is still recruiting participants, and its estimated completion date is August 2013. Hence, for the purposes of the current economic analysis, it was assumed that the 2 treatment options are identical in terms of effectiveness, survival, safety, and tolerability. This assumption was validated by local clinicians. In addition, an incremental cost-minimization analysis was undertaken,18 and thus, variables that implied an equal economic burden in the 2 arms (such as the cost or the probability of adverse events) were excluded to avoid the additional calculation burden. To collect information concerning the management of the disease and the resource use, a structured questionnaire was completed by 3 expert oncologists. The experts were selected based on their knowledge and experience with issues related to the hospital management of patients experiencing CRC. To avoid delays, a telephone survey was conducted as a valid but cheaper and quicker alternative to a face-to-face survey. The questionnaire determined the frequency of use for several costly components used in the model. To avoid omissions or other nonmedical-related errors, the final review of the model inputs was undertaken by the last author, who is a leading medical expert in the country. Table I22–25 contains information on the following data inputs: patient weight and body surface area; the loading dose and the dose administered in each cycle of therapy; the number of vials administered; the number of total therapy cycles during the 20 weeks of therapy; the utilization of a chemotherapy suite for administration of therapy; and data related to time required for premedication administration and infusion and posttreatment surveillance, which were key input parameters in the model. The treatment duration for the model was set at 20 weeks (5 months) of therapy based on median PFS from 2 studies.8,9 Moreover, the model was designed to estimate the total therapy cost per alternative treatment under 2 different approaches. In the first approach, which is based on milligrams, it is assumed that there is no drug wastage. In this scenario, surplus drug may be used for the next patient, a common practice in many hospitals

Volume 34 Number 10

V. Fragoulakis et al.

Table I. Key input parameters used in the analysis. Parameter

Panitumumab

Weight/body surface area, kg/1.7 m2 72.0 Dose Loading (initial), mg/m2 — Per cycle (maintenance) 6 mg/kg2 4 Administrated in first cycle 432 mg Administrated in subsequent cycles 432 mg No. of administered vials First cycle Subsequent cycles Cost per milligram No. of total cycles (20 weeks) Cost of day case

Cetuximab 72.022 40023 250 mg/m 680 mg 425 mg

Time for premedication

5 mg/kg 5 of 100 mg 3.62 €25 108,24 €280 from NHS perspective €73 from SSSF perspective NA

Time for infusion

1 h24

Time for posttreatment surveillance

0 min

2 23

7 of 100 mg 5 of 100 mg 1.77 €25 2023 €280 for NHS perspective €73 for SSSF perspective 15 min loading dose and each additional dose23 2 h for loading dose23 1 h for each additional dose23 1 h as minimum

NHS ⫽ National Health Service; SSSF ⫽ Social Security Sickness Fund; NA ⫽ not available.

in Greece and in other countries.22 By contrast, in the second approach, which is based on vial usage, drug wastage was taken into account. In this scenario, it is assumed that it may not be: (1) possible to treat many patients on the same day; or (2) permitted to use the drug surplus in a patient for whom it was not designated (particularly in cases of patient-specific reimbursement).

Costing Methods The aforementioned model was customized to reflect the resource utilization and treatment costs in Greece. In particular, data on resources used for treatment administration were collected from 2 large oncology units (Papageorgiou General Hospital in Thessaloniki and Attikon University Hospital in Athens). These resources, as well as their unit costs, are presented in Table II. Data regarding dose and time required for premedication and infusion were obtained by using the Summary of Product Characteristics,23,24 and they were confirmed, together with other resource utilization data, from the data obtained from the retrospective chart review and from expert advice.

October 2012

Patient-level resource utilization data were combined with unit cost data and then aggregated to compute total treatment cost for each patient. In the current model, 2 perspectives were considered: that of the National Health System (NHS) and of the Social Security Sickness Fund (SSSF). There has been an NHS in place in Greece since 1983 that provides free or lowcost health care services to the population. The financing of health care provision in Greece consists of a mix of 2 systems: the Beveridge system of financing through public taxation that relates to the funding of hospitals and the Bismarck system of financing through compulsory social insurance that relates to the reimbursement of outpatient services and products and the reimbursement of part of the hospital services. In our analysis, the NHS perspective takes into account the economic burden for hospitals aiming to maximize the health of the population covered. More specifically for patients with cancer, the health care system offers specific benefits, including free laboratory services, free in-patient hospital care, medical-related appliances, and transportation services in the case of an emergency. At the same time, for outpatient services, there is no copay-

2135

Clinical Therapeutics

Table II. Resources consumed in the management of metastatic colorectal cancer in Greece and unit costs. Unit Cost (€) Resource Outpatient office visit, specialist Hospital admission (in days) Chemotherapy session, day case Laboratory tests Other diagnostic aids Laboratory tests/other diagnostic aids Hemoglobin ESR AST ALT GGT LHD Creatine Albumin Bilirubin Sodium Potassium LHD ALP

NHS/SSSF

Frequency of Use

3.00/ 280.00/73.00 280.00/73.00 36.00 36.00

20 5 20 20 20

4.00 2.00 9.00 9.00 5.00 4.80 4.10 5.90 2.90 5.20 5.20 4.80 5.00

Drugs Erbitux Sol. Inf. 5 mg/mL Btx 1 Vial ⫻ 20 mL Vectibix Cs. Sol. Inf 100 mg/5 mL vial (20 mg/mL) Btx 1 vial ⫻ 5 mL Solu-Medrol Ps. Inj. Sol. 40 mg/vial Btx 1 Act-O-Vial (1 mL) Clarityne Tabl Bt 21 ⫻ 10 mg

20 20 20 20 20 20 20 20 20 20 20 20 20

Cost per Unit (€) 177.10 361.55 1.91 3.54

NHS ⫽ National Health Service; SSSF ⫽ Social Security Sickness Fund; ESR ⫽ erythrocyte sedimentation rate; AST ⫽ aspartate aminotransferase; ALT ⫽ alanine aminotransferase; GGT ⫽ ␥-glutamyl transpeptidase; LHD ⫽ lactate dehydrogenase; ALP ⫽ alkaline phosphatase.

ment that applies to drugs that are used to treat cancer or cancer-related symptoms in Greece. Hence, due to the current system, no out-of-pocket costs were considered in the study model. In addition, indirect costs (ie, productivity loss) were not taken into account because the majority of patients are in an unproductive age. The perspective of the SSSF was selected to explore the economic burden for the health insurance system only in terms of charges (official tariffs), an approach that is useful for policy makers in charge of SSSF that reimburse health care products and services. This separation is due to the fact that reimbursement tariffs of SSSF in

2136

many cases are much lower than true hospital production cost; hence, the necessity for different perspectives in the analysis. The unit tariffs and costs were obtained from the Government Gazette and correspond to €2011 (Table II). The prices of drugs (therapy and premedication) used in the model are set in the last price bulletin issued by the Greek Ministry of Economy, Competitiveness and Shipping, and are common across all public hospitals in Greece.25 The total drug cost reflects the premedication, initiation, and maintenance drug usage, and it is calculated by multiplying the number of units of drugs required with their corresponding unit price.

Volume 34 Number 10

V. Fragoulakis et al.

Table III. Distributions used in probabilistic analysis. Variable

Distribution

%CV

Outpatient office visit, general practitioner Outpatient office visit specialist Emergency department medical care Hospital admission (in days) Laboratory and other diagnostic tests Weight Height Time For premedication administration For infusion For posttreatment surveillance

Log-normal Log-normal Log-normal Log-normal Log-normal Normal Normal

5 5 5 5 5 10 10

Dealing With Uncertainty It must be noted that only in few cases, a simple series of 1-way sensitivity analyses may be sufficient to determine (in an appropriate manner) the cost differences among comparators. Frequently, simple sensitivity analyses cannot provide enough insight into the scale of decision uncertainty.26 In this context, several health technology assessment agencies (eg, National Institute for Health and Clinical Excellence) strongly recommend the use of probabilistic techniques for a robust conclusion.27 In several cases, such as discussed here, in which different inputs were combined and synthesized, probabilistic sensitivity analysis may be the most prominent way to capture the uncertainty of a model. Hence, the model used in the analysis includes many parameters that are subject to variation and are based on some assumptions. In addition, the cost data are not normally distributed and are truncated at zero. Therefore, using conventional statistical techniques is problematic. As a result, the model used for the current analysis was made probabilistic. In this context, probability distributions were assigned to all of its parameters. Cost variables were assumed to follow a log-normal distribution around the mean, and the rest of the parameters were assumed to follow a uniform or a normal distribution using a reasonable 5% to 10% %CV in each experiment (Table III). Nonparametric bootstrapping (using 1000 replications) was then undertaken, and bias-corrected uncertainty intervals of costs were calculated. In this context, given the probabilistic

October 2012

Uniform Uniform Uniform

5 5 5

nature of all parameters, the cost of the comparators can be slightly different in each bootstrap experiment, and the mean of the 1000 estimates is expected to be a good approximation of the true population mean cost per therapy arm.

Budget Impact Apart from cost-minimization analysis, which compares alternative therapeutic agents regarding their treatment costs (given an equivalent efficacy outcome), it is often useful to consider the budget impact of shifting patients from 1 drug to the alternative. This is considered in the context of budget impact analysis, in which different rates of utilization for existing therapies are applied to alternative comparators to estimate the economic implication of various scenarios to the budgets of providers or payers. In the current study, a hypothetical cohort of 1000 patients was used to run such a hypothetical budget impact analysis. The implications of alternative options were estimated for the 2 different scenarios (based on vial or milligram for the treated patients) in a per-patient basis.

RESULTS The total cost for 2 alternative regimens were calculated separately for the SSSF and NHS perspective.

SSSF Perspective Table IV presents the results of the cost-minimization analysis for an SSSF perspective. The per-milligram analysis revealed that the mean 20-week total cost of therapy was €16,349 per patient for panitu-

2137

Clinical Therapeutics

Table IV. Economic evaluation (€) from the perspective of the Social Security Sickness Fund (SSSF) and the National Health Service (NHS). Mean (95% CI) Perspective SSSF Per milligram Panitumumab Cetuximab Difference % Difference Per vial Panitumumab Cetuximab Difference % Difference NHS Per milligram Panitumumab Cetuximab Difference % Difference Per vial Panitumumab Cetuximab Difference % Difference

First Dose

Additional Doses

1562 1303 258 —

1562 815 747 —

15,619 (15,305 to 15,897) 730 (720 to 747) 16,782 (16,450–17,095) 1460 (1432 to 1487) ⫺1,163 ⫺730 ⫺6.9% ⫺50.0%

16,349 (16,036 to 16,638) 18,242 (17,916 to 18,562) ⫺1894 (⫺2323 to 1474) ⫺10.4%

1808 1417 391 —

1,808 886 922 —

18,078 (17,704 to 18,431) 730 (716 to 745) 18,242 (18,437 to 19,162) 1460 (1430 to 1490) ⫺164 ⫺730 ⫺9% ⫺50.0%

18,808 (18,437 to 19,162) 19,701 (19,358 to 20,053) ⫺893 (⫺1404 to 413) ⫺4.5%

1562 1303 259 —

1562 815 747 —

15,619 (14,169 to 17,137) 2800 (2524 to 3054) 16,782 (14,899 to 18,406) 5600 (5082 to 6129) ⫺1163 ⫺2800 ⫺6.9% ⫺50.0%

18,419 (16,907 to 19,937) 22,382 (20,517 to 24,121) ⫺3963 (⫺6293 to 1702) ⫺17.7%

1808 1417 391 —

1808 886 922 —

18,078 (16,452 to 19,712) 2800 (2520 to 3068) 18,241 (16,473 to 20,014) 5600 (5093 to 6133) 163 ⫺2800 0.9% ⫺50.0%

20,878 (19,187 to 22,568) 23,841 (21,998 to 25,717) ⫺2963 (⫺5408 to 374) ⫺12.4%

Medication Cost

mumab and €18,242 per patient for cetuximab. The main component of total treatment total was medication, which accounted for 96% and 92% of total cost for the panitumumab and cetuximab groups, respectively. In the per-vial analysis, the mean total cost of panitumumab was €18,808 per patient, whereas the cost of cetuximab was €19,701 per patient. Probabilistic results confirm the deterministic ones reaching a statistically significant difference in 95% level of significance. Figure 1 depicts the distribution of savings based on 1000 bootstrap replications. In the model, 100% of iterations in both the per-milligram and the per-vial analysis were cost saving for panitumumab in relation to cetuximab.

Administration Cost

Total Cost

the medication cost in both treatment groups. In particular, the medication cost accounted for 87% and 77% of the total cost in the panitumumab and cetuximab groups, respectively. Nonetheless, chemotherapy administration cost was higher for cetuximab compared with panitumumab (€5800 vs €2800, respectively). The per-milligram analysis revealed that the mean total cost of panitumumab was €18,419 per patient, whereas the cost of cetuximab was €22,382 per patient. Figure 2 depicts the distribution of savings based on 1000 bootstrap replications. In the Greek model, ⬎99% of iterations were cost-saving in favor of panitumumab.

Budget Impact NHS Perspective The per-vial analysis showed that the mean 20week total cost of therapy was €20,878 per patient for panitumumab and €23,841 per patient for cetuximab (Table IV). The total cost was mainly driven by

2138

Budget impact analysis revealed that panitumumab was less costly compared with cetuximab, in both perspectives, based on the per-milligram and per-vial analyses. Table V shows in detail the significant cost-saving profile of panitumumab for a treated patient in Greece.

Volume 34 Number 10

V. Fragoulakis et al.

250 Based on milligrams Based on vial

200 150 100 50

0 15 0 30 0 45 0 60 0 75 0 90 0 10 50 12 00 13 50 15 00 16 50 18 00 19 50 21 00 22 50 24 00 25 50 27 00

0 Savings (€)

Figure 1. Distribution of savings from the use of panitumumab compared with cetuximab, per patient, based on 1000 bootstrap replications from the Social Security Sickness Fund perspective.

This finding is particularly important in light of the current financial crisis and the effort to curtail costs in the production and finance of services. In particular, assuming that there is no drug wastage (per-milligram analysis), the use of panitumumab was associated with a 10.4% and 17.7% reduction in total cost per patient for the 20-week treatment in the SSSF and NHS, respectively. The corresponding cost saving from the pervial analysis was found to be 4.5% and 12.4%, respectively. The administration cost was the key driver for

DISCUSSION In this study, a comparison of the total cost between panitumumab and cetuximab was undertaken, using a probabilistic model. The comparison was conducted based on a cost-minimization model from NHS and SSSF perspectives in Greece. According to the results, panitumumab, despite its higher acquisition unit cost, constitutes a cost-saving alternative over its comparator because it is associated with lower total costs both for SSSF and NHS.

140 Based on milligrams Based on vial

120 100 80 60 40 20

7400

7000

6600

6200

5800

5400

5000

4600

4200

3800

3400

3000

2600

2200

1800

1400

1000

600

200

0 Savings (€)

Figure 2. Distribution of savings from the use of panitumumab compared with cetuximab, per patient, based on 1000 bootstrap replications from the National Health Service perspective.

October 2012

2139

Clinical Therapeutics

Table V. Budget impact analysis per treated patient (per month). Total Cost (€) Perspective

Panitumumab

Cetuximab

Difference (€)

Social Security Sickness Fund Per milligram Per vial

3503 4030

3909 4223

406 193

National Health Service Per milligram Per vial

3947 4473

4796 5109

849 636

the cost differences between treatments; for each patient, the savings of panitumumab laid at €750 to €2800 (or ⫺50% on average) for a 20-week time horizon depending on the model assumption. Conclusively, medication acquisition cost accounted for the largest part of the total cost for panitumumab and cetuximab therapies. However, the administration cost was substantially higher for cetuximab compared with panitumumab. This was considered to be mainly due to the different delivery mode for each agent (eg, panitumumab does not require premedication). Thus, comparing acquisition costs of panitumumab and cetuximab would not be representative and pragmatic. Ultimately, drug acquisition costs should be considered in conjunction with ancillary costs (direct or indirect) related to the delivery of treatments to determine optimal cost savings. These study results are partially in agreement with those presented in related literature; thus, further research is needed to establish certain conclusions. For instance, a cost analysis that was undertaken to evaluate the safety of panitumumab compared with cetuximab in third-line mCRC identified total savings per patient of €851 per month.28 Treatment cost savings associated with panitumumab were related to a low rate of severe infusion reaction versus cetuximab, reduced hospital costs associated with a less frequent dosing regimen, and patients’ weight savings (savings in drug costs because of patient weight and associated drug administration amounts) across a normally distributed patient population.29 In another study, conducted in Mexico, the authors compared the direct cost of cetuximab and panitumumab30 in patients diagnosed with mCRC. Similar to the current analysis, a probabilistic model was con-

2140

structed considering 2 scenarios: per vial of drug (assuming wastage of unused medication) and per milligram of drug (assuming no wastage). When the analysis was restricted to the acquisition drug costs, monthly savings of panitumumab compared with cetuximab were estimated to be 19.1% and 11.2% in per-vial and per-milligram scenarios, respectively. However, in a Markov model submitted to the National Institute for Clinical Excellence by the manufacturer of cetuximab (Merck Serono, Geneva, Switzerland), cetuximab seemed to be a cost-effective choice compared with panitumumab (ie, £15,000 per qualityadjusted life-year gained). The model included individuals with EGFR-expressing Kirsten rat sarcoma viral oncogene homolog wild-type mCRC who had received second- or subsequent-line chemotherapy for metastatic disease. The model had a 10-year time horizon and considered the NHS perspective. Nonetheless, the Appraisal Committee of the National Institute for Clinical Excellence agreed that the results of the analysis were produced under specific methodologic assumptions and should be interpreted with caution because there is not yet any head-to-head comparisons between treatment arms.19 Finally, to the best of our knowledge, our study is the first conducted in Greece aiming to assess the economic implications of the 2 therapeutic alternatives, except for a previous work that evaluated only the budget impact of panitumumab on the health care system.31 It is important to consider that the current analysis has certain limitations, which are common in studies using similar methodologies. Firstly, the analysis does not represent experimental research but is instead based on a model populated from data reported in the literature and on various assumptions. As a result, it may suffer from biases. Second, it must be noted that

Volume 34 Number 10

V. Fragoulakis et al. the results apply only to the Greek setting, at the present time and using current drug prices. If any of the underlying parameters change, the results and the conclusions of the analysis may also change. To limit possible sources of bias, standard recommendations for undertaking such studies were followed. A systematic review and assessment of the evidence was performed to ascertain the effectiveness of the treatments, and stochastic analysis was used to draw robust conclusions. This methodology, however, cannot substitute for direct real-life comparisons between these treatments. Indeed, we confined the analysis to the health care system and the SSSF and did not consider the financial impact at the socioeconomic level. Fewer hospital visits associated with panitumumab versus cetuximab may allow indirect costs savings for patients through reduced transportation costs to and from the clinic, and reduced loss of income due to absenteeism from employment for the relatives of patients.

CONCLUSIONS The current analysis indicates that, in Greece, based on the current prices and resource utilization, panitumumab may represent a cost-saving therapy option compared with cetuximab in the management of mCRC. This finding holds true from the perspective of both the NHS and SSSF. Health resource utilization studies would be required to confirm our findings.

ACKNOWLEDGMENT This study was sponsored by Amgen Hellas, Athens, Greece. All authors contributed equally to the literature review, data interpretation, and writing of the manuscript.

CONFLICTS OF INTEREST The authors have indicated that they have no conflicts of interest regarding the content of this article.

REFERENCES 1. American Cancer Society. Global Cancer Facts & Figures. 2nd ed. Atlanta, Ga: American Cancer Society; 2011. 2. Ferlay J, Shin HR, Bray F, et al. GLOBOCAN 2008. Cancer incidence and mortality worldwide. 2008; http://globocan. iarc.fr/. Accessed March 15, 2011. 3. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010;46:765–781.

October 2012

4. Benson AB 3rd. Epidemiology, disease progression, and economic burden of colorectal cancer. J Manag Care Pharm. 2007;13(Suppl C):S5–S18. 5. Lynch HT, Riley BD, Weissman SM, et al. Hereditary nonpolyposis colorectal carcinoma (HNPCC) and HNPCClike families: problems in diagnosis, surveillance, and management. Cancer. 2004;100:53– 64. 6. Cercek A, Saltz LB. First-line treatment of patients with metastatic colorectal cancer: an overview of recent data on chemotherapy plus targeted agents. Clin Colorectal Cancer. 2008;7(Suppl 2):S47–S51. 7. Tol J, Punt CJ. Monoclonal antibodies in the treatment of metastatic colorectal cancer: a review. Clin Ther. 2010;32: 437– 453. 8. Amado RG, Wolf M, Peeters M, et al. Wild-type KRAS is required for panitumumab efficacy in patients with metastatic colorectal cancer. J Clin Oncol. 2008;26:1626 –1634. 9. Karapetis CS, Khambata-Ford S, Jonker DJ, et al. K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med. 2008;359:1757–1765. 10. McCormack PL, Keam SJ. Bevacizumab: a review of its use in metastatic colorectal cancer. Drugs. 2008;68:487–506. 11. Tappenden P, Jones R, Paisley S, Carroll C. Systematic review and economic evaluation of bevacizumab and cetuximab for the treatment of metastatic colorectal cancer. Health Technol Assess. 2007;11:1–128, iii–iv. 12. Welch S, Spithoff K, Rumble RB, Maroun J. Bevacizumab combined with chemotherapy for patients with advanced colorectal cancer: a systematic review. Ann Oncol. 2010;21: 1152–1162. 13. Kendall A, Lord R, Maisey N. Anti-epidermal growth factor receptor antibodies in the treatment of metastatic colorectal cancer. Recent Pat Anticancer Drug Discov. 2010;5:142–151. 14. Spano JP, Fagard R, Soria JC, et al. Epidermal growth factor receptor signaling in colorectal cancer: preclinical data and therapeutic perspectives. Ann Oncol. 2005;16:189 –194. 15. Peeters M, Balfour J, Arnold D. Review article: panitumumab—a fully human anti-EGFR monoclonal antibody for treatment of metastatic colorectal cancer. Aliment Pharmacol Ther. 2008;28:269 –281. 16. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecanrefractory metastatic colorectal cancer. N Engl J Med. 2004;351:337–345. 17. Van Cutsem E, Peeters M, Siena S, et al. Open-label phase III trial of panitumumab plus best supportive care compared with best supportive care alone in patients with chemotherapy-refractory metastatic colorectal cancer. J Clin Oncol. 2007;25:1658 –1664. 18. World Health Organization. Introduction to drug utilization research. http://apps.who.int/medicinedocs/en/d/ Js4876e/5.2.html. Accessed March 15, 2011.

2141

Clinical Therapeutics 19. National Institute for Health and Clinical Excelence. Cetuximab, bevacizumab and panitumumab for the treatment of metastatic colorectal cancer after first-line chemotherapy, 2011. http://www.nice.org.uk/ nicemedia/live/13651/57924/ 57924.pdf. Accessed March 15, 2011. 20. Petrelli F, Borgonovo K, Cabiddu M, et al. Cetuximab and panitumumab in KRAS wild-type colorectal cancer: a meta-analysis. Int J Colorectal Dis. 2011;26:823– 833. 21. Clinical Trials. Available at http:// clinicaltrials.gov/ct2/show/NCT 01001377. Accessed March 15, 2011. 22. Maniadakis N, Pallis A, Fragoulakis V, et al. Economic analysis of a multicentre, randomised, phase III trial comparing FOLFOXIRI with FOLFIRI in patients with metastatic colorectal cancer in Greece. Curr Med Res Opin. 2007;23:2251–2257. 23. European Medicines Agency. Vectibix: summary of product characteristics. http://www.ema.europa.eu/docs/ en_GB/document_library/EPAR_-_ Summary_for_the_public/human/ 000741/WC500047704.pdf. Accessed March 2011. 24. European Medicines Agency. Erbitux: summary product characteristics. http://www.ema.europa.eu/ docs/en_GB/document_library/ EPAR_-_Product_Information/ human/000558/WC500029119.pdf. Accessed March 2011. 25. Greek Ministry of Economy. Competitiveness and shipping drug price bulletin. http://www.gge.gr/37/ index.asp. Accessed January 11, 2011. 26. Meltzer D. Addressing uncertainty in medical cost-effectiveness analysis implications of expected utility maximization for methods to perform sensitivity analysis and the use of cost-effectiveness analysis to set priorities for medical research. J Health Econ. 2001;20:109 –129.

2142

27. Claxton K, Sculpher M, McCabe C, et al. Probabilistic sensitivity analysis for NICE technology assessment: not an optional extra. Health Econ. 2005;14:339 –347. 28. Bracco A, Farrimond BJ, Fitzgibbon JW, et al. A model to demonstrate the comparative costs between panitumumab and cetuximab for thirdline metastatic colorectal cancer patients in Italy. ISPOR Annual European Congress. Athens, Greece. 2008. 29. Berardi R, Onofri A, Pistelli M, et al. Panitumumab: the evidence for its use in the treatment of metastatic colorectal cancer. Core Evid. 2010; 5:61–76.

30. Arocho R, Hurtado R. Economic evaluation of panitumumab vs cetuximab In patients with colorectal carcinoma (Crcm) with non-mutated (wild-type) Kras after failure of chemotherapy. ISPOR 3rd Latin Conference. Mexico City, Mexico 2011. 31. Papagiannopoulou V, Christodoulopoulou A, Yfantopoulos I. A budget impact model for the introduction of panitumumab: a new therapy for the management of metastatic colorectal cancer (Mcrc) In Greece. Poster presented at: ISPOR Annual European Congress. Athens, Greece 2008.

Address correspondence to: Fragoulakis Vassilis, MSc, Department of Health Services and Management, National School of Public Health, Alexandras Av. 196, Athens 11521 Greece. E-mail: [email protected]

Volume 34 Number 10