Evaluating the Long-Term Cost-Effectiveness of Liraglutide Versus Exenatide BID in Patients With Type 2 Diabetes Who Fail to Improve With Oral Antidiabetic Agents

Clinical Therapeutics/Volume 33, Number 11, 2011

Evaluating the Long-Term Cost-Effectiveness of Liraglutide Versus Exenatide BID in Patients With Type 2 Diabetes Who Fail to Improve With Oral Antidiabetic Agents William J. Valentine, PhD1; Andrew J. Palmer, MBBS2; Morten Lammert, MSc3; Jakob Langer, MSc4; and Michael Brändle, MD, MSc5 1

Ossian Health Economics and Communications GmbH, Basel, Switzerland; 2Menzies Research Institute, University of Tasmania, Hobart, Australia; 3Novo Nordisk Scandinavia AB, Copenhagen, Denmark; 4 Novo Nordisk Region Europe A/S, Zürich, Switzerland; and 5Kantonsspital St. Gallen, St. Gallen, Switzerland ABSTRACT Background: The global clinical and economic burden of type 2 diabetes is substantial. Recently, clinical trials with glucagon-like peptide-1 (GLP-1) receptor agonists (liraglutide and exenatide) have shown a multifactorial clinical profile with the potential to address many of the clinical needs of patients and reduce the burden of disease. Objective: The goal of this study was to evaluate the long-term cost-effectiveness of once-daily liraglutide versus exenatide BID in patients with type 2 diabetes who failed to improve with metformin and/or sulfonylurea, based on the results of a previous clinical trial in 6 European countries (Switzerland, Denmark, Norway, Finland, the Netherlands, and Austria). Methods: A validated computer simulation model of diabetes was used to predict life expectancy, qualityadjusted life years (QALYs), and incidence of diabetesrelated complications in patients receiving liraglutide (1.8 mg once daily) or exenatide (10 ␮g BID). Baseline cohort characteristics and treatment effects were derived from the Liraglutide Effect and Action in Diabetes 6 trial. Country-specific complication costs were taken from published sources. Simulations were run over 40 years from third-party payer perspectives. Future costs and clinical benefits were discounted at country-specific discount rates. Sensitivity analyses were performed. Results: Liraglutide was associated with improvements of 0.12 to 0.17 QALY and a reduced incidence of most diabetes-related complications versus exenatide in all settings. Evaluation of total direct medical costs (treatment plus complication costs) suggest that liraglutide was likely to cost between Euro (€) 1023 and €1866 more than exenatide over patients’

1698

lifetimes, leading to incremental cost-effectiveness ratios per QALY gained versus exenatide of: Switzerland, CHF (Swiss francs) 10,950 (€6902); Denmark, Danish krone [kr] 88,160 (€11,805); Norway, Norwegian krone [kr], 111,916 (€13,546); Finland, €8459; the Netherlands, €8119; and Austria, €8516. Conclusions: Long-term projections indicated that liraglutide was associated with benefits in life expectancy, QALYs, and reduced complication rates versus exenatide. Liraglutide was cost-effective from a health care payer perspective in Switzerland, Denmark, Norway, Finland, the Netherlands, and Austria. Clinicaltrials.gov identifier: NCT 00518882. (Clin Ther. 2011;33:1698 –1712) © 2011 Elsevier HS Journals, Inc. All rights reserved. Key words: costs, cost-effectiveness, exenatide, liraglutide, type 2 diabetes.

INTRODUCTION Type 2 diabetes is one of the largest health care challenges facing many developed and developing countries around the world, with current estimates suggesting that diabetes is responsible for between 5% and 13% of total health care expenditures in most developed countries.1,2 Diabetes is a complex and progressive disease that is associated with significant morbidity and mortality. Due to its nature, diabetes benefits from a multifactorial approach to disease manageAccepted for publication September 20, 2011. Express Track online publication October 21, 2011. doi:10.1016/j.clinthera.2011.09.022 0149-2918/$ - see front matter © 2011 Elsevier HS Journals, Inc. All rights reserved.

Volume 33 Number 11

W.J. Valentine et al. ment, as illustrated by the findings of the Steno-2 study, which compared conventional treatment for multiple risk factors versus intensive multifactorial treatment.3–5 The key challenges in the successful treatment of type 2 diabetes include maintaining tight glycemic control, minimizing the risk of hypoglycemia, and controlling cardiovascular risk factors, including reducing or controlling weight. Many existing oral antidiabetic drugs (OADs) address key aspects of the disease, but not all available treatments represent a true multifaceted solution, when used with lifestyle (diet and exercise) modification, that meets each of the clinical needs of a patient with diabetes.6 Meeting these complex clinical needs has been a powerful driver in the search for new therapeutic options in type 2 diabetes. Attention has focused on glucagon-like peptide 1 (GLP-1), a gut-derived incretin hormone that stimulates insulin and suppresses glucagon secretion, inhibits gastric emptying, reduces appetite and food intake, and improves ␤-cell function.7 Currently available therapeutic approaches include degradation-resistant GLP-1 receptor agonists, such as liraglutide and exenatide, and inhibitors of dipeptidyl peptidase-4 (DPP-4) (incretin enhancers, such as sitagliptin and vildagliptin, that inhibit the protease that rapidly degrades GLP-1). Clinical trials with the GLP-1 receptor agonists (liraglutide and exenatide) found substantial reductions in fasting and postprandial glucose concentrations and glycated hemoglobin (HbA1c) (0.5%–1.6%), with associated weight loss (2–5 kg).8 Data from published studies indicate that GLP-1 receptor agonists may be associated with a more substantial reduction in HbA1c compared with DPP-IV inhibitors (0.5%–1.6% vs 0.5%–1% reduction).6 The greater HbA1c reduction associated with GLP-1 receptor agonists is supported by data from recent randomized controlled trials. For example, in a recent study by Pratley et al,9 the reduction in HbA1c with liraglutide 1.8 mg and 1.2 mg was significantly larger (⫺1.50% and ⫺1.24%, respectively) in comparison with sitagliptin (⫺0.90%). The adverse-effect profile of GLP-1 receptor agonists also differs from that of DPP-4 inhibitors; GLP-1 receptor agonists have been associated with frequent, but transient, nausea and vomiting, whereas DPP-4 inhibitors have been associated with minor adverse effects such as upper respiratory tract infection and headache. Further differentiation between the GLP-1 receptor agonists and DPP-4 inhibitors can be seen in their effects on weight. Weight loss has been

November 2011

shown to be associated with liraglutide and exenatide treatment,10 –15 whereas DPP-4 inhibitors are weight neutral and have been associated only with the prevention of weight gain.16 –19 Evidence that liraglutide and exenatide may offer a multifaceted solution, in combination with lifestyle modification, to the challenge of treating type 2 diabetes has been published in a number of randomized controlled trials.20 The Liraglutide Effect and Action in Diabetes 6 (LEAD 6) trial21 was a head-to-head, openlabel, parallel-group, multinational (15 countries) trial of liraglutide versus exenatide in patients with type 2 diabetes whose disease was inadequately controlled with maximally tolerated doses of metformin, sulfonylurea, or both. Patients were randomly allocated to receive either liraglutide 1.8 mg once a day (n ⫽ 233) or exenatide 10 ␮g BID (n ⫽ 231) in addition to their current therapy over 26 weeks of follow up. The primary end point was change from baseline in HbA1c at week 26. Secondary end points included change from baseline in fasting plasma glucose, 7-point plasma glucose profile, weight change, systolic blood pressure (SBP) change, ␤-cell function, and hypoglycemia rates. Liraglutide was associated with a statistically significant improvement in mean (SE) HbA1c versus exenatide BID at week 26 (⫺1.12% [0.08%] vs ⫺0.79% [0.08%]; P ⬍ 0.0001 for difference). Liraglutide was also associated with statistically significant benefits in terms of the percentage of patients reaching the target HbA1c of ⱕ7% (54% vs 43%; P ⫽ 0.0015). There was a trend toward greater reduction in weight (⫺3.24 vs ⫺2.87 kg) and improved SBP with liraglutide, but these findings did not reach statistical significance. Both agents were well tolerated, but nausea was less persistent (estimated treatment rate ratio, 0.448; P ⬍ 0.0001) and minor hypoglycemia less frequent with liraglutide than with exenatide BID (1.93 vs 2.60 events/patient/year; P ⫽ 0.0131 for difference). Major hypoglycemic events were rare (2 patients taking both exenatide and a sulfonylurea). Given the multifactorial clinical profile of liraglutide and exenatide and their potential to address many of the clinical needs of patients with type 2 diabetes as well as to reduce the burden of the disease, we sought to address the question of which of these agents is more cost-effective. The aim of our analysis was to evaluate the long-term clinical and cost outcomes associated with once-daily liraglutide versus exenatide BID in patients with type 2 diabetes who failed to improve with

1699

Clinical Therapeutics metformin and sulfonylurea, on the basis of results of the LEAD 6 trial,21 across 6 different European countries.

MATERIALS AND METHODS Short-term changes in cardiovascular risk factors from the 26-week LEAD 6 trial21 were used as the basis for long-term projections of clinical and cost outcomes using the CORE Diabetes Model. For all 6 countries, cohort characteristics and treatment effects were based on the LEAD 6 trial, with countryspecific cost data used to estimate economic outcomes. A summary of the model and details of the input data used in the analysis are provided in the following sections.

Model Description The previously published CORE Diabetes Model was used to make long-term estimates of life expectancy, quality-adjusted life expectancy, complication rates, and direct medical costs.22 The architecture, assumptions, features, and capabilities of the CORE Diabetes Model have been previously published.21 The model is a validated, nonproduct-specific diabetes policy analysis tool and is based on a series of interdependent submodels that simulate the complications of diabetes (angina, myocardial infarction, congestive heart failure, stroke, peripheral vascular disease, diabetic retinopathy, macular edema, cataract, hypoglycemia, ketoacidosis, lactic acidosis, nephropathy and endstage renal disease, neuropathy, foot ulcer and amputation, and nonspecific mortality). Each submodel has a semi-Markov structure and uses time, state, time-instate, and diabetes type-dependent probabilities derived from published sources. Monte Carlo simulation using tracker variables overcomes the memory-less properties of the standard Markov model, and allows interconnectivity and interaction between individual complication submodels. Long-term outcomes projected by the model have been validated against real-life data.23 In short, 66 internal and external validation analyses were performed across a range of complications and outcomes simulated by the CORE Diabetes Model. Correlation analysis, designed to evaluate how closely the model outcomes matched real-life data, produced a correlation coefficient (R2) of 0.9224 (perfect fit ⫽ 1). Secondorder validation analyses (model predictions vs observed outcomes reported in studies used to construct

1700

the model) generated an R2 value of 0.9574, and the third-order analyses (model predictions vs observed outcomes reported in studies not used to construct the model) R2 value was 0.9023. These data indicate that the CORE Diabetes Model provides an accurate representation of patient outcomes when compared with real-life clinical and epidemiologic studies of diabetes and its complications. Additional data on validation and comparison with other diabetes models were published in the proceedings of the Fourth Mount Hood Challenge meeting.24

Simulation Cohort For all 6 country settings, baseline cohort characteristics were taken from the 464 patients in the LEAD 6 trial21 receiving either liraglutide or exenatide in addition to metformin and/or sulfonylurea. Baseline characteristics were calculated as a weighted average of both arms (Table I). The trial population was taken from 15 different countries in northern and central Europe and North America, and as such can be considered comparable overall to the population eligible for liraglutide therapy in most northern and central European countries. The proportion of patients taking concomitant medication (where it affects the risk of complications) and being screened regularly for renal disease and retinopathy were included in the modeling analysis. Previously published default settings for the CORE Diabetes Model were used, with country-specific data used for mortality (sourced from World Health Organization life tables), treatment modalities for end-stage renal disease, and management practices in terms of the use of concomitant medications (aspirin, statins, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers).21

Treatment Effects Applied In the Simulations After 26 weeks of follow up in the LEAD 6 trial,21 treatment with liraglutide was associated with statistically significant reductions in HbA1c and minor hypoglycemia versus exenatide BID, as well as favorable trends in terms of weight and blood pressure. For the long-term projections using the CORE Diabetes Model, these differences between treatments were applied in the first year of the modeling simulation in all 6 country settings (Table II), after which patients in both treatment arms followed a natural course of risk factor progression seen in patients with type 2 diabetes.21,25 Simulated patients were assumed to continue

Volume 33 Number 11

W.J. Valentine et al.

Table I. Simulated cohort characteristics on the basis of the Liraglutide Effect and Action in Diabetes 6 trial population.* Characteristic Demographic and risk factors Start age, y Duration of diabetes, y Percentage male HbA1c, % SBP, mm Hg Total cholesterol, mmol/L HDL-cholesterol, mmol/L LDL-cholesterol, mmol/L Triglycerides, mmol/L Body mass index, kg/m2 Ethnic group, % White Black Hispanic Other

Mean

Costs and Perspective 56.7 8.0 51.9 8.15 133.0 4.65 1.19 2.92 2.14 32.9 80.6 5.5 12.5 1.5

Baseline cardiovascular disease/history of, % Myocardial infarction Angina Peripheral vascular disease Stroke Heart failure

2.6 1.9 0.6 0.2 1.5

Baseline renal disease/history of, % Microalbuminuria Gross proteinuria End-stage renal disease

1.1 0.2 0.6

Baseline eye disease/history of, % Background diabetic retinopathy Proliferative diabetic retinopathy Severe vision loss Macular edema Cataract

7.1 0.6 0.4 0.2 6.9

Baseline neuropathy and amputation/ history of, % Neuropathy Amputation

7.8 0.2

HbA1c ⫽ glycated hemoglobin; SBP ⫽ systolic blood pressure; HDL ⫽ high-density lipoprotein; LDL ⫽ lowdensity lipoprotein. *Values are shown as means as used in the modeling simulations for all 6 countries.

November 2011

therapy with liraglutide or exenatide for 5 years, before switching to an insulin regimen to manage their diabetes (the same insulin regimen was used in both treatment arms).

For the base case analysis, costs were estimated from a health care payer perspective and as such captured direct medical costs (defined as the sum of pharmacy costs for each treatment arm and complication costs). Pharmacy costs were calculated for each of the long-term cost-effectiveness evaluations based on medication use and associated costs of needles and selfmonitoring of blood glucose in line with data from the LEAD 6 trial.21 Unit costs were taken from countryspecific published sources to calculate the annual costs of treatment. Complication costs were accounted for in the year in which an event was recorded, as well as in the following years, where appropriate, and were derived from country-specific published sources. All costs were expressed in local currency (CHF, Swiss francs; kr, Danish krone; kr, Norwegian krone; and €, Euros in Finland, the Netherlands, and Austria) in year 2008 values.

Estimation of Quality-Adjusted Life Expectancy As diabetes progresses, patients develop complications influencing their overall health-related quality of life. It was therefore important to evaluate both mortality and morbidity and address the utility levels associated with each of the complications modeled. Comprehensive country-specific health utility data for type 2 diabetes are lacking for most countries. As a proxy, health state utilities were used from populations of type 2 diabetes patients wherever possible, with data taken primarily from the United Kingdom Prospective Diabetes Study.26 Data gaps were filled using utilities from the Australian Institute of Health and Welfare “Burden of Disease and Illness in Australia” report,27 Tengs and Wallace,28 Carrington et al,29 Levy et al,30 and Bagust and Beale.31 Annual utility scores in the modeling analysis were calculated on the basis of health state utilities (to capture quality-of-life decrements due to previous events and conditions) and event utilities (to capture the occurrence of complications in any given year of simulation) as previously described.21

1701

Clinical Therapeutics

Table II. Treatment effects applied in the modeling analysis on the basis of the Liraglutide Effect and Action in Diabetes 6 trial.* Effect† Change from baseline in HbA1c, % Change from baseline in SBP, mm Hg Change from baseline in total cholesterol, mmol/L Change from baseline in HDL-cholesterol, mmol/L Change from baseline in LDL-cholesterol, mmol/L Change from baseline in triglycerides, mmol/L Change from baseline in weight, kg Major hypoglycemic event rate, events/100 patient-years Minor hypoglycemic event rate, events/100 patient-years

Liraglutide

Exenatide BID

P for Difference

⫺1.12 (0.08) ⫺2.51 (1.15) ⫺0.20 (0.07) ⫺0.04 (0.02) ⫺0.44 (0.06) ⫺0.20 (0.07) ⫺3.24 (0.33) 0 193.2

⫺0.79 (0.08) ⫺2.00 (1.18) ⫺0.09 (0.07) ⫺0.05 (0.02) ⫺0.40 (0.06) ⫺0.09 (0.07) ⫺2.87 (0.33) 2 260

⬍0.0001 0.6409 0.0946 0.5105 0.4412 0.0485 0.2235 NA 0.0131

HbA1c ⫽ glycated hemoglobin; SBP ⫽ systolic blood pressure; HDL ⫽ high-density lipoprotein, LDL ⫽ low-density lipoprotein; NA ⫽ not available. *Values shown are least square means with SEs in parentheses as used in the modeling simulations for all 6 countries. † Treatment effects were applied in the first year of the simulation and then risk factors followed a natural progression over time.21 Hypoglycemia rates remained constant for the duration of treatment with exenatide or liraglutide.

Statistical Approach and Other Model Settings For each country-specific analysis, a simulated cohort of 1000 patients was run through the model 1000 times for each simulation (base case and sensitivity analysis) using a nonparametric bootstrapping approach, and mean (SD) values were generated for longterm outcomes.32 One thousand mean values (each from 1000 patients) of incremental direct medical costs and incremental effectiveness in terms of quality-adjusted life expectancy were plotted (scatter plots) on a cost-effectiveness plane. Subsequently, acceptability curves were generated by calculating the proportion of points below a range of willingness-to-pay thresholds (up to €60,000 per QALY gained for each country). The time horizon was set to patient lifetimes in the base case (40 years) to capture all relevant long-term complications and associated costs to assess their impact on life expectancy and quality-adjusted life expectancy in all 6 base case evaluations of cost-effectiveness. Future costs and clinical benefits were discounted in line with published guidance for Switzerland (3% per annum), Denmark (3.0% per annum), Norway (4% per annum), Finland (3% per annum), the Netherlands (1.5% on clinical outcomes and 4% on cost outcomes per annum), and Austria (3.5% per annum).33 For purposes of concision, base case complication data are only presented for the Swiss setting. Similarly,

1702

an incremental cost-effectiveness scatter plot is only presented for Switzerland (scatter plots or additional modeled outcomes data from the other settings are available on request from the corresponding author). All cost-effectiveness evaluations were performed in local currencies. To aid comparison, all local costs were converted into 2008 Euros using purchase power parity indices for 2008 from the Organization for Economic Co-operation and Development Web site34 as follows: €0.682675 ⫽ Danish kr 5.09813 ⫽ Norwegian kr 5.64000 ⫽ CHF 1.08309.

Sensitivity Analysis The extrapolation of clinical results by modeling the long-term consequences is associated with uncertainty. Sensitivity analysis was performed on key parameters in the model to assess the robustness of the base case findings. The influence of time horizon on the outcomes projected by the model was investigated by running analyses over 5 and 10 years. The effect of discounting on base case outcomes was investigated by varying the annual discount rate on future costs and clinical outcomes. For the Swiss analysis, discount rates were varied between 0% and 6% per annum in line with published recommendations. Country-specific ranges were used in other settings (Denmark [0%– 6%], Norway [0%– 8%], Finland [0%– 6%], the Netherlands [0%– 8%], and Austria [0%–7%]). In the

Volume 33 Number 11

W.J. Valentine et al. LEAD 6 trial,21 treatment with liraglutide was associated with statistically significant improvements in HbA1c and minor hypoglycemia versus exenatide, as well as favorable trends in terms of weight, SBP, and a comparable serum lipid profile. All of these effects were captured in the base case analysis. A sensitivity analysis was performed to evaluate the long-term outcomes applying only those benefits shown to be statistically significant in the LEAD 6 trial (ie, HbA1c and minor hypoglycemia). All other treatment effects were set to the same level as in the exenatide BID arm in this sensitivity analysis. HbA1c is known to be a key driver of long-term outcomes in patients with type 2 diabetes. A sensitivity analysis was performed to investigate the effect on long-term outcomes of abolishing the HbA1c benefit associated with liraglutide (ie, change from baseline in HbA1c was the same in the liraglutide and exenatide BID treatment arms). Based on data from the LEAD 6 trial,21 liraglutide was associated with a significant reduction in minor hypoglycemic events versus exenatide, and this finding was captured in the base case analysis. To investigate the influence of minor hypoglycemia as a driver of quality-adjusted life expectancy (and therefore cost-effectiveness), a sensitivity analysis was performed where no quality-of-life disutility was applied for each minor hypoglycemic event (the influence of minor hypoglycemia on the results was abolished). In the base case analysis, liraglutide and exenatide BID were associated with benefits in terms of weight reduction from baseline, which was captured using a previously published quality-of-life utility score from the Cost of Diabetes in Europe Type 2 study.30 These improvements in utility were only applied in the first 5 years of the simulation, after which time patients were switched to insulin, and weight was assumed to increase (with the same utilities applied in both treatment groups [ie, the utility scores associated with body mass index for patients receiving insulin were assumed to be the same regardless of whether the patient previously received liraglutide or exenatide]). The influence of the health-related quality-of-life utility score for body mass index applied in the base case analysis was investigated in a sensitivity analysis by setting it to zero (ie, no utility score relating to body mass index was applied). Sensitivity analysis was performed to assess the impact of overestimating or underestimating the unit costs of complications used in the analysis by increas-

November 2011

ing and decreasing the values used by ⫾10%. The annual treatment costs were not altered in these sensitivity analyses. To estimate the cost-effectiveness of liraglutide versus exenatide from a societal perspective, a sensitivity analysis was performed whereby patient co-pay (at pharmacy-selling prices excluding valueadded tax) and the costs associated with lost productivity due to complications were included. A human capital approach was used, in preference to the friction cost approach (which depends on strong assumptions), whereby days off work associated with diabetes-related complications resulted in lost productivity based on average annual salaries for men and women of working age in each country.35

RESULTS Long-term Clinical Outcomes Based on the results of the LEAD 6 trial,21 model projections indicated that liraglutide 1.8 mg once daily was associated with improvements in mean life expectancy of 0.09 to 0.17 year versus treatment with exenatide 10 ␮g BID (Table III). Capturing health-related quality of life in the estimation showed a slightly greater benefit, with incremental quality-adjusted life expectancy values of 0.12 to 0.17 QALY (discounted) for liraglutide versus exenatide. Benefits in terms of life expectancy and quality-adjusted life expectancy were driven largely by a reduced incidence of diabetes-related complications with liraglutide (Table IV). The cumulative incidence of all microvascular complications, including renal complications, eye disease, and diabetic foot complications, was reduced in liraglutidetreated patients compared with the exenatide group. For instance, the cumulative incidence of myocardial infarction, one of the most costly macrovascular complications of diabetes, was ⬃6% lower with liraglutide (relative difference vs exenatide). The reductions in the long-term incidence of diabetes-related complications associated with liraglutide was driven primarily by improvements in risk factors, most notably HbA1c, associated with liraglutide based on data from the LEAD 6 trial. In addition to reducing the projected cumulative incidence of complications, liraglutide was associated with a delayed mean time to onset of diabetes-related complications. Mean time to onset of any diabetesrelated complication in the modeling analysis was ⬃0.32 year longer with liraglutide than with exenatide. With regard to individual complications,

1703

Clinical Therapeutics

Table III. Projected life expectancy and quality-adjusted life expectancy for liraglutide and exenatide BID.* Life Expectancy (y) Setting Switzerland Denmark Norway Finland The Netherlands Austria

Quality-Adjusted Life Expectancy (QALYs)

Liraglutide

Exenatide BID

Incremental

Liraglutide

Exenatide BID

Incremental

12.23 (0.17) 11.82 (0.17) 11.35 (0.14) 12.57 (0.17) 13.84 (0.22) 11.84 (0.16)

12.09 (0.17) 11.73 (0.18) 11.22 (0.15) 12.41 (0.18) 13.67 (0.21) 11.70 (0.16)

0.14 (0.24) 0.09 (0.24) 0.13 (0.19) 0.16 (0.25) 0.17 (0.30) 0.13 (0.23)

7.93 (0.11) 7.83 (0.11) 7.53 (0.09) 8.28 (0.11) 9.09 (0.14) 7.82 (0.10)

7.78 (0.11) 7.71 (0.12) 7.39 (0.10) 8.12 (0.12) 8.92 (0.13) 7.68 (0.11)

0.15 (0.16) 0.12 (0.16) 0.14 (0.13) 0.16 (0.16) 0.17 (0.19) 0.14 (0.15)

QALYs ⫽ quality-adjusted life years. *Values shown are means with SDs in parentheses from the base case modeling analyses. Outcomes were discounted based on country-specific guidance at the following annual rates: Switzerland, 3%; Denmark, 3%; Norway, 4%; Finland, 3%; the Netherlands, 1.5% (4% on costs); and Austria, 3.5%. Differences in projected life expectancy and QALYs reflect inclusion of country-specific mortality tables, concomitant medications and complication screening rates, and end-stage renal disease– related outcomes data as well as different discount rates for QALYs.

mean time to onset was delayed by between 0.22 year (stroke) and 0.41 year (neuropathy). Benefits were observed across all micro- and macrovascular complications. Interestingly, the delay in the mean time to onset of stroke associated with liraglutide is consistent with the effect of the survival paradox for this complication. The delayed onset of stroke provides evidence that liraglutide is not associated with an increased risk of this complication; rather, the occurrence of events late in the simulation in patients of advanced age is the reason behind the projected increase in cumulative incidence of stroke with liraglutide versus exenatide.

Long-term Costs and Cost-effectiveness Evaluation of direct medical costs suggested that the total costs (pharmacy and complication costs) for the liraglutide 1.8 mg treatment arm were likely to be between €1023 and €1866 more than those of the exenatide arm over patients’ lifetimes (Table V). Pharmacy costs, which captured the treatment of diabetes over patients’ lifetimes (including insulin, administration devices and needles, as well as self-monitoring of blood glucose devices, needles, and test strips), were the biggest driver of direct costs. Increased pharmacy costs in the liraglutide treatment group were partially offset by the reduced costs of complications in all 6 countries investigated. The total lifetime mean costs of diabetes-related complications were lower with liraglutide 1.8 mg versus exenatide (between €447 [Nor-

1704

way] and €934 [CHF 1482, Switzerland]). The biggest component of complication costs was cardiovascular complications in both treatment groups, accounting for between 34% and 70% of total complication costs depending on treatment and setting. Evaluation of cost-effectiveness based on qualityadjusted life expectancy produced incremental cost-effectiveness ratios (ICERs) in the range of €6902 to €13,546 per QALY gained across the 6 countries (Table V). For each setting, the ICER was in the range commonly quoted as representing good value for money in each country. The incremental cost-effectiveness scatter plot for Switzerland is shown in Figure 1. The majority of points fall in the upper-right quadrant, indicating both increased effectiveness (ie, incremental quality-adjusted life expectancy) and increased total costs for liraglutide compared with exenatide treatment. Similar patterns were observed for the other 5 countries included in the analysis. The data from the scatter plots were used to generate acceptability curves by estimating the percentage of points below a range of set willingness-to-pay values (Figure 2). Based on this analysis, assuming a willingness to pay of €40,000 per QALY gained, the analysis indicated that there was a 80% probability that liraglutide would be cost-effective versus exenatide in Switzerland. These values ranged between 71% and 79% in the other countries (Denmark, 71%; Norway, 75%; Finland, 79%; the Netherlands, 77%; and Austria, 78%).

Volume 33 Number 11

W.J. Valentine et al.

Table IV. Cumulative incidence of long-term complications for liraglutide and exenatide BID (Switzerland).* Complications

Liraglutide

Exenatide

Difference

Eye disease, % Diabetic retinopathy Severe vision loss Macular edema Cataract

20.0 (1.3) 7.9 (0.9) 16.0 (1.1) 9.8 (1.0)

21.0 (1.3) 8.3 (0.9) 16.8 (1.2)) 9.9 (1.0)

1.0 0.4 0.8 0.2

Renal disease, % Microalbuminuria Gross proteinuria End-stage renal disease

24.8 (1.4) 5.5 (0.7) 0.8 (0.3)

25.7 (1.4) 5.9 (0.7) 0.9 (0.3)

1.0 0.4 0.1

Diabetic foot complications, % Ulcer Amputation

24.8 (1.5) 6.5 (0.9)

25.4 (1.4) 6.8 (0.9)

0.6 0.3

Neuropathic complications, % Neuropathy

49.3 (1.7)

50.9 (1.7)

1.6

Cardiovascular disease, % Myocardial infarction Stroke Heart failure Angina Peripheral vascular disease

27.4 (1.4) 26.1 (1.4) 29.3 (1.5) 13.8 (1.1) 11.5 (1.0)

29.1 (1.5) 25.8 (1.4) 30.0 (1.5) 14.6 (1.1) 12.0 (1.1)

1.7 0.3 0.7 0.8 0.5

Hypoglycemia, % Major hypoglycemia Minor hypoglycemia

0 25.1 (0.4)

0.30 (0.02) 27.9 (0.4)

0.30 2.8

*Values shown are cumulative incidences over patient lifetimes from the base case modeling simulation for Switzerland expressed as a percentage of patients experiencing events.

Sensitivity Analysis Sensitivity analysis indicated that the base case findings were most sensitive to variation in the HbA1c benefit of liraglutide and changing time horizon (data from the Swiss setting are shown in Table VI). Shortening the time horizon to 5 and 10 years limited the clinical benefits associated with liraglutide in terms of complications avoided, as many diabetes-related complications take time to develop and consequently are not captured in short-term analyses. As a result, the ICERs for liraglutide 1.8 mg versus exenatide were higher in these scenarios than in the base case. Variation in the annual discount rates between 0% and 6% had little impact on the incremental findings.

November 2011

Sensitivity analysis in which only those treatment effects that were (statistically) significantly different from LEAD 6 were captured in the modeling analysis provided cost-effectiveness findings very similar to those in the base case. In this analysis, assuming that liraglutide was only associated with benefits in HbA1c and minor hypoglycemia produced an ICER of approximately CHF 18,573 or €11,707 per QALY gained, provides evidence that these 2 parameters are the main clinical drivers in the modeling study. Abolishing the HbA1c benefit associated with liraglutide notably reduced the long-term clinical benefits (incremental quality-adjusted life expectancy, 0.05 QALY) and produced an ICER of approximately CHF 43,378 or

1705

Clinical Therapeutics

Table V. Projected direct costs and cost-effectiveness for liraglutide versus exenatide BID.* Setting

Direct Costs (Local Currency Unless Otherwise Indicated) Incremental (€)

ICER (Cost/ QALY Gained)

CHF 1623 (3363)

1023 (2120)

262,365 (9293)

kr 10,186 (11,960)

1364 (1602)

256,485 (7153)

241,071 (7018)

kr 15,414 (9763)

1866 (1182)

43,173 (1557) 32,806 (1221) 32,448 (1141)

41,805 (1488) 31,412 (1208) 31,282 (1119)

CHF 10,950 €6902 kr 88,160 €11,805 kr 111,916 €13,546 €8459 €8119 €8516

Liraglutide

Exenatide

80,727 (2242)

79,105 (2398)

Denmark

272,552 (8179)

Norway

Switzerland

Finland The Netherlands Austria

Incremental (Local)

– – –

1368 (2103) 1395 (1678) 1207 (1569)

ICER ⫽ incremental cost-effectiveness ratio; QALY ⫽ quality-adjusted life year; CHF ⫽ Swiss francs; € ⫽ Euros; kr ⫽ Danish or Norwegian krone. *Values shown are means with SDs in parentheses from the base case modeling analyses. Outcomes were discounted based on country-specific guidance at the following annual rates: Switzerland, 3%; Denmark, 3%; Norway, 4%; Finland, 3%; the Netherlands, 4%; and Austria, 3.5%. Local costs were converted to Euros using purchase power parity indices for 2008 from the Organization for Economic Co-operation and Development Web site34 as follows: €0.682675 ⫽ Danish kr 5.09813 ⫽ Norwegian kr 5.64000 ⫽ CHF 1.08309.

15,000

Incremental Cost (CHF)

10,000 5000

–0.4

–0.2

0

0.2

0.4

0.6

0.8

–5000 –10,000 –15,000

Incremental Effectiveness (QALYs)

Figure 1. Scatter plot of incremental costs versus incremental effectiveness of liraglutide versus exenatide BID from the modeling analysis (Swiss setting). The scatterplot shows mean incremental costs (in Swiss francs [CHF]) versus mean incremental effectiveness (expressed in quality-adjusted life years [QALYs]) of the Switzerland base case modeling analysis, for 1000 iterations, each involving 1000 simulated patients in the CORE Diabetes Model.

1706

Volume 33 Number 11

Probability that Liraglutide would be Cost-Effective Versus Exenatide (%)

W.J. Valentine et al.

100 80 60

Switzerland Denmark Norway Finland Netherlands Austria

40 20 0 0

10,000

20,000

30,000

40,000

50,000

60,000

Willingness to Pay (EUR/QALY Gained)

Figure 2. Cost-effectiveness acceptability curves of liraglutide versus exenatide BID from the modeling analysis. The acceptability curve shows the proportion of model iterations (each representing mean values from 1000 simulated patients) from the base case modeling simulations that are below willingness-to-pay thresholds between 0 and Euro (€) 60,000 per quality-adjusted life year (QALY) gained for all 6 countries in the analysis. Incremental cost-effectiveness ratios below the willingness-to-pay threshold were considered cost-effective.

€27,341 per QALY gained, providing further evidence that HbA1c is a key driver of long-term outcomes in this modeling study. Exclusion of disutilities for minor hypoglycemia or utilities associated with changes in weight from the analysis did not notably change the long-term clinical outcomes. Similarly, increasing and decreasing the cost of diabetes-related complications by ⫾10% had little impact on the base case incremental findings. However, including indirect costs based on a human capital approach had a notable effect on the long-term cost outcomes. Total costs were approximately CHF 97,433 (€61,412) with liraglutide and CHF 97,778 (€61,630) with exenatide (difference, CHF ⫺346 [€218]), making liraglutide dominant (ie, cost and lifesaving) versus exenatide from a societal perspective.

DISCUSSION The aim of the pharmacoeconomic evaluation described in this article was to evaluate the long-term outcomes associated with liraglutide versus exenatide in the treatment of type 2 diabetes failing to improve with metformin and/or sulfonylurea. Because many of the costs incurred by patients with diabetes are associated with end-stage complications that only occur many years after diagnosis, a computer simulation modeling approach was used to make long-term pro-

November 2011

jections based on data from published clinical and epidemiologic studies. This approach is in line with published guidance on modeling diabetes.36 The analysis provides evidence that liraglutide is likely to be associated with improvements in life expectancy and qualityadjusted life expectancy, as well as reduced incidence rates of diabetes-related complications compared with exenatide. Liraglutide 1.8 mg was associated with higher direct medical costs over patient lifetimes, leading to ICERs between €6902 and €13,546 per QALY gained across all 6 countries investigated. Sensitivity analysis indicated that HbA1c change was the main driver of the long-term benefits associated with liraglutide. This is a reassuring observation, given that change from baseline in HbA1c was the primary end point of the LEAD 6 trial21 (on which the modeling analysis was based) and improvement in HbA1c associated with liraglutide was shown to be significantly greater than that observed with exenatide. One of the main limitations of this type of modeling approach is related to the inherent uncertainty of projecting long-term clinical outcomes on the basis of data from short-term clinical trials. Every effort was made to minimize this uncertainty in the present analysis by performing a sensitivity analysis to investigate costeffectiveness using only inputs with a statistically significant difference between treatments, and by using a

1707

Clinical Therapeutics

Table VI. Sensitivity analysis results from the modeling analysis on the basis of the Liraglutide Effect and Action in Diabetes 6 trial in the Swiss setting.* Quality-Adjusted Life Expectancy (QALYs) Analysis

Liraglutide

Exenatide

Base case

7.93 (0.11)

5-year time horizon 10-year time horizon

Difference

Lifetime Direct Costs (CHF) Liraglutide

Exenatide

7.78 (0.11) 0.15 (0.16)

80,727 (2242)

79,105 (2398)

3.07 (0.02)

3.04 (0.02) 0.03 (0.03)

20,589 (779)

18,209 (776)

5.24 (0.05)

5.18 (0.05) 0.06 (0.07)

36,850 (1164)

34,873 (1217)

10.30 (0.18) 0.23 (0.24) 121,602 (3899)

119,715 (4044)

Difference

ICER†/Outcome

0% discount rate

10.53 (0.18)

6% discount rate

6.25 (0.08)

6.14 (0.08) 0.10 (0.11)

57,662 (1512)

56,088 (1623)

Only HbA1c and minor hypoglycemia benefits for liraglutide included

7.88 (0.11)

7.78 (0.11) 0.10 (0.15)

81,031 (2257)

79,105 (2398)

1623 (3363) CHF 10,950 €6902 2379 (1093) CHF 72,790 €45,880 1976 (1628) CHF 34,330 €21,638 1887 (5670) CHF 8233 €5189 1574 (2288) CHF 15,254 €9615 1926 (3155) CHF 18,573

No HbA1c benefit for liraglutide included

7.83 (0.11)

7.78 (0.11) 0.05 (0.15)

81,409 (2153)

79,105 (2398)

€11,707 2304 (3197) CHF 43,378

Exclusion of disutilities for minor hypoglycemia

7.99 (0.11)

7.85 (0.12) 0.14 (0.16)

80,727 (2242)

79,105 (2398)

€27,341 1623 (3363) €11,662

Exclusion of utilities for weight

8.55 (0.12)

8.40 (0.12) 0.15 (0.17)

80,727 (2242)

79,105 (2398)

EUR 7351 1623 (3363) CHF 10,681

10% increase in complication costs

7.93 (0.11)

7.78 (0.11) 0.15 (0.16)

85,838 (2449)

84,363 (2621)

€6732 1475 (3676) CHF 9951

10% decrease in complication costs

7.93 (0.11)

7.78 (0.11) 0.15 (0.16)

75,617 (2035)

73,846 (2176)

€6272 1771 (3051) CHF 11,949

Inclusion of indirect costs

7.93 (0.11)

7.78 (0.11) 0.15 (0.16)

97,433 (3327)

€7531 97,778 (3584) ⫺346 (3770) Liraglutide dominant

QALY ⫽ quality-adjust life year; CHF ⫽ Swiss francs; ICER ⫽ incremental cost-effectiveness ratio; HbA1c ⫽ glycated hemoglobin. *Values shown are means with SDs in parentheses from the sensitivity analysis simulations for Switzerland. † ICERs are expressed in CHF or Euros (€) per QALY gained. Local costs were converted to Euros using purchase power parity indices for 2008 from the Organization for Economic Co-operation and Development Web site34 as follows: €0.682675 ⫽ CHF 1.08309.

computer simulation model that has been previously validated against real-life clinical data and has been accepted by a number of reimbursement authorities around the world. Another potential limitation of the study was that, because it was based on the LEAD 6 clinical trial,21 the modeling evaluation investigated the cost-effectiveness of the 1.8-mg dose of liraglutide rather than the 1.2-mg dose, which is the standard treatment dose for most patients.37 Given that data from the LEAD 1– 438 studies indicate that the clinical effectiveness of liraglutide 1.2 mg is ⬃80% to 90% of the 1.8-mg dose and that the lower dose costs less, liraglutide 1.2 mg is likely to be a very attractive treat-

1708

ment option in terms of both clinical and cost outcomes versus exenatide in the settings investigated in this analysis. Previous cost-effectiveness analysis using methods similar to that in the present analysis have revealed that exenatide BID is cost-effective versus insulin glargine from a health care payer perspective in several countries (although one study provided contradictory evidence).39 – 42 In most of these studies, improvements in weight were a key driver of improved quality-adjusted life expectancy with exenatide versus insulin glargine and were greater than the quality-of-life decrements associated with transient adverse effects such as nau-

Volume 33 Number 11

W.J. Valentine et al. sea. Evidence from the LEAD 6 clinical trial21 indicates that liraglutide may be associated with increased weight loss and that nausea was more transient compared with exenatide. The impact of transient nausea and vomiting associated with GLP-1 receptor agonists was not directly captured in the present modeling analysis. Although this may be considered a limitation of the study, it should be noted that the transient and nonserious nature of these adverse effects means that they would have very little effect on long-term qualityadjusted life expectancy or direct medical costs (and therefore would not notably alter cost-effectiveness). Any impact on compliance that these events may have had was captured in the modeling analysis, because the data used from LEAD 6 were derived from the intentto-treat population. This feature can also be considered a conservative approach, as any impact of noncompliance in the study population would be captured in terms of effectiveness (changes from baseline in risk factors) but not in terms of costs (it was assumed that all patients received a full course of study medication until treatment failure). The present analysis is associated with a number of limitations. First, the generalizability of the findings of the analysis are limited to those patients with relatively advanced disease (mean duration of diabetes in the simulation cohort was 8.0 years) who had failed to achieve adequate glycemic control with metformin, sulfonylurea, or both. It is also possible that owing to the relatively advanced disease stage of the simulated cohort, that the benefit of HbA1c reduction may be lower in these patients compared with patients with less advanced disease. In addition, although the findings of the present analysis suggest that liraglutide is cost-effective compared with exenatide, this finding is again limited to patients with relatively advanced disease. Both exenatide and liraglutide are generally associated with higher treatment costs compared with sulfonylureas and/or metformin. Further analyses in patients with less advanced disease would be necessary to ascertain whether liraglutide would be considered a cost-effective treatment option in a patient population with less advanced disease. There have been reports in the medical literature regarding concerns over the long-term safety of GLP-1 receptor agonists. In 2008, a summary was published of 30 cases of individuals taking exenatide who developed acute pancreatitis.43 This was followed in 2009 by analysis of insurance claims that showed no in-

November 2011

creased risk of acute pancreatitis in 27,667 initiators of exenatide treatment; it did show that the overall risk of pancreatitis for subjects followed up to 1 year was 0.13%.44 The risk of acute pancreatitis was comparable for initiators of exenatide (relative risk: 1.0 [95% CI: 0.6 –1.7]) relative to the comparison cohort of metformin or glyburide initiators. To the best of our knowledge, there are presently no conclusive data establishing a risk of pancreatitis with liraglutide treatment. Patients with type 2 diabetes have a 3-fold higher risk than the general population, equivalent to ⬃1.5 to 4.5 cases per 1000 patient-years.45 Data from the manufacturers have suggested that there is a low incidence of acute pancreatitis (0.8 case/1000 patient-years).46 Medullary thyroid carcinoma has also been discussed as a potential long-term adverse effect of treatment with GLP-1 receptor agonists on the basis of preclinical data in rodents.45,47 However, differences between rodent and human C-cell biology in terms of response to GLP-1 receptor activation raise important caveats around how these data should be interpreted.48 –50 Medullary thyroid carcinoma is an extremely rare condition in humans, with a reported lifetime risk of 0.013%, making the design of clinical trials to detect an increased risk challenging.45,51 Based on currently available data, several reviews on the risks and benefits of GLP-1 receptor agonists have been published in recent years. These studies have offered consensus that the use of incretin-based therapies is supported for patients requiring effective control of glycemia and weight while minimizing the risk of hypoglycemic events, and the risk-benefit ratio continues to be favorable.46,49,52,53 However, long-term safety studies (eg, the Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results trial on liraglutide54 and the Exenatide Study of Cardiovascular Event Lowering Trial on exenatide55) and ongoing focused pharmacovigilance programs in the years ahead will provide additional evidence on the safety of these agents. The evidence provided by this long-term cost-effectiveness evaluation suggests that liraglutide is highly cost-effective versus exenatide. In addition, liraglutide is likely to offer benefits for patients in terms of improved life expectancy, quality of life, and fewer diabetes-related complications. For the health care payer and society, liraglutide has the potential to deliver an attractive return on investment (especially given that 1.2 mg is likely to be the standard dose for most pa-

1709

Clinical Therapeutics tients) by reducing the burden of diabetes-related complications on the health care system and reducing lost workplace productivity and premature retirement. 10.

CONCLUSIONS Long-term projections indicated that liraglutide was associated with benefits in life expectancy, QALYs, and reduced complication rates versus exenatide. Liraglutide was cost-effective from a health care payer perspective in Switzerland, Denmark, Norway, Finland, the Netherlands, and Austria.

11.

12.

ACKNOWLEDGEMENTS The study was supported by funding from Novo Nordisk Pharma AG (Küsnacht, Switzerland). The authors have indicated that they have no conflicts of interest with regard to the content of this article.

REFERENCES 1. Wild S, Roglic G, Green A, et al. Global prevalence of diabetes. Estimates for the year 2000 and estimated for 2030. Diabetes Care. 2004;27:1047–1053. 2. International Diabetes Federation Diabetes Atlas, Third Edition. http://www.eatlas.idf.org/media. Accessed March 17, 2011. 3. Gaede P, Vedel P, Larsen N, et al. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348:383–393. 4. Gaede P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. N Engl J Med. 2008;358:580 –591. 5. Gaede P, Valentine WJ, Palmer AJ, et al. Cost-effectiveness of intensified versus conventional multifactorial intervention in type 2 diabetes: results and projections from the Steno-2 study. Diabetes Care. 2008;1:1510 –1515. 6. Nathan DM, Buse JB, Davidson MB, et al. Management of hyperglycemia in type 2 diabetes: A consensus algorithm for the initiation and adjustment of therapy: a consensus statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. 2006;29:1963–1972. 7. Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet. 2006;368:1696 –1705. 8. Monami M, Marchionni N, Mannucci E. Glucagon-like peptide-1 receptor agonists in type 2 diabetes: a metaanalysis of randomized clinical trials. Eur J Endocrinol 2009; 160:909 –917. 9. Pratley RE, Nauck M, Bailey T, et al, for the 1860-LIRADPP-4 Study Group. Liraglutide versus sitagliptin for patients with type 2 diabetes who did not have adequate

1710

13.

14.

15.

16.

17.

18.

19.

20.

21.

glycaemic control with metformin: a 26-week, randomised, parallel-group, open-label trial. Lancet. 2010;375: 1447–1456. Nauck MA, Hompesch M, Filipicazak R, et al. Five weeks of treatment with the GLP-1 analogue liraglutide improves glycaemic control and lowers body weight in subjects with type 2 diabetes. Exp Clin Endocrinol Diabetes. 2006;114:417– 423. Nauck M, Frid A, Hermansen K, et al. Efficacy and safety comparison of liraglutide, glimepiride, and placebo, all in combination with metformin in type 2 diabetes mellitus (LEAD-2 Met). Diabetes Care. 2009;32:84 –90. Garber A, Henry R, Ratner R, et al. Liraglutide versus glimepiride monotherapy for type 2 diabetes (LEAD-3 Mono): a randomised, 52-week, phase III, double-blind, parallel-treatment trial. Lancet. 2009;373:473– 481. DeFronzo RA, Ratner RE, Han J, et al. Effects of exenatide (exendin-4) on glycemic control and weight over 30 weeks in metformin-treated patients with type 2 diabetes. Diabetes Care. 2005;28:1092–1100. Buse JB, Henry RR, Han J, et al. Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. Diabetes Care. 2004;27:2628 –2635. Kendall DM, Riddle MC, Rosenstock J, et al. Effects of exenatide (exendin-4) on glycemic control over 30 weeks in patients with type 2 diabetes treated with metformin and a sulfonylurea. Diabetes Care. 2005;28:1083–1091. Åhren B, Gomis R, Standl E, et al. Twelve- and 52-week efficacy of the dipeptidyl peptidase IV inhibitor LAF237 in metformin-treated patients with type 2 diabetes. Diabetes Care. 2004;27:2874 –2880. Åhren B, Landin-Olsson M, Jansson PA, et al. Inhibition of dipeptidyl peptidase-4 reduces glycemia, sustains insulin levels, and reduces glucagon levels in type 2 diabetes. J Clin Endocrinol Metab. 2004;89:2078 –2084. Scott R, Herman G, Zhao P, et al. Twelve-week efficacy and tolerability of MK-0431, a dipeptidyl peptidase IV (DPPIV) inhibitor, in the treatment of type 2 diabetes (T2D). Diabetes. 2005;54(Suppl 1):10 –11. Hanefeld M, Herman G, Mickel C, et al. Effect of MK-0431, a dipeptidyl peptidase IV (DPP-IV) inhibitor, on glycemic control after 12 weeks in patients with type 2 diabetes. Diabetologia. 2005;48(Suppl 1):287–288. Hansen KB, Knop FK, Holst JJ, Vilsbøll T. Treatment of type 2 diabetes with glucagon-like peptide-1 receptor agonists. Int J Clin Pract. 2009;63:1154 –1160. Buse JB, Rosenstock J, Sesti G, et al, for the LEAD-6 Study Group. Liraglutide once a day versus exenatide twice a day for type 2 diabetes: a 26-week randomised, parallelgroup, multinational, open-label trial (LEAD-6). Lancet. 2009;374:39 – 47.

Volume 33 Number 11

W.J. Valentine et al. 22. Palmer AJ, Roze S, Valentine WJ, et al. The CORE Diabetes Model: projecting long-term clinical outcomes, costs and cost-effectiveness of interventions in diabetes mellitus (Types 1 and 2) to support clinical and reimbursement decision-making. Curr Med Res Opin. 2004;20(Suppl 1):5–26. 23. Palmer AJ, Roze S, Valentine W, et al. Validation of the CORE Diabetes Model against epidemiological and clinical studies. Curr Med Res Opin. 2004;20(Suppl 1):27– 40. 24. Mount Hood 4 Modeling Group. Computer modeling of diabetes and its complications: a report on the Fourth Mount Hood Challenge Meeting. Diabetes Care. 2007;30: 1638 –1646. 25. Clarke PM, Gray AM, Briggs A, et al. A model to estimate the lifetime health outcomes of patients with type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (UKPDS no. 68). Diabetologia. 2004;47:1747– 1759. 26. Clarke P, Gray A, Holman R. Estimating utility values for health states of type 2 diabetic patients using the EQ-5D (UKPDS 62). Med Decis Making. 2002;22:340 –349. 27. Australian Institute of Health and Welfare. The Burden of Disease and Injury in Australia. 2003. http:// www.aihw.gov.au/publication-detail/ ?id⫽6442467990. Accessed March 17, 2010. 28. Tengs TO, Wallace A. One thousand health-related quality-of-life estimates. Med Care. 2000;38:583– 637. 29. Carrington AL, Mawdsley SK, Morley M, et al. Psychological status of diabetic people with or without lower limb disability. Diabetes Res Clin Pract. 1996;32:19 –25. 30. Levy AR, Christensen TL, Johnson JA. Utility values for symptomatic non-severe hypoglycaemia elicited from persons with and without diabetes in Canada and the United

November 2011

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

Kingdom. Health Qual Life Outcomes. 2008;29:73. Bagust A, Beale S. Modelling EuroQol health-related utility values for diabetic complications from CODE2 data. Health Econ. 2005;14:217– 230. Briggs AH, Wonderling DE, Mooney CZ. Pulling cost-effectiveness analysis up by its bootstraps: a nonparametric approach to confidence interval estimation. Health Econ. 1997;6:327–340. Tarn T, Smith M. Pharmacoeconomic guidelines around the world. ISPOR Connections 2004;10:5–15. Organization for Economic Co-operation and Development Web site. http://stats.oecd.org/Index.aspx? datasetcode⫽SNA_TABLE4. Accessed March 3, 2011. Liljas B. How to calculate indirect costs in economic evaluations. Pharmacoeconomics. 1998;13:1–7. American Diabetes Association Consensus Panel. Guidelines for computer modeling of diabetes and its complications. Diabetes Care. 2004; 27:2262–2265. Schöffski O, Mentrup S, Lund N. Prescribed daily doses of oncedaily liraglutide in the German Statutory Health Insurance (SHI). Poster presented at: ISPOR 13th Annual European Congress; November 6 –9, 2010; Prague, Czech Republic. Blonde L, Russell-Jones D. The safety and efficacy of liraglutide with or without oral antidiabetic drug therapy in type 2 diabetes: an overview of the LEAD 1-5 studies. Diabetes Obes Metab. 2009; 11(Suppl 3):26 –34. Mittendorf T, Smith-Palmer J, Timlin L, et al. Evaluation of exenatide vs. insulin glargine in type 2 diabetes: cost-effectiveness analysis in the German setting. Diabetes Obes Metab. 2009;11:1068 –1079. Brändle M, Erny-Albrecht KM, Goodall G, et al. Exenatide versus insulin glargine: a cost-effectiveness

41.

42.

43.

44.

45.

46.

47.

48.

evaluation in patients with Type 2 diabetes in Switzerland. Int J Clin Pharmacol Ther. 2009;47:501–515. Ray JA, Boye KS, Yurgin N, et al. Exenatide versus insulin glargine in patients with type 2 diabetes in the UK: a model of long-term clinical and cost outcomes. Curr Med Res Opin. 2007;23:609 – 622. Woehl A, Evans M, Tetlow AP, McEwan P. Evaluation of the cost effectiveness of exenatide versus insulin glargine in patients with sub-optimally controlled type 2 diabetes in the United Kingdom. Cardiovasc Diabetol. 2008;11:24. Ahmad SR, Swann, J. Exenatide and rare adverse events. N Engl J Med. 2008;358: 1970 –1971. Dore DD, Seeger JD, Arnold Chan K. Use of a claims-based active drug safety surveillance system to assess the risk of acute pancreatitis with exenatide or sitagliptin compared to metformin or glyburide. Curr Med Res Opin. 2009;25:1019 –1027. Noel RA, Braun DK, Patterson RE, Bloomgren GL. Increased risk of acute pancreatitis and biliary disease observed in patients with type 2 diabetes: a retro- spective cohort study. Diabetes Care. 2009;32:834 – 838. Drucker DJ, Bergenstal RM, Sherman SI, et al. Incretin-based therapies for the treatment of type 2 diabetes: evaluation of the risks and benefits. Diabetes Care. 2010;33: 428 – 433. Parola A. FDA Advisory Committee Non-clinical Briefing Document. NDA 2009;22–34. http://www.fda. gov/ohrms/dockets/ac/09/briefing/ 2009-4422b2-01-FDA.pdf. Accessed October 06, 2011. Bjerre Knudsen L, Madsen LW, Andersen S, et al. Glucagon-like peptide-1 receptor agonists activate rodent thyroid C-cells causing calcitonin release and C-cell proliferation. Endocrinology. 2010;151:1473– 1486.

1711

Clinical Therapeutics 49. Parks M, Rosebraugh C. Weighing risks and benefits of liraglutide:. the FDA’s review of a new antidiabetic therapy. N Engl J Med. 2010;362: 774 –776. 50. Hegedüs L, Moses AC, Zdravkovic M, et al. GLP-1 and calcitonin concentration in humans: lack of evidence of calcitonin release from sequential screening in over 5000 subjects with type 2 diabetes or nondiabetic obese subjects treated with the human GLP-1 analog, liraglutide. J Clin Endocrinol Metab. 2011; 96:853– 860. 51. Horner MJ, Ries LAG, Krapcho M, et al, eds. SEER Cancer Statistics Review, 1975–2006, 2009. Bethesda, MD, National Cancer Institute. http://seer.cancer.gov/csr/1975_ 2006/. (November 2008 data submission) Accessed August 31, 2011. 52. Drucker DJ, Sherman SI, Bergenstal RM, Buse JB. The safety of incretinbased therapies—review of the scientific evidence. J Clin Endocrinol Metab. 2011;96:2027–2031. 53. Madsbad S, Kielgast U, Asmar M, et al. An overview of once-weekly glucagon-like peptide-1 receptor agonists—available efficacy and safety data and perspectives for the future. Diabetes Obes Metabol. 2011;13:394 – 407. 54. Bergenstal R, Daniels G, Mann J, et al. Liraglutide Effect and Action in Diabetes: Evaluation of Cardiovascular Outcome Results (LEADERTM) Trial: rationale and study design. Diabetes. 2011;60(Suppl 1):A612. Abstract 2303-PO. 55. US Institutes of Health Clinical Trial Database: Exenatide Study of Cardiovascular Event Lowering Trial (EXSCEL): A Trial To Evaluate Cardiovascular Outcomes After Treatment With Exenatide Once Weekly In Patients With Type 2 Diabetes Mellitus. http://clinicaltrials.gov/ ct2/show/NCT01144338. Accessed September 15, 2011.

1712

Address correspondence to: William J. Valentine, PhD, Ossian Health Economics and Communications GmbH, Bäumleingasse 20, 4051 Basel, Switzerland. E-mail: [email protected]

Volume 33 Number 11