Cost-effectiveness of Canakinumab for Prevention of Recurrent Cardiovascular Events | Acute Coronary Syndromes | JAMA Cardiology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  One-Way Sensitivity Analysis of Yearly Cost of Canakinumab vs Incremental Cost-effectiveness Ratio
One-Way Sensitivity Analysis of Yearly Cost of Canakinumab vs Incremental Cost-effectiveness Ratio

Yearly cost of canakinumab (x-axis) vs the incremental cost-effectiveness ratio in the base case analysis (y-axis). The current annual cost of canakinumab ($73 000) is indicated by the vertical dashed line, and the base case incremental cost-effectiveness ratio ($6.4 million per quality-adjusted life-years gained) is indicated by the horizontal dashed line. The inset presents the annual cost of canakinumab needed for the drug to be considered cost-effective at commonly accepted benchmarks for cost-effectiveness ($150 000, $100 000, and $50 000 per quality-adjusted life-years gained).

Figure 2.  One-Way Sensitivity Analysis of Yearly Cost of Canakinumab vs Incremental Cost-effectiveness Ratio in Base Case and Treatment Responders–Only Scenario
One-Way Sensitivity Analysis of Yearly Cost of Canakinumab vs Incremental Cost-effectiveness Ratio in Base Case and Treatment Responders–Only Scenario

The association of annual cost of canakinumab with the incremental cost-effectiveness ratio for the base case analysis (orange line) and for the scenario in which only responders continue treatment (brown line). In the treatment responders–only scenario, only those with a high-sensitivity C-reactive protein level less than 2 mg/L after the first dose of canakinumab continued treatment, and they achieved a 31% reduction in all-cause mortality. At the current annual cost of canakinumab ($73 000; vertical dashed line), the incremental cost-effectiveness ratio (horizontal dashed lines) is $6.4 million per quality-adjusted life-year gained in the base case and $819 000 per quality-adjusted life-year gained in the treatment responders–only scenario, both well above standard benchmarks for cost-effectiveness.

Figure 3.  Proportion of Cost-effective Simulations as a Function of the Annual Cost of Canakinumab
Proportion of Cost-effective Simulations as a Function of the Annual Cost of Canakinumab

The proportion of simulations from a probabilistic sensitivity analysis (y-axis) in which canakinumab was cost-effective at willingness-to-pay thresholds of $150 000 (orange line), $100 000 (blue line), and $50 000 (navy line) per quality-adjusted life-year (QALY) added in relation to different yearly drug costs (x-axis).

Table 1.  Base Case and Lung Cancer Scenario Parameters and Sources
Base Case and Lung Cancer Scenario Parameters and Sources
Table 2.  Total Costs, Quality-Adjusted Life-Years (QALYs), and Incremental Cost-effectiveness Ratios (ICERs) for Base Case, Sensitivity, and Scenario Analyses
Total Costs, Quality-Adjusted Life-Years (QALYs), and Incremental Cost-effectiveness Ratios (ICERs) for Base Case, Sensitivity, and Scenario Analyses
1.
Libby  P.  Interleukin-1 beta as a target for atherosclerosis therapy: biological basis of CANTOS and beyond.  J Am Coll Cardiol. 2017;70(18):2278-2289. doi:10.1016/j.jacc.2017.09.028PubMedGoogle ScholarCrossref
2.
Ridker  PM, Everett  BM, Thuren  T,  et al; CANTOS Trial Group.  Antiinflammatory therapy with canakinumab for atherosclerotic disease.  N Engl J Med. 2017;377(12):1119-1131. doi:10.1056/NEJMoa1707914PubMedGoogle ScholarCrossref
3.
Ridker  PM, MacFadyen  JG, Thuren  T, Everett  BM, Libby  P, Glynn  RJ; CANTOS Trial Group.  Effect of interleukin-1β inhibition with canakinumab on incident lung cancer in patients with atherosclerosis: exploratory results from a randomised, double-blind, placebo-controlled trial.  Lancet. 2017;390(10105):1833-1842. doi:10.1016/S0140-6736(17)32247-XPubMedGoogle ScholarCrossref
4.
US Office of Procurement, Acquisition and Logistics (OPAL). Pharmaceutical prices. https://www.va.gov/oal/business/fss/pharmPrices.asp. Accessed March 14, 2018.
5.
Ridker  PM, MacFadyen  JG, Everett  BM, Libby  P, Thuren  T, Glynn  RJ; CANTOS Trial Group.  Relationship of C-reactive protein reduction to cardiovascular event reduction following treatment with canakinumab: a secondary analysis from the CANTOS randomised controlled trial.  Lancet. 2018;391(10118):319-328. doi:10.1016/S0140-6736(17)32814-3PubMedGoogle ScholarCrossref
6.
Myerson  M, Coady  S, Taylor  H, Rosamond  WD, Goff  DC  Jr; ARIC Investigators.  Declining severity of myocardial infarction from 1987 to 2002: the Atherosclerosis Risk in Communities (ARIC) study.  Circulation. 2009;119(4):503-514. doi:10.1161/CIRCULATIONAHA.107.693879PubMedGoogle ScholarCrossref
7.
Peterson  ED, Dai  D, DeLong  ER,  et al; NCDR Registry Participants.  Contemporary mortality risk prediction for percutaneous coronary intervention: results from 588,398 procedures in the National Cardiovascular Data Registry.  J Am Coll Cardiol. 2010;55(18):1923-1932. doi:10.1016/j.jacc.2010.02.005PubMedGoogle ScholarCrossref
8.
Surveillance, Epidemiology, and End Results Program. Previous version: SEER Cancer Statistics Review, 1975-2013. https://seer.cancer.gov/csr/1975_2013/. Accessed December 1, 2017.
9.
Arias  E, Heron  M, Xu  J.  United States life tables, 2014.  Natl Vital Stat Rep. 2017;66(4):1-64.PubMedGoogle Scholar
10.
Bleumink  GS, Knetsch  AM, Sturkenboom  MC,  et al.  Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk and prognosis of heart failure: the Rotterdam Study.  Eur Heart J. 2004;25(18):1614-1619. doi:10.1016/j.ehj.2004.06.038PubMedGoogle ScholarCrossref
11.
Lassman  D, Hartman  M, Washington  B, Andrews  K, Catlin  A.  US health spending trends by age and gender: selected years 2002-10.  Health Aff (Millwood). 2014;33(5):815-822. doi:10.1377/hlthaff.2013.1224PubMedGoogle ScholarCrossref
12.
Dieleman  JL, Baral  R, Birger  M,  et al.  US spending on personal health care and public health, 1996-2013.  JAMA. 2016;316(24):2627-2646. doi:10.1001/jama.2016.16885PubMedGoogle ScholarCrossref
13.
US Centers for Medicare and Medicaid Services. Physician fee schedule search. https://www.cms.gov/apps/physician-fee-schedule/license-agreement.aspx. Accessed March 14, 2018.
14.
Kazi  DS, Moran  AE, Coxson  PG,  et al.  Cost-effectiveness of PCSK9 inhibitor therapy in patients with heterozygous familial hypercholesterolemia or atherosclerotic cardiovascular disease.  JAMA. 2016;316(7):743-753. doi:10.1001/jama.2016.11004PubMedGoogle ScholarCrossref
15.
Murphy  LB, Cisternas  MG, Pasta  DJ, Helmick  CG, Yelin  EH.  Medical expenditures and earnings losses among US adults with arthritis in 2013.  Arthritis Care Res (Hoboken). 2018;70(6):869-876. doi:10.1002/acr.23425PubMedGoogle ScholarCrossref
16.
Voigt  J, Sasha John  M, Taylor  A, Krucoff  M, Reynolds  MR, Michael Gibson  C.  A reevaluation of the costs of heart failure and its implications for allocation of health resources in the United States.  Clin Cardiol. 2014;37(5):312-321. doi:10.1002/clc.22260PubMedGoogle ScholarCrossref
17.
Cipriano  LE, Romanus  D, Earle  CC,  et al.  Lung cancer treatment costs, including patient responsibility, by disease stage and treatment modality, 1992 to 2003.  Value Health. 2011;14(1):41-52. doi:10.1016/j.jval.2010.10.006PubMedGoogle ScholarCrossref
18.
US Centers for Medicare and Medicaid Services. Details for title: FY 2018 final rule, correction notice, and notice tables. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/FY2018-IPPS-Final-Rule-Home-Page-Items/FY2018-IPPS-Final-Rule-Tables.html. Accessed March 14, 2018.
19.
Schweikert  B, Hunger  M, Meisinger  C, König  HH, Gapp  O, Holle  R.  Quality of life several years after myocardial infarction: comparing the MONICA/KORA registry to the general population.  Eur Heart J. 2009;30(4):436-443. doi:10.1093/eurheartj/ehn509PubMedGoogle ScholarCrossref
20.
Gold  MR, Franks  P, McCoy  KI, Fryback  DG.  Toward consistency in cost-utility analyses: using national measures to create condition-specific values.  Med Care. 1998;36(6):778-792. doi:10.1097/00005650-199806000-00002PubMedGoogle ScholarCrossref
21.
Hanmer  J, Lawrence  WF, Anderson  JP, Kaplan  RM, Fryback  DG.  Report of nationally representative values for the noninstitutionalized US adult population for 7 health-related quality-of-life scores.  Med Decis Making. 2006;26(4):391-400. doi:10.1177/0272989X06290497PubMedGoogle ScholarCrossref
22.
Tsevat  J, Goldman  L, Soukup  JR,  et al.  Stability of time-tradeoff utilities in survivors of myocardial infarction.  Med Decis Making. 1993;13(2):161-165. doi:10.1177/0272989X9301300210PubMedGoogle ScholarCrossref
23.
Pettersen  KI, Kvan  E, Rollag  A, Stavem  K, Reikvam  A.  Health-related quality of life after myocardial infarction is associated with level of left ventricular ejection fraction.  BMC Cardiovasc Disord. 2008;8:28. doi:10.1186/1471-2261-8-28PubMedGoogle ScholarCrossref
24.
Black  WC, Gareen  IF, Soneji  SS,  et al; National Lung Screening Trial Research Team.  Cost-effectiveness of CT screening in the National Lung Screening Trial.  N Engl J Med. 2014;371(19):1793-1802. doi:10.1056/NEJMoa1312547PubMedGoogle ScholarCrossref
25.
US Bureau of Labor Statistics. CPI inflation calculator. https://www.bls.gov/data/inflation_calculator.htm. Accessed March 14, 2018.
26.
Sanders  GD, Neumann  PJ, Basu  A,  et al.  Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: Second Panel on Cost-Effectiveness in Health and Medicine.  JAMA. 2016;316(10):1093-1103. doi:10.1001/jama.2016.12195PubMedGoogle ScholarCrossref
27.
US Congressional Budget Office. Prices for brand-name drugs under selected federal programs. https://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/64xx/doc6481/06-16-prescriptdrug.pdf. Accessed February 21, 2018.
28.
Cipriano  LE, Romanus  D, Earle  CC,  et al.  Lung cancer treatment costs, including patient responsibility, by disease stage and treatment modality, 1992 to 2003.  Value Health. 2011;14(1):41-52. doi:10.1016/j.jval.2010.10.006PubMedGoogle ScholarCrossref
29.
Eddy  DM, Hollingworth  W, Caro  JJ, Tsevat  J, McDonald  KM, Wong  JB; ISPOR−SMDM Modeling Good Research Practices Task Force.  Model transparency and validation: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force–7.  Value Health. 2012;15(6):843-850. doi:10.1016/j.jval.2012.04.012PubMedGoogle ScholarCrossref
30.
Kazi  DS, Penko  J, Coxson  PG,  et al.  Updated cost-effectiveness analysis of PCSK9 inhibitors based on the results of the FOURIER trial.  JAMA. 2017;318(8):748-750. doi:10.1001/jama.2017.9924PubMedGoogle ScholarCrossref
31.
Fonarow  GC, Keech  AC, Pedersen  TR,  et al.  Cost-effectiveness of evolocumab therapy for reducing cardiovascular events in patients with atherosclerotic cardiovascular disease.  JAMA Cardiol. 2017;2(10):1069-1078. doi:10.1001/jamacardio.2017.2762PubMedGoogle ScholarCrossref
32.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
33.
Ridker  PM.  How common is residual inflammatory risk?  Circ Res. 2017;120(4):617-619. doi:10.1161/CIRCRESAHA.116.310527PubMedGoogle ScholarCrossref
34.
Ridker  PM, Everett  BM, Pradhan  A,  et al; CIRT Investigators.  Low-dose methotrexate for the prevention of atherosclerotic events  [published online November 10, 2018].  N Engl J Med. doi:10.1056/NEJMoa1809798PubMedGoogle Scholar
35.
Hlatky  MA, Owens  DK, Sanders  GD.  Cost-effectiveness as an outcome in randomized clinical trials.  Clin Trials. 2006;3(6):543-551. doi:10.1177/1740774506073105PubMedGoogle ScholarCrossref
Original Investigation
January 16, 2019

Cost-effectiveness of Canakinumab for Prevention of Recurrent Cardiovascular Events

Author Affiliations
  • 1Department of Cardiovascular Epidemiology and Research, The Danish Heart Foundation, Copenhagen, Denmark
  • 2Department of Cardiology, Copenhagen University Hospital Herlev and Gentofte, Copenhagen, Denmark
  • 3Department of Health Research and Policy, Stanford University School of Medicine, Stanford, California
  • 4Department of Medicine, Stanford University, Stanford, California
  • 5Department of Management Science and Engineering, Stanford University, Stanford, California
  • 6VA Palo Alto Health Care System, Palo Alto, California
  • 7Stanford Health Policy, Centers for Health Policy and Primary Care and Outcomes Research, Stanford University, Stanford, California
JAMA Cardiol. 2019;4(2):128-135. doi:10.1001/jamacardio.2018.4566
Key Points

Question  What is the cost-effectiveness of canakinumab for the secondary prevention of cardiovascular events?

Findings  In this cost-effectiveness analysis, the outcomes of patients with a previous myocardial infarction treated either with canakinumab therapy added to standard of care or with standard of care alone were simulated. The addition of canakinumab was estimated to increase quality-adjusted life-years (QALYs) by 0.13 and costs by $832 000, yielding an incremental cost-effectiveness ratio of $6.4 million per QALY gained.

Meaning  Canakinumab is not cost-effective at its current market price.

Abstract

Importance  In the Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) trial, the anti-inflammatory monoclonal antibody canakinumab significantly reduced the risk of recurrent cardiovascular events in patients with previous myocardial infarction (MI) and high-sensitivity C-reactive protein (hs-CRP) levels of 2 mg/L or greater.

Objective  To estimate the cost-effectiveness of adding canakinumab to standard of care for the secondary prevention of major cardiovascular events over a range of potential prices.

Design, Setting, and Participants  A state-transition Markov model was constructed to estimate costs and outcomes over a lifetime horizon by projecting rates of recurrent MI, coronary revascularization, infection, and lung cancer with and without canakinumab treatment. We used a US health care sector perspective, and the base case used the current US market price of canakinumab of $73 000 per year. A hypothetical cohort of patients after MI aged 61 years with an hs-CRP level of 2 mg/L or greater was constructed.

Interventions  Canakinumab, 150 mg, administered every 3 months plus standard of care compared with standard of care alone.

Main Outcomes and Measures  Lifetime costs and quality-adjusted life-years (QALYs), discounted at 3% annually.

Results  Adding canakinumab to standard of care increased life expectancy from 11.31 to 11.36 years, QALYs from 9.37 to 9.50, and costs from $242 000 to $1 074 000, yielding an incremental cost-effectiveness ratio of $6.4 million per QALY gained. The price would have to be reduced by more than 98% (to $1150 per year or less) to meet the $100 000 per QALY willingness-to-pay threshold. These results were generally robust across alternative assumptions, eg, substantially lower health-related quality of life after recurrent cardiovascular events, lower infection rates while receiving canakinumab, and reduced all-cause mortality while receiving canakinumab. Including a potential beneficial effect of canakinumab on lung cancer incidence improved the incremental cost-effectiveness ratio to $3.5 million per QALY gained. A strategy of continuing canakinumab selectively in patients with reduction in hs-CRP levels to less than 2 mg/L would have a cost-effectiveness ratio of $819 000 per QALY gained.

Conclusions and Relevance  Canakinumab is not cost-effective at current US prices for prevention of recurrent cardiovascular events in patients with a prior MI. Substantial price reductions would be needed for canakinumab to be considered cost-effective.

Introduction

The effectiveness of reducing inflammation as a treatment of atherosclerotic disease has been debated for decades.1 The recent Canakinumab Anti-inflammatory Thrombosis Outcome Study (CANTOS) trial provided proof-of-concept evidence that targeting inflammation can reduce cardiovascular events.2 Patients with elevated high-sensitivity C-reactive protein (hs-CRP) levels after myocardial infarction (MI) in the CANTOS trial were randomly assigned to either placebo or treatment with the human monoclonal antibody canakinumab at 1 of 3 doses.2 Patients in the CANTOS trial treated with canakinumab, 150 mg, every 3 months had a 15% lower risk of the primary outcome (the composite of nonfatal MI, nonfatal stroke, or cardiovascular death) compared with placebo, which was mostly driven by reduced risks of recurrent MI, as there was no reduction in mortality or stroke. The trial also found a significant increase in fatal infections among canakinumab-treated patients but suggested that canakinumab may reduce the incidence of lung cancer, an intriguing finding that remains to be confirmed.3

Canakinumab, which inhibits interleukin 1β in the inflammatory cascade, is currently approved to treat uncommon autoimmune disorders, such as systemic juvenile idiopathic arthritis and periodic fever syndromes. Its current US market price is approximately $73 000 per year for 150 mg administered every 3 months.4 If canakinumab were approved by the US Food and Drug Administration (FDA) as a life-long therapy for coronary artery disease, the budget impact would be enormous. Canakinumab’s cost-effectiveness for this indication is unknown.

In this study, we sought to assess the cost-effectiveness of adding canakinumab to standard of care for post-MI management and to explore its value over a range of potential prices. Further, we aimed to estimate the cost-effectiveness of a selective strategy of limiting canakinumab therapy to patients who have a substantial reduction in hs-CRP levels after the first dose, as suggested by a post hoc analysis of the CANTOS trial.5 Finally, we wished to investigate the effect of the potential reduction in lung cancer on the cost-effectiveness of canakinumab.3

Methods
Model Structure, Treatments, and Target Population

We developed a state-transition Markov model to evaluate the long-term costs and outcomes of canakinumab using a US health care sector perspective and a lifetime horizon (eFigure 1 in the Supplement). The model simulates outcomes for a hypothetical cohort of patients aged 61 years with previous MI and an hs-CRP level of 2 mg/L or greater who receive post-MI standard of care with or without canakinumab, 150 mg, subcutaneously every 3 months. Institutional review board approval was not needed, as all analyses were conducted on publicly available data and no human participants were involved in the study.

In the model, each month, a patient can either remain stable, experience a recurrent MI, undergo a coronary revascularization procedure, develop lung cancer, experience an infection, or die of causes other than MI, revascularization, lung cancer, or infection. After recovery from either revascularization or infection, patients are assumed to return to their prior health state, with the same quality of life and risk of subsequent events. To account for lung cancer’s substantial effect on mortality, quality of life, and costs, individuals with lung cancer who survive recurrent MI, revascularization, or infection return to the stable with lung cancer state.

We based model inputs on the CANTOS trial and the literature (Table 1), adjusted costs to 2018 US dollars using the Consumer Price Index,25 and applied a 3% annual discount rate and half-cycle correction to all outcomes. We used TreeAge Pro 2017 (TreeAge Software) and R version 3.4.3 (The R Foundation), and completed the Impact Inventory from the Second Panel on Cost-effectiveness in Health and Medicine (eTable 1 in the Supplement).26

Mortality, Treatment Effects, and Adverse Events

We used the age-specific and sex-specific mortality rates from the US Centers for Disease Control and Prevention 2014 life tables, adjusted by an annual rate ratio to match the background rates of death from causes other than MI, revascularization, lung cancer, and infection in the placebo arm of the CANTOS trial.9 As the CANTOS trial did not show a significant effect on mortality (hazard ratio, 0.92; 95% CI, 0.78-1.09), we assumed no direct effect of canakinumab on all-cause mortality but did account for the indirect effects on mortality resulting from changes in other clinical outcomes (eg, cardiovascular events, infection, and lung cancer).

We assumed canakinumab reduced the risk of recurrent MI or requiring a revascularization procedure according to the hazard ratios reported in the CANTOS trial.2 We corrected for the overlap between the events of MI and any revascularization procedure in the CANTOS trial by subtracting the incidence rate of MI from the incidence rate for any revascularization.2 We assumed no effect on stroke. To account for an assumed increased prevalence of heart failure following a recurrent MI and consequently higher long-term mortality rates, we applied a multiplier of 2.21 to the mortality rate in this state. This was derived by assuming heart failure would develop in 20% of patients with recurrent MI with a 5-year survival probability of 35%, based on data from the Rotterdam Study.10 We also included additional short-term (30-day) mortality risks following MI admission and from undergoing a revascularization procedure.6,7

We included the rates of serious infections and infection mortality as reported by the CANTOS trial as well as the effects on rates of arthritis (including osteoarthritis and gout).2 We excluded the proposed beneficial effect of canakinumab on preventing incident lung cancer from the base case analysis3 and instead assumed the rates of lung cancer would be the same in both arms, corresponding to the rate in the placebo arm. In scenario analyses, we explored the trial-observed lung cancer benefits, using time-dependent lung cancer death rates based on Surveillance, Epidemiology, and End Results Program (SEER) data.8 We extrapolated beyond the trial follow-up by assuming all rates would remain constant and that the effects of ongoing canakinumab treatment in reducing these rates would continue over a patient’s remaining lifetime.

Costs

We derived background monthly medical costs from age-specific and sex-specific data on US health spending by subtracting the proportion of costs spent on treating ischemic heart disease in a population with the same sex distribution as the CANTOS trial population (approximately 25% women).11,12 We estimated the price for canakinumab, 150 mg, by adjusting Veteran Affairs costs reported in the US Federal Supply Schedule by 152%,4,27 further adding costs for an outpatient physician visit and a subcutaneous injection every 3 months, based on the US Centers for Medicare and Medicaid Services (CMS) physician fee schedule.13 We based costs of hospital admission for MI, revascularization procedure, and infection on 2018 CMS reimbursement rates (eTable 2 in the Supplement).

We applied previously published costs for standard of care treatment of coronary heart disease,14 adding an increase in costs following recurrent MI to account for the assumed development of heart failure in 20% of patients after MI.16 To capture the costs of improved arthritis symptoms (including osteoarthritis and gout, as defined in the CANTOS trial2) in the canakinumab-treated group, an average incremental monthly cost of arthritis treatment was included only in the standard of care group.15 We obtained costs of treating lung cancer from a study linking SEER statistics to CMS reimbursements.28 Since the CANTOS trial did not provide data on lung cancer stages, we assumed an average treatment cost corresponding to radiotherapy and chemotherapy for non–small cell lung cancer, the most prevalent type of lung cancer in the United States, with weighting on stages corresponding to SEER reports.8

Utilities

On the basis of age-weighted and sex-weighted EQ-5D (EuroQol Research Foundation) index scores from a community registry, we used baseline utility values of 0.872 for patients treated with canakinumab to reflect the quality-of-life decrement of living with ischemic heart disease.19 Baseline utility in the standard of care arm was further reduced to 0.868 to account for increased rates of arthritis symptoms based on the difference in incidence rates of arthritis between the standard of care and canakinumab, 150 mg, arms in the CANTOS trial2 combined with the utility decrement of living with arthritis.20 For individuals experiencing a recurrent MI, we added a permanent reduction in quality of life of −0.134, including a disutility of −0.07 in the 20% of patients assumed to develop heart failure.22,23 We applied short-term quality-of-life reductions for the duration of hospitalization for MI, revascularization, and infection and obtained time-dependent utilities after lung cancer diagnosis from the National Lung Screening Trial.24 Finally, to account for the decrease in quality of life associated with aging, we applied disutilities for patients older than 70 years and 80 years, based on a large US study.21

Model Validity

To assess the validity of our model, we compared the risk of all-cause mortality and nonfatal events from the CANTOS trial with the risk from our simulated cohort at 3.7 years of elapsed time (the median follow-up in the CANTOS trial).29 Additionally, we compared the relative risk of mortality between the 2 treatment strategies in our simulated cohort with the relative risk reported in the CANTOS trial.

Sensitivity and Scenario Analyses

We performed 1-way deterministic sensitivity analyses on all input parameters by varying each individually through plausible ranges, particularly the cost of canakinumab and its association with recurrent MI, revascularization, and infection. Additionally, we used a 2-way sensitivity analysis to explore a direct mortality risk reduction vs the cost of canakinumab.

While our base case was constructed to match the CANTOS trial population, we performed several scenario analyses to evaluate the effect of canakinumab in real-world population settings and within subpopulations of the trial. First, we investigated the cost-effectiveness of limiting canakinumab therapy to responders, defined as patients who had a reduction in hs-CRP level to less than 2 mg/L at 3 months following treatment initiation compared with standard of care. A post hoc analysis of the CANTOS trial suggested that this group (55% of the patients receiving the 150-mg dose) had a 31% risk reduction in all-cause mortality compared with patients receiving standard of care only.5 Importantly, responders and nonresponders did not differ significantly in baseline clinical characteristics. For this scenario analysis, we assumed that all canakinumab-treated patients were responders and included a 31% all-cause mortality reduction in the canakinumab arm compared with standard of care. We assumed that the rates of MI, revascularization, infections, and lung cancer to be the same as in the base case, as no data on these rates among responders have been reported individually.5 To account for the initial drug cost of determining whether a patient is a responder, we added an extra 45% cost of the first canakinumab dose, corresponding to the proportion of nonresponders reported in the CANTOS trial. We also repeated deterministic 1-way sensitivity analyses, particularly on the cost of canakinumab. Second, we performed a scenario analysis including the proposed beneficial effect of canakinumab on lung cancer incidence. Third, we simulated multiple cohorts with different risks of recurrent MI, ranging from 1% to 10% per year, as well as different rates of developing heart failure following a recurrent MI, ranging from 10% to 30%. Fourth, we simulated cohorts with different ages at initiation of therapy, ranging from 40 to 70 years. Fifth, we also performed a sensitivity analysis varying the quality-of-life decrement after recurrent MI between 0 and 0.20. Lastly, we performed a sensitivity analysis excluding the decrement in quality of life associated with aging.

Finally, we conducted probabilistic sensitivity analysis for both the base case and for the scenario of only treating responders. We ran 10 000 second-order Monte Carlo simulations, simultaneously sampling from the uncertainty distributions of all model inputs (eTable 3 in the Supplement). To describe the remaining model uncertainty at different potential prices for canakinumab, we performed 10 additional probabilistic sensitivity analyses of 1000 simulations each, varying the canakinumab price from $100 to $10 000 per year.

Results
Model Validity

The cumulative mortality estimates from our model were within the 95% CIs of the trial estimates at 3.7 years of follow-up for both the standard of care group and the canakinumab-treated groups. The model mortality estimate for the canakinumab group was 10.1% in our model vs 9.6% (95% CI, 8.4-10.8) in the CANTOS trial. For the standard of care group, the model mortality estimate was 9.9% in our model vs 10.4% (95% CI, 9.4-11.4) in the CANTOS trial. For the scenario analysis including an effect on lung cancer, the model mortality estimate was 9.7% and 9.9% for canakinumab and standard of care, respectively.

Base Case Analysis

Adding canakinumab to standard of care increased discounted life expectancy from 11.31 to 11.36 years, which was driven principally by the decreased risk of recurrent MI in the canakinumab group. After accounting for the effect on quality of life, canakinumab increased the discounted lifetime quality-adjusted life-years (QALYs) from 9.37 to 9.50, again mainly driven by limiting decrements in health-related quality of life resulting from recurrent MI. At canakinumab’s current price, mean discounted lifetime costs were $242 000 for standard of care and $1 074 000 for standard of care plus canakinumab, yielding an incremental cost-effectiveness ratio of $6.4 million per QALY gained (Table 2).

Sensitivity and Scenario Analyses

The price of canakinumab had the strongest impact on its cost-effectiveness relative to standard of care. For canakinumab to be cost-effective at the $150 000 per QALY willingness-to-pay threshold, its annual price would have to be reduced by more than 95% to less than $1725 per year (Figure 1). The annual price would have to be reduced to less than $1150 per year to meet the $100 000 per QALY threshold and to less than $575 per year to meet the $50 000 per QALY threshold.

Model results were generally robust to alternative assumptions (Table 2). For example, canakinumab was not cost-effective at current prices even if averted recurrent cardiovascular events were associated with a much lower quality of life or if the rates of infection with canakinumab were substantially lower than observed in the CANTOS trial. Including the suggested effect of canakinumab on lung cancer increased the QALYs gained from 0.13 to 0.24 and resulted in a more favorable incremental cost-effectiveness ratio (ICER) of $3.5 million per QALY gained.

In the scenario in which canakinumab treatment was continued only in responders to the first dose, mortality would be reduced by 31%, QALYs would improve from 9.37 to 10.56 QALYs, lifetime costs per patient would be increased by $974 000, and the ICER would improve to $819 000 per QALY gained (Table 2) (Figure 2). Assuming these results were confirmed and achievable in general practice, the current price of canakinumab would need to be lowered to $6575 per year to achieve a $100 000 per QALY threshold (eTable 4 in the Supplement).

In 2-way sensitivity analysis, canakinumab was not cost-effective at any level of mortality reduction with an annual cost greater than $40 000, and substantial price reductions would be needed even if canakinumab reduced mortality modestly (eFigure 2 in the Supplement). Excluding quality-of-life decrements associated with aging did not alter the ICER (eTable 5 in the Supplement). Older patients had higher incremental benefits and more favorable ICERs from canakinumab therapy than younger patients, but these results were especially sensitive to changes in the risk of fatal infections (eTable 5 in the Supplement). The CANTOS trial reported that patients dying from infection tended to be older.2 Similarly, treating populations with a higher incidence of recurrent MI than reported in the CANTOS trial would make canakinumab’s ICER more favorable, although still far from commonly accepted thresholds even with an annual MI risk of 8% (eFigure 3 in the Supplement).

Probabilistic Sensitivity Analysis

At current prices, canakinumab was not cost-effective in any of the 10 000 simulations, even at a willingness-to-pay threshold of $150 000 per QALY (eFigure 4 in the Supplement). The strategy of limiting treatment to canakinumab responders was also not cost-effective in any of the 10 000 simulations. However, changing the drug price would, as expected, change the outcome of the probabilistic sensitivity analysis; at a yearly drug cost of $1125 per year, canakinumab was cost-effective at the $100 000 per QALY threshold in approximately 50% of simulations (Figure 3).

Discussion

The addition of canakinumab to standard of care treatment after an acute MI appears to modestly increase life expectancy and QALYs but does not meet accepted benchmarks for cost-effectiveness because of its high cost of $73 000 per year. We project that canakinumab might provide an acceptable level of cost-effectiveness if its price were less than $1725 per year (a reduction of more than 95%). These results were consistent across a range of assumptions regarding the potential health benefits of canakinumab.

The price of canakinumab is the most important factor determining its cost-effectiveness. If canakinumab is approved by the FDA for secondary prevention of cardiovascular disease, it would have a much larger market, so a high orphan drug price would no longer be appropriate. The annual cost of proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors, another monoclonal antibody used to reduce atherosclerotic cardiovascular events in high-risk populations, is much lower, approximately $14 000, but even at this price, PCSK9 inhibitors are not cost-effective at standard willingness-to-pay thresholds.14,30,31

The cost-effectiveness of medications can be improved by identifying subgroups of patients who have greater than average reductions in clinical events. The CANTOS investigators reported that patients who had a robust response to the anti-inflammatory effects of canakinumab (indicated by a decrease in hs-CRP level to less than 2 mg/L) had a larger clinical benefit, with a potential 31% decrease in all-cause mortality.5 This finding suggests that a clinical strategy of limiting canakinumab therapy to responders might improve both its clinical effectiveness and its cost-effectiveness. This strategy would also reduce the budget impact of canakinumab because fewer patients would continue long-term treatment. While the post hoc analysis of the CANTOS trial should be interpreted with caution, our model suggests that limiting canakinumab treatment to responders would still not be cost-effective ($819 000 per QALY gained) at its current price. The price of canakinumab would still need to be reduced substantially to reach the accepted benchmarks for cost-effectiveness in the selective treatment strategy, despite the plausibility of using a targeted treatment strategy to improve value.

An unexpected finding in the CANTOS trial was the favorable effect of canakinumab on incident lung cancer. This unforeseen finding has been attributed to the anti-inflammatory effect of canakinumab reducing the rate of progression, invasiveness, and metastatic spread of lung cancers that were present but undiagnosed at trial entry.3 Including the proposed effect on lung cancer in a scenario analysis almost doubled the projected QALYs gained from canakinumab and improved the cost-effectiveness ratio from $6.4 million to $3.5 million per QALY gained, which is still far above accepted willingness-to-pay thresholds. Thus, further research is needed to assess whether canakinumab affects cancer progression and mortality.

Since approximately 1.4 million Americans and Europeans experience an MI each year32 and a large part of this population has residual inflammatory risk,33 many patients would be potentially eligible to receive canakinumab. Treatment of this large population with canakinumab would have substantial budgetary impact at the current price. Even with a 90% price reduction, the budget impact of canakinumab used in secondary prevention would be substantial for a US or European health care system. Such a major budgetary impact can only be justified by robust evidence of health benefits and pricing that offers good value for money in health care.

The positive results from the CANTOS trial provide support for the hypothesis that therapies aimed specifically at reducing inflammation might improve cardiovascular outcomes. It is likely that additional treatments will be developed that target inflammation in patients with atherosclerosis. Of note, the recently published National Heart, Lung, and Blood Institute–funded Cardiovascular Inflammation Reduction Trial34 tested whether low-dose methotrexate reduced major cardiovascular events in patients with coronary disease and either type 2 diabetes or metabolic syndrome. This trial was terminated early and found no difference between methotrexate and placebo.

Limitations

Our model had several important limitations, and its results should be interpreted within the context of data inputs and modeling assumptions. First, to our knowledge, only 1 randomized clinical trial2 has evaluated the clinical effectiveness of canakinumab in treating cardiovascular disease, which limits generalizability of our results. Furthermore, the median follow-up in the CANTOS trial was 3.7 years, but we assumed therapy would be effective for a much longer period, a potentially optimistic estimate of the long-term cardiovascular benefits of canakinumab.35 Conversely, rates of adverse events, such as fatal infections, were also extrapolated over a life-time, without any assumption of termination of therapy in case of a serious infection. In the 3.7-year base case model mortality validation, canakinumab and standard of care had similar mortality rates. Nevertheless, these numbers were still within the CANTOS trial’s 95% CIs, and during the total duration of the model, the canakinumab strategy had a beneficial effect on mortality.

Most importantly, while awaiting potential FDA approval and a possible reduction in drug price, we used the current high orphan drug price of canakinumab in our base case analysis. We addressed this issue by repeating our analyses at a wide range of potential prices of canakinumab and by including multiple scenarios with a reduced drug cost. In the scenario limiting canakinumab treatment to responders only, we assumed all event rates besides mortality would be identical to the base case since specific rates of MI, revascularization, and lung cancer were not reported in the post hoc analysis of the CANTOS trial.5 While this approach might underestimate the effect of canakinumab in the responders scenario if its effects on MI, revascularization, and lung cancer prove to be more beneficial than in the principal analysis, the 31% reduction in mortality would remain the predominant driver in reducing the ICER compared with the base case analysis.

Conclusions

At the current price of $73 000 per year, adding canakinumab to standard therapy for secondary prevention of cardiovascular events in patients after MI with an elevated hs-CRP level is not cost-effective. To provide value at generally accepted standards for cost-effectiveness, the price of canakinumab would need to be decreased substantially.

Back to top
Article Information

Accepted for Publication: November 15, 2018.

Corresponding Authors: Thomas S. G. Sehested, MD (tsg.sehested@gmail.com), and Jenny Bjerre, MD (jennybjerre@gmail.com), Department of Cardiology, Copenhagen University Hospital Herlev and Gentofte, Kildegaardsvej 28, 8, 3 Post 635, DK-2900 Copenhagen, Denmark.

Published Online: January 16, 2019. doi:10.1001/jamacardio.2018.4566

Author Contributions: Drs Sehested and Bjerre had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Sehested and Bjerre contributed equally.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Sehested, Bjerre, Ku, Chang, Jahansouz, Owens, Goldhaber-Fiebert.

Drafting of the manuscript: Sehested, Bjerre, Ku, Chang.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Sehested, Bjerre, Ku, Chang, Jahansouz, Goldhaber-Fiebert.

Administrative, technical, or material support: Hlatky.

Study supervision: Owens, Hlatky, Goldhaber-Fiebert.

Conflict of Interest Disclosures: Dr Sehested has received grants from the Lundbeck Foundation. Dr Bjerre has received grants from the Danish Society of Cardiology and the Danish Heart Foundation. No other disclosures were reported.

References
1.
Libby  P.  Interleukin-1 beta as a target for atherosclerosis therapy: biological basis of CANTOS and beyond.  J Am Coll Cardiol. 2017;70(18):2278-2289. doi:10.1016/j.jacc.2017.09.028PubMedGoogle ScholarCrossref
2.
Ridker  PM, Everett  BM, Thuren  T,  et al; CANTOS Trial Group.  Antiinflammatory therapy with canakinumab for atherosclerotic disease.  N Engl J Med. 2017;377(12):1119-1131. doi:10.1056/NEJMoa1707914PubMedGoogle ScholarCrossref
3.
Ridker  PM, MacFadyen  JG, Thuren  T, Everett  BM, Libby  P, Glynn  RJ; CANTOS Trial Group.  Effect of interleukin-1β inhibition with canakinumab on incident lung cancer in patients with atherosclerosis: exploratory results from a randomised, double-blind, placebo-controlled trial.  Lancet. 2017;390(10105):1833-1842. doi:10.1016/S0140-6736(17)32247-XPubMedGoogle ScholarCrossref
4.
US Office of Procurement, Acquisition and Logistics (OPAL). Pharmaceutical prices. https://www.va.gov/oal/business/fss/pharmPrices.asp. Accessed March 14, 2018.
5.
Ridker  PM, MacFadyen  JG, Everett  BM, Libby  P, Thuren  T, Glynn  RJ; CANTOS Trial Group.  Relationship of C-reactive protein reduction to cardiovascular event reduction following treatment with canakinumab: a secondary analysis from the CANTOS randomised controlled trial.  Lancet. 2018;391(10118):319-328. doi:10.1016/S0140-6736(17)32814-3PubMedGoogle ScholarCrossref
6.
Myerson  M, Coady  S, Taylor  H, Rosamond  WD, Goff  DC  Jr; ARIC Investigators.  Declining severity of myocardial infarction from 1987 to 2002: the Atherosclerosis Risk in Communities (ARIC) study.  Circulation. 2009;119(4):503-514. doi:10.1161/CIRCULATIONAHA.107.693879PubMedGoogle ScholarCrossref
7.
Peterson  ED, Dai  D, DeLong  ER,  et al; NCDR Registry Participants.  Contemporary mortality risk prediction for percutaneous coronary intervention: results from 588,398 procedures in the National Cardiovascular Data Registry.  J Am Coll Cardiol. 2010;55(18):1923-1932. doi:10.1016/j.jacc.2010.02.005PubMedGoogle ScholarCrossref
8.
Surveillance, Epidemiology, and End Results Program. Previous version: SEER Cancer Statistics Review, 1975-2013. https://seer.cancer.gov/csr/1975_2013/. Accessed December 1, 2017.
9.
Arias  E, Heron  M, Xu  J.  United States life tables, 2014.  Natl Vital Stat Rep. 2017;66(4):1-64.PubMedGoogle Scholar
10.
Bleumink  GS, Knetsch  AM, Sturkenboom  MC,  et al.  Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk and prognosis of heart failure: the Rotterdam Study.  Eur Heart J. 2004;25(18):1614-1619. doi:10.1016/j.ehj.2004.06.038PubMedGoogle ScholarCrossref
11.
Lassman  D, Hartman  M, Washington  B, Andrews  K, Catlin  A.  US health spending trends by age and gender: selected years 2002-10.  Health Aff (Millwood). 2014;33(5):815-822. doi:10.1377/hlthaff.2013.1224PubMedGoogle ScholarCrossref
12.
Dieleman  JL, Baral  R, Birger  M,  et al.  US spending on personal health care and public health, 1996-2013.  JAMA. 2016;316(24):2627-2646. doi:10.1001/jama.2016.16885PubMedGoogle ScholarCrossref
13.
US Centers for Medicare and Medicaid Services. Physician fee schedule search. https://www.cms.gov/apps/physician-fee-schedule/license-agreement.aspx. Accessed March 14, 2018.
14.
Kazi  DS, Moran  AE, Coxson  PG,  et al.  Cost-effectiveness of PCSK9 inhibitor therapy in patients with heterozygous familial hypercholesterolemia or atherosclerotic cardiovascular disease.  JAMA. 2016;316(7):743-753. doi:10.1001/jama.2016.11004PubMedGoogle ScholarCrossref
15.
Murphy  LB, Cisternas  MG, Pasta  DJ, Helmick  CG, Yelin  EH.  Medical expenditures and earnings losses among US adults with arthritis in 2013.  Arthritis Care Res (Hoboken). 2018;70(6):869-876. doi:10.1002/acr.23425PubMedGoogle ScholarCrossref
16.
Voigt  J, Sasha John  M, Taylor  A, Krucoff  M, Reynolds  MR, Michael Gibson  C.  A reevaluation of the costs of heart failure and its implications for allocation of health resources in the United States.  Clin Cardiol. 2014;37(5):312-321. doi:10.1002/clc.22260PubMedGoogle ScholarCrossref
17.
Cipriano  LE, Romanus  D, Earle  CC,  et al.  Lung cancer treatment costs, including patient responsibility, by disease stage and treatment modality, 1992 to 2003.  Value Health. 2011;14(1):41-52. doi:10.1016/j.jval.2010.10.006PubMedGoogle ScholarCrossref
18.
US Centers for Medicare and Medicaid Services. Details for title: FY 2018 final rule, correction notice, and notice tables. https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatientPPS/FY2018-IPPS-Final-Rule-Home-Page-Items/FY2018-IPPS-Final-Rule-Tables.html. Accessed March 14, 2018.
19.
Schweikert  B, Hunger  M, Meisinger  C, König  HH, Gapp  O, Holle  R.  Quality of life several years after myocardial infarction: comparing the MONICA/KORA registry to the general population.  Eur Heart J. 2009;30(4):436-443. doi:10.1093/eurheartj/ehn509PubMedGoogle ScholarCrossref
20.
Gold  MR, Franks  P, McCoy  KI, Fryback  DG.  Toward consistency in cost-utility analyses: using national measures to create condition-specific values.  Med Care. 1998;36(6):778-792. doi:10.1097/00005650-199806000-00002PubMedGoogle ScholarCrossref
21.
Hanmer  J, Lawrence  WF, Anderson  JP, Kaplan  RM, Fryback  DG.  Report of nationally representative values for the noninstitutionalized US adult population for 7 health-related quality-of-life scores.  Med Decis Making. 2006;26(4):391-400. doi:10.1177/0272989X06290497PubMedGoogle ScholarCrossref
22.
Tsevat  J, Goldman  L, Soukup  JR,  et al.  Stability of time-tradeoff utilities in survivors of myocardial infarction.  Med Decis Making. 1993;13(2):161-165. doi:10.1177/0272989X9301300210PubMedGoogle ScholarCrossref
23.
Pettersen  KI, Kvan  E, Rollag  A, Stavem  K, Reikvam  A.  Health-related quality of life after myocardial infarction is associated with level of left ventricular ejection fraction.  BMC Cardiovasc Disord. 2008;8:28. doi:10.1186/1471-2261-8-28PubMedGoogle ScholarCrossref
24.
Black  WC, Gareen  IF, Soneji  SS,  et al; National Lung Screening Trial Research Team.  Cost-effectiveness of CT screening in the National Lung Screening Trial.  N Engl J Med. 2014;371(19):1793-1802. doi:10.1056/NEJMoa1312547PubMedGoogle ScholarCrossref
25.
US Bureau of Labor Statistics. CPI inflation calculator. https://www.bls.gov/data/inflation_calculator.htm. Accessed March 14, 2018.
26.
Sanders  GD, Neumann  PJ, Basu  A,  et al.  Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: Second Panel on Cost-Effectiveness in Health and Medicine.  JAMA. 2016;316(10):1093-1103. doi:10.1001/jama.2016.12195PubMedGoogle ScholarCrossref
27.
US Congressional Budget Office. Prices for brand-name drugs under selected federal programs. https://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/64xx/doc6481/06-16-prescriptdrug.pdf. Accessed February 21, 2018.
28.
Cipriano  LE, Romanus  D, Earle  CC,  et al.  Lung cancer treatment costs, including patient responsibility, by disease stage and treatment modality, 1992 to 2003.  Value Health. 2011;14(1):41-52. doi:10.1016/j.jval.2010.10.006PubMedGoogle ScholarCrossref
29.
Eddy  DM, Hollingworth  W, Caro  JJ, Tsevat  J, McDonald  KM, Wong  JB; ISPOR−SMDM Modeling Good Research Practices Task Force.  Model transparency and validation: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force–7.  Value Health. 2012;15(6):843-850. doi:10.1016/j.jval.2012.04.012PubMedGoogle ScholarCrossref
30.
Kazi  DS, Penko  J, Coxson  PG,  et al.  Updated cost-effectiveness analysis of PCSK9 inhibitors based on the results of the FOURIER trial.  JAMA. 2017;318(8):748-750. doi:10.1001/jama.2017.9924PubMedGoogle ScholarCrossref
31.
Fonarow  GC, Keech  AC, Pedersen  TR,  et al.  Cost-effectiveness of evolocumab therapy for reducing cardiovascular events in patients with atherosclerotic cardiovascular disease.  JAMA Cardiol. 2017;2(10):1069-1078. doi:10.1001/jamacardio.2017.2762PubMedGoogle ScholarCrossref
32.
Benjamin  EJ, Blaha  MJ, Chiuve  SE,  et al; American Heart Association Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics-2017 update: a report from the American Heart Association.  Circulation. 2017;135(10):e146-e603. doi:10.1161/CIR.0000000000000485PubMedGoogle ScholarCrossref
33.
Ridker  PM.  How common is residual inflammatory risk?  Circ Res. 2017;120(4):617-619. doi:10.1161/CIRCRESAHA.116.310527PubMedGoogle ScholarCrossref
34.
Ridker  PM, Everett  BM, Pradhan  A,  et al; CIRT Investigators.  Low-dose methotrexate for the prevention of atherosclerotic events  [published online November 10, 2018].  N Engl J Med. doi:10.1056/NEJMoa1809798PubMedGoogle Scholar
35.
Hlatky  MA, Owens  DK, Sanders  GD.  Cost-effectiveness as an outcome in randomized clinical trials.  Clin Trials. 2006;3(6):543-551. doi:10.1177/1740774506073105PubMedGoogle ScholarCrossref
×