Chest ACCP Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (13)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Highland, K. B.
Right arrow Articles by Simpson, K. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Highland, K. B.
Right arrow Articles by Simpson, K. N.
(Chest. 2003;124:2087-2092.)
© 2003 American College of Chest Physicians

Treatment of Pulmonary Arterial Hypertension*

A Preliminary Decision Analysis

Kristin B. Highland, MD; Charlie Strange, MD, FCCP; Joe Mazur, PharmD and Kit N. Simpson, DrPH

* From the Division of Pulmonary, Critical Care, Allergy, and Clinical Immunology (Drs. Highland and Strange), College of Pharmacy Practice (Dr. Mazur), and College of Pharmacy (Dr. Simpson), Medical University of South Carolina, Charleston, SC.

Correspondence to: Kristin B. Highland, MD, Assistant Professor of Medicine Division of Pulmonary, Critical Care, Allergy, and Clinical Immunology, 96 Jonathan Lucas St, Suite 812 CSB, Charleston, SC 29425; e-mail: highlakb{at}musc.edu


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objective: New therapies for pulmonary arterial hypertension (PAH) improve functional status, quality of life (QOL), and survival. Clinicians must chose between very different therapies without the availability of comparison studies. We constructed a "virtual" clinical trial to help inform these treatment choices.

Design: We compare key outcomes related to survival, costs, and QOL using a Markov-type decision model to estimate the expected outcomes and costs for PAH patients treated for 1 year with bosentan and treprostinil compared to patients treated with epoprostenol, as well as patients treated with bosentan compared to those treated with treprostinil. The allowed transitions in the model were between World Health Organization functional class I to IV and death. Transition probabilities were based on observed transitions for bosentan. Treatment effect was estimated using 6-min walk data for treprostinil and epoprostenol. Utilities were calculated from estimated EuroQol health states. Cost was estimated from average wholesale price and Medicare reimbursement data. The effects of changing values of input variables on the key outcomes were calculated

Results: Treatment with bosentan compared to treatment with either epoprostenol or treprostinil was less costly and resulted in a greater gain in quality-adjusted life years (QALYs). Conversely, treprostinil was significantly more expensive than epoprostenol, without an appreciable gain in QALYs. These findings were not substantially affected by the reasonable adjustments of transition probabilities, utility values, or tachyphylaxis to epoprostenol.

Conclusion: Treatment with bosentan is more cost-effective than treatment with either treprostinil or epoprostenol. In addition, a net improvement in quality-adjusted survival may be expected.

Key Words: bosentan • decision analysis • epoprostenol • pulmonary arterial hypertension • treprostinil


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Pulmonary arterial hypertension (PAH) is characterized by a progressive elevation in pulmonary arterial pressures that eventually leads to right ventricular failure and death.1 2 3 Before the advent of newer medications to treat this disorder, a median survival time as short as 2.8 years has been recorded for patients with primary (ie, idiopathic) pulmonary hypertension (PPH).4 There is no known cure, and until recently, the conventional treatment for PAH

For editorial comment see page 2045

consisted of therapy with vasodilators,5 6 7 8 9 10 11 12 anticoagulant agents,13 inotropic agents,7 diuretics,14 oxygen,14 and transplantation.15 16 17

With better understanding of the pathobiological mechanisms of PAH, new therapies have become available that have been shown to improve functional status, quality of life (QOL), and survival. However, clinicians and patients now must choose between epoprostenol, treprostinil, and bosentan, three drugs with quite different modes of administration, costs, and risk of adverse events. This choice is complicated by the fact that there are no direct comparison studies for these three drugs. The objective of this study was to construct a "virtual" clinical trial to help inform these treatment choices, given what we currently know about PAH and the effects of these drugs. We compare key outcomes related to survival, costs, and QOL for cohorts of 100 patients using a Markov-type decision model to estimate the expected outcomes and costs for PAH patients treated for 1 year with bosentan and treprostinil compared to those treated with epoprostenol, as well as PAH patients treated with bosentan compared to those treated with treprostinil.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Epidemiologic Estimation
The estimation of life expectancy was carried out by Markov modeling.18 Five Markov states (ie, classes I to IV and death) were defined according to the functional classification developed by the World Health Organization (WHO) to allow comparisons of patients with respect to the clinical severity of the disease process. This classification was patterned after the New York Heart Association functional classification for heart disease.19

A 3-month Markov cycle period was selected for this study. The assumption was made that state transition probabilities are time-independent. The model was based on a cohort of 100 hypothetical individuals. The number of individuals in each state was calculated for more than four cycles (ie, 1 year). Figure 1 illustrates the transitions allowed in the model.



View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.. Allowed transitions between WHO functional classes and death for the model.

 
Survival Analysis
The survival analysis was performed by populating the Markov model with estimates of the 3-month transition probabilities. These probabilities were based on observed 3-month transitions between health states for bosentan (Table 1 ) reported by the authors.20 We assumed that these probabilities would be stable for the subsequent 9 months.


View this table:
[in this window]
[in a new window]

  		Table 1.. Observed Transition Probabilities for Bosentan*

 
Estimating Treatment Effect
The 6-min walk test (6MWT) distance21 22 and hemodynamic variables4 have been shown previously to be independent predictors of mortality in patients with PAH. The 6MWT correlates with hemodynamic variables21 23 and New York Heart Association/WHO functional class,22 and is commonly used to evaluate treatment efficacy for PAH.4 10 24 25

Table 2 compares the 6MWT data obtained from the primary studies of bosentan,20 treprostinil,26 and epoprostenol.21 Although the beginning 6MWT distances in the placebo and epoprostenol groups were not statistically different,21 when compared to those in the other two placebo groups (ie, the bosentan and treprostinil groups), the placebo group in the epoprostenol study started off with a lower 6MWT distance and deteriorated to a much greater degree than those of the other placebo groups. Therefore, we created an adjusted placebo group that was a weighted average of the 6MWT distance for the placebo groups in the bosentan and treprostinil studies, and we used this adjusted value in our model.


View this table:
[in this window]
[in a new window]

  		Table 2.. 6MWT Data for Drug vs Placebo and Relative Risk in Improvement in 6MWT after 3 mo of Drug Therapy*

 
Transition probabilities for treprostinil and epoprostenol were not available from the primary studies.21 26 Therefore, transition probabilities for treprostinil (Table 3 ) and epoprostenol (Table 4 ) were estimated based on the observed transition probabilities for bosentan20 adjusted by the relative risk of improvement in 6MWT observed for each respective drug (Table 2) .


View this table:
[in this window]
[in a new window]

  		Table 3.. Estimated Transition Probabilities for Treprostinil

 

View this table:
[in this window]
[in a new window]

  		Table 4.. Estimated Transition Probabilities for Epoprostenol

 
Utility Estimation
EuroQol health states27 were estimated for the various WHO functional classes. A group of clinical experts achieved consensus on the extent of limitations in the five dimensions of health, described by the EuroQol instrument,27 for each functional class. These health-state descriptions then were adjusted for the expected side effects associated with various treatments.

Dolan28 estimated utility values based on data from a cohort of individuals who had both standard gamble and EuroQol data available. We used a sample of the values from the Dolan study28 to calculate the utility weights for each health state in the model. An alternative set of utility values was estimated by increasing the state I estimates by 0.04 and each subsequent estimate by this factor plus 0.02. This allows minimum values of 0.10 for patients in state IV (Table 5 ).


View this table:
[in this window]
[in a new window]

  		Table 5.. Baseline and Alternative Utility Values for Each Functional Class per Treatment

 
Cost Estimation
The average wholesale price listed in the 2002 Drug Topics Red Book29 was used for drug and diluent costs. Costs for treprostinil and epoprostenol were based on the dose (9.3 and 9.2 ng/kg/min, respectively) used in their respective comparative study.21 26 The per diem cost of pumps for epoprostenol and treprostinil were obtained from their respective distributors. The Medicare reimbursement for pulmonary hypertension (DRG 144) was obtained by querying the hospital database of the Medical University of South Carolina. Provider and laboratory fees were obtained from average national Medicare reimbursement data published in 199630 and were adjusted for the year 2002 using the medical care portion of the consumer price index.31 A 3-day hospitalization for purposes of intensive teaching for patients receiving either treprostinil or epoprostenol was built into the model. Additional provider fees for the placement of a Hickman catheter during this hospitalization were added for patients receiving epoprostenol therapy. In addition, the cost of six home health visits was added to the costs of patients receiving treprostinil and epoprostenol, and the cost of monthly liver function tests30 31 was added to the costs of patients receiving bosentan. The costs per patient are shown in Table 6 .


View this table:
[in this window]
[in a new window]

  		Table 6.. Values Used for Cost

 
Economic Analysis
The economic analysis performed brought together the epidemiologic, cost, and treatment estimates described above. Cost savings, life-years saved, quality-adjusted life years (QALYs) expected were calculated for each of the three therapies.

The effects of changing the values of input variables on the outcomes of cost, life expectancy, cost-effectiveness, and cost utility ratios were calculated. The following sensitivity analyses were conducted:

  1. The original relative risk of improvement in 6MWT for epoprostenol was used instead of the adjusted relative risk (ie, the transition probabilities for epoprostenol were calculated using a relative risk of 1.52 instead of 1.16).
  2. An alternative set of utility values (Table 5) was used for all three drugs.
  3. Utility values for treprostinil and epoprostenol were assumed to be equivalent to that for bosentan; thus the "hassle factor" of drug delivery and potential additional side effects of the subcutaneous and IV delivery systems were ignored.
  4. The occurrence of tachyphylaxis as a result of epoprostenol therapy is well-known. In the epoprostenol study,21 patients initially received 5.3 ng/kg/min, with an increase to 9.2 ng/kg/min by the end of the 3-month study. Assuming a linear increase in dose to continue throughout the first year, the costs of the drug used in the model were adjusted accordingly.


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Results of the original model are shown in Table 7 . Treatment with bosentan compared to that with either epoprostenol or treprostinil was less costly and resulted in a greater gain in QALYs. Conversely, treatment with treprostinil was more expensive than that with epoprostenol without an appreciable gain in QALYs.


View this table:
[in this window]
[in a new window]

  		Table 7.. Base Estimate of Expected Annual Differences in Costs, Life-Years, and QALYs*

 
Results of the sensitivity analyses are shown in Table 8 . When using a relative risk of 6MWT improvement for epoprostenol therapy of 1.52 (calculated using the potentially biased original placebo group) instead of 1.16, a cost and QALYs savings persists for bosentan compared to treatment with either epoprostenol or treprostinil and for epoprostenol when compared to treprostinil. When using alternative utility values, a cost and QALYs savings is identical to the original model. The results of the model using utility values that were equalized across treatment groups show that patients receiving bosentan and epoprostenol continue to have cost and QALY savings compared to those receiving treprostinil. However, although therapy with bosentan remains less expensive than that with epoprostenol, there is essentially no difference in QALYs saved. When the model is adjusted for the occurrence of tachyphylaxis in response to epoprostenol, a cost and QALYs savings persists for treatment with bosentan compared to that with either epoprostenol or treprostinil, and for treatment with epoprostenol when compared to that with treprostinil.


View this table:
[in this window]
[in a new window]

  		Table 8.. Sensitivity of Cost, Life-Years, and QALYs Comparison Among Treatment Arms in Model*

 

   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Until recently, epoprostenol was the only available therapy for PAH that had been shown to improve functional status, QOL, and survival. Consequently, this therapy is perceived by many clinicians to be the "gold standard" to which other therapies should be compared. We developed a model to estimate, under varying assumptions, the expected differences in cost, survival, and QOL between epoprostenol and the two newer Food and Drug Administration-approved medications (ie, bosentan and treprostinil).

Our model showed that treatment with bosentan is more cost-effective than treatment with either treprostinil or epoprostenol. In addition, there was a gain in QALYs. Likewise, our model showed that treprostinil treatment is substantially more expensive than epoprostenol treatment, without an appreciable gain in QALYs. These findings were not substantially affected by reasonable adjustments of transition probabilities for epoprostenol treatment, utility values, or tachyphylaxis to epoprostenol.

However, medical decision models are simplified pictures, and will reflect only what is known about diseases and the effects of therapies. Furthermore, models are only as good as the data on which they are based, and the current data for the drugs compared here are sparse indeed. The specific limitations of this model center around the subjectivity of the WHO functional classes and the assignment of utility values to these classes for each therapy. The comparability of 6MWT distance improvements when the baseline 6MWT distance was lower for the epoprostenol study may have underestimated the treatment effect with this drug. This study was also limited by the different inclusion criteria of the three comparative studies as follows: the epoprostenol study21 included only patients with PPH; the bosentan study20 included patients with PPH and patients with PAH due to scleroderma and lupus; and the treprostinil study26 included patients with PPH, PAH due to connective tissue diseases, and PAH due to congenital systemic to pulmonary shunts. Another limitation of the study was our inability to utilize the survival data reported in the epoprostenol trial because these data were not presented in such a way that could be used to estimate transition probabilities. As a result, the population included in our model may not have the exact risk characteristics as those reported in the epoprostenol trial.21 This study is also limited by the lack of transition data on patients who withdrew (possibly due to clinical worsening) from the bosentan study.20 As a result, our model does not allow for the transition from WHO functional class III to IV. The transition probabilities estimated from the bosentan study20 are also limited by the small number of subjects (14 of a total of 144 subjects) who began the study in WHO functional class IV. Only additional trials will be able to test whether these transitions are adequate. Finally, the present study is limited by the lack of long-term mortality and functional class transition data for each therapy. However, until further research is available, we conclude that bosentan may be expected to offer the average PAH patient the most cost-effective treatment.

Our findings illustrate the fundamental tradeoffs that must be considered in choosing between these drugs. Individual patients may have very different utility values attached to living with PAH with continuous IV or subcutaneous infusions. To some patients, it is a nuisance, while to others it may seriously affect their QOL. Furthermore, the cost of the more expensive therapy may in some cases place a ruinous economic burden on patients and their families, while other patients with adequate insurance may be relatively unaffected by the added cost. Given the differences in the basic features of the treatments, different patients may have substantially different preferences for these therapies. With the minute differences in actual survival predicted by our model, it appears that until more and stronger medical evidence for choosing one specific drug is found, physicians and patients must discuss the relative advantages of these therapies. Our modeling results may serve as a means of focusing such discussions on the key tradeoffs involved.


   Footnotes
 
Abbreviations: 6MWT = 6-min walk test; PAH = pulmonary arterial hypertension; PPH = primary pulmonary hypertension; QALY = quality-adjusted life year; QOL = quality of life; WHO = World Health Organization

Received for publication December 17, 2002. Accepted for publication June 2, 2003.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Hughes, JD, Rubin, LJ (1986) Primary pulmonary hypertension: an analysis of 28 cases and a review of the literature. Medicine (Baltimore) 65,56-72[Medline]
  2. Rich, S Primary pulmonary hypertension. Prog Cardiovasc Dis 1988;31,205-238[CrossRef][ISI][Medline]
  3. Rich, S, Dantzker, DR, Ayres, SM, et al Primary pulmonary hypertension: a national prospective study. Ann Intern Med 1987;107,216-223[CrossRef][ISI][Medline]
  4. D’Alonzo, GE, Barst, RJ, Ayres, SM, et al Survival in patients with primary pulmonary hypertension: results from a national prospective registry. Ann Intern Med 1991;115,343-359[ISI][Medline]
  5. Reeves, JT, Groves, BM, Turkevich, D The case for treatment of selected patients with primary pulmonary hypertension. Am Rev Respir Dis 1986;134,342-346[ISI][Medline]
  6. Weir, EK, Rubin, LJ, Ayres, SM, et al The acute administration of vasodilators in primary pulmonary hypertension: experience from the National Institutes of Health Registry on Primary Pulmonary Hypertension. Am Rev Respir Dis 1989;140,1623-1630[ISI][Medline]
  7. Rich, S, Brundage, BH High-dose calcium channel-blocking therapy for pulmonary arterial pressure and regression in right ventricular hypertrophy. Circulation 1987;76,135-141[Abstract/Free Full Text]
  8. Rich, S, Kaufmann, E, Levy, PS The effect of high doses of calcium-channel blockers on survival in primary pulmonary hypertension. N Engl J Med 1992;327,76-81[Abstract]
  9. Rubin, LG, Nicod, P, Hillis, LD, et al Treatment of primary pulmonary hypertension with nifedipine: a hemodynamic and scintigraphic evaluation. Ann Intern Med 1983;99,433-438[ISI][Medline]
  10. D’Alonzo, GE, Gianotti, L, Dantzker, DR Noninvasive assessment of hemodynamic improvement during chronic vasodilator therapy in obliterative pulmonary hypertension. Am Rev Respir Dis 1986;133,380-384[ISI][Medline]
  11. Chan, NS, McLay, J, Kenmure, AC Reversibility of primary pulmonary hypertension during six years of treatment with oral diazoxide. Br Heart J 1987;57,207-209[Abstract/Free Full Text]
  12. Barst, RJ Pharmacologically induced pulmonary vasodilatation in children and young adults with primary pulmonary hypertension. Chest 1986;98,497-503
  13. Fuster, V, Steele, PM, Edwards, WD, et al Primary pulmonary hypertension: natural history and the importance of thrombosis. Circulation 1984;70,580-587[Abstract/Free Full Text]
  14. Morgan, JM, Griffiths, M, du Bois, RM, et al Hypoxic pulmonary vasoconstriction in systemic sclerosis and primary pulmonary hypertension. Chest 1991;99,551-556[Abstract/Free Full Text]
  15. Reitz, BA, Wallwork, JL, Hunt, SA, et al Heart-lung transplantation: successful therapy for patients with pulmonary vascular disease. N Engl J Med 1982;306,557-564[Abstract]
  16. Pasque, MK, Trulock, EP, Kaiser, LR, et al Single lung transplantation for pulmonary hypertension: three-month hemodynamic follow-up. Circulation 1991;84,2275-2279[Abstract/Free Full Text]
  17. Glanville, AR, Burke, CM, Theodore, J, et al Primary pulmonary hypertension: length of survival in patients referred for heart-lung transplantation. Chest 1987;91,675-681[Abstract]
  18. Beck, JR, Pauker, SG The Markov process in medical prognosis. Med Decis Making 1983;3,419-458[Abstract/Free Full Text]
  19. Rich, S Primary pulmonary hypertension: executive summary from the world symposium. 1998 World Health Organization. Geneva, Switzerland:
  20. Rubin, LJ, Badesch, DB, Barst, RJ, et al Bosentan therapy for pulmonary arterial hypertension. N Engl J Med 2002;346,896-903[Abstract/Free Full Text]
  21. Barst, RJ, Rubin, LJ, Long, WA, et al A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. N Engl J Med 1996;334,296-302[Abstract/Free Full Text]
  22. Miyamoto, S, Nagaya, N, Satoh, T, et al Clinical correlates and prognostic significance of six-minute walk test in patients with primary pulmonary hypertension: comparison with cardiopulmonary exercise testing. Am J Respir Crit Care Med 2000;161,487-492[Abstract/Free Full Text]
  23. Barst, RJ, Rubin, LJ, McGoon, MD, et al Survival in primary pulmonary hypertension with long-term continuous intravenous prostacyclin. Ann Intern Med 1994;121,409-415[Abstract/Free Full Text]
  24. Rhodes, J, Barst, RJ, Garofano, RP, et al Hemodynamic correlates of exercise function in patients with primary pulmonary hypertension. J Am Coll Cardiol 1991;18,1738-1744[Abstract]
  25. Janicki, JS, Weber, KT, Likoff, MJ, et al Exercise testing to evaluate patients with pulmonary vascular disease. Am Rev Respir Dis 1984;129,93-95
  26. Simonneau, G, Barst, RJ, Galie, N, et al Continuous subcutaneous infusion of treprostinil, a prostacyclin analogue, in patients with pulmonary artery hypertension. Am J Respir Crit Care Med 2002;165,800-804[Abstract/Free Full Text]
  27. The EuroQol Group. EuroQol: A new facility for the measurement of health related quality of life. Health Policy 1990;16,199-208[CrossRef][ISI][Medline]
  28. Dolan, P Modeling valuations for EuroQol health states. Med Care 1997;35,1095-1108[CrossRef][ISI][Medline]
  29. Drug topics red book. 2002 Medical Economics Co. Oradell, NJ:
  30. Practice Management Information Corporation. Physician fees: a comprehensive guide for fee schedule review and management. 1996 Practice Management Information Corporation. Los Angeles, CA:
  31. US Department of Labor. Consumer price index, medical care portion. Available at: http://www.bls.gov/cpi/home.htm. Accessed November 5, 2003.



This article has been cited by other articles:


Home page
 The Annals of PharmacotherapyHome page
S. E Baker and R. H. Hockman
Inhaled Iloprost in Pulmonary Arterial Hypertension
Ann. Pharmacother., July 1, 2005; 39(7): 1265 - 1274.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (13)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Highland, K. B.
Right arrow Articles by Simpson, K. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Highland, K. B.
Right arrow Articles by Simpson, K. N.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS