Pulmonary arterial hypertension (PAH) is a rare disease of the pulmonary vasculature that leads to right ventricular dysfunction, right ventricular failure, and premature death. There are a number of medications already on the market, representing different therapeutic classes and possessing multiple mechanisms of action. Three new agents were approved by the US Food and Drug Administration in 2013, and others are currently in development. Recent advancements in PAH have resulted in increased survival and improved quality of life; however, no therapy provides a cure. Experts in the field are now utilizing clinical trial designs and end points that better reflect the disease progression among patients with this chronic disease. Although randomized placebo-controlled monotherapy trials are considered the strongest design, ethical and practical considerations have led to an increasing number of randomized trials designed to compare a PAH-specific treatment with placebo as an add-on to standard therapy. As many patients who enroll in clinical trials are already being treated for their condition, it may be unethical to withdraw or delay lifesaving therapies. The most widely used primary end point for PAH trials, change in 6-minute walk distance (6MWD) from baseline, has substantial limitations. Although it is generally reproducible, inexpensive, and relatively easy to conduct, the 6MWD is not designed to assess disease progression. Recent data have shown that 6MWD has inconsistent correlation with key indicators of disease progression such as hospitalization due to PAH, worsening right-sided heart failure, and death. The Task Force on End Points and Clinical Trial Design that met at the 4th World Symposium on Pulmonary Hypertension (WSPH) in 2008 in Dana Point, California, questioned the clinical relevance of the 6MWD as a primary end point and recommended the use of a composite end point—time to clinical worsening (TTCW)—in phase 3 or pivotal trials. TTCW may include time from randomization to PAH-related hospitalization, need for interventional procedures (ie, lung transplantation or balloon atrial septostomy), and mortality. More recently, at the 5th WSPH, held in 2013 in Nice, France, experts reiterated these recommendations. They further noted that, as clinical trials increasingly allow background therapies and are longer in duration, it may be more meaningful to use primary end points that measure "clinical worsening" rather than 6MWD. This paradigm shift will not only lead to a clearer demonstration of efficacy and safety as new agents come on the market, but will provide important information on long-term benefits (ie, the effects of drugs on clinical deterioration) that can assist payers as they strive to make value-based formulary decisions and provide cost-effective high-quality care.
Am J Manag Care. 2014;20:S115-S122
For author information and disclosures, see end of text.
Pulmonary arterial hypertension (PAH) is a rare disease, one of a larger group of pulmonary hypertensive disorders that have been categorized by the World Health Organization (WHO) and updated as recently as 2013 at the 5th World Symposium on Pulmonary Hypertension (WSPH) held in Nice, France.1 Registry data indicate a prevalence of 15 to 25 cases per million adults in France, and an incidence of 2.4 new cases per million inhabitants per year.2 PAH is far more common in patients with certain comorbidities. For instance, as many as 8% of patients with scleroderma and systemic sclerosis have the condition.3 US claims data suggest that overall prevalence may be higher in the United States.4 PAH is classified as a WHO Group 1 disease. It is hemodynamically defined by a right heart catheterization (RHC) measured mean pulmonary artery pressure greater than or equal to 25 mm Hg at rest and is additionally characterized by a pulmonary artery wedge pressure less than or equal to 15 mm Hg, and pulmonary vascular resistance (PVR) greater than 3 Wood units.5 The pathophysiology of PAH is complex and involves multiple mechanisms, including endothelial dysfunction, proliferation of vascular intima and media resulting in obstructive remodeling of the pulmonary vessel wall, vasoconstriction, and thrombosis, all of which contribute to increasing PVR. Progressively increasing PVR, in turn, results in right ventricular (RV) overload. This overload manifests as RV hypertrophy and RV dilation, which may result in failure of the right ventricle and ultimately a premature death.6,7
In a study of mortality outcomes in a US prospective registry (194 patients in 32 clinical centers from 1981 to 1985) established by the National Institutes of Health, D'Alonzo et al identified that the estimated median survival for these PAH patients was 2.8 years (95% confidence internal [CI], 1.9-3.7 years). Single-year survival rates were estimated as follows: 1 year, 68% (CI, 61%- 75%); 3 years, 48% (CI, 41%-55%); and 5 years, 34% (CI, 24%- 44%).8 Despite decades of research following D'Alonzo's mortality data and the availability of multiple therapies for PAH, morbidity and mortality rates remain high.9 The prognosis for PAH varies according to several criteria, including comorbid conditions and etiology.10 Recent long-term data involving 55 US centers from the Registry to Evaluate Early and Long-term Pulmonary Arterial Hypertension Disease Management (REVEAL), the largest registry of PAH patients in the world, shows a 1-year survival rate of 85% for patients newly diagnosed between 2006 and 2009. Three-year survival was 68%, 5-year survival was 57%, and 7-year survival was 49%.11,12 Of interest, and potentially partially explaining the poor prognosis of PAH, is the fact that data from REVEAL and from a French national PAH registry show that the time from symptom onset to PAH diagnosis remains long, averaging 2.58 and 2.25 years, respectively.2,12
In addition to the clinical burden, PAH is associated with significant healthcare utilization and cost. A few studies have examined expenditures for PAH in the managed care setting. One study by Said and colleagues used MarketScan data from 2004 to 2009 to compare resource utilization and cost increases over time for the 12-month pre-index period (index date = time of diagnosis) and at follow-up (at least 12 months post index) for PAH patients (n = 1647) versus a matched control group (n = 6352). The average follow-up was 20.8 months for the PAH group and 23.9 months for the control group. From pre-index to post index, per patient per month (PPPM) increases were significantly greater for the PAH cohort compared with the control group across all resource utilization measures (ie, inpatient visits, outpatient visits, emergency department visits, and the number of prescriptions; P <.001 for all). During this same period, the PPPM direct medical costs increased by $1957 (from $2064 to $4021) for the PAH cohort. This is compared with an increase of $439 (from $1094 to $1533; P <.001) for the matched control group. These findings showed that, compared with the control group, PAH patients had higher healthcare utilization rates and costs before and after their diagnosis.13
A study by Kirson et al examined the excess costs associated with PAH using US medical claims data from 2002 to 2007 for 471 privately insured PAH patients aged 18 to 65 years versus the same number of matched controls. At baseline, patients with PAH had significantly higher rates of certain comorbidities (eg, essential hypertension, congestive heart failure, and diabetes mellitus; P <.0001) and a higher Charlson Comorbidity Index score (P <.0001). Results showed that mean direct costs per patient-month adjusted to 2007 US dollars were $2023 for PAH patients versus $498 for controls (P <.0001), yielding excess costs of $1525 per patient-month. The inpatient services category was the largest cost driver. Approximately 45% of the total direct costs for PAH patients were for inpatient services, 38% were for outpatient and other services, and 15% were for prescription drugs. A sensitivity analysis restricting the sample to patients diagnosed via RHC yielded a 64% increase in excess costs relative to the original sample. The results showed the estimated direct costs for PAH patients to be 4 times greater than the costs for patients who do not have PAH, highlighting the substantial economic burden of the disease.14
The Current State of PAH and Its Management
Treating patients with PAH has changed substantially since the first PAH-specific continuous intravenous agent, epoprostenol, was approved in 1995.9,15,16 Current pharmacologic treatment of PAH involves several different drug classes based on different therapeutic approaches that target, among other things, vasodilation, anti-proliferation effects, and antiremodeling effects.17 As shown in the Figure,18 PAH-specific treatments target 3 pathways: a) the endothelin pathway (treated with endothelin-receptor antagonists [ERAs]); b) the prostacyclin pathway (targeted by prostanoids, or more specifically, prostacyclin analogues); and c) the nitric oxide pathway (historically treated with phosphodiesterase type 5 [PDE5] inhibitors). A recently approved soluble guanylate cyclase stimulator has provided an additional treatment to modulate the nitric oxide pathway.19 In addition to newer agents, patients with PAH may receive background treatment with diuretics to reduce edema, and with anticoagulants such as warfarin.10 Calcium channel blockers (amlodipine, diltiazem, nifedipine) are indicated only for the relatively small percentage of PAH patients who have a positive acute vasodilator response during RHC. More advanced therapies such as prostacyclin analogues (epoprostenol, iloprost, treprostinil), ERAs (bosentan, ambrisentan), and PDE5 inhibitors (sildenafil, tadalafil) currently represent the mainstay of treatment for the majority of PAH patients, either as monotherapy or in combination.4,9,20 US Food and Drug Administration (FDA) approvals in 2013 include the ERA macitentan, the soluble guanylate cyclase stimulator, riociguat, and the prostacyclin analogue, oral treprostinil.21-23
The American College of Cardiology Foundation (ACCF)/American Heart Association treatment guidelines, published in 2009, support the use of combination therapy in patients who have an inadequate response to monotherapy.24 In 2010, Sitbon and Galié advocated for a treat-to-target strategy that supports early diagnosis to initiate treatment early, while monitoring patients closely to allow for prompt escalation of therapy rather than waiting for signs of clinical worsening.25 The most recent guidelines, published as a result of the 2013 WSPH in Nice, reiterate support for combination therapy when clinical response is inadequate. The Nice guidelines noted that, as there are 3 distinct pathways involved in PAH, combination therapy remains an attractive option. Treatment with 2 or more classes of drugs in combination has been successfully used in the management of other disease states (eg, systemic hypertension and heart failure). In addition, the REVEAL registry data suggest that this PAH treatment approach is commonly used by clinicians.24,26 The usage statements within the labels for 2 recent FDA approvals, macitentan and riociguat, include drug combinations with other PAH-specific therapies.27,28
Evolution of Clinical Trials in PAH
As the body of scientific knowledge in PAH has expanded, the methods by which the new therapeutic agents are studied and approved have evolved as well. However, there are inherent challenges in designing clinical trials in PAH.15,29 The heterogeneous nature of PAH, comorbidities, and the orphan nature of the disease are among the factors contributing to the difficulty.29 Historically, PAH clinical trials have been short in duration (12 to 16 weeks) and have utilized the 6-minute walk distance (6MWD) as the primary end point.30 Earlier studies also utilized a placebo-controlled monotherapy design. Given that PAH has become a chronic disease with increased survivability, it is increasingly difficult to find and recruit patients who are naive to therapy. Additionally, there are ethical considerations involved in randomizing patients in a monotherapy trial to the placebo group.29
Short-term trials provide little information related to long-term disease course or progression in a chronic disease such as PAH. Therefore, clinical trial designs have evolved to include trials that are longer in duration, may involve an add-on therapy, and to more adequately assess disease progression (ie, clinical deterioration), utilize composite end points as advocated by the Dana Point and Nice clinical trial design task forces.1,15 The SERAPHIN phase 3 trial for the recently approved macitentan, the ongoing AMBITION trial of ambrisentan/tadalafil combination therapy, and the GRIPHON phase 3 trial of selexipag each used a composite end point.31-33 To assess the treatment effect on disease progression, each of these trials was also significantly longer in duration than previous PAH studies. The SERAPHIN trial, the largest PAH trial with reported results to date, included patients with approximately 4 years of drug exposure.33 The trial is unique in that it is the longest trial to evaluate morbidity/ mortality as a primary end point in PAH1 (Table).30-43
It is important for managed care decision makers to be aware of these shifts in trial design as new agents come on the market. Payers will be able to more effectively assess drug efficacy based on important clinical outcomes, rather than on functional end points only.
Placebo-Controlled Monotherapy Versus Add-on Trials
Although placebo-controlled studies are generally considered the ideal design for assessing new therapies, this design is less feasible in PAH because these patients are generally already undergoing treatment for their condition at the time of enrollment. An alternative approach to conducting a placebo-controlled monotherapy trial is a study design in which a new treatment or placebo is added on to an established advanced PAH treatment. Another possible trial design alternative is the non-inferiority study, although this approach is limited by the need for a large study population, often in excess of 500 patients. The rarity of the disease makes this approach less practical. Withdrawal trials (ie, temporarily withdrawing treatment in a group of patients in lieu of having a placebo group) constitute an additional study design approach; however, in PAH their use may be unethical.15
Clinical Trial End Points
The 2008 Dana Point Task Force on End Points and Clinical Trial Design recognized the challenges in designing appropriate clinical trials that truly assess the efficacy and safety of new therapeutic choices in PAH. The task force questioned the clinical relevance of the 6MWD, especially when used earlier in the course of the disease, and recommended that other, more sensitive methodologies be explored. The task force also recommended that TTCW be used as a primary end point in phase 3 and pivotal trials in PAH. The Dana Point expert panel recommended that primary end points be 1) clinically relevant; 2) sensitive to treatment effect; and 3) measurable and interpretable. Secondary end points, according to the panel, must complement primary end points. They can be clinically relevant toward determining a drug's indication, or more of a "feel good" end point, which is unlikely to inform decisions about indications or label changes, but may offer reassurance regarding the outcome of the primary end point. The panel further noted that in some cases, secondary end points may also be "exploratory" or hypothesis-generating. Common secondary end points include cardiopulmonary exercise testing, functional class, and hemodynamics.15
Most recently, the 2013 Nice Task Force on New Trial Designs and Potential Therapies for PAH stated that the selection of the primary end point in a registration trial is one of the most important steps in trial development. The Nice task force recommended that the primary end point be clinically meaningful, which may be defined as a direct measure of how a patient feels, functions, or survives.44
6MWD as a Primary End Point
Change in 6MWD, the most common primary end point in PAH clinical trials, involves a form of exercise testing that is widely used in clinical practice. The 6MWD measures total distance walked in 6 minutes, typically along with Borg dyspnea score and oxygen saturation during exertion.15,45 Apart from the ease of its administration and low cost, the 6MWD is useful for evaluating activities of daily living, and distance walked at baseline has been associated with survival prediction in some trials.45,46 The 6MWD is widely accepted by regulatory agencies,47,48 but has a number of drawbacks, including variability in the conduct of the test.49,50 Age, height, weight, and comorbidities, such as joint pain, can also impact 6MWD results.45-47,51-54 Recent studies have found modest between-group differences, where effective background therapies are permitted, compared with earlier monotherapy trials.7,45 Also, change in 6MWD is a less reliable marker of treatment effect in patients who are at the upper end of the functional scale. This may be related to a "ceiling effect," or the level at which improvements are difficult to detect because the patient walked at or close to maximum speed at baseline.55
Change in 6MWD has been shown to have a modest to poor correlation with key indicators of PAH disease progression. Gabler et al examined pooled data from 2404 patients in 10 placebo-controlled randomized trials with PAH-specific therapy to assess whether 6MWD could predict the relationship between treatment (or placebo) and the occurrence of a predefined event, including death, lung or heart-lung transplantation, atrial septostomy, hospitalization due to worsened PAH, withdrawal for worsened right-sided heart failure, and addition of other PAH medications. A large proportion of the treatment effect on clinical events could not be explained by 6MWD changes, leading Gabler and colleagues to conclude that 6MWD by itself may be an inadequate surrogate end point for clinical events in PAH.50
In another meta-analysis, Savarese and colleagues examined pooled data from 3112 patients in 22 PAH studies.49 Active treatment with advanced therapies in monotherapy or combination therapy resulted in a significant risk reduction for a composite outcome of clinical events (including mortality, hospitalization, lung or heart-lung transplantation, or initiation of rescue therapy) at a mean follow-up of 14.5 weeks. However, the results showed that change in 6MWD did not correlate well with the composite end point or with any of its components. Thus, change in 6MWD failed to predict the treatment effect on clinical outcomes. As the data were derived from short-term trials, no statements can be made about the correlation of the 6MWD with long-term clinical outcomes.49
However, improvement in function, as reflected by 6MWD, remains clinically valuable in PAH. The inclusion of the 6MWD as an end point will likely continue, although it is more likely to be used as a secondary end point.
Composite End Points
Composite end points have been increasingly used in clinical trial designs to more accurately reflect disease progression. Both the 2008 Dana Point and the 2013 Nice clinical trial design task forces recommended this approach. In PAH, the composite end point of time to clinical worsening (TTCW) has previously been used as a secondary end point in registration trials. TTCW may be useful for detecting disease progression even in patients with relatively mild disease.56 TTCW may include various components designed to measure morbidity and mortality, such as time to PAHrelated hospitalization, all-cause mortality, clinical disease progression (eg, increase in New York Heart Association/ WHO functional class), and time to requirement of an interventional procedure (ie, lung transplantation or balloon atrial septostomy). One of the concerns with TTCW is that there is no uniform or consistent definition. This variability makes efficacy evaluation challenging. Also, disease progression may be subjective.15
To address these issues, the Dana Point task force recommended that a uniform definition of TTCW be used in future pivotal and phase 3 trials, and that the composite end point include:
The task force also strongly suggested that the adjudication of events be mandatory in clinical trials that employ TTCW as the primary end point and that the studies be longer in duration.15
Recently completed trials in PAH suggest that using primary end points that reflect clinical worsening may be more meaningful, as new trials will involve patients who are already on background therapy.44 The 2013 Nice task force noted that the TTCW composite end point, which measures time to first event, could be enhanced to include 4 symptom categories as the basis for one of the components (ie, dyspnea, chest pain, dizziness/syncope, and fatigue/activity level). Measuring symptoms is important, as death is rarely the first clinical event. The task force suggested that the TTCW composite end point might be defined as44:
2. Lung transplantation;
3. Hospitalization for worsening PAH (including atrial septostomy);
4. Initiation of IV therapy due to worsening PAH;
5. Worsening of function (worsening functional class and exercise capacity); or
6. Worsening of PAH symptoms (worsening of at least 2 of the 4 symptoms listed above).
The further benefit of such a TTCW composite end point is that clinicians, patients, and payers are likely to agree that the above-listed criteria better reflect "real-world" events that they are already focused on reducing.
Development of clinical end points for trials of PAH therapies has evolved in recent years after a long period of reliance on the 6MWD, which has increasingly demonstrated its limitations in this patient population. The 6MWD and other short-term functional end points now may be superseded by composite morbidity and mortality end points that demonstrate a more comprehensive evaluation of disease progression, with fewer concerns regarding correlation to indicators of disease progression.15,50 This end point change has now been advocated by experts at the WSPH in 2008 and reaffirmed in 2013. Using this new approach will require longer clinical trials with larger patient populations in order to be properly powered, and their design will be driven more by the total occurrence of clinically meaningful events. The use of trial designs that examine the PAH-specific agent added on to standard therapy will also become increasingly common and will more accurately reflect the real-world treatment of these patients that has been described in the longitudinal REVEAL registry.
Trends in PAH Clinical Trial Design15,50
6MWD indicates 6-minute walk distance; PAH, pulmonary arterial hypertension; TTCW, time to clinical worsening.
Thus, this paradigm shift signals the evolution of trial design from treatment effect on symptoms to one that reflects clinical outcomes. This evolution has occurred in other disease states, such as systemic hypertension, hyperlipidemia, and congestive heart failure. Given the substantial clinical and economic burden of PAH, the shift to outcomes-based end points will likely help payers in their evaluation of new PAH-specific therapies. Further, in light of healthcare reform and its emphasis on quality and costeffectiveness, the availability of clinical outcomes data is increasingly important.
Author affiliation: New York University, New York, NY (SMS); Gemini Healthcare, LLC, Westbrook, CT (RJG).
Funding source: This supplement was sponsored by Actelion Pharmaceuticals US, Inc.
Author disclosure: as a consultant or paid advisory board member for and receipt of lecture fees from Actelion Pharmaceuticals, Bayer, Gilead, and United Therapeutics Corporation. Mr Gilkin reports serving as a consultant for Actelion Pharmaceuticals and Pfizer.
Authorship information: Concept and design (RJG, SMS); analysis and interpretation of data (RJG); drafting of the manuscript (RJG, SMS); critical revision of the manuscript for important intellectual content (RJG, SMS); administrative, technical, or logistic support (RJG, SMS); and supervision (RJG, SMS).
Address correspondence to: Sean.Studer@woodhullhc.nychhc.org.