Introduction
Pulmonary hypertension (PH) is defined by an abnormal elevation of mean pulmonary
arterial pressure.[1] It can occur in different situations, either autonomously (as in pulmonary arterial
hypertension [PAH], which may be associated with others conditions), or as a complication
of other cardiac or respiratory clinical entities: left heart diseases (postcapillary
PH), pulmonary embolism (PE; chronic thromboembolic PH, CTEPH), chronic lung diseases,
and/or hypoxia. Thus, there is an international clinical classification of PH, separating
precapillary (group 1, 3, 4, and 5) from postcapillary causes (group 2).[2] In every situation, the development of PH is associated with a worse prognosis,
by the development of right heart failure as the main cause of death.
During the past decade, significant improvements have been achieved in patients with
group 1 (PAH) and group 4 (CTEPH). Historically, PAH has aroused the greatest interest,
particularly after iatrogenic causes had been identified, such as anorexigens.[3] PAH may be idiopathic, heritable, drug- and toxin- induced, or associated with conditions
such as connective tissue diseases (CTD-PAH). Several targeted therapies for PAH have
been developed in the past two decades.[4] Before the advent of these drugs, treatment of PAH was mainly based on diuretics,
oxygen, and oral anticoagulants. This conventional therapy remains important as the
initial step of the treatment algorithm of patients with PAH.[5] During the same time, there have been two major advances in the treatment of CTEPH,
with the development of medical and interventional treatment, in patients ineligible
to surgery (pulmonary endarterectomy), which is the treatment of choice to relieve
pulmonary artery obstructions when possible. However, whatever the treatment option
retained for a CTEPH patient, lifelong anticoagulation therapy remains the cornerstone
of medical therapy.
In parallel with these advances in the management of PAH and CTEPH, anticoagulant
treatment modalities in venous thromboembolism (VTE) and atrial fibrillation (AF)
have been profoundly modified with the emergence of direct oral anticoagulants (DOACs),
gradually replacing vitamin K antagonists (VKAs) in most of their indications.[6] The arrival of these drugs, as well as their pharmacological particularities, justifies
the need to redefine the role of anticoagulants in the treatment of patients suffering
from PAH or CTEPH.
General Review and Clinical Implications
Why Is Anticoagulant Therapy Prescribed to Patients with Pulmonary Arterial Hypertension
or Chronic Thromboembolic Pulmonary Hypertension?
Patients with PAH
Rationale for Treating PAH with Warfarin
From a pathophysiological point of view, PAH is characterized by the association of
lesions in the distal muscular-type arteries: medial hypertrophy/hyperplasia, intimal
and adventitial fibrosis, and (in situ) thrombotic lesions/plexiform lesions.[7] The frequency of thrombotic lesions in autopsy studies,[8] and the demonstration of both an associated activation of the coagulation system[9]
[10]
[11]
[12] and a decreased in physiological fibrinolysis,[13] paved the way for the rationale of use of anticoagulant drugs for the treatment
of PAH. These hypotheses were comforted by the sub-analysis of a landmark trial evaluating
the use of calcium blockers in patients with PAH,[14] in which a better prognosis was described for patients who received warfarin. A
summary of this study and of other prospective trials regarding the efficacy and safety
of anticoagulant therapy in patients with PAH is provided in [Table 1]. Of note, the allocation of warfarin was not provided by the randomization, but
was advised to the investigators, in the case of the presence of an abnormal perfusion
lung scan. Then, anticoagulant therapy with warfarin was prescribed in the majority
of patients with PAH, for a potential benefit on survival despite the absence of clear
validation.[15]
[16]
[17] This may explain the high prevalence of anticoagulant therapy (∼90%) in the epidemiological
studies from the current era.[18]
Table 1
Published and ongoing studies on anticoagulant therapy in PAH and/or CTEPH patients
|
Study
|
Population
|
Design
|
Outcomes
|
Results
|
Comments/limits
|
|
Rich et al(1992)
|
64 PAH
|
Post-hoc analysis of aprospective cohort
|
5-y survival
|
Improved survival in the 35 patients who received VKA
|
Post-hoc analysis of prospective cohort Allocation to VKA let to the investigator,
iflung perfusion scan was abnormal
|
|
Kawut et al (2011)
|
92 PAH
|
Randomized, double blind, placebo controlled
|
6-min walking distance at 6 mo
|
Trial terminated prematurely by the steering committee for futility
|
|
|
Ngian et al (2012)
|
117 CTD-PAH, incident
|
Prospective multi center cohort
|
3-y survival
|
Improved survival in the patients with CTD-PAH who received VKA
|
Cohort study
Lack of information on length of therapy, as for presence of concomitant VTE or AF
|
|
Olsson et al (2014)
|
1283 PAH, incident
|
Posthoc analysis of aprospective cohort (COMPERA)
|
3-y survival
|
Improved survival in the patients withi PAH who received VKA, but not in other forms
of PAH
|
Posthoc analysis of prospective cohort
Attribution of VKA let the investigator
Lack of information on length of therapy, as for presence of concomitant VTE or AF
|
|
Preston et al (2015)
|
187 PAH(144 iPAH and 43
SSc-PAH) who received VKA at anytime of the study, matched with 187 who did not
|
Posthoc analysis of aprospective cohort (REVEAL)
|
3-y survival
|
Similar survival in patients with iPAH who have taken VKA at any time of the study,
compared with those who did not
Lower survival in patients with SSc-PAH who have taken VKA at any time of the study
|
Post-hoc analysis of prospective cohort
Attribution of VKA let to the investigator
Lack of information on length of therapy as for presence of concomitant VTE or AF
Mix of prevalent and incident cases
|
|
SPHInX
Study
|
SSc-PAH
|
Multicenter randomized placebo-controlled trial of oral anticoagulation with apixaban
|
Composite of time to death or clinical worsening of PAH at 3 y
|
Ongoing trial, ACTRN12614000418673, 170 patients expected
|
|
HEMA-HTP
|
PAH or CTEPH, receiving anti coagulant therapy
|
Multicenter prospective cohort
|
Major bleeding according to ISTH definition
|
Ongoing trial, NCT02800941, 203 patients expected
|
Abbreviations: AF, atrial fibrillation; CTD, connective tissue disease; CTEPH, chronic
thromboembolic pulmonary hypertension; iPAH, idiopathic pulmonary arterial hypertension;
ISTH, International Society on Thrombosis and Haemostasis; PAH, pulmonary arterial
hypertension; SSc, systemic sclerosis; VKA, vitamin K antagonists; VTE, venous thromboembolism.
Note: The table shows a summary of data obtained from prospective studies, regarding
the efficacy and safety of anticoagulant therapy in patients with pulmonary arterial
hypertension (PAH) orchronic thromboembolic pulmonary hypertension (CTEPH).
In parallel, some studies indicated that patients with PAH experience a dysfunction
of platelets,[19]
[20] suggesting a potential place for antiplatelets in the treatment of PAH. This hypothesis
was the primary aim of a randomized controlled trial ([Table 1]), evaluating aspirin (and simvastatin, in a crossover design) efficacy in patients
with PAH.[21] Although aspirin confirmed its efficacy on a surrogate endpoint (level of thromboxane
B2), it failed to demonstrate any effect on the 6-minute walk distance (which was
the main endpoint for any PAH-targeted therapy at that time).
Concomitant Validated Indication for Anticoagulant Therapy
The phenotype of patients diagnosed with PAH changed deeply in the past two decades.[22] While patients included in the historic premodernmanagement National Institutes
of Health registry were most frequently young females,[23] current patients included in contemporary registries are older and with a more balanced
sex ratio at time of PAH diagnosis.[24]
[25]
[26] These moving phenotypes may increase the prevalence of acquired cardiovascular diseases,
such as VTE or AF,[26]
[27] two clinical situations in which the indication for anticoagulant therapy is indubitable.[28]
[29] In a multicenter study from four PH centers in France, we found that two-thirds
of patients with precapillary PH were receiving anticoagulant therapy.[30] Of note, anticoagulant therapy was prescribed for secondary prevention of VTE or
AF in one half of the patients.[30] This significant proportion was subsequently confirmed in another retrospective
cohort study from the Unites States and Spain in which 60.2% of PAH patients were
treated with oral anticoagulants.[31]
Patients with Congenital Heart Disease-Related PAH
Patients with PAH and cyanotic congenital heart diseases, especially those with Eisenmenger
physiology, are quite susceptible to the development of thrombosis. The main manifestation
is insitu pulmonary artery thrombosis rather than PE and may be present in up to one-fifth
of patients with Eisenmenger syndrome. Age was the only factor associated with pulmonary
thrombosis in a cohort of 98 patients with cyanotic congenital heart disease.[32] Administration of prophylactic antithrombotic therapy in these patients may lead
to an increased risk of bleeding complications, which sometimes can be fatal in this
population.[33] Thus, the routine use of anticoagulation in patients with Eisenmenger syndrome is
under debate.
Patients with CTEPH
CTEPH is defined by the presence of precapillary PH associated with persistent perfusion
defects at lung perfusion scan, after at least 3 months of effective anticoagulant
therapy.[5] In the majority of cases, CTEPH is a complication of acute PE. Hence, anticoagulant
therapy is prescribed to avoid a recurrent VTE/PE, and to prevent in situ pulmonary
artery thrombosis.[34]
What Do the Guidelines Say?
Every 5 years, the world's experts in PH meet to discuss scientific advances and update
management proposals. This work updates the scientific knowledge in the field of PAH
from experts invited by the European Society of Cardiology (ESC) and the European
Respiratory Society (ERS), leading to the development of European recommendations
(the latest dating from 2015).[5]
According to 2015 ESC/ERS guidelines,[5] anticoagulant therapy may be considered in patients with idiopathic PAH, heritable
PAH, and PAH due to use of anorexigens (class of recommendation: IIb; level of evidence:
C). For patients with CTD-PAH (like scleroderma), oral anticoagulation may be considered
on an individual basis and in the presence of thrombophilic predisposition (class
of recommendation: IIb; level of evidence: C).
In patients with CTEPH, lifelong anticoagulant therapy is recommended as the first
step of treatment (class of recommendation: I; level of evidence: C), according to
2015 ESC/ERS PH guidelines.[5] Recently, European experts proposed an update of the scientific guidelines on acute
PE.[35] These guidelines included a dedicated chapter on CTEPH management, confirmed the
recommendation of lifelong anticoagulation, and indicated that VKA was the recommended
regimen. This recommendation is in agreement with national guidelines from European
countries.[36]
[37]
Finally, in patients with congenital heart-disease-related PAH, oral anticoagulant
treatment may be considered in patients with pulmonary arterial thrombosis or signs
of heart failure, in the absence of significant hemoptysis (class of recommendation:
IIb; level of evidence: C).[5]
Unmet Needs, Recent Advances, Ongoing Studies
Determining the Efficacy of Anticoagulant Therapy on the Survival of Patients with
PAH
As previously discussed, although anticoagulant therapy is recommended by the guidelines
as a potential treatment of PAH, the level of evidence is low. Putting aside the results
of the ancillary analysis of a therapeutic trial performed before the appearance of
specific PAH drugs,[14] we have mainly epidemiological data with discordant results. A cohort study of patients
with CTD-PAH (mainly scleroderma) suggested that patients who were receiving anticoagulation
therapy with VKAs had better survival than those who did not receive anticoagulant
therapy ([Table 1]).[38] To date, two central studies have been conducted based on national or international
clinical registries: the COMPERA[39] and the REVEAL[40] registries ([Table 1]). The COMPERA registry is an international European PAH registry (82% of patients
coming from Germany). Among 2,414 patients with PAH, the authors analyzed the clinical
evolution of 1,283 incident cases, and compared the survival of two groups: those
who never received anticoagulant therapy during the 3 years of follow-up versus those
who received anticoagulant therapy at any time during the follow-up. Of the whole
cohort of patients with incident PAH, 58% received anticoagulant therapy (66% idiopathic
PAH subgroup, 44% other forms of PAH). Patients exposed to anticoagulant therapy at
any time of the follow-up were older, and had more severe diseases (more frequently
with a World Health Organization functional class III/IV, a need for combination targeted
PAH therapy, and worse hemodynamics). Multivariable regression analysis showed that
patients with idiopathic PAH who received anticoagulant therapy had a lower risk of
death (hazard ratio [HR] =0.79, 95% confidence interval [CI]: 0.66–0.94). This finding
was not replicated in other forms of PAH (mainly scleroderma), in whom a trend toward
a lower survival was reported in patients exposed to anticoagulant therapy. This work
suffers from some limitations, such as the absence of randomization, the lack of data
on the prevalence of validated indication for anticoagulant therapy, and the absence
of prospective monitoring of bleeding events.[29]
The REVEAL[40] registry is an ongoing U.S. clinical registry of patients with PAH, which differs
from other registries by accepting a higher threshold of pulmonary capillary wedge
pressure (≤18 mm Hg). As was the case with the COMPERA registry, investigators from
the REVEAL registry assessed the association between chronic anticoagulation (with
warfarin) and outcomes in patients with PAH. To avoid the “immortal time bias,” the
authors performed a sensitivity analysis on patients in whom warfarin was initiated
during the study. Controls consisted of patients who were never exposed to anticoagulant
therapy during the follow-up. Among the 2,197 included patients, 1,116 started warfarin
before or at enrollment. Of these, 218 patients (163 idiopathic PAH patients and 55
PAH patients associated with scleroderma) started warfarin after enrollment and were
compared with matched controls in the sensitivity analysis. The authors did not find
any positive effect of warfarin (mean length of exposure: 1 year) in patients with
idiopathic PAH. When analysis was modified to include patients who received warfarin
at any time, the survival was lower in patients with PAH associated with scleroderma
who received warfarin at any time of the study. Similar to the COMPERA trial, the
results of the REVEAL study were suffer from lack of randomization, lack of data on
the prevalence of validated indication for anticoagulant therapy, and absence of prospective
monitoring of bleeding events.
These retrospective analyses of prospective epidemiological cohorts were pooled to
conduct a meta-analysis which assessed the effect of anticoagulant therapy on survival
in patients with PAH.[41] The authors concluded that anticoagulant therapy significantly reduced mortality
in the overall PAH cohort (HR = 0.73, 95%CI: 0.57–0.93), with a similar effect in
patients with idiopathic PAH, while it was associated with an increase in mortality
in patients with PAH associated with scleroderma (HR = 1.58, 95%CI: 1.08–2.31). Besides
the large number of patients and the homogeneous effect of anticoagulant therapy in
the whole cohort and in patients with idiopathic PAH, the results of the meta-analysis
suffer from the low quality of data, with a moderate risk of bias, as well as the
absence of randomized controlled trials.
Currently, only two trials with anticoagulation drugs are ongoing in patients with
PAH, both focusing on patients with PAH associated with scleroderma. The SPHYNX trial
aims to assess the efficacy, safety, and cost-effectiveness of treatment with apixaban
in PAH associated with scleroderma over 3 years, in a multicenter, double-blind, placebo-controlled
randomized controlled trial.[42] Of note, the SPHYNX trial includes a cost-effectiveness analysis, and will be able
to infirm the results of a previous medico-analytic study which suggested a positive
effect of anticoagulant therapy in patients with idiopathic PAH but not PAH associated
with scleroderma.[43] The Pragmatic Clinical Trials in Scleroderma program (PCTS) is also registered on
Clinicaltrial.gov (NCT03610217), and includes an arm which will evaluate the efficacy
of anticoagulant therapy (warfarin, rivaroxaban, or apixaban) on PAH.
Confirming the Safety of Anticoagulant Therapy in Patients with PAH or CTEPH
The amount of available data to assess the safety of anticoagulant therapy in patients
with PAH or CTEPH is less compared with that of efficacy data. This is a problem,
since anticoagulant treatment increases the risk of bleeding, including life-threatening
bleeding associated with PAH and CTEPH per se, such as hemoptysis,[44] a situation which may particularly affect patients with PAH associated with congenital
heart diseases.[33]
All the published data come from retrospective analyses[31]
[45]
[46] of patients with PAH or CTEPH who received anticoagulant therapy. The first paper
was a single-center retrospective study from Amsterdam.[45] The authors reported the annual rate of major bleeding (defined by the International
Society on Thrombosis and Haemostasis criteria[47]) in 198 patients with precapillary PH (idiopathic PAH [n = 99], PAH associated with connective-tissue disease [n = 39], and CTEPH [n = 60]). Of note, the international normalized ratio (INR) target range was above
2.5 and 3.5 for more than 80% of cases. Interestingly, they reported a lower rate
of major bleeding in patients with CTEPH (2.4 for 100patient-years) than in patients
with idiopathic PAH (5.4 for 100patient-years), despite CTEPH patients being older.
The rate of major bleeding was markedly elevated in patients with PAH associated with
connective-tissue disease (19 for 100patient-years). No factors were found to be associated
with the risk of major bleeding. However, there was a significant association between
the use of prostacyclin analogues as PAH-targeted therapy and major bleeding, suggesting
a potential increased rate of alveolar bleeding in PAH patients with concomitant use
of prostanoids and VKA reported also in a study from Japan.[48] Bleeding from the gastrointestinal tract was the most frequent type of bleeding
event (50%). As with all retrospective single-center studies, the results are weakened
by a center bias (as illustrated by the INR target higher than usual, as well as the
age which was younger than the age of patients included in multicenter prospective
registries). Roldan et al confirmed these results in a retrospective study of 201
PAH patients, with an incidence rate for major bleeding of 4.7 per 100 patient-years
(5.0 per 100 patient-years for idiopathic PAH),[31] but without the excess in major bleeding events for patients with CTD-PAH (6.3 per
100 patient-years). Baseline factors which were associated with the occurrence of
bleeding were a HAS-BLED score equal or greater than 3, history of prior bleeding,
poor anticoagulation control, and increased number of concomitant medications.[31] Finally, Jujo-Sanada et al[46] evaluated the risk of clinically relevant bleeding in 72 patients with CTEPH under
warfarin in a retrospective single-center study. The incidence of major bleeding was
to 5.0% per person-year (two times more than in the Henken et al's study), which seems
to be more in agreement with the current literature in the field of unselected patients
with acute VTE.[49] All these three studies suffer from their retrospective design, and their single-center
nature for two of them. One prospective multicenter cohort is ongoing (NCT02800941),[50] which aims to estimate the incidence of major bleeding in patients with PAH or CTEPH
receiving anticoagulant therapy, with a sufficient power also to provide some predictive
factors.
Tailoring the Best Anticoagulant Option for Patients with CTEPH, as well as for PAH
Patients Who Have an Indication for Anticoagulation
The emergence of DOACs deeply modified the landscape of anticoagulant therapy.[6] In patients with VTE as well as in nonvalvular AF, they have demonstrated a better
safety, with at least noninferior efficacy compared with warfarin.[51]
[52]
[53] Conversely, negative results were shown in patients with a mechanical heart valve
or antiphospholipid syndrome.[54]
[55]
In patients with PAH or CTEPH who receive VKA, there is a significant risk of pharmacological
interactions,[56]
[57]
[58] but this may be monitored with INR measurements. We also have data suggesting that
the quality of anticoagulation with warfarin, as evaluated with the time in therapeutic
range, is low, exposing these patients to complications.[59] These points may support the role of DOACs, at least in patients with a validated
indication such as secondary prevention of VTE, or AF. Basic research suggested that
an anti-Xa factor (rivaroxaban) may be efficient in an experimental model of PAH.[60] In the German cohort, about half of patients with CTEPH were receiving rivaroxaban
for secondary VTE prevention.[61] Unfortunately, no information on bleeding complications was provided. Very recently,
investigators from CTEPH centers in the United Kingdom shared their experience with
the use of DOACs in CTEPH patients following pulmonary endarterectomy.[62] In their retrospective analysis, a similar rate of bleeding was found between patients
under VKA and those under DOACs, while more recurrent VTE events were diagnosed in
patients receiving DOACs. These results are striking, as available evidence in VTE
studies has led to the opposite findings: a lower rate of bleeding with similar efficacy.[63] This may be explained by pharmacological issues between DOACs and the targeted PAH
drugs, as some of them (bosentan, phosphodiesterase-5 inhibitors, riociguat) may provide
significant drug–drug interaction with DOACs.[64]
[65]
[66] Hence, a modified bioaccumulation of DOACs in patients receiving treatment for PAH
or CTEPH may be part of the absence of better safety compared with VKAs. In terms
of efficacy, it has been demonstrated that VTE recurrence may be an explanation of
PH recurrence after surgery,[67] highlighting the need for a highly efficient therapy. Historically, the antiphospholipid
syndrome has been recognized as a significant thrombotic risk factor of CTEPH. Its
prevalence in CTEPH patients ranges from 2.5 to 7.7%.[68]
[69] The therapeutic impact is important, as a triple-positive antiphospholipid antibody
profile is a condition in which DOACs are contraindicated,[55] and patients should be prescribed VKAs. This issue has been raised as a potential
explanation of reported failure of DOACs in CTEPH patients from the United Kingdom
study.[70] Moreover, the fact that cases were identified through the routine screening of asymptomatic
patients as well as low adherence of treatment with DOACs might also explain this
finding. However, all these results should be confirmed in future randomized trials
or larger multicenter registries. In addition, the reduceddosage of DOACs does not
seem a good option in patients with CTEPH.
