You are on page 1of 25

Clin Chest Med 28 (2007) 91–115

Standard Therapies for Pulmonary

Arterial Hypertension
Shoaib Alam, MDa,*, Harold I. Palevsky, MDb,c,d
Division of Pulmonary, Allergy and Critical Care Medicine, Penn State University-Hershey Medical Center,
500 University Drive, Hershey, PA 17033, USA
University of Pennsylvania School of Medicine, 3400 Spruce Street, Philadelphia, PA 19104-4283, USA
Pulmonary, Allergy and Critical Care, Penn Presbyterian Medical Center, 51 N. 39th Street,
Philadelphia, PA 19104, USA
Pulmonary Vascular Disease Program, Penn Presbyterian Medical Center, 51 N. 39th Street,
Philadelphia, PA 19104, USA

Pulmonary arterial hypertension (PAH) was use of diuretics) or physiologic observations (eg,
first described in 1891 in a case report by hypoxemia in emphysema and use of oxygen sup-
Romberg [1]. The term ‘‘primary pulmonary hy- plementation therapy, low cardiac output in left
pertension’’ (PPH) was first used in 1951 when ventricular [LV] systolic dysfunction and use of di-
Dresdale and colleagues [2] reported clinical fea- goxin, high blood pressure in essential systemic hy-
tures and hemodynamics of 39 patients. Until pertension and use of vasodilators, prothrombotic
approximately two decades ago, before the de- tendency in venous thromboembolism and use
velopment of specific PAH therapies, such as of anticoagulation). These practices were subse-
prostacyclin analogs, endothelial receptor antag- quently adopted in the treatment of PAH and right
onists, and phosphodiestrase-5 inhibitors, PAH heart dysfunction and failure. None of these thera-
(especially idiopathic [IPAH] or PPH form of pies is supported by well-designed placebo-con-
PAH) was considered a disease that was univer- trolled trials. Some of these practices are based on
sally fatal with a median survival of 2.8 years marked symptomatic relief or improvement with
[3]. An explosion of research and drug develop- short-term use (eg, diuretics for edema) or by obser-
ment has resulted in the development of several vations based on acute testing in the laboratory set-
specific PAH therapies (see other articles else- ting (eg, improvement in cardiac output by single
where in this issue) [4–21]. Therapies such as ox- administration of digoxin) [23] or from autopsy
ygen supplementation, calcium channel blockers findings [24–26] (eg, in situ microthrombi in lungs
(CCBs), anticoagulation, digoxin, and diuretics of patients who have PAH and use anticoagula-
have been in use since long before the develop- tion). Therapies such as anticoagulation have
ment of the newer specific PAH agents. These been supported by a few retrospective studies
therapies have been referred to as ‘‘standard [27,28] and an uncontrolled, single-center prospec-
PAH therapies’’ or ‘‘conventional PAH thera- tive study [29], especially in IPAH [30]. This raises
pies’’ [22]. The concept of these standard thera- the ethical question of the appropriateness of per-
pies originated from the experience with other forming a placebo-controlled trial in these patients.
pulmonary and cardiac diseases with similard On the other hand, the short-term benefits from di-
but not identicaldmanifestations in terms of symp- uretics are so obvious that it virtually obviates the
toms (eg, edema in congestive left heart failure and need for a clinical trial to validate such benefit.
For these reasons it seems unlikely that any
* Corresponding author. large, prospective, clinical trials will examine the
E-mail addresses:; role of the standard therapies used in PAH in the (S. Alam). near future. Such trials are needed, however.
0272-5231/07/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved.

Especially lacking are data about use of standard therapies are poorly tolerated or adverse events
therapies in PAH other than IPAH (associated occur.
[APAH]). It may be possible that in specific popu-
lations, these therapies may not be helpful,
Oxygen therapy
that they may be less helpful than they are
currently considered, or that they may be even The value of long-term oxygen supplementa-
harmful. tion therapy in patients with PAH has not been
In a prospective, controlled study of patients evaluated by well-designed clinical trials. Recom-
who had PAH and Eisenmenger syndrome, the mendations and guidelines for oxygen therapy in
use of oxygen supplementation in hypoxemic patients who have PAH have been extrapolated
patients was not associated with any improve- from the clinical data available for chronic ob-
ment in hematologic variables, quality of life, or structive pulmonary disease (COPD). Pathophy-
survival [31]. A recent retrospective chart review siologically, hypoxemia is a potent stimulus for
of French patients who had IPAH showed that pulmonary vasoconstriction [38]. In COPD pa-
only 6.8% of all patients who had IPAH could tients who have hypoxemia, two well-designed,
be maintained on CCB therapy alone without prospective, randomized, non–placebo-controlled
any need for augmentation of therapy by trials [39,40] have shown that there is marked im-
a newer agent [32]. This is an example in which provement in long-term survival with the use of
retrospective [33] and noncontrolled prospective supplemental oxygen (Table 1) [31,39,40].
[29,34] data led to enthusiasm, first about the One trial [40] studied 87 patients who had COPD
use of acute vasoreactivity testing for assessing and a history of right heart failure (RHF) along
prognosis in PPH and second about the exces- with arterial oxygen partial pressure !60 mm Hg.
sive use of CCBs in patients who have PAH. Patients were randomly assigned to receive 15 h/d
It should be noted that the initial studies of oxygen supplementation therapy or no therapy.
[29,34], which reported that the ‘‘responders’’ Five-year mortality rate in the oxygen treatment
had an excellent (94%) 5-year survival on group was 46% versus 67% in control group (num-
CCBs, had mean fall in mean pulmonary artery ber needed to treat to save one life in approximately
pressure (mPAP) of 39% to 48% and mean fall five). The study also showed an increase in PVR in
in pulmonary vascular resistance (PVR) of 53% the control group over the study period but no
to 60%. The acceptance of relatively less strin- increase in PVR in the oxygen therapy group. The
gent criteria for favorable acute vasoreactivity study was not powered to evaluate this parameter,
response [35–37] was at least partly caused by however.
the fact that at that time the only alternative The other prospective, randomized, controlled
to CCB therapy was continuous intravenous study of 203 patients who had chronic hypoxemic
prostacyclin therapy. Because currently several COPD showed that in patients who were hypox-
oral and inhaled therapies are available and it emic at rest during the daytime, the use of
is known that as many as 46% of ‘‘acute nocturnal oxygen supplementation (12 h/d) was
responders’’ eventually will fail, CCB monother- associated with a higher 3-year mortality rate
apy [32], the need for and emphasis on perform- (42% versus 22%) when compared with the group
ing acute vasoreactivity testing, and the use of that used continuous (at least 19 h/d) supplemen-
CCBs as primary therapy for IPAH has changed tal oxygen [39]. The number needed to treat to
[32]. Reliance on CCBs in such patients and not save one life was approximately five. The survival
using specific PAH therapies may or may not be advantage was more pronounced in patients with
in patients’ best interest. less severe pulmonary hypertension at baseline
This article reviews current recommendations (mean pulmonary arterial pressure !27 mm
and widely accepted practices regarding the use Hg). Continuous oxygen supplementation therapy
of standard therapies in PAH and presents the improved long-term survival in patients who had
evidence available to support these practices. COPD with significant hypoxemia (PaO2 !55
The primary purpose of this article is to pro- mm Hg at rest) even if there was no significant im-
vide clinicians with relevant information to be provement in pulmonary hemodynamics with
able to make informed therapeutic decisions in acute oxygen supplementation. Patients with low
clinical situations in which data regarding baseline PVR had an improved mortality on con-
standard therapies are lacking, relative contra- tinuous oxygen supplementation therapy, but the
indications to these therapies exist, or the patient with high baseline PVR did not experience
Table 1
Clinical studies: oxygen therapy
Study [reference] (N) Design, method Follow-up duration, site Study findings Comments
Nocturnal Oxygen Prospective, randomized, non– At least 12 months, at Overall mortality in 12-h therapy No placebo control and no blinding
Therapy Trial placebo-controlled trial, six centers in United States group was 1.94 times greater than was used
Group, multicenter continuous group (P ¼ .01). 1-y and NOTE: the trial involved COPD pa-
1980 [39] (203) Patients with chronic, hypoxic COPD 2-y mortality was 21% and 41%, tients and not WHO I PAH
Patients received either continuous respectively, in nocturnal therapy patients; hence the results should
oxygen therapy or 12-h oxygen group versus 12% and 22%, be extrapolated with caution
therapy respectively, in continuous therapy The survival benefit was also present
group in patients with low mean pulmo-
nary artery pressure and PVR and
with relatively preserved exercise
Survival advantage was pronounced in


hypercapnic patients but was also
present with relatively poor lung
function, low mean nocturnal
oxygen saturation
Medical Research Prospective, randomized, non– Five years, at three 5-y mortality in treated group was No placebo control and no blinding
Council Working placebo-controlled trial, multicenter centers in U.K. significantly less than control was used
Party, 1981 Patients with chronic, hypoxic cor (45% versus 67%) NOTE: the trial involved COPD
[40] (87) pulmonale secondary to COPD Survival advantage from oxygen did patients and not WHO I PAH
42 patients received oxygen at 2 L/min not emerge until 500 days of patients; hence the results should
at least 15 h/d and 45 control therapy be extrapolated with caution
patients received no oxygen
Sandoval et al, Prospective, randomized, non– Two years, at a center Nocturnal oxygen therapy did not It is not known whether continuous
2001 [31] (23) placebo-controlled trial, single center in Mexico improve hematologic variables, 6- oxygen therapy will be helpful in
Patients with congenital heart disease minute walk distance, quality of this patient population
and Eisenmenger’s syndrome. 12 life, or survival at 2-y follow-up Two patients died in oxygen therapy
randomly assigned patients received groups and three patients died in
nocturnal oxygen for at least 8 h at control group; probably the
flow rate (mostly 2–3 L/min) that numbers were too small to reliably
increased the oxyhemoglobin conclude ‘‘no difference’’ between
saturation from 79  6.5% to the two groups
88  6.0% This study clearly highlights that
similar interventions may have
entirely different impact on different
disease processes

N ¼ Number of subjects.

a survival benefit. Patients who showed a large !60% of predicted) at rest should be tested for
decrease in PVR on repeat right heart catheteri- possibility of oxyhemoglobin desaturation with ac-
zation after 6 months of continuous oxygen tivity and while sleeping [42,43]. It should be noted
therapy had greater mortality compared with pa- that all such recommendations are based on expert
tients with smaller decrease in PVR [39]. This find- opinion in the absence of any direct evidence in
ing suggests that mortality benefit in hypoxemic IPAH or any form of APAH except pulmonary hy-
patients who have COPD and are on continuous pertension (PH) associated with COPD.
long-term oxygen therapy is probably not derived Many patients with congenital heart disease
from any reduction in PVR. This observation is have hypoxemia secondary to right-to-left shunt.
important from the standpoint of oxygen therapy Such hypoxemia is relatively refractory to oxygen
in patients who have PAH. supplementation therapy. The use of oxygen sup-
Because of the substantial survival (and hemo- plementation therapy in patients who have con-
dynamic) benefit observed in COPD patients who genital heart disease and Eisenmenger syndrome
have hypoxemia, the oxygen supplementation remains controversial. A study of 15 pediatric
therapy is widely used in PAH patients who patients with PAH associated with congenital
have hypoxemia. The principles and criteria used heart disease and hypoxemia initially reported
for oxygen therapy in PAH are derived from the improved mortality with oxygen supplementation
studies and practice guidelines for treatment of for a minimum of 12 h/d for 5 years [44]. A well-de-
hypoxemic patients who have COPD [41]. signed, prospective, randomized, controlled study
Many patients who have PAH have hypoxemia of 23 adult patients who had hypoxemia and Eisen-
at rest, with exertion, or during sleep. Hypoxemia menger syndrome showed that nocturnal oxygen
in PAH is thought to be caused by low mixed- supplementation had no effect on survival, quality
venous oxygen tension secondary to low cardiac of life, hematocrit, or 6-minute walking distance
output, altered diffusion capacity, and ventilation- [31]. Some studies suggest that oxygen supplemen-
perfusion mismatch. Some patients who have PAH tation in these patients may reduce need for
present with relatively rapid worsening of hypox- phlebotomy and may reduce neurologic complica-
emia secondary to opening of a patent foramen tions [35]. Nocturnal oxygen supplementation in
ovale, which results in right-to-left shunt. In such children who have PAH associated with congenital
patients, hypoxemia is relatively refractory to heart disease has been shown to decrease the rate of
oxygen supplementation. In most other patients, progression of polycythemia. Similarly in the PAH
however, with the exception of patients with pediatric population, oxygen therapy has been
advanced disease, the hypoxemia at rest and at shown to improve symptoms [44,45].
night can be corrected by oxygen supplementation PaO2 during sleep is almost always lower than
at 2 to 6 L/min via nasal cannula, although PaO2 while awake [46], which is probably second-
increased oxygen supplementation is frequently ary to sleep-induced hypoventilation and the resul-
required during activity or exertion. tant rise in PCO2. In normal individuals with
Most current guidelines [29,32–34] from Amer- a normal wake PaO2, nocturnal oxyhemoglobin de-
ican or European professional societies recom- saturation is not seen because the oxyhemoglobin
mend that all patients with PAH whose PaO2 is dissociation curve is relatively flat for PaO2 above
consistently !55 mm Hg or in whom oxyhemoglo- 90 mm Hg. Patients who are hypoxemic at rest
bin saturation is %88% at rest, during sleep, or while awake are always more hypoxemic during
with ambulation should be prescribed sufficient sleep, however [47]. This reduction in arterial oxy-
supplemental oxygen therapy to keep the pulse genation is even more pronounced during rapid
oximetry oxyhemoglobin saturation O90% at all eye movement sleep. The long-term value of treat-
times. In patients with laboratory or clinical find- ing nocturnal hypoxemia in patients who have
ings that suggest chronic hypoxemia (eg, hemato- PAH remains unclear, however. Screening with
crit O55%), clinical signs of RHF, or suggestion nocturnal oximetry may be performed to evaluate
of RHF on EKG or echocardiography, long-term possibility of nocturnal hypoxemia when clinically
oxygen supplementation therapy should be initi- suspected. Oxyhemoglobin saturation !90% for
ated at PaO2 !60 mm Hg or oxyhemoglobin satu- less than 5% of the recording time may be consid-
ration of %89%. Such patients should be retested ered clinically insignificant regardless of the lowest
for oxygen requirement 3 months after the initia- recorded oxyhemoglobin saturation value. After
tion of oxygen therapy. All patients with moderate initiation of oxygen supplementation, it is sug-
to severely decreased diffusion capacity (DLCO gested that a repeat nocturnal oximetry study be

obtained to assess the adequacy of nocturnal Oxygen supplementation or CPAP therapy or

oxygen supplementation therapy. both combined are not adequate treatments for
In most patients without significant right-to- nocturnal hypoxemia in such patients; noninva-
left shunt, a nocturnal oxygen supplementation at sive positive pressure ventilation or bilevel posi-
2 to 3 L/min above the resting oxygen requirement tive pressure ventilation should be used. In
is sufficient to maintain adequate oxygenation patients who cannot tolerate this therapy (eg, be-
during sleep. The exception to this rule is presence cause of claustrophobia), tracheostomy should be
of obstructive sleep apnea (OSA), in which nasal offered along with chronic outpatient mechanical
continuous positive airway pressure (CPAP) (or ventilation. Near complete reversal of hypercap-
BiPAP) therapy is needed to correct the sleep- nia and signs of RHF within a few months with
disordered breathing and hypoxemia. An over- invasive or noninvasive mechanical ventilation
night polysomnogram should be ordered for therapy usually are seen in patients who do not
patients in whom OSA is suspected because of have concurrent factors contributing to the devel-
history of snoring, excessive daytime sleepiness, or opment of PAH [65]. Weight loss after gastric sur-
witnessed apneas. Routine ordering of polysom- gery for morbid obesity also has been shown to
nogram in all patients who have PAH is not substantially decrease the PCO2 (mean, 52 mm
indicated [48]. Hg to 42 mm Hg) and mPAP (mean, 36 mm Hg
Prevalence of PH in OSA is much higher than to 23 mm Hg) at right heart catheterization [66].
the prevalence of IPAH in the general population; In patients who have PH and a body mass index
the reported prevalence of PH in OSA is 17% to O34, an arterial blood gas analysis obtained on
53% [48–60]. In contrast, the estimated incidence room air, should be performed to rule out possibil-
of IPAH in the general population is 1 to 2 new ity of obesity-hypoventilation syndrome (OHS)
cases per million [61], which suggests a prevalence because the management principles of OHS associ-
of !10 to 20 per million. Most of the studies [48– ated PAH are different than those of OSA [65,66].
60] that estimated prevalence of PH in OSA used There is reluctance on the part of some physi-
right heart catheterization for defining PAH. cians and many patients to consider oxygen
Data are difficult to interpret, however, because supplementation as part of therapy. Oxygen
many of these studies defined PH as mPAP R20 supplementation is a major lifestyle change for
mm Hg (less than the standard used to define patients (with the exception nocturnal oxygen
PAH) or used estimated pressures by echocardio- supplementation therapy alone). Factors such as
gram. Generally PH associated with OSA is mild cost, the type of equipment required, and the
in the absence of obesity-hypoventilation syn- inconvenience of many ambulatory oxygen systems
drome [47–59]. Any moderate or severe PH impact on patients’ acceptance and use of pre-
should not be attributed to OSA alone, no matter scribed supplemental oxygen. In clinical practice
how severe the OSA. Modest improvement in PA most patients agree to the use of oxygen if pre-
pressures is expected from the treatment of OSA; scribed by physician but do not like the oxygen
the expected median decrease in pulmonary artery tubing they need to wear in public. This has an
pressure is 3 to 6 mm Hg and expected median impact on their perceptions of themselves and
decrease in PVR is approximately 0.5 wood units others’ perceptions of them. Falls associated with
[59,62,63]. Patients who have OSA and PH with long cords and tubing attached to oxygen concen-
hypoxemia who are treated with CPAP and oxy- trators and other oxygen delivery systems can
gen supplementation have a more pronounced occur.
decrease in mPAP and PVR than patients who
are treated with oxygen therapy alone. Resolution
of PH is not an expected outcome of even the
most successful therapy by CPAP, regardless of The long-term role of diuretics in RHF and
the severity of OSA and the duration of the PAH has not been systematically studied. The use
CPAP therapy [59,62,63]. of diuretics in patients who have PAH with
Obesity-hypoventilation syndrome (usually as- peripheral edema secondary to RHF is wide-
sociated with body mass index O34 and always spread and universally accepted. Almost all
associated with the presence of daytime hyper- PAH patients with peripheral edema secondary
capnia, ie, PCO2 of R45 mm Hg) can be associ- to RHF are treated with diuretics, and the effect
ated with severe PAH and RHF, regardless of of the diuretics on edema is usually obvious. Most
presence or absence of concurrent OSA [64]. patients with edema report some improvement

after ‘‘effective diuresis’’ as assessed by change in animal models [69]. The proposed mechanism is
weight or pedal edema over relatively short inhibition of sodium-proton (Naþ/Hþ) exchange
periods of time. For many patients, maximal to prevent intracellular alkalinization, which seems
effects are only obtained when salt and fluid to play a permissive role in pulmonary artery
intake restrictions are used with diuresis. Whether smooth muscle cell proliferation in the process of
diuretics alter mortality or morbidity in patients vascular remodeling [24]. There are no clinical
who have PAH with RHF is not known, however. studies in patients who have PAH to speculate the
Diuretics also can decrease the sensation of clinical significance of these observations.
dyspnea in patients with enlarged right heart In most recent trials of pharmacologic therapy
chambers compromising left heart function (see for PAH [4–21], diuretics were used as adjunct
later discussion). therapy in a large proportion of patients, but the
The effect of most diuretics (eg, furosemide) on effect of diuretics was not tested as an end point
mortality or morbidity either in left or right heart in any of these trials. In severe RHF, large doses
failure has not been evaluated by long-term trials. of diuretics may be required for effective diuresis.
Spironolactone, an aldosterone antagonist used as Some patients may require doses as high as furose-
a potassium-sparing diuretic, decreased mortality mide, 600 mg/d, or bumetanide, 10 mg/d, or the
in a long-term prospective, randomized, placebo- addition of metolazone (up to 20 mg/d). Edema
controlled trial of patients with recent hospitali- in many patients with severe RHF and impaired
zation for severe left heart failure with NYHA III renal function may be disabling and can be refrac-
or IV symptoms [67]. In this study, a relatively low tory to even such high doses of diuretics (Box 1).
dose of spironolactone (12.5 mg/d taken orally) Right ventricular (RV) performance is preload
was used as an add-on therapy to angiotensin- dependant. Because of ventricular interdepen-
converting enzyme inhibitor therapy. Two-year dence (ie, shared pericardium and septum), how-
mortality rate in the treatment group was 35% ever, extreme RV dilatation may compromise LV
compared with 46% in controls, indicating 24% filling in part by displacing the interventricular
relative risk reduction (number needed to treat septum into the LV, which results in LV diastolic
to save one life in approximately nine). The study dysfunction and a decrease in cardiac output [70–
also showed a 35% decrease in hospitalizations 72]. Patients who have PAH are likely to develop
secondary to worsening of heart failure and im- interstitial edema at relatively lower pulmonary
provement in functional class in treated patients. capillary wedge pressures because the effective
No studies have evaluated the long-term role of hydrostatic pressure to favor interstitial edema
aldosterone antagonists in patients who have formation is equal to pulmonary capillary wedge
PAH and RHF. Secondary to convincing data re- pressure plus 40% of the gradient between
garding left heart failure, however, spironolactone mPAP and pulmonary capillary wedge pressure
is widely used in patients who have PAH and (Gaar’s equation). Relatively minor fluctuations
RHF as an adjunct to loop diuretics. Long-term in fluid status may result in significant change in
value of spironolactone in such patients remains cardiac output or gas exchange, resulting in arte-
unknown. It should be noted that many interven- rial hypoxemia. Caution should be observed
tions known to have a beneficial effect in LV fail- when adjusting diuretic dose. There is no evidence
ure (eg, angiotensin-converting enzyme inhibitors or scientific rationale to urgently diurese patients
and beta adrenergic antagonists) have not been or remove large volumes of fluid over a short
proven to be beneficial in patients who have period of time, except when there is cardiogenic
PAH and RHF. Spironolactone should not be pulmonary edema or acute worsening of arterial
used in patients with serum creatinine O2.5 mg/dL hypoxemia. ‘‘Optimal fluid status’’ in patients
or potassium O5.0 mEq/L or in patients with who have PAH and RHF is often a relatively nar-
history of severe refractory hyperkalemia. Extreme row window in which patient’s cardiac output is
caution should be observed in patients who have maintained, arterial hypoxemia is minimized,
impaired renal function or diabetes, in elderly and signs of tissue perfusion (as assessed by renal
persons, and in patients with concurrent use of function, presence or absence of peripheral cyano-
angiotensin-converting enzyme inhibitors or non- sis, or altered mentation) are optimal.
steroidal anti-inflammatory agents [68]. Severe passive hepatic congestion and bowel
Amiloride (a potassium-sparing diuretic) ana- wall edema and ischemia may be present when
logs have been shown to inhibit development of patients who have severe RHF and reduced
hypoxia-induced pulmonary hypertension in cardiac output present with significantly deranged

liver function tests or a sepsis-like picture second-

Box 1. Management options ary to translocation of colonic micro-organisms as
for refractory edema in pulmonary a consequence of bowel wall edema and bowel
arterial hypertension and severe chronic ischemia. Patients may present with abdominal
right heart failure pain, mild diffuse abdominal tenderness (without
rebound or rigidity), and occult blood in the
History and evaluation stools with or without gram-negative bacteremia.
 Carefully review compliance with salt This may be a terminal event in patients who have
and water intake restriction. PAH and severe RHF. The appropriate therapy
 Consider concurrent medications may include use of diuretics or IV fluids. Depend-
(eg, CCB, bosentan, sitaxsentan, or ing on the individual patient, low-dose dopamine
prostacyclin analogs) and temporal or dobutamine (2–3 mg/kg/min) and inhaled NO
relation to the worsening of edema. (10–40 ppm) or inhaled prostacyclin (started at
 Review possibility of worsening of 50 ng/kg/min and titrated down to lowest effective
underlying disease by other dose) IV antibiotics covering colonic flora (ie,
parameters (eg, 6-minute walking covering gram-negative and anaerobic bacteria)
distance, oxygen requirement, RV should be considered. There are no published data
hypertrophy on echocardiography). to define the optimal way to manage acutely
 Reassess renal function. decompensated patients who have severe PAH
 Consider discontinuing all and acute RHF.
nonsteroidal anti-inflammatory drugs. Diuretic dosing should be revised if patients
develop orthostatic dizziness, which temporally
Options correlates with decreased edema, decreased weight,
 Increase the dose of current and recently increased diuretic dose. Charting of
loop diuretic (may use up to daily weights, use of written sliding scale for
furosemide, 600 mg/d, or bumetanide, diuretics, and potassium use, based on edema or
10 mg/d). weight, may be useful strategies. Many patients
 Add a diuretic from another class who have severe PAH and RHF have relatively low
(metolazone, 2.5–5 mg twice daily, or systemic blood pressure. Diuretics should be used
spironolactone, 12.5–50 mg twice with caution in these patients. Concern over low
daily). Do not use spironolactone if blood pressure and subsequent underdosing of
serum creatinine is above 2.5 mg/dL or diuretics actually may result in worsening of edema
above 1.6 in patients older than or overall clinical worsening, however [68].
70 years. Do not use in patients who Potassium supplementation should be man-
have diabetes or potassium above aged carefully, especially if there is worsening or
4.5 or history of severe hyperkalemia. fluctuation of renal function, addition or with-
 Change current loop diuretic to drawal of potassium-sparing diuretics, or concur-
another loop diuretic. rent use of digoxin. Arrhythmias, prolonged
 Consider hospitalization and palpitations, or syncopal episodes may be second-
intravenous diuretics. ary to PAH and RHF; however, in patients on
 Consider hospitalization and diuretics, electrolyte and digoxin level testing
ionotropic support by very low dose should be considered.
(2–3 ng/kg/min; avoid higher doses) Edema is a relatively common side effect of
IV dopamine or dobutamine. many PAH therapies (eg, CCBs [2%–15%],
 Consider atrial septostomy (if bosentan [4%–8%], sitaxsentan [7%], and less
appropriate for patient). commonly, prostacyclin). Adjusting the dose of
 Consider lung transplantation (if diuretics often can ameliorate such edema. In
appropriate for patient). some instances, however, this medication-related
 Consider right ventricular assist device edema is refractory to diuretics. Patients may
(if appropriate for patient). benefit from scheduled elevation of legs, com-
 Tolerate current level of edema. pression stockings, or, if possible, switching
 Discuss end-of-life issues (if from the suspected causative agent to other
appropriate for patient). alternatives. In other instances in which the
PAH therapy is otherwise believed to be

effective, patients are asked to tolerate increased output and decrease circulating norepinephrine
pedal edema. levels in a small, short-term, uncontrolled study
(Table 2) [23]. This study included 17 patients
who received a single dose of 1 mg IV digoxin
over a 30-minute infusion; hemodynamic mea-
Cardiac glycosides such as digoxin increase myo- surements were made 2 hours after the infusion
cardial contractility by blocking sodium-potassium and were compared with baseline hemodynamic
adenosine triphosphatase pump (Naþ-Kþ ATPase) values. A prospective, randomized, double-blind,
[73], which is present in the cell membrane of placebo-controlled study [74] of 15 patients who
myocardial cells. In patients who have PAH had COPD and RHF and no clinical LV failure
and RHF, a single intravenous administration found that 8 weeks administration of digoxin
of digoxin has been shown to increase cardiac did not improve RV ejection fraction in any

Table 2
Clinical studies: pulmonary arterial hypertension and digoxin
Study Follow-up
[reference] duration,
(N) Design method site Study findings Comments
Mathur et al, Prospective, randomized, 8 wk, At baseline, RV It is not clear whether the
1981 [74] placebo-controlled McMaster ejection fraction was reduced LVEF at baseline
(15) trial, single center University reduced (on actually represents LV
Patients who have in equilibrium disease or simply reflects
RV failure secondary Canada radionuclide severely dilated RV causing
to COPD and no angiography) in all LV dysfunction or coexisting
known LV failure patients and reduced LV disease; conversely,
LVEF was seen normalization of LV systolic
in 4 patients function by 8-wk digoxin
8 wk digoxin use, all therapy in all 4 patients
the abnormal LVEF with low baseline LVEF is
were normal; unusual for LV systolic
however, Digoxin dysfunction
increased RVEF NOTE: the trial involved
only in patients COPD patients with RV
who had failure and not WHO I PAH
reduced LVEF patients; hence the results
at baseline should be extrapolated with
Rich et al, Prospective, single-center Short-term Single administration No placebo control was used
1998 [23] study; patients’ baseline study of 1 mg IV digoxin NOTE: the study was
(17) was used as control involving was associated with short-term and no definitive
Patients with PPH diagnosis hemodynamic significant increase conclusion can be drawn
by NIH registry criteria measurements in cardiac out put regarding effect on cardiac
and normal LV function before and (3.49  1.2 to output after long-term use
on echocardiogram after single 3.81  1.2 L/min) NOTE: the trials in LV systolic
administration measured 2 hours dysfunction did not show
of digoxin after the completion any prolongation of survival
1 mg IV over on digoxin infusion with long-term digoxin use
30 min, (subgroup analysis, however,
University of did show reduced mortality
Illinois and hospitalizations in
patients with LVEF below
25%, NYHA class III or IV
functional class, or patients
with cardiomegaly on chest
N ¼ Number of subjects.

patient who did not have reduced left ventricu- toxicity is generally not seen below a plasma di-
lar ejection fraction (LVEF) on baseline equilib- goxin concentration of 2.0 ng/mL (although di-
rium radionuclide angiography (see Table 2) goxin toxicity may rarely be seen at lower
[23,74]. The safety, efficacy, and impact on mor- concentrations) [78,79]. Digoxin should not be
bidity or mortality with long-term use of digoxin used in cases of suspected coronary ischemia or
in patients who have PAH and RVF are not in patients with recent acute coronary syndrome
known. It should be noted that in patients because of increased risk of death from arrhyth-
with systolic LV dysfunction and NYHA class mia or myocardial infarction [68,75,80,81]. Con-
II and III symptoms, treatment with digoxin current use of clarithromycin, amiodarone,
for 2 to 5 years did not have any impact on itraconazole, cyclosporine, and many other drugs
mortality [75]; there was some beneficial effect can increase digoxin levels and increase the poten-
on rate of congestive heart failure-related tial for toxicity [82–84].
hospitalizations, however.
Currently, digoxin is used in patients who
have PAH and signs of RHF and in patients
who have atrial rhythm abnormalities (eg, multi- Currently no prospective, randomized, pla-
focal atrial tachycardia, atrial flutter, or atrial cebo-controlled trials are available to provide
fibrillation). Some experts advocate using digoxin evidence for or against the use of anticoagulation
with calcium channel antagonists to counteract in patients who have PAH. Currently, it is a widely
negative ionotropic effects of CCBs. Because accepted practice to use anticoagulation in pa-
most patients who have PAH and RHF are on tients who have PAH (WHO group I [35]) unless
digoxin and diuretics, it is important to take there is a contraindication.
extreme caution to monitor digoxin levels and Patients who have significant PAH may have
promptly recognize toxicity, especially in associ- a sedentary lifestyle. Venous engorgement and
ation with electrolyte disturbances that result stasis (as a result of elevated right atrial pressures)
from the use of diuretics. Paroxysmal atrial and poor flow through pulmonary and systemic
tachycardia with variable atrioventricular block circulations as a consequence of low cardiac output
and bidirectional tachycardia are well-known place them at increased risk for developing venous
arrhythmias associated with digoxin toxicity. thromboembolism. The pulmonary vascular bed is
The half-life of digoxin is prolonged with wors- already significantly compromised in patients who
ening of renal function. Digoxin is renally cleared have PAH by the time they become symptomatic
and is used in most US centers, whereas digitoxin and the diagnosis is made. As a consequence, even
is hepatically cleared and is mostly used in a relatively minor pulmonary embolic event has
European centers. more significant hemodynamic and gas exchange
There are no specific recommendations re- consequences and may be life threatening.
garding how to monitor digoxin therapy in Substantial evidence from human studies of
patients who have PAH, so guidelines developed biochemical and serologic markers suggests pres-
for monitoring of digoxin therapy in patients with ence of a prothrombotic state in patients who
LV failure are relied upon [68]. Digoxin therapy have PAH. Three key elements that influence
may be started at 0.125 to 0.25 mg orally daily; existence of a prothrombotic state are endothelial
however, many centers only use low-dose (ie, function, platelet activation, and plasma proteins
0.125 mg/d) therapy. In patients with normal related to coagulation and fibrinolysis. Certain
renal function the digoxin levels may be first components of these three elements have been
checked after 1 week. In patients above age 70 shown to have aberrant behavior favoring a pro-
with impaired renal function or low body weight, thrombotic state in several human studies in-
a reduced starting dose [76] (0.125 mg every other volving patients with idiopathic and associated
day) is more appropriate. When digoxin is used forms of PAH (Table 4) [85–93]. The evidence of
for positive ionotropic effect, it is generally recom- these prothrombotic abnormalities is more ro-
mended to keep the digoxin trough level between bust in IPAH and chronic thromboembolic pul-
0.5 and 1.0 ng/mL [68]. At a plasma concentration monary hypertension (CTE-PHT) than most
above 1.0 ng/mL, the risk-adjusted mortality in- other forms of PAH. Some of these abnormali-
creases, as shown by the Digitalis Investigation ties have been shown to reverse with specific
Group Trial [77]. In the absence of hypokalemia, PAH therapies, such as intravenous prostacyclin
hypomagnesemia, or hypothyroidism, digitalis [94–97] or bosentan [98].
Table 3
Clinical studies: pulmonary arterial hypertension and calcium channel antagonists
[reference] Follow-up
(N) Design method duration, site Study findings Comments
Rich et al, Prospective, single center, 1 y, University At 1 y, 5 patients returned for follow-up; 4 of Nifedipine was used if heart rate was O100/min,
1987 uncontrolled of Illinois these 5 patients had sustained reduction in PVR diltiazem was used if heart rate !100/min,
[34] (13) PPH patients who were and PAP and regression of RV hypertrophy by starting
‘‘responders’’ as defined by echocardiogram and electrocardiogram dose 20 mg or 60 mg, respectively. This was
patients who had a fall in NOTE: In the study once the effective dose was followed by repeated doses every 1 h (all patients
PAP and PVR by 20% determined, the drug was administered in had PA catheter in place) until 50% fall in PVR,
after acute CCB challenge divided doses (every 4–8 h) to achieve total desired 33% fall in mPAP, or an untoward side effect
(8 out of 13 patients) and daily dose. More frequent and small doses (every 4 [eg, hypotension, dizziness, or worsening of


were treated with high-dose h) were used in patients who did not tolerate hypoxemia])
CCB in long-term higher, less frequent dosing (every 8 h). Some Long-term dose of CCB: Up to 720 mg/d diltiazem
NOTE: acute vasoreactivity patients were started at low dose of CCB or 240 mg/d nifedipine
testing with CCB has been (nifedipine, 20 mg, tid or diltiazem, 60 mg, tid) as No patient was treated with anticoagulants
abandoned secondary to outpatients and the dose was slowly titrated to The responders in this study had fall in PAP by 48%
the high frequency of desired dose (as determined at the time of acute and in PVR by 60%, which is much more
untoward events such as CCB testing) over 6 wk. Digoxin was also started pronounced than most current criteria for
profound hypotension. in these patients to counteract negative defining a positive acute vasoreactivity test
Currently the testing is ionotropic effects of CCB.
performed with inhaled
nitric oxide, IV adenosine,
IV or inhaled prostacyclin
Rich et al, Prospective, single center, Up to 5 y, Patients who were ‘‘responders’’ as defined by This study also involved use of warfarin in all
1992 uncontrolled (historical University patients who had a fall in PAP and PVR by PPH patients with nonuniform perfusion on V/Q
[29] (64) control from NIH PPH of Illinois 20% after acute CCB challenge (26% of all scan (55% of all)
registry) patients) were treated with high-dose CCB. 47% of responders and 57% of nonresponders had
High-dose CCBs were used in Treated patients had much better 5-y survival nonuniform perfusion of V/Q scans
patients who were rate (94% in treated responders versus 55% The responders in this study had fall in PAP by 39%
‘‘responders’’ in nonresponders) and in PVR index by 53%, which is much
(Note: Mean dose of nifedipine Survival of responders was much better than more pronounced than most current criteria
was 172  41 mg/d, historical control from NIH PPH registry for defining a positive acute vasoreactivity test
and of diltiazem was patients (in which vasoreactivity status was
720  208 mg/d) not known) at the same institution
Barst et al, Retrospective, single center All patients from Patients who were ‘‘responders’’ (31 responders out ’’Responders’’ were defined as patients who had
1999 Review of medical records 1982–1987, of 74 tested; ie, 42%) were treated with high-dose a fall in PAP and PVR by 20%, no change or
[33] (77) of pediatric patients with Columbia CCB. Survival of responders was better than increase in cardiac index, no change or decrease
PPH (mean age 7 years) University, nonresponders (1-, 3-, and 5-y survival rates were in PVR/SVR ratio, after acute testing with
(range: 7 mo–13 y) New York 97%, 97%, and 97%, respectively, in responders prostacyclin
Median versus 66%, 52%, and 35%, respectively, in Warfarin was started in all the patients in mid-1980s
follow-up: nonresponders). 6 nonresponders were also when the adult studies showed survival benefit
47 mo (range: treated with CCB in this study
24–166 mo)
Sitbon et al, Retrospective, single center All IPAH patients Only 6.8% of all IPAH patients undergoing acute ‘‘Responders’’ were defined as patients who had
2005 [32] Review of medical records from vasoreactivity testing are long-term CCB a fall in both PAP and PVR by 20%, after acute
(557) of patients with IPAH 1984–2001, responders, defined as patients who were in testing with IV prostacyclin (until 1994) on
Universite NYHA class I or II after 1 y of CCB inhaled NO (after 1994); 70 of 557 (12.6%)
Paris-Sud, monotherapy were vasoreactive; 38 of 70 (ie, 6.8% of all 557)
Clamart Predictors of long-term CCB response: were long-term CCB responders
France On multivariate analysis: (although the numbers The long-term CCB responders had less severe


Median are too small to reach sufficient power): Higher baseline PAH (PVR 10.3  4.6 versus
follow-up: baseline mixed venous oxyhemoglobin 14.9  5.3 wood units, CO 5.0  1.5 versus
30  28 mo in saturation (69  8 versus 61  8), and lower 3.9  1.1 L/min). At acute testing they had
nonresponders PVR absolute value reached after acute more pronounced fall in mPAP (39  11 versus
and 7  4.1 y vasoreactivity testing (5.2  2.7 versus 26  7 mm Hg) and their mPAP reached
in responders 8.6  3.3 wood units) a lower absolute value (33  8 versus
On univariate analysis: baseline SVO2 !65% 46  10 mm Hg) compared with nonresponders
(OR: 19.18), absolute value of PVR reached
post- acute vasoreactivity testing of !6.7 wood
units (OR: 7.35), post- acute vasoreactivity
testing fall in mPAP fall of O31% (OR: 7.35),
absolute value of post-acute vasoreactivity
testing on mPAP of !37 mm Hg (OR: 6.13),
baseline PVR of !11.5 wood units (OR: 4.24),
post-acute vasoreactivity testing fall in PVR of
O45% (OR: 3.27), baseline cardiac index of
!2.5 L/min/m sq. (OR: 3.21), absence of class
III or IV dyspnea at baseline (OR: 3.02)
N ¼ Number of subjects.


Table 4
Prothrombotic abnormalities in pulmonary arterial hypertension
Abnormalities Subjects Reference
I Platelets aggregation related
Urinary thromboxane metabolite (11 dehydroxy TxB2) IPAH [115]
Urinary PGI2 metabolite (PGI-M) IPAH [115]
Thromboxane A2 IPAH [115]
Prostacyclin IPAH [115]
Nitric oxide (exhaled or urinary excretion) IPAH, APAH [85]
Circulating platelet aggregates APAH [86]
Plasma serotonin, platelet serotonin IPAH [87,88]
Plasma-P selectin IPAH [94]
Thrombomodulin IPAH [89,94,99]
II Endothelial function related
Endothelial NO synthase IPAH [90]
Prostacyclin synthase expression IPAH [116]
Urinary PGI2 metabolite (PGI-M) IPAH [115]
Urinary thromboxane metabolite (11 dehydroxy IPAH [115]
Thrombomodulin IPAH [89,94,99]
VWF IPAH, APAH [91,99]
Fibrinogen inhibitor plasminogen activator 1 IPAH [92,99]
III Coagulation and fibrinolytic proteins related
Prevalence of antiphospholipid IPAH, CTE-PHT [93]
antibody/lupus anticoagulant
vWF antigen level IPAH [91,99]
Fibrinogen inhibitor IPAH [92,99]
plasminogen activator 1
Euglobulin lysis time IPAH [99]
Fibrinogen level APAH [99]

Theoretical concerns about importance of among various anticoagulation agents such as

thromboembolism are supported by various early warfarin, unfractionated heparin, or low molec-
observations (Table 5) [24–26,99,100]. ular weight heparins. Warfarin is the most
Retrospective [27,28,101] clinical studies and a commonly used agent for anticoagulation in
small prospective [29], noncontrolled, nonrandom- patients who have PAH. Substantial data from
ized clinical study of PPH patients has shown that animal studies suggest that heparin may have
use of anticoagulation with warfarin seems to some additional therapeutic advantage in subjects
confer a survival advantage in patients who have with PAH. Human studies, however, have not
PPH (Table 6) [27–29,101–103]. In PAH, other been performed to support this theoretical supe-
than IPAH, the evidencedeither basic science or riority. Heparin has been shown to prevent
clinicaldfor or against the use of anticoagulation development of PAH and RV hypertrophy in
is even scarcer (with the exception of CTE-PHT, animal models (hypoxic mice or guinea pig model
in which the need for chronic anticoagulation is of PAH) [105,106]. The mechanism of action of
well established). Because there are many similari- heparin on pulmonary vasculature is not com-
ties in the clinical course, hemodynamics, and his- pletely understood. Proposed mechanisms include
topathology of these and IPAH patients, the inhibition of platelet-derived growth factor [105]
practice of anticoagulation has been extended to and inhibition of pulmonary artery smooth mus-
other forms of PAH. It should be noted, however, cle cell growth, probably by upregulation of
that the risk of bleeding complication (Table 7) expression of cell cycle regulation of gene p27,
and need for anticoagulation (Box 2) may differ in which influences the level of cyclin-dependent ki-
other causes of APAH [104]. nase inhibitor [107]. Cyclin-dependent kinase and
There are no data from human studies as to cyclin-dependent kinase inhibitor play key roles
whether there is any difference in terms of efficacy in the balance between cell proliferation and

cell quiescence. Research also has shown in ani- least one subsegmental mismatched defect
mal models that anticoagulation with warfarin or diffusely decreased tracer uptake in certain
does not have the same effect as that of heparin subsegments or in whom PA angiograms are
to protect animals from developing PAH upon unequivocal for the absence of CPTED
prolonged exposure to hypoxia [108]. O-hexanoyl 4. In patients with history of ischemic stroke or
low-molecular-weight heparin derivatives have transient ischemic attack with known right-
been shown to be more effective in growth inhibi- to-left shunt
tion of bovine pulmonary artery smooth muscle
The role of antiplatelet agents such as aspirin
cells in culture than heparin [109]. Because of
and clopidogrel has not been thoroughly evaluated.
such observations, the authors concluded that
A recent preliminary randomized, double-blind,
heparin derivatives may be envisioned as poten-
placebo-controlled, crossover study of 19 patients
tial future PAH therapy. Currently, there are
who have IPAH explored the biochemical effects of
no human clinical data to support the use of hep-
clopidogrel and aspirin on inhibition of platelet
arin or heparin derivatives instead of warfarin in
aggregation eicosanoid metabolism [111]. The
patients who have PAH.
study represents the first well-structured, placebo-
In clinical practices, warfarin has been the
controlled trial to evaluate biochemical effects of
agent most frequently used. Target international
antiplatelet agents in patients who have IPAH.
normalized ratio (INR) in most US centers is
The study showed that both drugs inhibit platelet
between 1.5 and 2.5 and in many European
aggregation in patients who have IPAH. Arachi-
centers it is 2.0 to 3.0. Both approaches are based
donic acid–induced platelet inhibition was more
on expert opinions, weighing potential benefits of
completely blocked by aspirin, whereas ADP-
therapy versus risk of bleeding complications with
induced platelet aggregation was more effectively
higher target INR [110]. It may be appropriate to
blocked by clopidogrel. Only aspirin inhibited
keep INR O 2.0 in the following situations:
thromboxane metabolite production without af-
1. In any patient in whom there exists a clinical fecting prostaglandin I2 metabolite synthesis, thus
indication for anticoagulation (eg, atrial restoring normal eicosanoid balance.
fibrillation, CTE-PHT, acute recent venous Empiric anticoagulation may be considered
thromboembolism, or prosthetic valve) if clinical suspicion of PAH is moderate to high
2. In patients with remote history of idiopathic and an echocardiogram shows an estimated pul-
venous thromboembolism monary artery pressure O60 while the patient
3. In patients in whom ventilation-perfusion is undergoing evaluation [112]. When an interrup-
(V/Q) scans or PA angiograms are not con- tion in anticoagulation therapy is required, such
sistent with chronic pulmonary thromboem- as before a surgical procedure, in the absence of
bolic disease (CPTED) but who have at any other indication for anticoagulation warfarin
may be stopped 5 to 7 days before the procedure
and restarted after the procedure when the in-
creased risk of bleeding has resolved, without
Box 2. Pulmonary arterial any overlap with intravenous heparin or subcuta-
hypertension–associated conditions neous low molecular weight heparin. When the
in which anticoagulation management INR is !1.5, appropriate deep venous thrombo-
is dictated by principals sis prophylaxis should be instituted when patients
of the associated disease are hospitalized or as appropriate for the patient’s
clinical status [113].
CTE-PHT (also consider inferior
vena-caval [IVC] filter) Calcium channel blockers
Mitral or other valve prosthesis
High-risk atrial fibrillation The hallmark of PAH is increased PVR, which
Osler-Weber Rendu syndrome and compromises the ability of the right ventricle to
similar diseases with history of maintain cardiac output [3]. Increased RV stroke
ischemic strokes or transient ischemic work (volume of blood pumped  pressure
attacks against which the volume is pumped) against the
Lupus anticoagulant syndrome or other abnormally increased PVR has been thought to
procoagulant states be the cause of RV hypertrophy and dilatation
and, eventually, failure to pump effectively
Table 5
Observations suggesting role of thrombosis in patients who have idiopathic pulmonary arterial hypertension
[reference] (N) Design method Study findings Comments
Wagenvoort, Retrospective 12/40 (30%) patients with unexplained PAH had The criteria for calling a biopsy consistent with
1980 [24] (72) Review of lung biopsy specimens evidence of CPTED (ie, severe eccentric intimal CPTED may or may not correctly identify
from patients with pulmonary fibrosis) in the absence of concentric lesions CPTED cases
vascular disease It is not clear whether all of these patients were
evaluated by V/Q scan or pulmonary artery
angiogram before the biopsy
Bjornsson et al, Retrospective 56% of the patients had biopsies consistent with Mean age for primary pulmonary arteritis on


1985 [25] (85) Review of lung biopsy specimens thromboembolic disease biopsy was 16 years and 21–34 years for
from patients with clinical plexogenic pulmonary arteriopathy, primary
diagnosis of PPH from medial arteriopathy, and pulmonary
1930–1983 at Mayo Clinic veno-occlusive disease versus 41 years for
patients with thromboembolic disease
Biopsy specimens were taken from tissue registry
(73 autopsies, 6 open lung biopsies, 1 both)
Loyd et al, Retrospective 18/23 (78%) patients from 12/13 (92%) families The study represents one of the most insightful
1988 [100] Review of histopathology of had vessels with organized thrombi reviews of pathologic changes in PAH
(Familial postmortem lung sections 23 18/23 (78%) patients from 12/13 (92%) families Authors concluded that coexistence of thrombotic
PAH: 23) members of 13 families with had plexiform lesions and plexiform lesions within the same family
familial PAH; pathologists were 2.7% of all arteries had organized thrombi and individual represents that these lesions are
blinded to patient and family 3.4 of all arteries had plexiform lesions not specific but they represent different
identity; every artery and 15/23 (65%) patients had coexisting plexiform and manifestations of the same initial pathologic
abnormal vein was categorized by thrombotic lesions process
adopted WHO classification that There was no correlation between type of lesion or These lesions are probability two different
was used in pathology core NIH percent of arteries with plexiform lesions outcomes of a single starting pathologic process
PPH registry between age of patient or survival in this study in a disease caused by autosomal dominant
2516 vessels (28–264 with mean of inheritance pattern
109  62 SD per patient) were The differentiation of these lesions may depend on
described the anatomic location of the lesion, other
Quantitative description was reported factors in the local lung environment, or
a random phenomenon

(maintain cardiac output) against increasing resis-

Prognosis of patients with thrombotic lesions was

low probability V/Q scans or normal pulmonary

of or cause of elevated mean pulmonary artery

Whether decreased fibrinolytic activity is an effect
NOTE: the NIH registry only included patients with

Lower fibrinolytic activity correlated with higher

artery angiogram; all patients were required to
much better than for patients with plexiform

tance. The idea of attempting to decrease PVR in

PAH by vasodilation has always been tempting
have V/Q scan or pulmonary angiogram
[114]. Multiple basic science experiments and clin-
ical studies suggest that in animal models of PAH
mean pulmonary artery pressures
lesions or veno-occlusive disease

and in patients who have PAH, the mediators of

pulmonary vasodilation are decreased, whereas
the mediators of vasoconstriction are overex-
pressure is not clear pressed [115–117]. Clinical studies indicated that
the ability to successfully produce vasodilation
by certain agents with known vasodilator proper-
ties (eg, CCBs [118], IV prostacyclin [119,120],
inhaled NO [121–124], IV adenosine [125], or
inhaled iloprost [126]) in patients with PPH iden-
tifies the patients who have much better prognosis
[29,32–34,127], particularly if they are treated with
high-dose CCBs (Table 3) [29,32–34]. Such testing
(known as acute vasoreactivity testing [Table 8])
 Fibrinolytic inhibitor plasminogen activator-I
both recanalized thrombi and eccentric intimal
19/48 (40%) patients who had luminal or intimal

9/25 (36%) patients with plexiform lesions also

lesions had thrombotic lesions as defined by

became a routine for almost all patients who

 Thrombomodulin levels were decreased

have PAH and are undergoing initial diagnostic

 Von Willebrand factor was increased

right heart catheterization, except patients with

 Euglobulin lysis time was increased

low cardiac output or concurrent left heart failure,

fibrosis without plexiform lesions

 Fibrinogen level was increased

Compared with the control group

as indicated by elevated pulmonary capillary

wedge pressure. Patients who have significant
(see Table 6) pulmonary vasodilation without
had recacalized thrombi

a significant systemic vasodilation (as assessed

by change in PVR/SVR ratio or clinically unac-
was increased

ceptable fall in systemic blood pressure) in acute

In SPH patients:
In PPH patients:

vasoreactivity testing are called ‘‘responders’’

and are candidates for long-term CCB therapy.
The current definition of a responder is a decrease
in mean pulmonary arterial pressure of O10 mm
Hg to a mean pulmonary artery pressure of %40
mm Hg with a maintained (or increased) cardiac
Retrospective (although patients were

output. ‘‘Responders’’ comprise 10% to 26% of

collected and multiple coagulation
patients in NIH/NHLBI registry

Plasma samples of PPH, SPH, and

related parameters were studied

IPAH patients undergoing such testing; unfortu-

Review of histopathology of lung

autopsy specimens from PPH

nately, the proportion of patients who are re-

biopsy, pneumonectomy, or

age-matched controls were

sponders among APAH (eg, scleroderma or

identified prospectively)

CTE-PHT) is even lower [128]. The role of acute

vasoreactivity testing and the use of CCBs in pa-
of PPH patients

tients who have APAH who demonstrate acute


vasodilation are not known.

There are reports of patients who have IPAH
and who were initially nonresponders to acute
vasoreactivity testing, transforming to responders
N ¼ Number of subjects.

at repeat right heart catheterization after pro-

longed use of IV prostacyclin [129]. No data are
available to define the safety or role of CCBs in
1996 [99] (PPH:

such patients. A recent report suggests that only

1989 [26] (58)

12, SPH: 25,

Control: 15)

a small minority (6.8%) of all patients who have

Welsh et al,
Pietra et al,

PAH can be managed by CCBs in the long-term

without need for additional specific PAH therapy,
because as many as half of the patients on CCBs
deteriorate in the subsequent years [32]. Because
Table 6
Clinical studies: anticoagulation and pulmonary arterial hypertension
Study Follow-up duration,
[reference] (N) Design method site Study findings Comments
Fuster et al, Retrospective, single PPH patients from Of 56 patients who underwent autopsy, 57% Improved survival on anticoagulation was
1984 [27] center 1955–1977 who patients had changes consistent with only noted in thromboembolic and plexogenic
(120) Review of medical were followed up thromboembolic PH in the absence of arteriopathy, but numbers were too small
records of patients until 1983 at plexogenic pulmonary arteriopathy to reach statistical significance (to ascertain
with PPH diagnosis Mayo Clinic, with Patients who were treated with statistical power)
by clinical and minimum anticoagulation as defined by initiation The criteria for deciding not to anticoagulate
hemodynamic criteria follow-up of 5 y within 1 y of diagnosis had significantly are not clear and might have introduced
(median 14 y, better 3-y survival (49% versus 21%) a selection bias
longest 27 y) or up compared with those who were not


to death anticoagulated
Rich et al, Prospective, Up to 5 y, Warfarin use was associated with improved This study also involved use of high-dose
1992 [29] (64) uncontrolled, single University of survival, particularly in patients who were CCB in patients who were ‘‘responders’’ as
center (historical Illinois nonresponders to high-dose CCB (1-y defined by patients who had a fall in PAP
control from survival: 91% versus 62%, 3-y survival: and PVR by 20% after acute CCB
NIH PPH registry) 47% versus 31%) when compared with challenge
Warfarin was given to all historical control from NIH PPH registry 47% of responders and 57% of
PPH patients with nonresponders had nonuniform perfusion
nonuniform perfusion of V/Q scans Patients with high probability
on V/Q scan (55% of V/Q scans were excluded from the study
Frank et al, Retrospective, two Approximately up to Anticoagulated aminorex-treated patient had Contrary to most studies cited in this table,
1997 [101] centers were involved 10 y, University of better mean survival compared with this study showed no difference in survival
(173) Review of medical Vienna and nonanticoagulated aminorex-treated in first 5 y of follow-up of anticoagulated
records of patients University of Bern patients (8.3 versus 6.1 y) and nonanticoagulated PPH patients. In
with PPH (total 64, 24 Aminorex PHT patients who started receiving the following 5 y, some survival advantage
received warfarin) and anticoagulation therapy sooner had better was noted in anticoagulated PPH group
anorectic mean survival compared with patients who that did not reach statistical significance.
drug-induced PAH stated treatment 2 y after diagnosis (10.9 NOTE: the study included only 24 PPH
(total 104, 56 versus 5.9 y) patients who received anticoagulation
received warfarin) Overall, patients treated with aminorex had
better mean survival compared with PPH
patients (7.5 versus 3.9 y)

of such reports and the availability of newer safe

were 87%, 75%, and 61%, respectively, for

Note: There are other less structured retrospective studies or case series, some of which favor anticoagulation [102,127] and others [103] that reported no difference in
The criteria for deciding not to anticoagulate
1-, 3-, and 5-y transplant-free survival rates

are not clear and might have introduced

and effective oral and inhaled (rather than paren-
teral) PAH therapies, the enthusiasm to support
the necessity of acute vasoreactivity testing as an
essential part of PAH evaluation has somewhat
decreased. Currently, however, acute vasoreactiv-
ity testing continues to be part of the algorithm
for the evaluation of patients who have PAH (par-
ticularly IPAH). The predictors of long-term re-
the whole cohort

a selection biasa

sponse to CCBs without the need for additional

agent are listed in Table 3 [32].
It should be noted that 80% to 90% of all
patients who have IPAH and a greater percentage
of patients who have APAH are nonresponders at
acute vasoreactivity testing, and these patients are
treated with specific PAH therapies (eg, prostacy-
clin analogs, endothelial receptor antagonists, or
(hazard ratio: 0.35, 95% CI: 0.12–0.90,

5-phosphodiestrase inhibitors). Importantly, the

transplant-free survival was improved

lack of response to acute vasoreactivity testing (ie,

the administration of inhaled or IV prostacyclin,
Among the patients treated with
anticoagulation (94% of all),

NO, or adenosine) does not rule out the likelihood

of improvement with long-term use of prostacy-
clins or other specific PAH therapies [4–21].
The primary purpose of acute vasoreactivity
testing is to determine whether CCBs could be
used as PAH therapy. The CCBs are not used in
patients who have significant acute RHF because
P ¼ .05).

of cardiodepressor effects. There is no role of acute

vasoreactivity testing in patients who have PAH
survival among anticoagulated and non-anticoagulated patients who have PAH.

who are hospitalized with acute worsening of

RHF. Similarly, if at the time of right heart
catheterization (RHC) the cardiac output is de-
All consecutive adult

between 1994 and

follow-up, 764 d)
initial evaluation

University, New

termined to be low (CI !2.0) acute vasoreactivity

2002, Columbia

York (median

should not be performed because it will have no

patients with

impact on the treatment and it may be associated

with an undesirable fall in systemic blood pressure
or cardiac output. In any patient, acute vaso-
reactivity testing may precipitate acute pulmonary
edema and marked worsening of hypoxemia [130].
It should be done with caution in patients who
have a pulmonary capillary wedge pressure of
records of patients

familial PAH (14),

Retrospective, single

R15 cm H2O. In the event of any sudden worsen-

with IPAH (66),
Review of medical

ing in oxyhemoglobin saturation during acute vas-

oreactivity testing, acute pulmonary edema should
PAH (4)

be considered. The vasodilator agent should be


N ¼ Number of subjects.

discontinued and treatment with intravenous mor-

phine, intravenous nitroglycerine, and furosemide
may be considered [130]. It should be noted that
acute severe pulmonary edema during acute vaso-
2005 [28] (84)

reactivity testing is not a contraindication to the

Kawut et al,

use of pulmonary vasodilator agents, such as in-

travenous prostacyclin. This should be done with
extreme caution, and dose escalation should be

done slowly [130]. Some authorities favor testing

Table 7
Associated pulmonary arterial hypertensive conditions and clinical circumstances with increased rate of complications from anticoagulation
Diseases/circumstances Caution Management comments
Congenital heart disease Massive hemoptysis is cause of death in many patients Extreme caution; discontinue anticoagulation if more than minimal
Scleroderma Risk of gastrointestinal bleed secondary to telangiectesias Extreme caution; monitor for hypochromic microcytic anemia, occult
blood, or malena; discontinue anticoagulation if significant
gastrointestinal bleed; consider restarting anticoagulation after a few
months if bleed was minor and endoscopic findings are not alarming
Portopulmonary hypertension Many patients have coagulopathy and/or Extreme caution; do not anticoagulate patients with significant varices,
thrombocytopenia; many have varices history of significant variceal bleed, or moderate to severe
coagulopathy (INRO1.4) or thrombocytopenia (platelet count


!40, 000/mL); drug interactions may be more pronounced
HIV Many patients have thrombocytopenia; many have Similar cautions apply as in portopulmonary hypertension
concurrent advanced liver disease; HAART therapy
may have drug interaction
Bosentan therapy Increases warfarin dose requirement Increase warfarin dose by 30%–40%; recheck INR in 7–10 days and
monitor INR more frequently until INR becomes stable, only at the
initiation or termination of Bosentan therapy
Sitaxsentan therapy Decreases warfarin dose requirement Decrease warfarin dose by 80%, recheck INR in 7–10 days, and
monitor INR more frequently until INR becomes stable, only at the
initiation or termination of Sitaxsentan therapy
Intravenous prostacyclin or treprostinil Risk of central venous catheter–related clot and risk of It is generally recommended that all patients receiving PAH therapies
clogging of the catheter lumen secondary to low flow via central venous catheter should be anticoagulated. NOTE:
rate of infusion prostacyclins also have antiplatelet aggregation properties, and
direct evidence to support such recommendation is lacking
Severe RHF Passive hepatic congestion may cause coagulopathy Check INR and hemoglobin more frequently, hold anticoagulation if
or ischemic bowl necessary; patient may become bactremic secondary to ischemic
colonic ulcerations (as many terminal patients with severe RHF do);
consider broad-spectrum antibiotics covering enteric gram-negative
rods and anaerobes
Syncope Syncopal and presyncopal episodes are common Anticoagulation should be continued at any cost in patients with
in patients who have severe PAH CTE-PHT or in whom PE is suspected cause of an isolated syncope;
in all other patients, risks and benefits should be carefully weighed;
advising patients to pace themselves may be the initial advice

with IV nitroprusside [131] in patients with high

termination of pregnancy should be strongly advised in patients who
Pregnancy carries a high risk of mortality (30%–50%) in patients with
moderate to severe PAH; all female PAH patients of reproductive

pulmonary capillary wedge pressure. A fall in sys-

Consider transiently holding anticoagulation while using systemic

does insist on continuing pregnancy, LMWH should be used
temic arterial blood pressure with IV nitroprusside
potential should be advised to observe strict contraception;

causing increase in cardiac output and drop in pul-

compression devices (SCDs) for deep venous thrombosis
monary capillary wedge pressure highly suggests
prophylaxis; consider oral or subcutaneous vitamin K the possibility of concurrent LV diastolic dysfunc-
tion. Similarly, in patients with pulmonary veno-
occlusive disease, the acute vasoreactivity testing
do become pregnant; if, however, a patient

may result in massive pulmonary edema, which

may even be fatal. Death has been reported in a pa-
tient with pulmonary veno-occlusive disease as
a result of administration of IV prostacyclin
[132] at 2 ng/kg/min for only 5 minutes.
Inhaled NO, inhaled prostacyclin, IV prostacy-
clin, or IV adenosine may be used for acute testing
[35–37]. These agents are short acting, easily ad-

ministered and titrated, have minimal systemic

effects at doses that can result in pulmonary arterial
vasodilation, and demonstrate rapid reversal of ef-
fects if complications ensue. The CCBs should not
be used for this testing because they are longer act-
ing and their use may be associated with complica-
tions such as systemic hypotension and worsening
of hypoxemia. These reactions to CCBs were pri-
marily seen in patients who were actually nonre-
Relative vitamin K deficiency related to antibiotic

sponders. Because of little likelihood of benefit

and a risk of side effects and complications, empiric
use or poor oral intake may prolong INR;

use of CCBs in all patients who have PAH without

contraindicated (Class D) in pregnancy

sepsis may cause coagulopathy and/or

knowing acute vasoreactivity status is not indi-

cated, is dangerous, and may be fatal.
In only a small subset (10%–26%) of patients
who are responders and have no specific contrain-
Warfarin is teratogenic and is

dication to CCBs, these agents may be started as an

inpatient treatment with PA catheter in place (acute
CCB dose escalation) or by slowly increasing the

dose on an outpatient basis (see Table 8). The

choice of agent depends on the resting heart rate
of the patient [29,32–34]. If the patient’s heart rate
is O100 beats/min, diltiazem is used; if the heart
rate is !100 beats/min, nifedipine is used. Verapa-
mil is generally not used because of its strong cardi-
odepressor properties. Amlodipine is another
choice, especially for patients who cannot tolerate
Acute illness, antibiotic use, sepsis

the other CCBs secondary to side effects, such as

worsening of peripheral edema, systemic hypoten-
sion, or abnormally low or high heart rate.

After half a century of clinical experience and
research, PAH management remains a challenge.

Currently, data to support the use of standard

PAH therapies (eg, oxygen supplementation, di-
uretics, digoxin, anticoagulation, and CCBs) are

Table 8
Use of calcium channel blockers in pulmonary arterial hypertension
Indications Initiation of CCB
 PAH patients with WHO group I and CCB may be started in either inpatient setting or outpatient setting
 PAH diagnosis confirmed on right heart catheterization (there are no rigid recommendations,
(ie, PVR above 3.0 wood units by Fick’s method) and and approaches may vary among different centers)
 ‘‘Responder’’ on ‘‘acute vasoreactivity testing’’ A. Inpatient setting: rapid CCB dose escalation
 All rapid CCB dose escalations are generally performed with
Contraindications PA catheter in place
 Low cardiac output, cardiac index !2.0  Patients receive initial dose of nifedipine 10–20 mg orally or
 Severe RHF diltiazem 60 mg orally; hemodynamic measurements are
 Hypotension, systolic BP below 90 mm Hg obtained in 1 h
 History of adverse reaction or intolerance to CCB  The dose and hemodynamic measurements are repeated
Cautions every hour until a ‘‘threshold’’ response (fall in PVR by 50%
 Empiric trial of CCB without acute vasoreactivity and, not or, fall in mPAP by 33%) is achieved or significant
testing should not be performed. It is unsafe and may side effects are experienced (eg, hypotension [mBP !90 mm
cause hypotension [125] and even death [124] Hg], gastrointestinal upset [nausea, vomiting])
 Acute vasoreactivity testing should not be performed  Total daily dose is calculated by adding up total amount of
with CCB because it is unsafe and may cause profound drug administered during this testing. The goal is to achieve
hypotension, acute pulmonary edema, and potentially this in three divided doses (four to six divided doses if
death significant side effects)
 Patients who are acutely decompensated with RHF  Patient is then given nifedipine 20 mg orally three times daily
are not candidates for CCB therapy. There is no real or diltiazem 60 mg orally three times daily next day and is
reason to perform acute vasoreactivity testing in such discharged
patients  The dose is then gradually increased to the desired level
(as estimated previously) over a period of 6–12 wk while
Acute vasoreactivity testing frequently monitoring BP and heart rate
A. Agents used [121–127]
 Inhaled NO: 20 ppm for 6–10 minute by face mask B. Outpatient setting: slow CCB dose escalation
 IV prostacyclin: Start 1 ng/kg/min; increase by  Patients are started on initial dose of nifedipine 10–20 mg
1–2 ng/kg/min every 5–15 min to maximum orally three times daily or diltiazem 60 mg orally three times
of 12 ng/kg/min or untolerable side effects or daily
2.5 ng/kg/min, increase by 2.5 ng/kg/min every  The dose is then gradually increased with a goal to
10 min to maximum of 12 ng/kg/min; mean ultimately achieve the maximum dose (nifedipine
tolerated dose: 8 ng/kg/min 240 mg/d or diltiazem 720 mg/d, both in three divided doses;
 Inhaled prostacyclin:50 ng/kg/min via nebulizer amlodipine up to 5 mg twice daily) over a period of 6–12 wk,
for 15 min while frequently monitoring the BP and heart rate. If patient
 IV adenosine: fast IV bolus 50 mg/kg/min; increase experiences limiting side effects (as mentioned previously)
by 50 mg/kg/min every 2 min to maximum before the maximum dose is reached, either the dose is kept
of 500 mg/kg/min at that level or further increase is achieved by increasing the
dose frequency to every 4–6 h
B. Criteria for significant response (‘‘acute responder’’)
 Drop in mPAP O10 mm Hg Follow-up, when to use additional therapies
or Attaining an mPAP !40 mm Hg  Only 6.8% of all patients who have PAH are long-term
or Decrease in PVR by R20% responders to CCB (ie, who will be in NYHA I or II on
or Attaining a PVR !8 wood units monotherapy with CCB for 1 y) [32]
 and Increase or no decrease in cardiac output  Approximately half of the patients who are responders at
 and No or only clinically acceptable fall in systemic initial acute vasoreactivity testing and are placed on CCB
blood pressure require an additional PAH therapy within 1 y
 NOTE: there are no universally accepted criteria  Secondary to this risk of failure of CCB monotherapy, such
[35–37] patients should be closely followed (every 3–6 mo) and
Practical use of CCBs should be started on an additional therapy if there is clinical
Choice and dose of calcium channel blockers worsening or worsening of 6-minute walking distance.
 Chose nifedipine if baseline resting heart rate  Secondary to this risk of failure of CCB monotherapy, some
!100/min and diltiazem if O100/min experts consider adding a specific PAH agent at the
 Amlodipine may be used if significant side effects beginning, especially in patients who have poor predictors
from other agents (eg, worsening edema, significant of long-term response at baseline hemodynamics and
tachycardia, bradycardia, or hypotension) vasoreactivity testing [32]
 Verapamil should not be used secondary to strong
cardiodepressor effects
 Generally high doses are required (eg, nifedipine
up to 240 mg/d and diltiazem 720 mg/d, both in three
divided doses, amlodipine up to 5 mg twice a day
 Only oral administration is used

mostly retrospective, uncontrolled prospective, or clin analogue, in patients with pulmonary arterial
derived from other diseases with similar but not hypertension: a double-blind, randomized, pla-
identical manifestations. In the absence of any cebo-controlled trial. Am J Respir Crit Care Med
further prospective, controlled studies, it is rea- 2002;165(6):800–4.
[13] Sitbon O, Humbert M, Nunes H, et al. Long-term
sonable to use these therapies when they are
intravenous epoprostenol infusion in primary pul-
tolerated. When these therapies are poorly toler- monary hypertension: prognostic factors and sur-
ated, however, the threshold for discontinuation vival. J Am Coll Cardiol 2002;40(4):780–8.
should be low. [14] McLaughlin VV, Shillington A, Rich S. Survival in
primary pulmonary hypertension: the impact of
References epoprostenol therapy. Circulation 2002;106(12):
[1] Romberg E. Über sklerose der lungen arterie. [15] Ghofrani HA, Wiedemann R, Rose F, et al. Com-
Dtsch Arch Klin Med 1891;48:197–206 [in bination therapy with oral sildenafil and inhaled
German]. iloprost for severe pulmonary hypertension. Ann
[2] Dresdale DT, Schultz M, Michtom RJ. Primary Intern Med 2002;136(7):515–22.
pulmonary hypertension. I: clinical and hemody- [16] Sitbon O, Gressin V, Speich R, et al. Bosentan for
namic study. Am J Med 1951;11(6):686–705. the treatment of human immunodeficiency virus-
[3] Rich S, Dantzker DR, Ayres SM, et al. Primary associated pulmonary arterial hypertension. Am J
pulmonary hypertension: a national prospective Respir Crit Care Med 2004;170(11):1212–7.
study. Ann Intern Med 1987;107(2):216–23. [17] Mikhail GW, Prasad SK, Li W, et al. Clinical and
[4] Rubin LJ, Mendoza J, Hood M, et al. Treatment of haemodynamic effects of sildenafil in pulmonary
primary pulmonary hypertension with continuous hypertension: acute and mid-term effects. Eur
intravenous prostacyclin (epoprostenol): results of Heart J 2004;25(5):431–6.
a randomized trial. Ann Intern Med 1990;112(7): [18] Wilkins MR, Paul GA, Strange JW, et al. Sildenafil
485–91. versus Endothelin Receptor Antagonist for Pulmo-
[5] Barst RJ, Rubin LJ, McGoon MD, et al. Sur- nary Hypertension (SERAPH) study. Am J Respir
vival in primary pulmonary hypertension with Crit Care Med 2005;171(11):1292–7.
long-term continuous intravenous prostacyclin. [19] McLaughlin VV, Sitbon O, Badesch DB, et al. Sur-
Ann Intern Med 1994;121(6):409–15. vival with first-line bosentan in patients with pri-
[6] Barst R, Rubin LJ, Long WA, et al. A comparison mary pulmonary hypertension. Eur Respir J 2005;
of continuous intravenous epoprostenol (prostacy- 25(2):244–9.
clin) with conventional therapy for primary pulmo- [20] Galie N, Ghofrani HA, Torbicki A, et al. Sildenafil
nary hypertension: the Primary Pulmonary citrate therapy for pulmonary arterial hyperten-
Hypertension Study Group. N Engl J Med 1996; sion. N Engl J Med 2005;353(20):2148–57.
334(5):296–302. [21] Barst RJ, Langleben D, Badesch D, et al. Treat-
[7] Rich S, McLaughlin VV. The effects of chronic ment of pulmonary arterial hypertension with the
prostacyclin therapy on cardiac output and symp- selective endothelin-A receptor antagonist sitax-
toms in primary pulmonary hypertension. J Am sentan. J Am Coll Cardiol 2006;47(10):2049–56.
Coll Cardiol 1999;34(4):1184–7. [22] Naeije R, Vachiery JL. Medical therapy of pulmo-
[8] Badesch DB, Tapson VF, McGoon MD, et al. nary hypertension: conventional therapies. Clin
Continuous intravenous epoprostenol for pulmo- Chest Med 2001;22(3):517–27.
nary hypertension due to the scleroderma spectrum [23] Rich S, Seidlitz M, Dodin E, et al. The short-term
of disease: a randomized, controlled trial. Ann In- effects of digoxin in patients with right ventricular
tern Med 2000;132(6):425–34. dysfunction from pulmonary hypertension. Chest
[9] Hoeper MM, Schwarze M, Ehlerding S, et al. 1998;114(3):787–92.
Long-term treatment of primary pulmonary hyper- [24] Wagenvoort CA. Lung biopsy specimens in the
tension with aerosolized iloprost, a prostacyclin an- evaluation of pulmonary vascular disease. Chest
alogue. N Engl J Med 2000;342(25):1866–70. 1980;77(5):614–25.
[10] Channick RN, Simonneau G, Sitbon O, et al. [25] Bjornsson J, Edwards WD. Primary pulmonary hy-
Effects of the dual endothelin-receptor antago- pertension: a histopathologic study of 80 cases.
nist bosentan in patients with pulmonary hyper- Mayo Clin Proc 1985;60(1):16–25.
tension: a randomised placebo-controlled study. [26] Pietra GG, Edwards WD, Kay JM, et al. Histopa-
Lancet 2001;358(9288):1119–23. thology of primary pulmonary hypertension:
[11] Rubin LJ, Badesch DB, Barst RJ, et al. Bosentan a qualitative and quantitative study of pulmonary
therapy for pulmonary arterial hypertension. N blood vessels from 58 patients in the National
Engl J Med 2002;346(12):896–903. Heart, Lung, and Blood Institute, Primary Pulmo-
[12] Simonneau G, Barst RJ, Galie N, et al. Continuous nary Hypertension Registry. Circulation 1989;
subcutaneous infusion of treprostinil, a prostacy- 80(5):1198–206.

[27] Fuster V, Steele PM, Edwards WD, et al. Primary bronchitis and emphysema: report of the Medical
pulmonary hypertension: natural history and the Research Council Working Party. Lancet 1981;
importance of thrombosis. Circulation 1984;70(4): 1(8222):681–6.
580–7. [41] Global strategy for the diagnosis, management and
[28] Kawut SM, Horn EM, Berekashvili KK, et al. New prevention of chronic obstructive pulmonary dis-
predictors of outcome in idiopathic pulmonary ar- ease: NHLBI/WHO workshop report. In: Global
terial hypertension. Am J Cardiol 2005;95(2): initiative for chronic obstructive pulmonary lung
199–203. disease. Bethesda (MD): National Heart, Lung
[29] Rich S, Kaufmann E, Levy PS. The effect of high and Blood Institute; 2005 Available at: http://
doses of calcium-channel blockers on survival in¼2&l2¼
primary pulmonary hypertension. N Engl J Med 1&intId¼989. Accessed January 25, 2007.
1992;327(2):76–81. [42] Owens GR, Rogers RM, Pennock BE, et al. The
[30] Johnson SR, Granton JT, Mehta S. Thrombotic diffusing capacity as a predictor of arterial oxygen
arteriopathy and anticoagulation in pulmonary desaturation during exercise in patients with
hypertension. Chest 2006;130(2):545–52. chronic obstructive pulmonary disease. N Engl J
[31] Sandoval J, Aguirre JS, Pulido T, et al. Nocturnal Med 1984;310(19):1218–21.
oxygen therapy in patients with the Eisenmenger [43] Kelley MA, Panettieri RA Jr, Krupinski AV. Rest-
syndrome. Am J Respir Crit Care Med 2001; ing single-breath diffusing capacity as a screening
164(9):1682–7. test for exercise-induced hypoxemia. Am J Med
[32] Sitbon O, Humbert M, Jais X, et al. Long-term 1986;80(5):807–12.
response to calcium channel blockers in idio- [44] Bowyer JJ, Busst CM, Denison DM, et al. Effect of
pathic pulmonary arterial hypertension. Circula- long term oxygen treatment at home in children
tion 2005;111(23):3105–11. with pulmonary vascular disease. Br Heart J
[33] Barst RJ, Maislin G, Fishman AP. Vasodilator 1986;55(4):385–90.
therapy for primary pulmonary hypertension in [45] Widlitz A, Barst RJ. Pulmonary arterial hyper-
children. Circulation 1999;99(9):1197–208. tension in children. Eur Respir J 2003;21(1):
[34] Rich S, Brundage BH. High-dose calcium channel- 155–76.
blocking therapy for primary pulmonary hyper- [46] Douglas NJ, White DP, Pickett CK, et al. Respira-
tension: evidence for long-term reduction in tion during sleep in normal man. Thorax 1982;
pulmonary arterial pressure and regression of right 37(11):840–4.
ventricular hypertrophy. Circulation 1987;76(1): [47] Koo KW, Sax DS, Snider GL. Arterial blood gases
135–41. and pH during sleep in chronic obstructive pulmo-
[35] Badesch DB, Abman SH, Ahearn GS, et al. Medi- nary disease. Am J Med 1975;58(5):663–70.
cal therapy for pulmonary arterial hypertension: [48] Atwood CW Jr, McCrory D, Garcia JG, et al. Pul-
ACCP evidence-based clinical practice guidelines. monary artery hypertension and sleep-disordered
Chest 2004;126(1 Suppl):35S–62S. breathing: ACCP evidence-based clinical practice
[36] Galie N, Torbicki A, Barst R, et al. Guidelines on guidelines. Chest 2004;126(1 Suppl):72S–7S.
diagnosis and treatment of pulmonary arterial hy- [49] Podszus T, Bauer W, Mayer J, et al. Sleep apnea
pertension: the task force on diagnosis and treat- and pulmonary hypertension. Klin Wochenschr
ment of pulmonary arterial hypertension of the 1986;64(3):131–4.
European Society of Cardiology. Eur Heart J [50] Weitzenblum E, Krieger J, Apprill M, et al. Day-
2004;25(24):2243–78. time pulmonary hypertension in patients with ob-
[37] British Cardiac Society Guidelines and Medical structive sleep apnea syndrome. Am Rev Respir
Practice Committee, Gibbs JS, Higenbottam TW. Dis 1988;138(2):345–9.
Recommendations on the management of pulmo- [51] Krieger J, Sforza E, Apprill M, et al. Pulmonary
nary hypertension in clinical practice. Heart 2001; hypertension, hypoxemia, and hypercapnia in ob-
86(Suppl 1):11–3. structive sleep apnea patients. Chest 1989;96(4):
[38] Weissmann N, Hassell KL, Badesch DB, et al. 729–37.
Hypoxic vasoconstriction in intact lungs: a role [52] Apprill M, Weitzenblum E, Krieger J, et al. Fre-
for NADPH oxidase-derived H(2)O(2)? Am J quency and mechanism of daytime pulmonary hy-
Physiol Lung Cell Mol Physiol 2000;279(4): pertension in patients with obstructive sleep
L683–90. apnoea syndrome. Cor Vasa 1991;33(1):42–9.
[39] Nocturnal Oxygen Therapy Trial Group. Continu- [53] Laks L, Lehrhaft B, Grunstein R, et al. Pulmo-
ous or nocturnal oxygen therapy in hypoxemic nary hypertension in obstructive sleep apnoea.
chronic obstructive lung disease: a clinical trial. Eur Respir J 1995;8(4):537–41.
Nocturnal Oxygen Therapy Trial Group. Ann [54] Chaouat A, Weitzenblum E, Krieger J, et al. Pul-
Intern Med 1980;93(3):391–8. monary hemodynamics in the obstructive sleep ap-
[40] Long term domiciliary oxygen therapy in chronic nea syndrome: results in 220 consecutive patients.
hypoxic cor pulmonale complicating chronic Chest 1996;109(2):380–6.

[55] Sajkov D, Wang T, Saunders N, et al. Daytime pul- in the adult: a report of the American College of
monary hemodynamics in patients with obstructive Cardiology/American Heart Association Task
sleep apnea without lung disease. Am J Respir Crit Force on Practice guidelines (Writing Committee
Care Med 1999;159(5 Pt 1):1518–26. to Update the 2001 Guidelines for the Evaluation
[56] Sanner BM, Doberauer C, Konermann M, et al. and Management of Heart Failure). J Am Coll
Pulmonary hypertension in patients with obstruc- Cardiol 2005;46(6):e1–e82.
tive sleep apnea syndrome. Arch Intern Med [69] Quinn DA, Du HK, Thompson BT, et al. Amilor-
1997;157(21):2483–7. ide analogs inhibit chronic hypoxic pulmonary hy-
[57] Niijima M, Kimura H, Edo H, et al. Manifestation pertension. Am J Respir Crit Care Med 1998;157(4
of pulmonary hypertension during REM sleep in Pt 1):1263–8.
obstructive sleep apnea syndrome. Am J Respir [70] Krayenbuehl HP, Turina J, Hess O. Left ventricu-
Crit Care Med 1999;159(6):1766–72. lar function in chronic pulmonary hypertension.
[58] Bady E, Achkar A, Pascal S, et al. Pulmonary arte- Am J Cardiol 1978;41(7):1150–8.
rial hypertension in patients with sleep apnoea syn- [71] Dittrich HC, Chow LC, Nicod PH. Early improve-
drome. Thorax 2000;55(11):934–9. ment in left ventricular diastolic function after relief
[59] Alchanatis M, Tourkohoriti G, Kakouros S, et al. of chronic right ventricular pressure overload. Cir-
Daytime pulmonary hypertension in patients with culation 1989;80(4):823–30.
obstructive sleep apnea: the effect of continuous [72] Gan CT, Lankhaar JW, Marcus JT, et al. Impaired
positive airway pressure on pulmonary hemody- left ventricular filling due to right-to-left ventricular
namics. Respiration 2001;68(6):566–72. interaction in patients with pulmonary arterial hy-
[60] Yamakawa H, Shiomi T, Sasanabe R, et al. Pulmo- pertension. Am J Physiol Heart Circ Physiol 2006;
nary hypertension in patients with severe obstruc- 290(4):H1528–33.
tive sleep apnea. Psychiatry Clin Neurosci 2002; [73] Akera T, Baskin SI, Tobin T, et al. Ouabain: tem-
56(3):311–2. poral relationship between the inotropic effect
[61] Abenhaim L, Moride Y, Brenot F, et al. Appetite- and the in vitro binding to, and dissociation from,
suppressant drugs and the risk of primary pulmo- (Na þ þ K þ)- activated ATPase. Naunyn Schmie-
nary hypertension: International Primary Pulmo- debergs Arch Pharmacol 1973;277(2):151–62.
nary Hypertension Study Group. N Engl J Med [74] Mathur PN, Powles P, Pugsley SO, et al. Effect of
1996;335(9):609–16. digoxin on right ventricular function in severe
[62] Sajkov D, Wang T, Saunders NA, et al. Continu- chronic airflow obstruction: a controlled clinical
ous positive airway pressure treatment improves trial. Ann Intern Med 1981;95(3):283–8.
pulmonary hemodynamics in patients with ob- [75] The Digitalis Investigation Group. The effect of di-
structive sleep apnea. Am J Respir Crit Care Med goxin on mortality and morbidity in patients with
2002;165(2):152–8. heart failure. N Engl J Med 1997;336(8):525–33.
[63] Arias MA, Garcia-Rio F, Alonso-Fernandez A, [76] Jelliffe RW, Brooker G. A nomogram for digoxin
et al. Pulmonary hypertension in obstructive sleep therapy. Am J Med 1974;57(1):63–8.
apnoea: effects of continuous positive airway pres- [77] Rathore SS, Curtis JP, Wang Y, et al. Association
sure. A randomized, controlled cross-over study. of serum digoxin concentration and outcomes in
Eur Heart J 2006;27(9):1106–13. patients with heart failure. JAMA 2003;289(7):
[64] Kessler R, Chaouat A, Schinkewitch P, et al. The 871–8.
obesity-hypoventilation syndrome revisited: a pro- [78] Fogelman AM, La Mont JT, Finkelstein S, et al.
spective study of 34 consecutive cases. Chest 2001; Fallibility of plasma-digoxin in differentiating toxic
120(2):369–76. from non-toxic patients. Lancet 1971;2(7727):
[65] Masa JF, Celli BR, Riesco JA, et al. The obesity 727–9.
hypoventilation syndrome can be treated with non- [79] Ingelfinger JA, Goldman P. The serum digitalis
invasive mechanical ventilation. Chest 2001;119(4): concentration: does it diagnose digitalis toxicity?
1102–7. N Engl J Med 1976;294(16):867–70.
[66] Sugerman HJ, Baron PL, Fairman RP, et al. He- [80] Leor J, Goldbourt U, Rabinowitz B, et al. Digoxin
modynamic dysfunction in obesity hypoventilation and increased mortality among patients recovering
syndrome and the effects of treatment with surgi- from acute myocardial infarction: importance of
cally induced weight loss. Ann Surg 1988;207(5): digoxin dose. The SPRINT Study Group. Cardio-
604–13. vasc Drugs Ther 1995;9(5):723–9.
[67] Pitt B, Zannad F, Remme WJ, et al. The effect of [81] Eichhorn EJ, Gheorghiade M. Digoxin. Prog Car-
spironolactone on morbidity and mortality in pa- diovasc Dis 2002;44(4):251–66.
tients with severe heart failure: Randomized Aldac- [82] Hager WD, Fenster P, Mayersohn M, et al. Di-
tone Evaluation Study Investigators. N Engl J Med goxin-quinidine interaction pharmacokinetic eval-
1999;341(10):709–17. uation. N Engl J Med 1979;300(22):1238–41.
[68] Hunt SA. ACC/AHA 2005 guideline update for the [83] Bizjak ED, Mauro VF. Digoxin-macrolide drug
diagnosis and management of chronic heart failure interaction. Ann Pharmacother 1997;31(9):1077–9.

[84] Juurlink DN, Mamdani M, Kopp A, et al. Drug- hypertension: response to bosentan therapy. Am J
drug interactions among elderly patients hospital- Respir Crit Care Med 2005;172(3):352–7.
ized for drug toxicity. JAMA 2003;289(13): [99] Welsh CH, Hassell KL, Badesch DB, et al. Coagu-
1652–8. lation and fibrinolytic profiles in patients with se-
[85] Archer SL, Djaballah K, Humbert M, et al. Nitric vere pulmonary hypertension. Chest 1996;110(3):
oxide deficiency in fenfluramine- and dexfenflur- 710–7.
amine-induced pulmonary hypertension. Am J [100] Loyd JE, Atkinson JB, Pietra GG, et al. Heteroge-
Respir Crit Care Med 1998;158(4):1061–7. neity of pathologic lesions in familial primary pul-
[86] Lopes AA, Maeda NY, Almeida A, et al. Circulat- monary hypertension. Am Rev Respir Dis 1988;
ing platelet aggregates indicative of in vivo platelet 138(4):952–7.
activation in pulmonary hypertension. Angiology [101] Frank H, Mlczoch J, Huber K, et al. The effect of
1993;44(9):701–6. anticoagulant therapy in primary and anorectic
[87] Herve P, Launay JM, Scrobohaci ML, et al. In- drug-induced pulmonary hypertension. Chest
creased plasma serotonin in primary pulmonary 1997;112(3):714–21.
hypertension. Am J Med 1995;99(3):249–54. [102] Roman A, Rodes-Cabau J, Lara B, et al. Clinico-
[88] Kereveur A, Callebert J, Humbert M, et al. High hemodynamic study and treatment of 44 patients
plasma serotonin levels in primary pulmonary hy- with primary pulmonary hypertension. Med Clin
pertension: effect of long-term epoprostenol (pros- (Barc) 2002;118(20):761–6.
tacyclin) therapy. Arterioscler Thromb Vasc Biol [103] Storstein O, Efskind L, Muller C, et al. Primary
2000;20(10):2233–9. pulmonary hypertension with emphasis on its etiol-
[89] Cacoub P, Karmochkine M, Dorent R, et al. ogy and treatment. Acta Med Scand 1966;179(2):
Plasma levels of thrombomodulin in pulmonary 197–212.
hypertension. Am J Med 1996;101(2):160–4. [104] Ansell J, Hirsh J, Poller L, et al. The pharmacol-
[90] Giaid A, Saleh D. Reduced expression of endothe- ogy and management of the vitamin K antago-
lial nitric oxide synthase in the lungs of patients nists: the seventh ACCP conference on
with pulmonary hypertension. N Engl J Med antithrombotic and thrombolytic therapy. Chest
1995;333(4):214–21. 2004;126(3 Suppl):204S–33S.
[91] Collados MT, Sandoval J, Lopez S, et al. Charac- [105] Hales CA, Kradin RL, Brandstetter RD, et al. Im-
terization of von Willebrand factor in primary pul- pairment of hypoxic pulmonary artery remodeling
monary hypertension. Heart Vessels 1999;14(5): by heparin in mice. Am Rev Respir Dis 1983;
246–52. 128(4):747–51.
[92] Hoeper MM, Sosada M, Fabel H. Plasma coagu- [106] Hassoun PM, Thompson BT, Hales CA. Partial
lation profiles in patients with severe primary reversal of hypoxic pulmonary hypertension by
pulmonary hypertension. Eur Respir J 1998;12(6): heparin. Am Rev Respir Dis 1992;145(1):193–6.
1446–9. [107] Yu L, Quinn DA, Garg HG, et al. Gene expression
[93] Wolf M, Boyer-Neumann C, Parent F, et al. of cyclin-dependent kinase inhibitors and effect of
Thrombotic risk factors in pulmonary hyperten- heparin on their expression in mice with hypoxia-
sion. Eur Respir J 2000;15(2):395–9. induced pulmonary hypertension. Biochem Bio-
[94] Sakamaki F, Kyotani S, Nagaya N, et al. Increased phys Res Commun 2006;345(4):1565–72 [Epub
plasma P-selectin and decreased thrombomodulin 2006].
in pulmonary arterial hypertension were improved [108] Hassoun PM, Thompson BT, Steigman D, et al. Ef-
by continuous prostacyclin therapy. Circulation fect of heparin and warfarin on chronic hypoxic
2000;102(22):2720–5. pulmonary hypertension and vascular remodeling
[95] Boyer-Neumann C, Brenot F, Wolf M, et al. Con- in the guinea pig. Am Rev Respir Dis 1989;
tinuous infusion of prostacyclin decreases plasma 139(3):763–8.
levels of t-PA and PAI-1 in primary pulmonary hy- [109] Garg HG, Hales CA, Yu L, et al. Increase in
pertension. Thromb Haemost 1995;73(4):735–6. the growth inhibition of bovine pulmonary ar-
[96] Veyradier A, Nishikubo T, Humbert M, et al. Im- tery smooth muscle cells by an O-hexanoyl
provement of von Willebrand factor proteolysis af- low-molecular-weight heparin derivative. Carbo-
ter prostacyclin infusion in severe pulmonary hydr Res 2006;341(15):2607–12.
arterial hypertension. Circulation 2000;102(20): [110] Levine MN, et al. Hemorrhagic complications
2460–2. of anticoagulant treatment: the seventh ACCP
[97] Friedman R, Mears JG, Barst RJ. Continuous in- conference on antithrombotic and thrombolytic
fusion of prostacyclin normalizes plasma markers therapy. Chest 2004;126(3 Suppl):287S–310S.
of endothelial cell injury and platelet aggregation [111] Robbins IM, et al. A study of aspirin and clopidog-
in primary pulmonary hypertension. Circulation rel in idiopathic pulmonary arterial hypertension.
1997;96(9):2782–4. Eur Respir J 2006;27(3):578–84.
[98] Girgis RE, Champion HC, Diette GB, et al. De- [112] Rubin LJ. Pulmonary arterial hypertension. Proc
creased exhaled nitric oxide in pulmonary arterial Am Thorac Soc 2006;3(1):111–5.

[113] Geerts WH, et al. Prevention of venous thrombo- [123] Pepke-Zaba JH, Dinh-Xuan AT, Stone D, et al. In-
embolism: the seventh ACCP conference on antith- haled nitric oxide as a cause of selective pulmonary
rombotic and thrombolytic therapy. Chest 2004; vasodilatation in pulmonary hypertension. Lancet
126(3 Suppl):338S–400S. 1991;338(8776):1173–4.
[114] Dresdale DT, Michtom RJ, Schultz M. Recent [124] Ricciardi MJ, Knight BP, Martinez FJ, et al. In-
studies in primary pulmonary hypertension, includ- haled nitric oxide in primary pulmonary hyperten-
ing pharmacodynamic observations on pulmonary sion: a safe and effective agent for predicting
vascular resistance. Bull N Y Acad Med 1954;30(3): response to nifedipine. J Am Coll Cardiol 1998;
195–207. 32(4):1068–73.
[115] Christman BW, McPherson CD, Newman JH, [125] Schrader BJ, Inbar S, Kaufmann L, et al. Compar-
et al. An imbalance between the excretion of throm- ison of the effects of adenosine and nifedipine in
boxane and prostacyclin metabolites in pulmonary pulmonary hypertension. J Am Coll Cardiol 1992;
hypertension. N Engl J Med 1992;327(2):70–5. 19(5):1060–4.
[116] Tuder RM, Cool CD, Geraci MW, et al. Prostacy- [126] Opitz CF, Wensel R, Bettmann M, et al. Assess-
clin synthase expression is decreased in lungs from ment of the vasodilator response in primary pulmo-
patients with severe pulmonary hypertension. Am J nary hypertension: comparing prostacyclin and
Respir Crit Care Med 1999;159(6):1925–32. iloprost administered by either infusion or inhala-
[117] Stewart DJ, Levy RD, Cernacek P, et al. Increased tion. Eur Heart J 2003;24(4):356–65.
plasma endothelin-1 in pulmonary hypertension: [127] Ogata M, Ohe M, Shirato K, et al. Effects of
marker or mediator of disease? Ann Intern Med a combination therapy of anticoagulant and va-
1991;114(6):464–9. sodilator on the long-term prognosis of primary
[118] Olivari MT, Levine TB, Weir EK, et al. Hemody- pulmonary hypertension. Jpn Circ J 1993;57(1):
namic effects of nifedipine at rest and during exer- 63–9.
cise in primary pulmonary hypertension. Chest [128] Humbert M, Sitbon O, Chaouat A, et al. Pulmo-
1984;86(1):14–9. nary arterial hypertension in France: results from
[119] Higenbottam T, Wheeldon D, Wells F, et al. Long- a national registry. Am J Respir Crit Care Med
term treatment of primary pulmonary hypertension 2006;173(9):1023–30.
with continuous intravenous epoprostenol (prosta- [129] Ziesche R, Petkov V, Wittmann K, et al. Treatment
cyclin). Lancet 1984;1(8385):1046–7. with epoprostenol reverts nitric oxide non-respon-
[120] Groves B, Badesch DB, Turkevich D, et al. Corre- siveness in patients with primary pulmonary hyper-
lation of acute prostacyclin response in primary tension. Heart 2000;83(4):406–9.
(unexplained) pulmonary hypertension and efficacy [130] Preston IR, Klinger JR, Houtchens J, et al. Pulmo-
of treatment with calcium channel blockers and nary edema caused by inhaled nitric oxide therapy
survival. In: Weir K, editor. Ion flux in pulmonary in two patients with pulmonary hypertension asso-
vascular control. New York: Plenum Press; 1993. p. ciated with the CREST syndrome. Chest 2002;
317–30. 121(2):656–9.
[121] Sitbon O, Brenot F, Denjean A, et al. Inhaled nitric [131] Zakliczynski M, Zebik T, Maruszewski M, et al.
oxide as a screening vasodilator agent in primary Usefulness of pulmonary hypertension reversibil-
pulmonary hypertension: a dose-response study ity test with sodium nitroprusside in stratifica-
and comparison with prostacyclin. Am J Respir tion of early death risk after orthotopic heart
Crit Care Med 1995;151(2 Pt 1):384–9. transplantation. Transplant Proc 2005;37(2):
[122] Sitbon O, Humbert M, Jagot JL, et al. Inhaled nitric 1346–8.
oxide as a screening agent for safely identifying re- [132] Palmer SM, Robinson LJ, Wang A, et al. Massive
sponders to oral calcium-channel blockers in pri- pulmonary edema and death after prostacyclin in-
mary pulmonary hypertension. Eur Respir J 1998; fusion in a patient with pulmonary veno-occlusive
12(2):265–70. disease. Chest 1998;113(1):237–40.