Published ahead of print on August 18, 2004, doi:10.1164/rccm.200404-445OC
© 2004 American Thoracic Society doi: 10.1164/rccm.200404-445OC
Bosentan for the Treatment of Human Immunodeficiency Virusassociated Pulmonary Arterial HypertensionHôpital Antoine Béclère, Clamart; Hôpital de Bicêtre, le Kremlin-Bicêtre; Actelion Pharmaceuticals Ltd., Paris, France; University Hospital Zurich, Zurich, Switzerland; and St. Vincent's Hospital, Sydney, Australia Correspondence and requests for reprints should be addressed to Dr. Olivier Sitbon, M.D., Hôpital Antoine Béclère, 157 rue de la porte de Trivaux, 92140 Clamart, France. E-mail: olivier.sitbon{at}abc.ap-hop-paris.fr
Clinical studies have shown the importance of endothelin as a pathogenic mediator in pulmonary arterial hypertension (PAH). We describe the effects of bosentan, an oral dual endothelin receptor antagonist, in patients with PAH associated with human immunodeficiency virus (HIV) infection. In this prospective study, 16 patients with PAH associated with HIV infection in stable condition received bosentan for 16 weeks. Efficacy endpoints included exercise capacity, cardiopulmonary hemodynamics, Doppler echocardiography, New York Heart Association functional class, and quality of life (SF-36 and EQ-5D). Safety was assessed by laboratory tests, vital signs, and adverse events. Improvements were observed from baseline to Week 16 in all efficacy parameters: 6-minute walk distance (+91 ± 60 m, p < 0.001), New York Heart Association class (14 patients improved), hemodynamics (cardiac index: +0.9 ± 0.7 L/minute/m2, p < 0.001), Doppler echocardiographic variables, and quality of life. During the study, no patient died and none required epoprostenol treatment. Hepatic tolerability was similar to that reported in patients with PAH. Bosentan had no negative impact on control of HIV infection. Although limited by uncontrolled design, small sample size and short duration, this study suggests that bosentan may benefit patients with PAH associated with HIV infection, and that endothelin is an important pathogenic mediator in this disease.
Key Words: bosentan Doppler echocardiography HIV infection pulmonary arterial hypertension quality of life Pulmonary arterial hypertension (PAH) results from chronic obstruction of small pulmonary arteries due in part to endothelial and vascular smooth muscle cell dysfunction and proliferation. PAH may be idiopathic or associated with other conditions, such as infection with the human immunodeficiency virus (HIV) (1). The prevalence of PAH in patients with HIV is estimated to be 0.5% (2, 3). PAH is observed in all groups of patients with HIV, irrespective of the cause of the infection and the degree of immunodeficiency (4, 5). In patients with HIV-associated PAH, the median survival (300 days with conventional therapy [6]) is decreased in comparison with patients with HIV without PAH (2). In most cases, death is causally related to PAH (7). In patients with HIV infection, the HIV-1 envelope glycoprotein 120 may stimulate the production of endothelin (ET) by macrophages (8). ET, when produced in excess, is responsible for deleterious effects including vasoconstriction, cell proliferation, fibrosis, and inflammation. These effects are mediated through the ETA and ETB receptors. High serum concentrations of ET are present in patients with idiopathic PAH and correlate with disease severity. In addition, ET levels are significantly elevated in patients with vascular complications of HIV infection, such as encephalopathy or retinal microangiopathic syndrome, and are correlated with the severity of the lesions (9, 10). Thus, ET may contribute to the pathogenesis of vasculopathies associated with HIV infection, including PAH. In idiopathic PAH and PAH associated with other conditions, bosentan (an oral dual ET receptor antagonist) improves functional class, exercise capacity, hemodynamics, and time to clinical worsening (11, 12). We hypothesized that ET is an important pathogenic mediator in PAH associated with HIV infection and that the use of bosentan may improve the clinical status of patients by preventing the deleterious effects of ET. The purpose of this study was to assess the safety and efficacy of bosentan in PAH associated with HIV infection. Preliminary results of this study have been previously reported in the form of abstracts (1316).
Patients Enrolled patients (> 18 years old) had HIV infection and symptomatic severe PAH (New York Heart Association [NYHA] functional class IIIIV) with a baseline 6-minute walk distance lower than 450 m. PAH was diagnosed from the results of right heart catheterization, with a mean pulmonary artery pressure (mPAP) at rest more than 25 mm Hg, a pulmonary capillary wedge pressure (PCWP) less than 15 mm Hg, and pulmonary vascular resistance (PVR) greater than 240 dyn · s/cm5. Within the past 3 months, patients had to be in stable condition with respect to HIV infection. Exclusion criteria consisted of PAH related to other conditions, CD4 cell count less than or equal to 100 cells/mm3, treatment with prostacyclin analogs, portal hypertension, liver cirrhosis, moderate/severe liver impairment or liver enzymes above three times the upper limit of normal (ULN), anemia, or significant recent HIV-related opportunistic infections. Patients could not be enrolled if they were taking investigational PAH therapies, glibenclamide, or cyclosporine A.
Study Design The study was conducted in three centers in France, Switzerland, and Australia in accordance with the Declaration of Helsinki of 1975 and amendments, and in adherence to the International Conference of Harmonization Good Clinical Practice Guidelines and the US Federal register (1997). The study was approved by the local Ethics Committees. Written informed consent was obtained from all patients.
Outcome Measures Safety was assessed on the basis of recorded adverse events, vital signs, and clinical, laboratory, and ECG variables. Control of HIV infection was evaluated at baseline, Week 4, and Week 8 through CD4 cell count, plasma HIV-1 RNA, and number of patients with undetectable viral load (HIV-1 RNA < 400 copies/ml). Liver function was checked every 2 weeks for the first 2 months and then monthly. Women of childbearing potential had to use effective contraception and to have a negative baseline pregnancy test.
Statistical Analysis
Sixteen patients were enrolled from May 2002 to January 2003 and were treated with bosentan for 16 weeks. All patients completed the study and all but one were titrated to the dose of 125 mg twice daily.
Baseline Clinical Characteristics
Exercise Capacity The individual baseline and Week 16 values for the 6-minute walk distance are shown in Figure 1. The distance walked in 6 minutes after 16 weeks of treatment with bosentan improved significantly by 91 ± 60 m (n = 16) from 333 ± 79 m at baseline to 424 ± 57 m at Week 16 (p < 0.001), with a change ranging from 17 to 232 m. The two patients who demonstrated the smallest improvement in 6-minute walk distance (17 m) had relatively high baseline values (> 400 m). The Borg dyspnea index at Week 16 was significantly smaller (1.5 ± 1.6) compared with baseline (3.4 ± 2.5) (p = 0.013), indicating less severe dyspnea despite an increase in walk distance.
NYHA Functional Class All patients but two (who remained stable) improved by at least one class. Improvement was seen in nine patients as early as Week 4 (Figure 2). One patient was in class IV at baseline and was enrolled in the study because she was not a good candidate for epoprostenol due to active intravenous cocaine addiction. On bosentan, she improved to Class III, her 6-minute walk distance increased from 276 m at baseline to 350 m at Week 16, and her hemodynamics improved with a decrease in mPAP (from 50 to 36 mm Hg) and in PVR (from 917 to 725 dyn · s/cm5).
Cardiopulmonary Hemodynamics After 16 weeks of treatment with bosentan, cardiac index increased by an average of +39% and mPAP decreased by 21% (Table 3). PVR decreased in all patients (43% on average) (Figure 3).
Doppler Echocardiography At baseline, and in comparison with echocardiographic data in normal subjects (23), patients had severe dilation of the right ventricle (RV) and right atrium, with a depressed right ventricular function resulting in distorted left ventricular geometry (Table 4). At Week 16, left ventricular size was increased, the end systolic eccentricity index was reduced, and early diastolic filling parameters were improved. The Doppler-derived cardiac index increased from 2.2 ± 0.7 L/minute/m2 at baseline to 2.7 ± 0.6 L/minute/m2 at Week 16 (mean change 0.5 ± 0.4 L/minute/m2, p = 0.002). The parameters indicative of the right heart geometry and right ventricularpulmonary arterial coupling also improved.
Quality of Life Quality of life improved significantly from baseline to Week 16, as shown by the change in EQ-5D visual analog scale (from 44 ± 21 to 63 ± 17, p < 0.001), EQ-5D score (from 0.37 ± 0.43 to 0.63 ± 0.21, p = 0.036), and SF-36 health transition item (from 3.8 ± 1.3 to 1.8 ± 1.0, p 0.001). This last score improved in 13 patients (81%) and worsened in none. Significant improvements were seen in six out of the eight domains (except emotional limitation and pain) of the SF-36 questionnaire (Figure 4).
Safety The most frequent adverse events were peripheral edema (n = 5), headache (n = 3), abnormal liver function (n = 2), muscle cramps (n = 2), fluid retention (n = 2), and vomiting (n = 2). Four out of the five cases of edema were observed in patients not receiving diuretics at baseline and were corrected upon initiation of diuretic therapy. No change in body weight was observed (from 64.1 ± 18 kg to 65.9 ± 18 kg). A decrease from baseline in systolic (8 ± 15 mm Hg) and diastolic (12 ± 11 mm Hg) blood pressure with no change in heart rate (4 ± 15 beats/minute) was recorded. No patient experienced symptomatic hypotension. Asymptomatic increases (above 3 ULN in liver transaminases [alanine/aspartate aminotransferase]) were observed in two patients. Liver function tests returned to baseline values at maintenance dose in one patient. In the other patient (co-infected with hepatitis C virus), liver transaminases remained elevated (around 3 ULN) at Week 16. This patient remained on treatment with the 62.5-mg twice-daily dose; despite this low dose, a marked improvement in hemodynamic parameters from baseline to Week 16 was observed with a decrease in mPAP from 38 to 24 mm Hg and in PVR from 800 to 224 dyn · second/cm5. CD4 count and HIV viral load remained stable from baseline to Week 8 with medians of 333 to 393 cells/mm3 (median change of 20; n = 15; p = 0.52) and of 2.2 to 2.6 log 10 copies/ml (median change of 0.21; n = 13; p = 0.15), respectively. Seven patients out of the 13 in whom data were available (54%) had suppressed viral load (defined as HIV RNA < 400 copies/ml) at baseline and also at Week 8. Viral load increased at Week 8 in two patients (441 and 600 copies/ml) who were controlled at baseline with no concomitant decrease in CD4 count. Six patients were not controlled at baseline: their median viral load changed from 3.8 at baseline to 4.3 log 10 copies/ml at Week 8 (median change of 0.2 log 10 copies/ml, p = 0.69). Blood levels of antiretroviral drugs (nonnucleoside reverse transcriptase inhibitors and protease inhibitors) were measured at baseline, Week 4, and Week 8. Interpretation is limited due to the small sample size, heterogeneity of drug regimens, nonstandardized collection time, and high intra- and interindividual variability. However, no clinically relevant changes were observed from baseline. Two patients required hospitalization during the study: one for aggravated dyspnea due to bronchitis and one for fluid retention (peripheral edema and dyspnea) considered as PAH related. No patient died during the study.
Long-term Data
This uncontrolled prospective study showed that a 16-week treatment with bosentan leads to a marked improvement in all clinical and hemodynamic parameters of severe PAH associated with HIV infection. Furthermore, the hemodynamic improvement translates into improvement in right and left geometry and function assessed by Doppler echocardiography as previously reported (20). In these patients with PAH associated with HIV infection, bosentan seems at least as effective as previously observed in double-blind, placebo-controlled studies in patients with idiopathic PAH and PAH associated with connective tissue disease (11, 12). Overall, in the present study, significant improvements were reported for those parameters that are considered prognostic indicators of survival in idiopathic PAH, such as the 6-minute walk distance (24), the cardiac index (25), the PVR (26), and the right atrial area and eccentricity index on echocardiography (27). The Tei index, which is also an independent predictor of long-term prognosis in PAH (28), was not significantly improved in our study. It is to note that this index reflects global RV function and may be influenced by changes in afterload but also in preload. The main limitation of our study is its uncontrolled design, which was chosen for ethical reasons. With conventional therapy, the prognosis of HIV-associated PAH is poor with a faster evolution than for idiopathic PAH (7). Effective therapeutic options remain limited. Patients are rarely responsive to acute vasodilator testing (5) and are not good candidates for chronic treatment with calcium-channel blockers. Epoprostenol is an effective therapy but data in PAH associated with HIV infection are limited (5). In a retrospective uncontrolled study of 20 patients with severe HIV-associated PAH (5), significant improvement of 6-minute walk distance and hemodynamic parameters was observed after 3 months of treatment with epoprostenol. However, the risk of infection related to continuous intravenous administration is a significant limitation in these patients. Other therapies such as iloprost or sildenafil have only been evaluated for the treatment of HIV-associated PAH in case reports (29, 30). In addition, it has recently been suggested that HAART therapy may improve hemodynamics and survival in HIV-associated PAH (31). Although the potential beneficial effect of HAART cannot be excluded, all our patients but one had been treated with HAART for over 3 months, which renders this hypothesis less likely. Despite the limitation of a noncomparative design, the major improvements in this study were observed on functional tests (exercise capacity) and clinical symptoms (NYHA class) as well as on objective end-points such as invasive hemodynamic parameters. All findings were statistically significant, clinically relevant, and consistent. The present study is the first one conducted with bosentan in which quality of life was assessed using two complementary tools, the SF-36 questionnaire and the EQ5D. Baseline values of the SF-36 domains indicated that the quality of life of patients with HIV with PAH is poor and significantly lower than in a population of patients with HIV without PAH (32). Improvements in quality of life were observed with both tools and may be related to the improvement in cardiopulmonary hemodynamics. A major objective of the study was to assess the safety profile of bosentan in this high-risk population. The main concern with bosentan is increase in liver enzymes. Elevated aminotransferases were reported in 12% of patients on bosentan in the two double-blind, placebo-controlled studies in PAH (11, 12). These liver function abnormalities were dose-dependent with respect to incidence and severity. They usually developed gradually, remained asymptomatic, and were reversible either spontaneously or after dose reduction or discontinuation. Despite the enrolment of four patients co-infected with chronic hepatitis B or C, and co-treatment with potentially hepatotoxic antiretroviral therapies in 15 of the 16 patients, the number of cases of abnormal hepatic function observed in this study (2 out of 16 patients) is within the range of previous experience. In addition, these cases normalized or returned to values around 3 ULN, despite continuation of bosentan therapy. One other potential issue of administering bosentan to HIV-infected patients concerns possible pharmacokinetic interactions between bosentan and antiretroviral drugs because bosentan is a mild to moderate inducer of CYP 3A4 and CYP 2C9 (33). The impact of bosentan on the metabolism of other drugs has been specifically studied with warfarin and simvastatin. The interaction propensity of antiretrovirals is complex; particularly with protease inhibitors that can either induce or inhibit drug-metabolizing enzymes or both in a time-dependent fashion. Taking all facts into consideration, we hypothesized that the effect of bosentan on the blood levels of antiretroviral drugs metabolized by the CYP system was likely to be small and probably clinically insignificant. Therefore the usual monitoring of efficacy of antiretroviral drugs through CD4 count and viral load was considered appropriate to indirectly assess these possible pharmacokinetic interactions. No deleterious changes in CD4 counts and viral load were observed after 8 weeks of treatment with bosentan or during long-term follow-up, suggesting that bosentan did not reduce the efficacy of HAART treatment. In summary, this prospective clinical study conducted in patients with PAH associated with HIV infection suggests that treatment with the oral dual ET receptor antagonist bosentan improves exercise capacity, NYHA functional class, and quality of life as well as cardiopulmonary hemodynamics, cardiac geometry, and cardiac function. The combination of bosentan with antiretroviral therapy appears to be safe and to have no adverse impact on the control of HIV infection. This study indicates that ET could play a major role in PAH associated with HIV infection as in other forms of PAH. Due to the limitations of this trial (uncontrolled design, small sample size, short duration), further information on the effect of bosentan in PAH associated with HIV infection is needed to support the use of dual ET receptor antagonists in this patient population. Nevertheless, we believe that treatment with bosentan appears a suitable option in the management of patients with symptomatic PAH associated with HIV infection.
The authors acknowledge the collaboration and commitment of all the local investigators and their staff, without whom the present study would not have been possible. In particular they thank Annie Pelissier from Hôpital Antoine Béclère, Clamart, France; Ursula Treder from University Hospital Zurich, Switzerland; and Annette Pidoux from St. Vincent's Hospital, Sydney, Australia. The authors thank Jasper Dingemanse and Paul van Giersbergen from Actelion Pharmaceuticals Ltd, Allschwil, Switzerland for advice regarding drug interactions and John Wlodarczyk from John Wlodarczyk Consulting Services, Newcastle, Australia for advice regarding quality of life.
This study was supported by Actelion Pharmaceuticals Ltd., Allschwil, Switzerland.
Conflict of Interest Statement: O.S. has been involved in Scientific Advisory Boards for Actelion in 2001 ( Received in original form April 2, 2004; accepted in final form August 16, 2004
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||