Pulmonary Hypertension: Title and subTitle BreakAn Increasingly Recognized Complication of Hereditary Hemolytic Anemias and HIV Infection

A man with homozygous sickle cell disease (SCD) presented for evaluation of increasing dyspnea on exertion. During the prior year, episodes of vaso-occlusive pain crisis and acute chest syndrome increased and dyspnea and peripheral edema developed.

Physical examination was notable for normal vital signs, jugular venous distention, a loud P₂, a diastolic murmur at the left sternal border, and a holosystolic murmur at the apex. There was no leg edema. The hemoglobin was 10.5 g/dL (normal: 12.7-16.7 g/dL; to convert to g/L, multiply by 10.0).

Further evaluation revealed a low probability ventilation/perfusion scan, pulmonary function tests with a low diffusing capacity for carbon monoxide, and a positive serum enzyme-linked immunosorbent assay and Western blot for human immunodeficiency virus (HIV) with a CD4 T-cell count of 342/μL (normal is 300-800/μL) and a viral load of 49 500 copies/mL (normal is an undetectable level). He walked 240 m in 6 minutes (normal is 720 m). An echocardiogram showed right ventricular dilation and systolic dysfunction, a tricuspid regurgitant jet velocity (TRV) of 3.7 m/s (estimated pulmonary arterial systolic pressure of 60 mm Hg; normal is <30 mm Hg), and normal left ventricular function (Figure 1 and Figure 2) (see video here). Right heart catheterization showed a pulmonary arterial systolic blood pressure of 64 mm Hg and diastolic blood pressure of 39 mm Hg (mean, 47 mm Hg; normal is <25 mm Hg), pulmonary capillary wedge pressure of 14 mm Hg (normal is <15 mm Hg), and mixed venous oxygen saturation of 58% (normal is >75%). His cardiac output was 6 L / min (normal is 4-6 L /min) and his pulmonary vascular resistance was 440 dynes × s × cm⁻⁵ (normal is <240 dynes × s × cm⁻⁵). Therapeutic options were considered.

Pulmonary hypertension is a disease process associated with an increase in the mean pulmonary arterial pressure. It can be a consequence of several pathological states including pulmonary arterial hypertension, left heart disease, lung diseases associated with hypoxia, and chronic pulmonary embolism. Specifically, pulmonary arterial hypertension defines an increase in the mean pulmonary arterial pressure related to arteriopathy of the pulmonary vasculature. Pulmonary arterial hypertension can be idiopathic, familial, or secondary to a variety of conditions such as connective tissue disease, congenital systemic-to-pulmonary shunts, drugs and toxins, hemoglobinopathies, liver cirrhosis, or HIV infection.¹ Pulmonary arterial hypertension leads to a progressive increase in mean pulmonary arterial pressure and pulmonary vascular resistance, and a decrease in cardiac output. As the disease progresses, pulmonary arterial pressure may normalize or decrease as right heart failure occurs and cardiac output falls, ultimately leading to progressive exercise limitation and death. Because the etiology of elevated pulmonary pressures is mixed in SCD (pulmonary arterial hypertension and left heart dysfunction), it will be referred to as pulmonary hypertension for the remainder of this article.

Patients with pulmonary hypertension present with symptoms of dyspnea on exertion, fatigue, and syncope. Many patients in the early stages of the disease will attribute symptoms to deconditioning and not present for evaluation when symptoms are first noted. Once patients present, symptoms are often attributed to chronic illness, therefore delaying the diagnosis of pulmonary hypertension by months to years. Pulmonary hypertension is a an entity that is clinically important to recognize because it adversely affects quality of life and survival, and prognosis might be improved with therapeutic intervention.

Hemolytic diseases and HIV could represent major causes of pulmonary hypertension. There are 30 million individuals worldwide with SCD² and 40 million individuals worldwide with HIV infection.³ Given that pulmonary hypertension occurs in 10% to 30% of patients with hemoglobinopathies⁴ and 0.5% of patients with HIV,⁵ there may be as many as 3 million to 9 million patients with hemoglobinopathies and 200 000 patients with HIV infection affected by pulmonary hypertension worldwide. If the natural history of pulmonary hypertension is as ominous in SCD and HIV infection as it is in other patient populations, pulmonary hypertension will become a major health care problem in these populations.

Remarkably, virtually every cause of hemolytic anemia, including SCD, thalassemia intermedia and major, hereditary spherocytosis and stomatocytosis, paroxysmal nocturnal hemoglobinuria, hemoglobin-Mainz hemolytic anemia, microangiopathic hemolytic anemia, malaria, hemolysis from mechanical heart valves, left ventricular assist devices, and cardiopulmonary bypass procedures, has been associated with pulmonary hypertension (recently reviewed by Gladwin and Kato,⁶ complete list of references at http://www.cc.nih.gov/ccmd/resources/bibliography.html). In contrast, there have been no reports of pulmonary hypertension associated with anemia of chronic disease or iron-deficiency anemia. These observations emphasize the critical role of hemolysis in the development of pulmonary hypertension.

Retrospective studies of patients with SCD and thalassemia indicate that the prevalence of clinically evident pulmonary hypertension ranges from 10% to 50%.^{4 ,7} Histological findings of pulmonary hypertension are even more common in SCD—75% of patients have some evidence at autopsy.⁷ These data were corroborated by the prospective National Institutes of Health Pulmonary Hypertension Screening Study, which found that 32% of patients with SCD had elevated pulmonary arterial systolic pressures (defined by a TRV ≥2.5 m/s, which corresponds to a pulmonary arterial systolic pressure of approximately 30 mm Hg) and 9% had moderately to severely elevated pressures (defined as a TRV of ≥3.0 m/s, which corresponds to a pulmonary arterial systolic pressure of approximately 41 mm Hg).⁷ Similar rates were found in echocardiographic studies of patients with sickle β-thalassemia and SCD.^{8 - 9} Retrospective measurement of N-terminal pro-brain natriuretic peptide, a prohormone released by the right and left ventricular myocardium under pressure stress, showed that 30% of patients with SCD had elevated levels, suggesting the possible presence of pulmonary hypertension.¹⁰

Pulmonary hypertension increases the risk of death for patients with SCD. In the National Institutes of Health study, compared with patients with a TRV of less than 2.5 m/s (pulmonary arterial systolic pressure of 30 mm Hg), the rate ratio for death for a TRV of 2.5 to 2.9 m/s (pulmonary arterial systolic pressure of 30-37 mm Hg) and higher than 3.0 m/s (pulmonary arterial systolic pressure of 41 mm Hg) was 4.4 (95% confidence interval [CI], 1.6-12.2; P < .001) and 10.6 (95% CI, 3.3-33.6; P < .001), respectively, when followed up for a median of 27 months.^{7 ,11} In support of these findings, De Castro et al⁹ found that 6 of 42 patients (14%) with pulmonary hypertension and 2 of 83 patients (2%) without pulmonary hypertension died during a 2-year follow-up period. Similarly, in the study by Ataga et al,⁸ 9 of 36 patients with pulmonary hypertension and 1 of 57 patients without pulmonary hypertension died during the 2.5-year follow-up period (relative risk, 9.24; 95% CI, 1.20-73.30). To date, there are no studies evaluating the impact of pulmonary hypertension on survival in patients with thalassemia or other hemoglobinopathies.

For patients with SCD, the severity of hemolysis correlates with the severity of pulmonary hypertension.⁷ This suggests that hemolysis is related mechanistically to pulmonary hypertension. Such a relationship is biologically plausible because free hemoglobin inactivates the intrinsic vasodilator nitric oxide. Hemolysis also releases arginase, which depletes L-arginine, the substrate for nitric oxide synthesis,¹² resulting in a state of decreased nitric oxide bioavailability and “resistance” to nitric oxide–dependent vasodilatation.¹³ Hemolysis and decreased nitric oxide bioavailability also induce platelet activation,¹⁴ thrombin generation, and tissue factor activation.¹⁵ These factors all contribute to an increased risk of thrombosis. The accumulation of redox active heme and iron from lysed red blood cells further contributes to the generation of reactive oxygen species that can exacerbate ischemia-reperfusion injury, thrombosis, and endothelial and smooth muscle proliferative responses.¹⁶ Thus, hemolysis triggers multiple processes that could cause pulmonary hypertension.

Other factors could contribute to pulmonary hypertension in SCD. In patients with SCD, both at steady state and during vaso-occlusive pain crises, plasma endothelin-1 levels are increased.¹⁷ In vitro, sickle erythrocytes increase endothelin-1 production by cultured human endothelial cells. In addition, endothelin receptor A antagonism abolishes the vasoconstrictive effects of media from pulmonary endothelial cells exposed to sickled erythrocytes on aortic rings.¹⁷ These observations all suggest enhanced vasoconstrictive activity in SCD and other hemolytic disorders.

Loss of splenic function may trigger platelet activation, promoting pulmonary microthrombosis and red cell adhesion to the endothelium.¹⁸ The spleen also plays a critical function in the removal of senescent and damaged erythrocytes. Following splenectomy, the rate of intravascular hemolysis increases.^{19 - 20} Thus, splenic dysfunction can produce a series of processes that contribute to pulmonary hypertension.

Historically, pulmonary hypertension in patients with SCD was thought to derive from the repetitive episodes of pulmonary vaso-occlusive crisis and acute chest syndrome leading to pulmonary fibrosis, vascular obstruction, and chronic hypoxia. However, this association has not been supported by epidemiological studies because the number of episodes of vaso-occlusive crisis and acute chest syndrome is not associated with pulmonary hypertension.^{7 ,10} In addition, pulmonary hypertension occurs in other hemolytic disorders such as thalassemia, which is not associated with vaso-occlusive disorders or chest syndromes.^{4 ,13} Thus, the theory that vaso-occlusive disorders or chest syndrome contribute to pulmonary hypertension may not be valid.

Patients with chronic hemolytic disorders become symptomatic when mean pulmonary arterial pressures reach 30 to 40 mm Hg.²¹ In contrast, patients with other forms of pulmonary hypertension do not become symptomatic until mean pulmonary arterial pressure reaches 50 to 60 mm Hg.²² Patients with chronic hemolytic disorders also have mild elevations in pulmonary vascular resistance and a coexistent mild elevation in pulmonary capillary wedge pressure, suggesting left heart failure. Right heart catheterization data show that pulmonary arterial hypertension is present in 54% of patients with SCD while pulmonary venous hypertension secondary to left heart disease is present in 46%.²¹ Sachdev et al²³ found that 47% of patients had pulmonary hypertension, diastolic dysfunction, or both (29% had pulmonary hypertension alone, 11% had diastolic dysfunction and pulmonary hypertension, and 7% had diastolic dysfunction alone). Pulmonary hypertension and diastolic dysfunction were associated with a relative risk of death of 5.1 (95% CI, 2.0-13.3) and 4.8 (95% CI, 1.9-12.1), respectively, while the relative risk of death when both were present was 12.0 (95% CI, 3.8-38.1).²³ These data suggest that both pulmonary hypertension and diastolic dysfunction arise independently and carry additive mortality risk.

Patients with anemia maintain a high compensatory resting cardiac output to ensure adequate oxygen delivery. As such, they appear to be poorly tolerant of even small additional increases in pulmonary vascular resistance that may be associated with exercise or hemolytic crisis.²⁴ Compared with SCD alone, patients with SCD and pulmonary hypertension had a decreased 6-minute walking distance (mean [SE], 320 [20] m vs 435 [31] m; P = .002) and peak oxygen consumption on cardiopulmonary exercise testing (mean [SE], 41% [2%] of predicted vs 50% [3%]; P = .02).²¹ Patients with pulmonary arterial hypertension from other etiologies and a higher mean pulmonary arterial pressure (54 mm Hg) also performed better on 6-minute walk testing (398 m) and exercise testing (46% of predicted).²⁵

For patients with chronic hemolytic disorders, exercise intolerance is often attributed to anemia or lung disease rather than pulmonary hypertension. The diagnostic evaluation of exercise intolerance should include an aggressive search for other conditions that might contribute to pulmonary hypertension, such as iron overload, chronic liver disease, HIV, nocturnal hypoxemia, and thromboembolism. A diagnostic right heart catheterization is essential.²⁶ The 6-minute walk test inversely correlates with the severity of pulmonary hypertension,²¹ and pulmonary hypertension–specific therapy improves walking distance.²⁷ Thus, the 6-minute walk test is a useful surrogate for functional capacity in this patient population.

The prevalence of HIV-associated pulmonary arterial hypertension has been estimated at between 0.06% and 2%.⁵ The most comprehensive assessment of HIV-associated pulmonary arterial hypertension has been a prospective cohort study of 10 547 patients with HIV, which showed a prevalence of pulmonary arterial hypertension of 0.21% (95% CI, 0.12%-0.30%).²⁸ In this study, pulmonary arterial hypertension was diagnosed by right heart catheterization. Two studies have assessed the impact of antiretroviral therapy on the prevalence of pulmonary arterial hypertension, but have come to opposite conclusions regarding whether antiretroviral therapy is associated with reduction in pulmonary arterial hypertension.⁵

A retrospective echocardiographic study in patients treated at the National Institutes of Health HIV clinic showed that 9.3% of patients had a TRV of 2.5 m/s or higher (pulmonary arterial systolic pressure of 30 mm Hg) and 0.4% had a TRV of 3.0 m/s or higher (pulmonary arterial systolic pressure of 41 mm Hg).²⁹ When the maximum TRV for each patient during the observation period was considered, 16.4% had a TRV of 2.5 m/s or higher and 0.7% had a TRV of 3.0 m/s or higher on at least 1 occasion. Univariate analysis showed significant associations between elevated TRV and increased age (P = .05), male sex (P = .03), lower CD4 T-cell count (r = −0.14; P = .02), and absence of interleukin 2 exposure (P = .02). The prevalence of elevated TRV of 0.4% to 0.7% among the study population is consistent with that reported previously,⁵ as are associations with older age and male sex.

In a similar analysis of patients at San Francisco General Hospital, TRV and right atrial pressure were used to estimate pulmonary arterial systolic pressure in 186 individuals with HIV infection and 36 age-matched controls without HIV infection.³⁰ Individuals with HIV infection had a higher pulmonary arterial systolic pressure compared with controls (median [interquartile range], 26 [19-31] mm Hg compared with 21 [17-26] mm Hg; P = .001). A pulmonary arterial systolic pressure of greater than 30 mm Hg and 40 mm Hg was found in 30% of patients with HIV compared with 6% of controls (P = .003), and 6% of patients with HIV compared with 0% of controls (P = .28), respectively. After adjustment for age, sex, smoking, and intravenous drug use, individuals with HIV had 5.5-fold greater odds of having a pulmonary arterial systolic pressure higher than 30 mm Hg (P = .02).

The prevalence of increased pulmonary arterial systolic pressure among the San Francisco General Hospital cohort (6%) is significantly higher than among the National Institutes of Health cohort (0.4%-0.7%), but is similar to that found in a German screening study (7.5%).³¹ The reason for this difference could be related to demographics, mode of transmission, or intravenous drug use. Although the evidence linking drugs of abuse and the development of pulmonary hypertension in HIV infection is inconclusive,^{32 - 33} the prevalence of intravenous drug use in the San Francisco General Hospital cohort was 38% compared with 4% in the National Institutes of Health cohort.

Pulmonary arterial hypertension increases mortality in patients with HIV infection as it does in other patient populations. In a case-control study, the relative risk of death in individuals with HIV infection and pulmonary hypertension was 2.1 (95% CI, 1.0-4.5; P < .05) compared with individuals with HIV but without pulmonary hypertension.³⁴ In addition, the median survival for patients with pulmonary hypertension was 1.3 years compared with 2.6 years in the group without pulmonary hypertension (P < .05). Although these data are 10 years old, they are perhaps best representative of the natural history of the effects of pulmonary hypertension on mortality in patients with HIV infection because the population is relatively uniform. Only a small proportion of patients were treated with antiretroviral drugs (zidovudine only) and no patients received treatment targeting the pulmonary hypertension. Thus, although HIV-associated mortality is significantly improved by highly active antiretroviral therapy, mortality data from the era before highly active antiretroviral therapy and before pulmonary hypertension–specific therapy may provide the clearest description of the effects of HIV-associated pulmonary arterial hypertension on survival.

Patients with HIV and pulmonary arterial hypertension have plexogenic lesions similar to patients with other diseases associated with pulmonary arterial hypertension.³⁵ There has been no consistent association between pulmonary arterial hypertension and CD4 T-cell count (contrary to the findings of the National Institutes of Health cohort) or HIV viral load.³⁶

Human immunodeficiency virus has never been shown to directly infect pulmonary vascular endothelial cells,^{37 - 38} but HIV viral antigens are present in pulmonary endothelium and may directly stimulate abnormal apoptosis, growth, and proliferation.³⁹ Glycoprotein 120, a viral protein necessary for the binding and entry of HIV into macrophages, has been shown to target human lung endothelial cells, increase markers of apoptosis, and stimulate the secretion of endothelin-1.³⁷ The negative factor (Nef) antigen, critical for the maintenance of viral loads and for host cell signaling interactions, has been localized to multiple pulmonary and vascular cell types.⁴⁰ In vitro, human endothelial cells exposed to Nef demonstrate increased apoptosis followed by proliferation.⁴¹ Primates infected with a simian immunodeficiency virus expressing HIV Nef protein develop lesions resembling plexiform lesions, and Nef antigen is detectable in alveolar, vascular, perivascular, and lymphoid tissues.⁴⁰ Thus, HIV could plausibly be linked to the genesis of pulmonary arterial hypertension.

There are other mechanisms by which HIV could cause pulmonary arterial hypertension. Human immunodeficiency virus infection induces a chronic inflammatory state and persistent immune activation and dysregulation⁴² that could indirectly induce the release of proinflammatory cytokines and growth factors that could produce pulmonary arterial hypertension.⁴³ Increased expression of platelet-derived growth factor, a potent stimulus of smooth muscle cell and fibroblast growth and migration, also has been noted in lung tissue from patients with HIV-associated pulmonary arterial hypertension.³⁸ Similarly, vascular endothelial growth factor A induces vascular permeability and endothelial cell proliferation and is produced by T cells infected by HIV in vivo.⁴⁴

Bone morphogenetic protein receptor 2 (BMPR-2) mutations, associated with familial pulmonary arterial hypertension, result in decreased signaling through BMPR-2.⁴⁵ The HIV-1 tat protein (transcriptional transactivator) represses BMPR-2 gene expression in human macrophages in vitro, interfering with transcriptional regulation of BMP and BMPR-2.⁴⁵ Exogenous tat protein also has been shown to activate endothelial cells, resulting in the release of growth factors,⁴⁵ supporting the hypothesis that HIV viral proteins could induce aberrant endothelial function, leading to pulmonary arterial hypertension.

Human herpesvirus 8 has been reported to be associated with pulmonary arterial hypertension histologically.⁴⁶ This association has not been consistently confirmed by others.⁴⁷

Therapy for pulmonary arterial hypertension has classically included diuretics, cardiac glycosides, supplemental oxygen, warfarin anticoagulation, and high-dose calcium channel blockers in selected patients. Prostanoids, endothelin antagonists, and phosphodiasterase-5 inhibitors have been used in symptomatic patients to improve hemodynamics and increase exercise tolerance.⁴⁸

Because chronic intravascular hemolysis is a central mechanism in the development of pulmonary hypertension, the underlying hemoglobinopathy therapy should be optimized. For SCD, hydroxyurea and transfusion therapy should be optimized.^{49 - 50} A similar strategy seems rational for other types of hemolytic anemias. A recent study⁵¹ showed that pulmonary hypertension was completely prevented in patients with thalassemia major who were well transfused and iron-chelated.

Therapies that increase nitric oxide bioavailability are currently being assessed as treatments for pulmonary arterial hypertension. Chronic inhaled nitric oxide selectively dilates the pulmonary vasculature and oxidatively inactivates circulating plasma hemoglobin.⁵² Chronic inhaled nitric oxide is investigational, expensive, and requires a complicated delivery system.⁵³ L-arginine has been shown to decrease pulmonary arterial systolic pressure by 15% in a small study of patients with SCD⁵⁴ and seems promising as another approach to improving nitric oxide bioavailability. Sildenafil functions by inhibiting the metabolism of cyclic guanosine monophosphate, the second messenger that mediates the effects of nitric oxide. In a recently published case series,⁵⁵ 7 patients with various forms of thalassemia and severe pulmonary hypertension were treated with sildenafil from 4 weeks to 48 months with an improvement in TRV, New York Heart Association functional class, and 6-minute walking distance. Twelve patients in the National Institutes of Health cohort were treated with sildenafil for a mean of 6 months.²⁷ Mean pulmonary arterial systolic pressure decreased by 9 mm Hg (95% CI, 0.3-17.0 mm Hg; P = .05), 6-minute walking distance improved by 78 m (95% CI, 40-117 m; P = .003), mean N-terminal pro-brain natriuretic peptide decreased by 448 pg/mL (P = .002).²⁷ These results are encouraging and suggest the need for a larger trial sponsored by the National Heart, Lung, and Blood Institute.

There have been no large studies on the effects of antiretroviral therapy on HIV-associated pulmonary arterial hypertension. Epoprostenol and bosentan appear to be effective in terms of improving symptoms and hemodynamics, but studies to date have involved few patients.⁵⁶ Sildenafil has also been used with encouraging results.⁵⁶ However, interactions between sildenafil and antiretroviral agents, especially ritonavir, can lead to dangerously high sildenafil levels—symptomatic hypotension can occur if dose adjustments are not made.

Thus, HIV-associated pulmonary arterial hypertension has been successfully treated in some patients, but the effects of these agents on the natural history of pulmonary arterial hypertension remain to be determined in this patient population.

In the case presented, the patient's SCD was managed with exchange transfusions and combination antiretroviral therapy (stavudine, lamivudine, and efavirenz). The patient's hemoglobin remained stable, his CD4 T-cell count increased to 400/μL, and his viral load decreased to 400 copies/mL. His pulmonary hypertension was treated with sildenafil. He experienced symptomatic improvement and the 6-minute walking distance increased to 430 m. A follow-up echocardiogram showed persistent elevation in TRV of 3.2 m/s (pulmonary arterial systolic pressure of 46 mm Hg), but the right ventricle size and systolic function improved. His clinical condition remained stable for approximately 1.5 years when he developed worsening exertional dyspnea related to progression of pulmonary hypertension, which required the addition of bosentan to his regimen. He experienced initial improvement in symptoms, functional capacity, and pulmonary pressures. However, after 2 years he developed progressive right heart failure. Therapy with intravenous or subcutaneous prostacyclin was offered, which was refused by the patient. Inhaled iloprost was initiated, but shortly thereafter the patient died suddenly while at home. Thus, recognition of this patient's pulmonary hypertension, and appropriate therapy, led to symptomatic improvement. Whether survival is improved can only be determined by larger studies.

As patients with hemolytic disorders and HIV survive longer, new pathological processes that will threaten their improved quality of life and survival are likely to become apparent. Pulmonary hypertension appears such a threat. Screening for pulmonary hypertension in these populations is warranted. Because drugs are becoming available that can alter the progression of pulmonary hypertension, investigations should be expanded to understand how pulmonary hypertension and its therapy should be managed.