|Home | About | Journals | Submit | Contact Us | Français|
Non-invasively assessed pulmonary pressure elevations and left ventricular diastolic dysfunction (LVDD) are associated with increased mortality in adults with sickle cell disease (SCD), but their relationship to exercise intolerance has not been evaluated prospectively.
Echocardiography, six-minute walk distance, hemolytic rate, and serum concentrations of ferritin and erythropoietin were evaluated in a cohort of 483 subjects with homozygous hemoglobin S in the US and UK Walk-PHaSST study. Tricuspid regurgitation velocity (TRV), which reflects systolic pulmonary artery pressure, was 2.7 to <3.0 m/sec (mean±SD 2.8±0.1) in 26% of the subjects and ≥3.0 m/sec (3.4±0.4) in 11%. LV lateral E/e′ ratio, which has been shown to reflect LV filling pressure in other conditions but has not been studied in SCD, was significantly higher in the groups with TRV ≥2.7 m/sec. Increased hemolysis (P<0.0001), LV lateral E/e′ ratio (P=0.0001), BUN (P=0.0002) and erythropoietin (P=0.002) were independently associated with an increased TRV. Further, female gender (P<0.0001), older age (P<0.0001), LV lateral E/e′ ratio (P=0.014), and TRV (P=0.019) were independent predictors of a shorter six-minute walk distance.
Echocardiography-estimated elevated pulmonary artery systolic pressure and LV lateral E/e′ ratio were independently associated with poor exercise capacity in a large cohort of patients with sickle cell anemia. Controlled trials investigating whether strategies to prevent or delay pulmonary hypertension and/or LVDD will improve exercise capacity and long-term outcomes in sickle cell anemia should be considered.
Sickle cell anemia is caused by homozygote substitution of valine for glutamic acid at position six of the β-globin chain of hemoglobin, resulting in mutant hemoglobin S polymerization with deoxygenation. This is the central event that leads to the acute and chronic complications of sickle cell disease (SCD), including vaso-occlusive episodes and hemolysis. Similar to other conditions such as HIV, cardiopulmonary complications were not prominent in SCD when therapeutic options were limited and life expectancy was short. Now that treatment and survival have improved, patients are manifesting right and left heart dysfunction.1, 2 Clinical studies using echocardiography or right heart catheterization to assess cardiopulmonary function have shown abnormal elevation in pulmonary pressures3–6 and left ventricular diastolic dysfunction (LVDD),2, 6 with each independently associated with increased mortality in adults with SCD.2–5
Laboratory, transgenic mouse and clinical studies suggest that there is a clinical syndrome of hemolysis-associated increased pulmonary artery pressure.7–15 Cross-sectional and longitudinal studies report associations between the degree of hemolysis and markers of pulmonary artery pressure in SCD.3–5, 16–18 Nevertheless, the clinical importance of elevated pulmonary pressures in SCD and their relationship to hemolysis have been challenged in recent reports, raising the need for further study in large multi-center cohorts.19–22
Exercise intolerance occurs in patients with SCD,16, 23 and correlates with elevated pulmonary pressures in previously reported case series.18, 23 However, it is unclear how often exercise limitations are due to pulmonary abnormalities including pulmonary hypertension (PH) and resultant right ventricular dysfunction, impaired LV filling, an inadequate heart rate response to exercise, or additional factors not fully explored. In SCD patients evaluated for suspected PH, increased LV filling pressures are not infrequently observed by right heart catheterization (RHC).6 In the general population, several studies have shown that Doppler echocardiography can estimate LV filling pressures24, 25 and that abnormalities are associated with exercise intolerance.26, 27 An association with a putative echocardiographic marker of LV filling pressure and exercise intolerance in children with sickle cell anemia has also been reported.18
The walk-PHaSST study (treatment of Pulmonary Hypertension and Sicklecell disease with Sildenafil Therapy) was designed to develop a large, multi-center, observational cohort of patients with a tricuspid regurgitation velocity (TRV) ranging from normal to ≥3.0 m/sec and to conduct a smaller interventional trial of treatment with sildenafil in the subset of patients with a TRV ≥2.7 m/sec. This observational cohort is the largest prospective study to date evaluating echocardiography in adults and adolescents with SCD. The purpose of this study was to non-invasively assess the relationship of echocardiographic markers of PH and LVDD with functional impairment as assessed by the six-minute walk test.
The Walk-PHaSST study was designed and supervised by an Executive Committee composed of investigators. Data management and analyses were performed by a contract organization working with the Executive Committee. The authors wrote the manuscript and made the decision to submit it for publication. All authors contributed to the writing of the manuscript and had full access to the data and analyses. The authors vouch for the accuracy and completeness of this report. Subjects were recruited at nine United States Centers and one United Kingdom Center. Local institutional review boards or ethics committees approved the protocol, and written informed consent was obtained from all study subjects (Clinicaltrials.gov identifier NCT00492531). Subjects with sickle cell hemoglobinopathy (hemoglobin SS, SC, Sβ-thalassemia and other less common sickle cell hemoglobinopathies) aged ≥12 years and at ‘steady state’ were eligible for the observational study; 720 were enrolled from February 2008 to June 2009. ‘Steady state’ in the study subjects was defined by their usual state of health, and no participant was enrolled until at least three weeks had elapsed since any acute pain event, hospitalization or emergency room visit. Subjects were evaluated by self-reported history modeled after the NIH-PH screening survey,3 physical examination, laboratory screening, transthoracic Doppler echocardiography and the six-minute walk test. All patients were included in the observational study regardless of estimated pulmonary artery systolic pressure or six-minute walk distance. The subset of patients recruited into the separate intervention study have been reported elsewhere.28 To avoid confounders related to sickle genotype, the present report focuses on the 483 subjects with hemoglobin SS who underwent screening in the Walk-PHaSST study; 250 (52%) were females and the median age was 35 years at enrollment (range 12 to 69 years).
Routine laboratory tests (complete blood count, reticulocyte count, serum chemistry profile and lactate dehydrogenase) were performed in the local laboratories of the participating institutions. Serum erythropoietin (R&D Systems, Minneapolis, MN; reference range 3.3–16.6 IU/L) and ferritin (Ramco Laboratories Inc., Stafford, TX; reference range 20–300 ng/ml) were measured with enzyme immunoassays. Hemoglobin F percent was measured by HPLC (Ultra Resolution System, Trinity Biotech).
Echocardiography was performed at the participating institutions and read centrally in the NHLBI echocardiography core laboratory. Cardiac measurements were performed according to American Society of Echocardiography guidelines.29 The LV ejection fraction was estimated using the biplane Simpson’s method.29 Left atrial volume was measured using the biplane method of discs incorporating both apical 4- and 2-chamber views.30 LV mass and LV mass index were calculated using the formula described by Devereux et al.31 Mitral inflow recordings were obtained with the pulsed Doppler sample volume at the level of the mitral leaflet tips during maximal opening in diastole. Tissue Doppler recordings were measured with sample volumes at both the septal mitral annulus and lateral mitral annulus. The ratio of early diastolic LV inflow velocity to lateral mitral annular velocity (LV lateral E/e′) was calculated as an estimate of LV filling pressure.24 TRV was used to estimate pulmonary artery systolic pressures. TRV was assessed in the parasternal long and short-axis and apical four-chamber views as previously described.3
Principal component analysis was used to derive a hemolytic component from four markers of hemolysis: lactate dehydrogenase (site adjusted-values), aspartate aminotransferase (site-adjusted values), total bilirubin and reticulocyte percent. The new variable had a mean of 0 (SD=1.48) and predicted 54% of variation among all four variables (Eigenvalue = 2.18).). Linear regression analysis was used to adjust lactate dehydrogenase and aspartate aminotransferase for site. In this analysis the residuals (which are the source of variation after site adjustment) were used to calculate the new site adjusted values. Continuous variables with a skewed distribution were converted to a normal distribution using natural log or square root. Univariate relationships were determined with the Pearson or the Spearman correlation coefficient. The study prospectively defined three TRV groups: a) <2.7 m/sec, b) 2.7 to <3.0 m/sec and c) ≥3.0 m/sec. The non-parametric test for trend was used to assess the relationship of different variables to the three TRV groups. Multiple linear regression was used to assess the independent effect of predictors on TRV, LV lateral E/e′ ratio and six-minute walk distance. Variables were entered into the regression models if they had a significant univariate association with the dependent variable (higher TRV categories, LV lateral E/e′ ratio, six-minute walk distance). For continuous variables we tested whether a linear relationship or a non-linear relationship was the best approach to analysis. We used pair wise correlation between predictors and Variance Inflation Factor to assess co-linearity in each model. We checked for an interaction between site and predictors in each model. The final model was selected using a step-wise approach. Data are reported as mean (95% CI) or geometric mean (95% CI). Missing values were the result of lack of laboratory or echocardiographic observations in random subgroups of patients. The study outcomes of TRV, LV lateral E/e′ ratio and six-minute walk distance did not differ according to whether variables were missing or not. A p value <0.05 was considered statistically significant. Analyses were performed using Stata 10.1 software (StataCorp, College Station, TX).
Clinical and echocardiographic characteristics are shown in Table 1. Of the 483 participants, 52 (10.8%) were 12 to 20 years of age.
The TRV was measurable in 435 (90.1%) of the subjects with a median of 2.6 m/sec (range of 1.6 to 4.6 m/sec). Of these 435, the TRV was 2.7 to <3.0 m/sec in 124 (28.5%) and ≥3.0 m/sec in 53 (12.2%). Among the participants 12–20 years of age, the TRV was measurable in 44 (84.6%); of these the TRV was ≥3.0 m/sec in 4 (9.1%). Univariate analysis showed that patients in the higher TRV groups were older and had higher systemic systolic pressures and systemic pulse pressures, lower resting systemic arterial oxygen saturations, and lower hemoglobin concentrations (Table 2). They had higher serum concentrations of erythropoietin, BUN, and creatinine and higher values for the hemolytic component. By echocardiography, the left atrial volume index, LV mass index, and LV lateral E/e′ ratio were higher among patients with increased TRV. Figure 1a depicts the linear relationship of TRV with the hemolytic component, and Figure 1b depicts the relationship of TRV with the LV lateral E/e′ ratio.
In multiple linear regression analysis, a 1 SD increase in the hemolytic component was independently associated with an estimated 0.09 m/sec increase in the TRV (P<0.0001), a natural log increase in LV lateral E/e′ ratio with a 0.19 m/sec increase (P=0.001), a natural log increase in the BUN with a 0.10 m/sec increase (P=0.0002) and a natural log increase in the serum erythropoietin concentration with a 0.07 m/sec increase (P=0.002) (Table 3). The hemoglobin concentration did not have a significant independent relationship with elevated TRV.
We employed the LV lateral E/e′ ratio to estimate LVDD in this study, but its usefulness for this purpose in the setting of SCD has not been studied.
The LV lateral E/e′ ratio was reported in 436 (90%) of the subjects with a median of 6.4 (range of 2.2–29.5). Increasing age, higher BMI, higher systolic blood pressure, lower hemoglobin concentration, and higher serum creatinine concentration were associated with higher LV lateral E/e′ ratios (Table 4). Additionally, the natural log LV left lateral E/e′ ratio correlated with TRV (n=415, r=0.24, P<0.0001), left atrial volume index (n=421, r=0.27, P<0.0001) and LV mass index (n=427, r=0.17, P=0.0005), but not with LV ejection fraction (n=426, r = −0.04, P=0.4) or LV internal dimension in diastole (n=436, r = −0.02, P=0.7). Figure 2 depicts the relationship between the natural log LV lateral E/e′ ratio and hemoglobin concentration.
In a multiple linear regression analysis, lower hemoglobin concentration (P=0.0001), older age (P=0.001), higher BMI (P = 0.001), and higher systolic blood pressure (P=0.007) were each independently associated with a higher natural log LV lateral E/e′ ratio (Table 5).
The six-minute walk distance correlated inversely with the TRV categories. The mean ± SD distance was 458±91 m, 438±98 m and 409±96 m in patients with velocities of <2.7 m/sec, 2.7 to <3.0 m/sec and ≥3.0 m/sec respectively (P=0.001) (Figure 3a). Six-minute walk distance correlated positively with younger age, male gender, lower BMI and higher hemoglobin concentration (Table 6), and was inversely related to the natural log LV lateral wall E/e′ ratio (R= −0.22, P<0.0001) (Figure 3b).
In a multivariate analysis, both TRV (P=0.019) and LV lateral E/e′ (P=0.014) were independent predictors of a lower six-minute walk distance (Table 7). Additional predictors were older age and female gender.
The findings from this large prospective, multicenter study of patients with sickle cell anemia demonstrate an independent association between high hemolytic rate and an increased estimated pulmonary artery systolic pressure as assessed by Doppler echocardiography. The findings additionally demonstrate an independent association between worse anemia and LV lateral E/e′ ratio. Furthermore, in this study elevations of both TRV and LV lateral E/e′ ratio were associated with functional impairment as assessed by the six-minute walk test. The mean six-minute walk distance among the hemoglobin SS patients with TRV ≥3.0 m/sec in this study was similar to the distance reported in patients with chronic thromboembolic pulmonary hypertension.32 While we have not analyzed mortality in Walk-PHaSST, these findings relating exercise capacity to estimated pulmonary artery systolic pressure and LV lateral E/e′ ratio in this large cohort parallel our previous observations with regard to mortality from the NIH.2, 3 They also provide support in adults for our recent findings in children that echocardiographic estimates of elevated right ventricular systolic pressure and LVDD may be associated with reduced exercise capacity.18 Thus, our present findings are supportive of echocardiographic screening for PH and LVDD in SCD, and suggest that these observations may have functional relevance in terms of assessing exercise capacity.
Increased TRV as estimated by Doppler echocardiography is a useful screening test for suspected PH including patients with SCD6, 7, 23 and is a marker of increased mortality in adults with SCD.2–5 In confirmed PH (diagnosed by RHC), approximately one-half of the cases were attributable to increased pulmonary vascular resistance meeting criteria for pulmonary arterial hypertension with the majority of the remainder having pulmonary venous hypertension attributed to LVDD.23 This large international study confirms independent associations of increased TRV with hemolytic severity, elevated LV filling pressure, renal dysfunction, and high circulating erythropoietin concentrations in SCD patients. The independent association of increased BUN with an increased TRV is consistent with our previous report of an association of renal disease with elevated TRV.3 The independent association of increased erythropoietin with increased TRV is also consistent with some previous observations33–35 but in contrast to others.36, 37 Circulating erythropoietin concentrations reflect the degree of tissue hypoxia,38 and the association of a greater level of erythropoietin with a higher TRV may serve as a marker of the degree of tissue hypoxia, which appears to be associated with the development of PH in other conditions.39 Additionally, erythropoietin has functions other than the stimulation of erythropoiesis, such as regulation of the development of endothelial progenitor cells and angiogenesis;40, 41 the inducement of such processes could contribute to vascular remodeling and the risk of developing PH. The circulating erythropoietin concentrations could also be a marker for potential contributions of increased erythropoiesis to increased pulmonary artery pressure. Developing erythroblasts release placental growth factor into the circulation. Placental growth factor up-regulates the production of inflammatory mediators and the potent vasoconstrictor, endothelin-1. Studies in animals and humans suggest a potential causative role for placental growth factor in PH.42
LVDD as assessed by Doppler echocardiography is also a marker of increased mortality in adults with SCD2. The results of the present study confirm that echocardiographic markers of LVDD are prevalent in sickle cell anemia and point to independent associations of older age, worse anemia, higher systolic blood pressure and elevated BMI with an elevated LV lateral E/e′ ratio. The degree of anemia, rather than the rate of hemolysis, appeared to be the stronger independent predictor of LVDD as assessed by the LV lateral E/e′ ratio in the present study. This observation is in contrast to elevated TRV, for which the strongest independent predictor was the rate of hemolysis rather than the degree of anemia. Anemia and elevated BMI have been reported as risk factors for LVDD in other settings as well.43, 44
There are a number of limitations to this study. 1) Doppler estimation of LV filling pressure is challenging in the setting of preserved LV systolic function.24, 45, 46 SCD patients have chronic anemia, high cardiac output, and volume overload, which may affect the interpretation of non-invasive measures. Doppler parameters of LV diastolic function have not been reported in SCD patients undergoing invasive measurement of LV filling pressure. 2) The study did not include the measurement of inflammatory markers, such as high-sensitivity C-reactive protein. Sickle cell disease is associated with chronic inflammation, which itself has been implicated in the cardiopulmonary complications of this disorder.47 3) Reference ranges for markers of hemolysis including lactate dehydrogenase varied considerably among the sites participating in this study. The use of principal component analysis for developing a hemolytic component that reflects shared variability among several markers of hemolysis has the advantage of permitting adjustment for each participating center in developing the hemolytic component. 4) Our findings do not fully explain PH, LVDD and limited six-minute walk distance in sickle cell anemia. The multivariate models presented in Tables 3, ,55 and and77 could explain 25%, 14% and 34% respectively of these measurements. Further research is needed to more fully understand these complications.
Moving forward, clinical trials may be warranted to evaluate strategies to prevent or delay the development of PH and LVDD in sickle cell anemia, but there are challenges in targeting hemolysis and the degree of anemia. The interventions assessing a Gardos channel inhibitor as well as the phosphodiesterase type 5 inhibitor sildenafil were both associated with increased vaso-occlusive complications.48, 49 Hydroxyurea administration, which reduces hemolysis, raises hemoglobin concentration and increases hemoglobin F percentage, has not been associated with a lower TRV in a number of large cross-sectional studies.3, 16 Prospective studies are needed to evaluate whether hydroxyurea is effective in reducing the development of PH and/or LVDD. Blood transfusions could also be expected to improve left-sided cardiac function and the six-minute walk distance in hemoglobin SS patients. Chronic transfusion therapy may therefore be an ideal approach to prevent and treat both abnormalities, but would be complicated by iron overload and alloimmunization. We suggest that hydroxyurea, blood transfusion, and novel approaches be investigated in patients at increased risk of developing PH and/or elevated LV lateral E/e′ ratio, as these complications independently predict poor outcome in patients with sickle cell anemia.
We thank Ines Cabrita for assistance with performing echocardiograms and Cynthia Brenneman and Wen Li for assistance with reviewing echocardiograms.
This project was funded with funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under contract HHSN268200617182C. ClinicalTrials.gov Identifier: NCT00492531. This work was supported in part by NIH CTSA grant UL1 RR024131 (to CRM), grant nos. 2 R25 HL003679-08 and 1 R01 HL079912-02 from NHLBI (to VRG), by Howard University GCRC grant no 2MOI RR10284-10 from NCRR, NIH, Bethesda, MD, by NIH grants R01HL098032, RO1HL096973, and RC1DK085852 to Dr. Gladwin, by the Institute for Transfusion Medicine and the Hemophilia Center of Western Pennsylvania, by the intramural research program of the National Institutes of Health, and by a grant from the NIHR Biomedical Research Centre (to JSRG).
Robyn J. Barst has received support for research grants and/or consulting from Actelion, Eli Lilly, Gilead, Glaxo Smith-Kline (GSK), Medtronics, Bayer, Ikaria, Pfizer, Novartis, United Therapeutics, and NHLBI. Gregory J. Kato has received research support from a cooperative research and development agreement between the National Institutes of Health and Ikaria INO Therapeutics and from the Division of Intramural Research of the National Institutes of Health. Victor R. Gordeuk has received research support from Biomarin and TRF-Pharma, and Emmaus Pharmaceuticals. J. Simon R. Gibbs has received research support from Actelion and Bayer and has served on advisory boards and/or received lecture fees from Actelion, Bayer, GSK, Lilly, Pfizer and United Therapeutics. Lakshmanan Krishnamurti has received research support from the National Heart Lung and Blood Institute (NHLBI HB-06-06). Reda E. Girgis has served as an advisory board member for Actelion, Gilead, and United Therapeutics and has received research support from Actelion, Gilead, United Therapeutics, Pfizer, and ICOS. Erika Berman Rosenzweig has received research support from Pfizer. David B. Badesch has received research support from the NHLBI, has participated as a consultant or steering committee or advisory board member for Actelion/CoTherix, Gilead/Myogen, Encysive, Pfizer, Mondo-Biotech/Mondogen, Biogen IDEC, United Therapeutics/Lung Rx, GlaxoSmithKline, and Lilly/ICOS. Sophie Lanzkron has received award K23HL083089-03 from the NHLBI. Mark T. Gladwin has received research support in the form of a Collaborative Research and Development Agreement between the US Government and INO Therapeutics and is listed as a co-inventor on a US Government Patent for the use of nitrite salts for cardiovascular indications.
Clinical Trial Registration. www.clinicaltrials.gov identifier NCT00492531.
Roberto F. Machado, Kathryn L. Hassell, Jane A. Little, Dean E. Schraufnagel, Claudia R. Morris, Oswaldo L. Castro, Vandana Sachdev, and Jonathan C. Goldsmith report no relevant conflict of interest related to the subject of this manuscript.