Clinical differences between idiopathic and scleroderma-related pulmonary hypertension

Authors


Abstract

Objective

Pulmonary arterial hypertension related to scleroderma (PAH-Scl) is associated with high morbidity and mortality as well as poorer response to therapy and worse outcomes compared with the idiopathic form of PAH (IPAH). Scleroderma is an autoimmune disease that can affect left and right heart function directly through inflammation and fibrosis and indirectly through systemic and pulmonary hypertension. This study tested the hypothesis that an increased prevalence of left heart disease might explain the higher mortality in patients with PAH-Scl compared with patients with IPAH.

Methods

The study was designed as a retrospective cohort study comparing the baseline clinical data from 91 consecutive patients (41 with IPAH and 50 with PAH-Scl). Cox proportional hazards models were used to predict the effect of clinical covariates on patient survival.

Results

Patients with PAH-Scl had a lower mean pulmonary artery pressure (46.6 mm Hg versus 54.4 mm Hg in patients with IPAH; P = 0.002) despite similar levels of cardiac dysfunction (cardiac index 2.2 and 2.1 liters/minute/m2, respectively; P = 0.19). Echocardiography revealed similar degrees of right ventricular dysfunction in the 2 groups, whereas a predominance of left heart dysfunction was observed in patients with PAH-Scl. One- and three-year survival estimates were 87.8% and 48.9%, respectively, in patients with PAH-Scl and 95.1% and 83.6%, respectively, in those with IPAH. Patients with PAH-Scl were 3.06 times more likely to die than were patients with IPAH, after controlling for the presence of pericardial effusion; there was no significant change in increased risk of death in PAH-Scl after controlling for left heart disease.

Conclusion

The results confirm that there are significant clinical and survival differences between IPAH and PAH-Scl. The presence of left heart disease, although more common in PAH-Scl, was not predictive of the higher mortality in these patients.

The presence of pulmonary hypertension, defined as a mean pulmonary artery pressure of >25 mm Hg, is associated with significant morbidity and mortality (1–3). The recent consensus definition of pulmonary arterial hypertension (PAH) subdivides the condition into multiple categories, including idiopathic PAH (IPAH) and PAH associated with other diseases or conditions, such as infection with the human immunodeficiency virus, portal hypertension, anorexigen use, and connective tissue disease (4). IPAH (formerly known as primary pulmonary hypertension) is the prototypic form and most commonly studied subtype of PAH, both in phenotypic analyses and in treatment trials (2, 5, 6). Recently, there has been increasing interest in PAH related to systemic sclerosis/scleroderma (PAH-Scl), which is the connective tissue disease most often associated with PAH (7).

Scleroderma is an autoimmune disease with the potential for multiple organ system involvement, including the gastrointestinal, cardiac, renal, and pulmonary systems (8). It is a rare disease, with a reported prevalence in the US ranging from 138 cases per million to 286 cases per million (9–11). Estimates of the prevalence of PAH in patients with scleroderma vary on the basis of the criteria used for diagnosis of pulmonary hypertension and the method of obtaining the measurements (i.e., echocardiography or cardiac catheterization). Using cardiac catheterization for diagnosis, the prevalence of pulmonary hypertension appears to be 12% (12) when including patients with interstitial lung disease, and 7.85% when limiting the analysis to patients with PAH (13). However, most of the data on the prevalence of PAH come from university or tertiary care centers, where the tendency is toward severe disease, which could lead to bias and thus provide an underestimate of the prevalence of PAH. In a study assessing the echocardiographic diagnosis of pulmonary hypertension in community rheumatology centers, the prevalence of PAH was estimated to be 13.3% among a population of patients with scleroderma or mixed connective tissue disease (14).

Compared with IPAH, PAH-Scl demonstrates a poorer response to therapy, and patients with PAH-Scl have worse long-term survival (15, 16). The reason for these differences is poorly understood. Compared with patients with IPAH, patients with PAH-Scl have an older age at disease onset, and PAH is known to be a later manifestation of the disease (17). In addition, PAH is likely to be developing at a time when concomitant left heart disease is contributing to poorer outcomes. Echocardiography analysis of heart structure and function is able to provide some information about heart disease not assessed by catheterization.

The purpose of this study was to examine the hemodynamic and echocardiographic findings in well-characterized populations of IPAH and PAH-Scl patients at the time of diagnosis, and to evaluate whether differences exist that might affect survival. We hypothesized that patients with PAH-Scl would have more echocardiographic evidence of left heart disease (i.e., left atrial dilatation, left ventricular hypertrophy, and diastolic dysfunction) than patients with IPAH, and that this difference would be predictive of higher mortality in those with PAH-Scl.

PATIENTS AND METHODS

Study sample and measurements.

The Johns Hopkins University Institutional Review Board reviewed and approved the conduct of this retrospective cohort study. The cohort was derived from the Johns Hopkins Pulmonary Hypertension Program, which maintains a registry of all patients evaluated at the clinical center in Baltimore. We identified all patients in the registry with evidence of pulmonary hypertension (mean pulmonary artery pressure >25 mm Hg) based on the findings of right heart catheterization performed between January 1, 2000 and June 1, 2005, leading to a diagnosis of IPAH or PAH-Scl.

The diagnosis of scleroderma was based on 1 of 3 definitions: the American College of Rheumatology (formerly, the American Rheumatism Association) criteria (18); the presence of 3 of 5 features of the CREST (calcinosis, Raynaud's phenomenon, esophageal dysmotility, sclerodactyly, telangiectasias) syndrome; or definite Raynaud's phenomenon, abnormal nailfold capillaries typical of scleroderma, and the presence of a specific scleroderma-related autoantibody. Limited skin involvement was defined as skin tightening distal to the elbows and knees with or without facial involvement. Diffuse skin involvement was defined as tightening proximal to these joints or truncal involvement.

Patients were excluded from the study if their first right heart catheterization documenting pulmonary hypertension preceded the study entry date. In addition, patients were excluded if they had evidence of pulmonary venous hypertension (pulmonary capillary wedge pressure >15 mm Hg), significant chronic obstructive or interstitial lung disease, portal hypertension, severe obstructive sleep apnea, or chronic thromboembolic disease. Significant chronic obstructive lung disease was defined as a forced expiratory volume in 1 second (FEV1) to forced expiratory volume (FVC) ratio of <70% and an FEV1 of <60% of predicted. Interstitial lung disease was defined based on the results of a combination of pulmonary function tests and chest radiography. Patients were excluded if they had a total lung capacity (TLC) of <60% of predicted, and were included if the TLC was ≥70% of predicted. Patients with a TLC between 60% and 70% of predicted were included if their computed tomography scan showed only minimal interstitial fibrosis.

Patients were also excluded from the study if they tested positive for antibodies to the human immunodeficiency virus, had a history of anorexigen use including phen-fen, or had any other disease known to be associated with pulmonary hypertension. Portal hypertension was diagnosed based on the usual clinical criteria (19). Patients with chronic thromboembolic disease were excluded based on the results of ventilation and perfusion scanning, contrast-enhanced computed tomography, and, if necessary, pulmonary angiography.

Clinical records were reviewed to identify the echocardiogram reports, pulmonary function test results, World Health Organization (WHO) functional classification, and results of 6-minute walking tests done closest to the date of diagnosis. The primary outcome of death was determined by reviewing the clinical and hospital records as well as the Social Security Death Index.

Echocardiogram reports were reviewed for documentation of the abnormalities of interest. In the event that no comment was made regarding an abnormality, it was assumed that the abnormality was not present. For the purposes of analysis, the presence of a pericardial effusion was dichotomized between the report of no or minimal effusion present and the report of a small or larger effusion.

Treatment was established in accordance with the standards of practice of the Johns Hopkins Pulmonary Hypertension Center. There was no effort to randomize or assign specific therapy to either group.

Statistical analysis.

Continuous variables were compared using Student's t-test. Categorical variables were compared using the chi-square statistic. The Mantel-Haenszel statistical method was used to control for the possible effects of a history of systemic hypertension and age on the presence of left heart disease (20). Survival analysis was performed using the Kaplan-Meier method to compare the time from diagnosis of IPAH or PAH-Scl to death in the 2 groups. The date of diagnosis was defined as the date of the patient's first cardiac catheterization documenting pulmonary hypertension. Bivariate Cox proportional hazards modeling was performed to determine the variables associated with increased risk of death. We then constructed a multivariable Cox proportional hazards model to determine if patients with PAH-Scl had a higher risk of death than patients with IPAH, independent of demographic features, hemodynamic findings, and functional classification. A backward stepwise approach was utilized in these analyses (variable removed if P > 0.2, variable entered if P < 0.05). Variables indicating the presence of left heart disease on echocardiograms were then added to the model to determine their impact on hazard ratio (HR) estimates. A 2-tailed P value was used to indicate statistically significant differences between groups. All computations were performed using Stata statistical software (version 8.0; Stata, College Station, TX).

RESULTS

Demographics and pulmonary function.

From January 1, 2000 to June 1, 2005, 91 patients (41 with IPAH and 50 with PAH-Scl; all were predominantly white and female) met the eligibility criteria for the study. Table 1 presents the baseline demographic and clinical data, WHO functional classification, and 6-minute walk distance at the time of diagnosis for both groups of patients. The patients with PAH-Scl were significantly older (mean age difference 11.1 years) than the patients with IPAH at the time of diagnosis. Among the scleroderma patients, the limited variant of the disease was overwhelmingly represented (92.0%). There was no difference in baseline 6-minute walk distances between groups.

Table 1. Patients' demographic characteristics and results of functional assessments*
 IPAH (n = 41)PAH-Scl (n = 50)P
  • *

    Except where indicated otherwise, values are the mean ± SEM. P values were determined by Student's t-test. IPAH = idiopathic pulmonary arterial hypertension; PAH-Scl = scleroderma-related pulmonary arterial hypertension.

  • Chi-square statistic.

  • World Health Organization (WHO) functional class could not be determined at the time of diagnosis for 5 patients.

Age, years47.6 ± 2.458.7 ± 1.6<0.001
Female, no. (%)35 (85.4)43 (86.0)0.93
Race, no. (%)   
 White32 (78.0)42 (84.0)0.67
 African American7 (17.1)7 (14.0)
 Hispanic2 (4.9)1 (2.0)
Body mass index, kg/m229.4 ± 1.427.2 ± 1.00.20
Limited scleroderma, no. (%)46 (92.0)
History of systemic hypertension, no. (%)14 (34.1)20 (40.0)0.57
Mean arterial pressure, mm Hg90.6 ± 2.285.9 ± 1.70.08
WHO functional classification, no. (%)   
 I and II9 (22.5)14 (28)0.41
 III and IV31 (77.5)32 (64)
6-minute walk distance, meters337.5 ± 31.9323.3 ± 25.80.75

Pulmonary function data were available for 84 (92%) of the study participants. The FEV1 was mildly decreased and significantly lower in patients with PAH-Scl compared with patients with IPAH (mean ± SEM 75.5 ± 2.7% versus 85.5 ± 2.3% of predicted; P = 0.007). The diffusing capacity for carbon monoxide (DLCO) was moderately to severely reduced in patients with PAH-Scl but only mildly reduced in patients with IPAH (48.7 ± 2.4% versus 68.4 ± 3.4% of predicted; P < 0.001). The TLC tended to be lower in the patients with PAH-Scl compared with the patients with IPAH; however, there was no significant difference between groups (86.5 ± 2.9% versus 93.4 ± 2.4% of predicted; P = 0.08).

Hemodynamics.

The results of the right heart catheterization revealed severe pulmonary hypertension in both groups of patients (Table 2). However, the mean pulmonary artery pressure and pulmonary vascular resistance index were significantly lower in patients with PAH-Scl. Filling pressures indicated right ventricular failure with elevated right atrial pressures, moderately decreased cardiac indexes, and normal pulmonary capillary wedge pressures in both groups, and there were no significant differences between the groups.

Table 2. Baseline right heart catheterization findings*
 IPAH (n = 41)PAH-Scl (n = 50)P
  • *

    Values are the mean ± SEM. See Table 1 for definitions.

Right atrial pressure, mm Hg10.1 ± 0.911.2 ± 0.70.36
Pulmonary artery systolic pressure, mm Hg86.4 ± 2.975.6 ± 2.40.004
Pulmonary artery pressure, mm Hg54.4 ± 1.946.6 ± 1.50.002
Pulmonary capillary wedge pressure, mm Hg12.0 ± 0.811.4 ± 0.70.59
Cardiac index, liters/minute/m22.1 ± 0.12.2 ± 0.10.19
Pulmonary vascular resistance index, Wood units22.8 ± 1.817.5 ± 1.50.026

Echocardiography results.

Results from echocardiograms at the time of diagnosis were available for 87 of the 91 patients (96%). The median time between echocardiography and right heart catheterization was 28 days. As shown in Table 3, both groups of patients had a similar prevalence of right-sided chamber enlargement. In contrast, a significantly higher proportion of patients with PAH-Scl had left chamber morphologic and functional abnormalities despite normal left ventricular systolic function. Left atrial dimensions were significantly larger in the PAH-Scl group, and more patients with PAH-Scl had left atrial dilation. The presence of diastolic dysfunction was associated with the diagnosis of PAH-Scl (odds ratio [OR] 3.4, P = 0.027). The OR did not significantly change after controlling for age and history of systemic hypertension (OR 2.9, P = 0.06). In addition, there was an increased presence of pericardial effusion in patients with PAH-Scl (34.7% versus 13.2%; P = 0.022).

Table 3. Baseline echocardiographic findings*
 IPAH (n = 38)PAH-Scl (n = 49)P
  • *

    Except where indicated otherwise, values are the number (%). See Table 1 for definitions.

Right atrial dilation31 (81.6)36 (73.5)0.37
Right ventricular dilation34 (89.5)39 (79.6)0.21
Right ventricular hypertrophy7 (18.4)5 (10.2)0.27
Left atrial diameter, mean ± SEM cm3.3 ± 0.23.8 ± 0.10.004
Left atrial dilation4 (10.5)14 (28.6)0.039
Left ventricular hypertrophy5 (13.2)17 (34.7)0.022
Left ventricular ejection fraction, mean ± SEM57.3 ± 1.655.7 ± 1.40.44
Diastolic dysfunction5 (13.2)16 (32.7)0.035
Pericardial effusion5 (13.2)17 (34.7)0.022

Treatment.

Differences in treatment at the last date of followup revealed that patients with IPAH were more likely to be receiving anticoagulation therapy with warfarin, while patients with PAH-Scl were more likely to be receiving immunosuppressants and calcium channel blockers, the latter being prescribed for Raynaud's phenomenon. The dosages of calcium channel blockers prescribed for Raynaud's phenomenon were substantially lower (i.e., in the order of 30 mg of nifedipine daily) than the dosages commonly administered for IPAH in vasodilator-responsive patients. There were no significant differences between the groups with regard to the use of diuretics, oxygen, endothelin receptor antagonists, sildenafil, or prostanoid therapy.

Survival analysis.

Five patients (5% of the cohort [4 with IPAH and 1 with PAH-Scl]) were lost to followup, and there were 29 deaths. One patient with IPAH underwent lung transplantation. Of the 29 deaths, the cause of death was determined from the clinical records of 23 of the patients, of whom 16 died of right heart failure, 4 died of respiratory failure, 2 died of sepsis, and 1 died following a motor vehicle accident. The median duration of followup was 2.2 years. The 1-year and 3-year estimates of survival in patients with IPAH were 95.1% (95% confidence interval [95% CI] 81.9–98.8%) and 83.6% (95% CI 69.8–91.4%), respectively (Figure 1). Estimates of survival in the patients with PAH-Scl were significantly worse (P = 0.002 by log rank test), with 1- and 3-year estimates of 87.8% (95% CI 69.6–95.4%) and 48.9% (95% CI 31.4–64.2%), respectively.

Figure 1.

Kaplan-Meier survival estimates from the time of diagnosis to the time of death in patients with idiopathic pulmonary arterial hypertension (IPAH) compared with patients with scleroderma-related pulmonary arterial hypertension (PAH-Scl). Significantly poorer survival is evident in patients with PAH-Scl. The 1- and 3-year survival was estimated to be 97.4% and 88.6%, respectively, in patients with IPAH and 82.5% and 48.2%, respectively, in patients with PAH-Scl.

Bivariate Cox proportional hazard models indicated that patients with PAH-Scl were >3 times more likely to die compared with patients with IPAH (Table 4). Other significant predictors of mortality included older age, the presence of a pericardial effusion, and a lower DLCO. A multivariable Cox model using backwards stepwise regression analysis, which tested for the effects of hemodynamic variables, WHO functional classification, age, and the presence of pericardial effusion, revealed that only the diagnosis of PAH-Scl (HR 3.06, 95% CI 1.29–7.25, P = 0.011) and the presence of pericardial effusion (HR 2.35, 95% CI 1.06–5.20, P = 0.035) (Figure 2) remained as significant independent predictors of death. Inclusion of variables to control for the presence of echocardiographic findings suggestive of left heart disease in this model showed that the presence of left heart disease was not independently predictive of mortality (P > 0.05), nor did it substantially change the increased risk of death in patients with PAH-Scl (HR 5.09, 95% CI 1.81–14.30; P = 0.002).

Table 4. Bivariate Cox proportional hazards model estimates of risk*
VariableHazard ratio95% confidence intervalP
  • *

    mPAP = mean pulmonary artery pressure; FEV1 = forced expiratory volume in 1 second; TLC = total lung capacity; DLCO = diffusing capacity for carbon monoxide (see Table 1 for other definitions).

PAH-Scl vs. IPAH3.451.48–8.060.004
Age, per 10 years1.371.03–1.820.028
Female sex0.960.37–2.540.94
WHO functional class III and IV vs. I and II1.150.43–3.090.78
Right atrial pressure, mm Hg1.060.99–1.130.09
mPAP, per 10 mm Hg0.870.62–1.240.45
Cardiac index, liters/minute/m20.730.39–1.370.33
FEV1, per 10% predicted1.040.82–1.320.75
TLC, per 10% predicted1.311.03–1.660.027
DLCO, per 10% predicted0.800.66–0.970.027
Left atrial dilatation present0.740.26–2.140.58
Left ventricular hypertrophy present1.110.47–2.600.81
Diastolic dysfunction present1.440.64–3.260.38
Pericardial effusion present2.831.34–5.980.006
Figure 2.

Kaplan-Meier survival estimates from the time of diagnosis to the time of death by presence of pericardial effusion in patients with idiopathic pulmonary arterial hypertension (IPAH) compared with patients with scleroderma-related pulmonary arterial hypertension (PAH-Scl). Survival estimates in patients with PAH-Scl indicate higher mortality in those with a pericardial effusion noted on their baseline echocardiogram.

DISCUSSION

The present study represents one of the largest studies to compare baseline clinical data in patients with PAH-Scl and patients with IPAH and to examine the impact these factors have on long-term survival. The major finding of this study was that patients with PAH-Scl were almost 4 times more likely to die than were patients with IPAH, after controlling for the presence of pericardial effusion and presence of left heart disease. The 1- and 3-year survival estimates were 87.8% and 48.9%, respectively, in patients with PAH-Scl, and 95.1% and 83.6%, respectively, in patients with IPAH.

Although PAH is recognized as a major cause of death in patients with scleroderma, there are only a few descriptions of hemodynamics, left and right heart function, and predictors of survival in the literature (21, 22). Previous studies were limited by the small numbers of patients, lack of hemodynamic confirmation of disease or nonconsensus definitions of pulmonary hypertension, and absence of detailed phenotype data, including echocardiographic findings.

Several recent trials involving patients with PAH demonstrated a therapeutic benefit (i.e., improved survival and/or functional status) with the use of drugs such as prostaglandin derivatives and endothelin receptor antagonists (16, 23). However, it is noteworthy that in studies examining the outcomes in both subtypes of PAH, patients with PAH-Scl had a worse prognosis compared with patients with IPAH (16, 24). The reason for this difference between the 2 groups remains unclear but may be related to the severity of disease at diagnosis, direct microvascular involvement of the right ventricle, and coexisting left heart disease. In our study, we found differences in the use of glucocorticoids and coumadin, but stratified bivariate Cox proportional hazards analysis did not show any impact of their use on survival in patients with PAH-Scl.

The baseline hemodynamic data from this study indicate that, as compared with patients with IPAH, those with PAH-Scl had similarly depressed cardiac function despite having significantly lower mean pulmonary artery pressures. This finding, in conjunction with a similar prevalence of right ventricular dilatation in the 2 groups, suggests that in patients with PAH-Scl compared with those with IPAH, the right ventricle may have a reduced ability to adapt to the increased pulmonary vascular resistance. This maladaptation, although unexplained at this time, may be, in part, the result of inflammation and scarring, as supported by evidence on endomyocardial biopsy samples from patients with scleroderma (25), a finding not reported in IPAH or subclinical atherosclerosis. In advanced PAH, the mean pulmonary artery pressure may decrease as a function of a decrease in cardiac output (26). Therefore, it could be argued that the lower mean pulmonary artery pressure in patients with PAH-Scl is related to more advanced right ventricular dysfunction as compared with that in patients with IPAH. However, if this were the case, one would expect a more depressed cardiac output in patients with PAH-Scl compared with patients with IPAH, a finding that was not observed in this study. In addition, there was no difference in WHO functional class distribution between the 2 groups, which suggests that there were no significant differences in disease severity at presentation.

Another significant result of the present study is the increased presence of echocardiographic markers of left heart disease in the patients with PAH-Scl compared with the patients with IPAH. This difference was still present after controlling for the older age of the PAH-Scl patients and the tendency for PAH-Scl patients to develop systemic hypertension. This may potentially be related to an increased prevalence of diastolic dysfunction in the PAH-Scl group, given that there was no difference in the left ventricular ejection fraction. However, since patients with a capillary wedge pressure higher than 15 mm Hg were excluded from the study (in order to adhere to the definition of PAH), it is possible that associated left heart disease is indeed more prevalent in this patient population than was suggested from the current analysis. There was no significant correlation between left atrial size and pulmonary capillary wedge pressure (R = 0.177). However, most patients underwent aggressive diuresis before catheterization, which may explain why only 2 of the patients with PAH-Scl who were originally screened for the study were excluded for having an elevated capillary wedge pressure after catheterization. Alternatively, we cannot exclude the possibility that, instead of a true increased prevalence of stiff, nonrelaxing left ventricles in these patients, these findings may actually represent surrogate markers for right ventricular maladaption, in which pressure and volume overload leads to compression of the left ventricle and limited diastolic filling (27).

Although echocardiography is useful for diagnosing diastolic dysfunction and clinical nonsystolic heart failure events in the general population (28, 29), its utility and significance in the PAH population is unknown and requires further study. Although the mechanism has yet to be elucidated, it appears from these findings that patients with PAH-Scl have more generalized cardiac dysfunction, as compared with patients with IPAH, which is likely related to the systemic nature of the disease.

The long-term survival of the patients with PAH-Scl at our center is poor but better than previously reported for similar patients (15). This improved survival may represent a combination of recent advances in therapy, an integrated multidisciplinary approach to patient care, and early referral to the pulmonary hypertension clinic at our center. This cohort includes patients diagnosed after January 1, 2000, a time shortly preceding the widespread introduction of oral endothelin receptor antagonists. Although patients with PAH-Scl generally showed a poorer response to treatment compared with patients with IPAH in a randomized controlled trial (16), therapy with endothelin receptor antagonists did represent a viable alternative for patients who were not otherwise appropriate candidates for intravenous prostacyclin.

Our study included a significant number of patients whose condition was in WHO functional classes I and II, suggesting that these patients were diagnosed earlier in the course of their disease compared with patients in other reported studies. Patients who were in WHO functional class I and class II may have been followed up longer prior to decompensation and death, raising the possibility of lead-time bias as a contributor to higher survival estimates. However, mean cardiac indexes in the present study were very similar to the values reported by Kawut et al (15), suggesting similar levels of cardiac depression and stage of disease in the 2 studies. Despite similarities in hemodynamics, our survival rates for the PAH-Scl population were remarkably better than those reported by Kawut et al (<60% at 1 year) and are closer to the survival rates reported by Mukerjee et al, who included a population with a mean cardiac index in the normal range (2.6 liters/minute/m2) (12). The improved survival seen in our study is also consistent with that recently reported for another cohort of PAH-Scl patients (30).

The results of bivariate hazards modeling are consistent with prior studies but raise new questions about risk factors for higher mortality. Our finding that PAH-Scl patients have 4 times the risk of death compared with IPAH patients is consistent with previously reported risk estimates (15). Our data also demonstrate, using bivariate analysis, that a decreasing DLCO is associated with higher mortality. The risk estimate (HR value) did not change significantly after controlling for the underlying diagnosis, although it was no longer statistically significant (not shown). A severely altered DLCO may be an indicator of undiagnosed interstitial lung disease and/or significant pulmonary vascular disease. Further studies are needed to determine whether a fall in the DLCO is truly a prognostic factor. Contrary to our hypothesis, we did not find that echocardiographic markers of left heart disease were predictive of higher mortality in bivariate or multivariate analyses. However, the lack of a significant impact of left heart disease on mortality may be related to the limited number of deaths in our study.

The presence of a pericardial effusion in patients with PAH is thought to be due to increased right atrial pressure (31–33). In a previous study by Hinderliter et al, the presence of a pericardial effusion was shown to be predictive of higher mortality in patients with IPAH (31). Our analysis showed a significant difference in survival of patients with PAH-Scl who had a small or larger pericardial effusion (P = 0.043 by log rank test) on their baseline echocardiogram (Figure 2). Pericardial effusion, which was almost 3 times more common in patients with PAH-Scl, could potentially represent more severe right ventricular overload. Alternatively, pericardial effusion is a common echocardiographic finding among scleroderma patients, and is reported to be predictive of a poor outcome and prognosis (independent of PAH), and may also be a manifestation of serositis and systemic inflammatory disease (34, 35).

There are several limitations to this study. First, the possibility of a selection bias cannot be excluded, since this is a retrospective study. Second, although the initial evaluation for pulmonary hypertension was relatively standardized, the relative timing of the studies (i.e., catheterization, 6-minute walk test, echocardiography, and pulmonary function tests) was not homogeneous among the patients, but rather represented the usual sequence for diagnostic testing in a referral center, with screening echocardiography leading to referral to the pulmonary hypertension clinic and performance of right heart catheterization, often spread over several months. However, limiting the analysis to echocardiograms obtained within 120 days of the right heart catheterization did not significantly affect the conclusions of this study.

This study was also limited by the lack of centralized standardization of the tests and, in the case of the echocardiograms, central interpretation. Although most of the tests were performed at our center, a number of studies were performed at other institutions, where the techniques and interpretation of results may differ compared with the protocols at our center. These differences and lack of standardization apply both to the performance of the right heart catheterization and to echocardiography. However, since these methods were applied both to the patients with IPAH and to the patients with PAH-Scl, any misclassification would have biased the study results toward not finding a difference.

In conclusion, we have shown that patients with PAH-Scl have a worse prognosis and poorer survival compared with patients with IPAH. In addition, the presence of pericardial effusion independently contributes to mortality. Interestingly, when compared with the patients with IPAH, the patients with PAH-Scl had similar degrees of functional capacity and cardiac dysfunction despite having lower mean pulmonary artery pressures at the time of presentation. However, despite the increased prevalence of echocardiographic evidence of left heart disease in the PAH-Scl patients, this does not appear to be predictive of higher mortality. Further studies with more sensitive measures of right and left heart disease (e.g., magnetic resonance imaging) will be needed to confirm this finding. We speculate that the increased mortality among patients with PAH-Scl is secondary to widespread microvascular disease and fibrosis, in conjunction with myocardial dysfunction and failure of the right ventricle to adapt to an increased pulmonary vascular load.

  • 1

    Dr. Champion has received honoraria (less than $10,000 each) from Pfizer and Encysive. Dr. Girgis has received consulting or advisory fees (less than $10,000 each) from Encysive and Cotherix and has received speaking fees (less than $10,000) from Actelion. Dr. Wigley has received honoraria (less than $10,000 each) from Biogen-Idec, Actelion, Millennium, and Ashai-Kasei and has received honoraria (more than $10,000) from Mediquest.

Ancillary