Dr. Girgis has received consulting fees, speaking fees, and/or honoraria from Gilead Sciences, United Therapeutics, and Actelion (less than $10,000 each).
Survival in pulmonary hypertension associated with the scleroderma spectrum of diseases: Impact of interstitial lung disease
Article first published online: 29 JAN 2009
Copyright © 2009 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 60, Issue 2, pages 569–577, February 2009
How to Cite
Mathai, S. C., Hummers, L. K., Champion, H. C., Wigley, F. M., Zaiman, A., Hassoun, P. M. and Girgis, R. E. (2009), Survival in pulmonary hypertension associated with the scleroderma spectrum of diseases: Impact of interstitial lung disease. Arthritis & Rheumatism, 60: 569–577. doi: 10.1002/art.24267
- Issue published online: 29 JAN 2009
- Article first published online: 29 JAN 2009
- Manuscript Accepted: 23 OCT 2008
- Manuscript Received: 17 MAY 2008
- NIH. Grant Numbers: HL-084946, HL-70052
Pulmonary hypertension (PH) is an important cause of mortality in systemic sclerosis (SSc), where it can be isolated (pulmonary arterial hypertension [PAH]) or associated with interstitial lung disease (ILD). This study was undertaken to characterize determinants of survival among SSc patients with either type of PH who received PAH-specific therapy.
Consecutive SSc patients with PAH or ILD-associated PH confirmed by right heart catheterization were included in the study. Kaplan-Meier and Cox proportional hazards models were used to compare survival between SSc patients with PAH and those with ILD-associated PH and to identify predictors of survival.
Fifty-nine patients (39 with PAH and 20 with ILD-associated PH) were identified. The majority (15 of 20 with ILD-associated PH and 27 of 39 with PAH) received an endothelin receptor antagonist as initial therapy. Median followup time was 4.4 years (range 2.7–7.4 years). Survival was significantly worse in SSc patients with ILD-associated PH than in those with PAH (1-, 2-, and 3-year survival rates 82%, 46%, and 39% versus 87%, 79%, and 64%, respectively; P < 0.01 by log rank test). In a multivariable analysis, ILD-associated PH was associated with a 5-fold increase in risk of death compared with PAH. Pulmonary vascular resistance index was also an independent predictor of mortality in the overall cohort (hazard ratio 1.05, P < 0.01) and was a significant univariable risk factor in each group separately. Type of initial PAH therapy and the use of warfarin were not related to survival.
Survival in SSc complicated by PH remains poor despite currently available treatment options. While therapy may be associated with improved survival in PAH compared with historical controls, the prognosis for patients with ILD-associated PH is particularly grim. Early diagnosis and treatment may improve outcomes since worsening hemodynamic factors were associated with reduced survival.
Scleroderma (systemic sclerosis [SSc]) is an autoimmune disease characterized by multiorgan involvement resulting in significant morbidity and mortality (1). Prior to the advent of angiotensin-converting enzyme inhibitors, renal disease, specifically scleroderma renal crisis, was the most common cause of death in these patients (2). However, pulmonary complications are now the leading causes of death (3, 4). There are 2 major types of lung disease in scleroderma, with different clinical associations and pathogenesis (5). Interstitial lung disease (ILD) is more common in patients with diffuse cutaneous SSc (dcSSc) and is often complicated by the development of pulmonary hypertension (PH) (6). Isolated pulmonary arterial hypertension (PAH), histologically similar to idiopathic PAH (7), is more common in patients with limited cutaneous SSc (lcSSc). The reported prevalence of PAH in scleroderma varies by method of assessment (i.e., echocardiography or right heart catheterization), but appears to be ∼8% by cardiac catheterization (8).
Both types of lung involvement portend a poor prognosis in scleroderma. Despite therapy targeting the pulmonary vasculature, median survival time of SSc patients with PAH ranges from ∼1 to 3 years (9, 10). While scleroderma patients with ILD alone have a median survival of 5–8 years (11), the impact of combined PH and ILD is not well characterized. The results of a recent study by Trad et al (12), in which PH was assessed by right heart catheterization, suggested increased mortality among patients with ILD-associated PH compared with patients with ILD alone. Similarly, there are few data regarding outcomes in SSc patients with PAH compared with those with ILD-associated PH. Importantly, although currently available pharmacologic agents for PAH may be associated with improved survival (13, 14), the impact of PAH therapies on survival in patients with PH complicating SSc-related ILD is unknown. We reviewed our experience with PAH as well as ILD-associated PH in SSc to assess baseline determinants of survival among patients treated with PAH-specific agents.
PATIENTS AND METHODS
The cohort was identified from the Johns Hopkins Pulmonary Hypertension Program, which maintains a registry of all patients evaluated at our center. The Institutional Review Board approved the registry and this specific analysis. The registry was queried for patients who were diagnosed as having PH associated with scleroderma. Consecutive outpatients who were diagnosed as having SSc and either PAH or ILD-associated PH as confirmed by right heart catheterization and who were initially evaluated between September 2000 and September 2005 were included. The data set was locked for analysis in September 2007, so that the minimum followup since diagnosis of PH was 2 years. Patients in whom PAH therapy had been initiated prior to diagnostic right heart catheterization or who were referred to our center after starting PAH therapy were excluded.
Limited cutaneous and diffuse cutaneous SSc were defined as previously described (15). Mixed connective tissue disease was diagnosed in patients who had clinical features of limited scleroderma (3 of the following 5 findings: edema of the hands, synovitis, myositis, Raynaud's phenomenon, and acrosclerosis) with high titers of anti–U1 RNP antibodies in the absence of anti-Sm antibodies (16). Undifferentiated connective tissue disease was defined as previously described (17). PAH was defined as mean pulmonary artery pressure >25 mm Hg, pulmonary capillary wedge pressure ≤15 mm Hg, and pulmonary vascular resistance (PVR) index >3 Wood units (18) in the absence of other known causes of PH (19). Patients with a pulmonary capillary wedge pressure between 16 and 18 mm Hg were also included if there was no clinical or echocardiographic evidence of left heart disease, since elevations in pulmonary capillary wedge pressure may be related to impaired left ventricular filling due to right ventricular distention (20).
Results of pulmonary function tests (PFTs) and high-resolution computed tomography (CT) of the chest obtained closest to the date of the diagnostic right heart catheterization were recorded. Percent of predicted results for all PFT data were calculated using the method of Crapo et al (21). Patients with significant obstructive lung disease, defined as a forced expiratory volume in 1 second/forced vital capacity (FEV1/FVC) <0.5, or between 0.5 and 0.7 if accompanied by radiographic evidence of emphysema, were excluded (22, 23). ILD was defined by a combination of PFT and high-resolution CT criteria, as previously described in epidemiologic studies and clinical trials of PAH therapy (8, 24, 25). Patients were classified as having ILD-associated PH if they met all of the above criteria and had a total lung capacity (TLC) of <60% predicted or a TLC between 60% and 70% predicted combined with moderate to severe fibrosis (grade 3–4) evident on high-resolution CT (26). The onset of scleroderma was defined as the first non–Raynaud's phenomenon manifestation. Comorbid conditions, including gastrointestinal disease, renal disease, and myositis, were assessed using a previously described severity score (27).
The following variables were assessed as risk factors for time to death or lung transplantation: demographic characteristics, pulmonary hemodynamic features, pulmonary function, diagnosis (ILD-associated PH versus PAH), SSc disease type (lcSSc versus dcSSc), duration of scleroderma before PH diagnosis, comorbidities, World Health Organization (WHO) functional class (18), supplemental oxygen use, warfarin use, and initial PAH therapy. Survival was ascertained by reviewing medical records, phone contact, and through the Social Security Death Index.
Continuous variables were summarized as the mean ± SD or the median and range and compared using Student's t-test or Wilcoxon's rank sum test where appropriate. Categorical variables were compared using the chi-square statistic. P values less than 0.05 (2-tailed) were considered significant. Factors associated with transplant-free survival were ascertained by univariable Cox proportional hazards models. Variables found to be significant in univariable analysis (P ≤ 0.10) were incorporated into a multivariable model. The proportional hazards assumption was examined for all covariates using a continuous time-varying predictor and generalized linear regression of scaled Schoenfeld residuals on functions of time (28, 29). Time-to-event analyses were performed using the Kaplan-Meier product-limit estimator to compare PAH with ILD-associated PH.
The initial search of the registry yielded 91 patients. Thirty-two were excluded for the following reasons: 13 did not have a confirmatory right heart catheterization, 10 were referred after institution of therapy or had therapy initiated prior to right heart catheterization, 5 had evidence of significant left heart disease, 2 had significant obstructive lung disease, and 2 did not have a confirmed diagnosis of the scleroderma spectrum of diseases. A total of 59 patients meeting our study criteria were identified (39 with PAH and 20 with ILD-associated PH).
Table 1 summarizes the clinical and hemodynamic features of the entire cohort and of each group. Mean pulmonary artery pressure was severely elevated in the group of patients with PAH, averaging 49 mm Hg, which was significantly higher than in the group of patients with ILD-associated PH (mean 40 mm Hg). Cardiac index was comparably reduced in both groups, but right atrial pressure was higher in the PAH group, suggesting more advanced right ventricular dysfunction. All but 5 subjects had pulmonary capillary wedge pressure ≤15 mm Hg. The individuals with pulmonary capillary wedge pressure of 16–18 mm Hg were all in the PAH group, and had normal left ventricular ejection fraction, left atrial size, and no evidence of left-sided valvular heart disease, coronary atherosclerosis, or pulmonary congestion. One patient had an FEV1/FVC ratio <0.7, but no evidence of emphysema or ILD on high-resolution CT and thus was included in the PAH group. TLC was lower in the ILD group, as was the lung diffusing capacity for carbon monoxide (DLCO), which was severely decreased to 31% predicted, compared with 51% predicted in the PAH subset. A higher proportion of patients in the ILD group was African American, had diffuse skin involvement, and was positive for Scl-70 antibody. In contrast, all but 1 patient in the PAH group had lcSSc, and a higher proportion was positive for anticentromere antibodies. Comorbid conditions, including gastrointestinal disease, renal disease, and myositis, did not differ between groups.
|All patients (n = 59)||Patients with ILD-associated PH (n = 20)||Patients with PAH (n = 39)||P, ILD-associated PH vs. PAH|
|Age, years||60 ± 12||59 ± 9||60 ± 14||0.73|
|Sex, no. (%) female||52 (88)||18 (90)||34 (87)||0.75|
|No. (%) African American||12 (20)||7 (35)||5 (13)||0.04|
|WHO class at baseline, no. of patients||0.34|
|Disease duration before diagnosis of PH, median (range) years||4 (1–36)||3 (1–20)||4.5 (1–36)||0.73|
|Definite SSc, no. (%)||55 (93)||19 (95)||36 (92)||0.53|
|MCTD, no. (%)||2 (3)||0 (0)||2 (5)||–|
|UCTD, no. (%)||2 (3)||1 (5)||1 (3)||–|
|No. (%) with lcSSc||46 (78)||7 (35)||38 (97)||<0.01|
|ANA positive, no. (%)†||49 (88)||16 (85)||33 (89)||0.59|
|Scl-70 positive, no. (%)†||7 (17)||7 (41)||0 (0)||<0.01|
|ACA positive, no. (%)†||16 (35)||1 (6)||15 (52)||<0.01|
|Gastrointestinal tract severity score, no. (%)||0.18|
|0||9 (15)||4 (20)||5 (13)|
|1||39 (66)||11 (55)||28 (72)|
|2||8 (14)||5 (25)||3 (8)|
|3||2 (3)||0 (0)||2 (5)|
|4||0 (0)||0 (0)||0 (0)|
|Renal severity score, no. (%)||0.62|
|0||49 (83)||18 (90)||31 (79)|
|1||6 (10)||1 (5)||5 (13)|
|2||3 (5)||1 (5)||2 (5)|
|3||0 (0)||0 (0)||0 (0)|
|4||0 (0)||0 (0)||0 (0)|
|Myositis, no. (%)||5 (8)||1 (5)||4 (10)||0.48|
|FEV1, % predicted||71 ± 19||57 ± 14||78 ± 18||<0.01|
|FVC, % predicted||71 ± 20||54 ± 13||79 ± 18||<0.01|
|FEV1/FVC, % predicted||80 ± 7||85 ± 6||78 ± 5||<0.01|
|TLC, % predicted||74 ± 26||52 ± 17||84 ± 22||<0.01|
|DLCO, % predicted||44 ± 17||31 ± 12||51 ± 15||<0.01|
|Oxygen use at baseline, no. (%)||22 (37)||12 (60)||10 (26)||0.01|
|Right atrial pressure, mm Hg||10 ± 6||8 ± 5||11 ± 6||0.04|
|Mean pulmonary artery pressure, mm Hg||46 ± 11||40 ± 8||49 ± 11||<0.01|
|Cardiac index, liters/minute/m2||2.2 ± 0.7||2.3 ± 0.7||2.2 ± 0.6||0.51|
|Pulmonary capillary wedge pressure, mm Hg||11 ± 4||10 ± 3||11 ± 4||0.21|
|PVR index, Wood units/m2||18 ± 10||14 ± 7||19 ± 11||0.08|
|Pulmonary artery oxygen saturation, %||62 ± 9||64 ± 8||61 ± 9||0.24|
|No. (%) receiving warfarin||35 (59)||11 (55)||24 (62)||0.63|
|Initial PAH-specific therapy, no. (%)||0.55|
|Endothelin receptor antagonist||42 (71)||15 (75)||27 (69)|
|Phosphodiesterase V inhibitor||6 (10)||3 (15)||3 (8)|
|Prostanoid||10 (17)||2 (10)||8 (21)|
|Calcium-channel blocker||1 (2)||0 (0)||1 (3)|
All patients received pharmacologic therapy for PH (Table 1). The initial agent was an endothelin receptor antagonist, either bosentan or sitaxsentan, in 71%. The latter was available as part of investigational trials for PAH patients only (n = 7). Once sildenafil became available, this was the preferred initial drug for ILD, given the favorable acute effects previously reported in this population (30). The 1 patient who was initially treated with high-dose calcium-channel blockers was switched to bosentan after 6 months due to lack of hemodynamic response. In 8 of the 10 patients in whom epoprostenol treatment was initiated, treatment had been started prior to the approval of bosentan. Subsequently, epoprostenol was generally avoided in patients with severe ILD, due to the potential for worsening of ventilation/perfusion mismatch (30). Thirty-two patients (54%) ultimately received a second or third agent for PH, either instead of or in combination with the initial drug. Twenty patients received a continuous intravenous or subcutaneous prostanoid compound at some time during the course of treatment. Thirty-five patients (59%) received adjunctive anticoagulation therapy with warfarin.
None of the patients in the PAH group received immunosuppressants during the followup period. Fourteen of the patients in the ILD-associated PH group received immunosuppressants during the course of followup (8 received prednisone alone, 1 received cyclophosphamide alone, 2 received a combination of prednisone and cyclophosphamide, and 3 received a combination of mycophenolate mofetil and cyclophosphamide).
During a median observation period following the initial right heart catheterization of 4.4 years (range 2.7–7.4 years), 30 patients died and 3 underwent lung transplant. In the entire cohort, the actuarial 1-, 2-, and 3-year transplant-free survival rate was 86%, 68%, and 56%, respectively. Table 2 lists the effects of selected variables on survival in the univariable analysis. A diagnosis of ILD-associated PH was associated with a significantly increased risk of death (hazard ratio 2.87). Survival in this group at 1, 2, and 3 years was 82%, 46%, and 39%, compared with 87%, 79%, and 64%, respectively, in the PAH group (Figure 1). A lower DLCO was also associated with increased mortality. There was a strong trend toward reduced survival with increasing PVR index (P = 0.06). A weaker trend was observed for dcSSc. In the multivariable model, diagnosis remained a strong risk factor for mortality (Table 3). PVR index emerged as a significant predictor of survival, while DLCO and type of skin involvement were not significantly associated with survival.
|HR (95% CI)||P|
|Diagnosis of ILD vs. PAH||2.87 (1.39–5.92)||<0.01|
|Duration of SSc prior to diagnosis of PH||0.99 (0.95–1.04)||0.73|
|dcSSc vs. lcSSc||1.88 (0.88–4.01)||0.10|
|Gastrointestinal tract severity score||0.83 (0.48–1.44)||0.51|
|Renal severity score||1.23 (0.66–2.27)||0.51|
|African American vs. white or Asian race||1.41 (0.60–3.28)||0.42|
|Male vs. female sex||1.33 (0.51–3.48)||0.55|
|WHO class||1.30 (0.79–2.13)||0.29|
|Oxygen use||1.59 (0.79–3.20)||0.19|
|Warfarin use||0.98 (0.49–1.95)||0.95|
|Initial PAH therapy type||1.03 (0.71–1.50)||0.86|
|FVC, % predicted||1.00 (0.98–1.01)||0.54|
|FEV1, % predicted||1.00 (0.98–1.02)||0.89|
|TLC, % predicted||0.99 (0.98–1.01)||0.33|
|DLCO, % predicted||0.97 (0.95–0.99)||0.04|
|Right atrial pressure, mm Hg||0.98 (0.92–1.05)||0.64|
|Mean pulmonary artery pressure, mm Hg||1.00 (0.97–1.03)||0.97|
|Cardiac index, liters/minute/m2||0.66 (0.36–1.20)||0.17|
|Pulmonary capillary wedge pressure, mm Hg||0.95 (0.87–1.04)||0.27|
|PVR index, Wood units/m2||1.03 (0.99–1.06)||0.06|
|Pulmonary artery oxygen saturation, %||0.98 (0.95–1.02)||0.31|
|HR (95% CI)||P|
|Diagnosis of ILD vs. PAH||5.15 (1.73–15.3)||<0.01|
|PVR index, Wood units/m2||1.05 (1.01–1.09)||<0.01|
|DLCO, % predicted||0.99 (0.96–1.02)||0.57|
|lcSSc vs. dcSSc||0.64 (0.24–1.69)||0.36|
We then explored risk factors for mortality among the 2 subsets separately (Table 4). In both groups, a higher PVR index was significantly associated with reduced survival. Among PAH patients, reduced cardiac index was also a marker of increased mortality, whereas in the ILD-associated PH group, a trend was observed for increased mean pulmonary artery pressure. Decreasing DLCO and African American race were associated with a worse prognosis in the ILD group only.
|HR (95% CI)||P|
|African American race||2.76 (0.92–8.3)||0.07|
|DLCO, % predicted||0.93 (0.87–0.98)||0.02|
|Mean pulmonary artery pressure, mm Hg||1.07 (0.99–1.15)||0.09|
|PVR index, Wood units/m2||1.06 (1.00–1.19)||0.04|
|Cardiac index, liters/minute/m2||0.41 (0.19–0.97)||0.04|
|PVR index, Wood units/m2||1.05 (1.02–1.09)||0.02|
Causes of death.
Most of the deaths (11 [61%] of 18) in the PAH subset were attributed to right heart failure, and 3 (16%) were classified as respiratory failure, characterized by refractory hypoxemia. Two deaths in the PAH group were caused by hemorrhage (1 by lower gastrointestinal bleeding and 1 by intracranial hemorrhage). The latter patient had been taking warfarin. The remaining 2 deaths among PAH patients were due to bronchogenic carcinoma and automobile accident. In contrast, the majority of deaths (9 [75%] of 12) in the ILD group were considered to be primarily the result of respiratory failure. Two of these patients were receiving high-dose steroids as rescue therapy. The remaining 7 patients who died of respiratory failure were not taking immunosuppressants at the time of death. The remaining deaths were due to sudden death in 2 patients and to right heart failure in 1 patient. All 3 patients who received lung transplants were in the ILD-associated PH group. Excluding these individuals did not appreciably alter the results.
Despite the availability of therapeutic agents from 3 pharmacologic classes, PH continues to exact a heavy toll on the long-term survival of patients with SSc. The 3-year survival rate of 64% observed in the PAH group in this study is comparable with the ∼48% observed in scleroderma patients with PAH who were enrolled in a clinical trial of continuous intravenous epoprostenol (31), and the 2-year survival rate of 79% in the present cohort is similar to the 73% reported among 66 PAH patients with connective tissue disease (52 [79%] with SSc and 8 [12%] with systemic lupus erythematosus) who were enrolled in clinical trials of bosentan (32).
Relatively little information is available regarding determinants of survival in PAH associated with SSc (33). Among the baseline variables tested in this study, only cardiac index and PVR index were significantly associated with survival. In a larger cohort of 148 patients, including 40 with advanced pulmonary fibrosis, from the Royal Free Hospital (London, UK), worsening pulmonary hemodynamic features were also associated with increased mortality, with right atrial pressure being the strongest predictor (34). Reduced functional capacity, as assessed by 6-minute walk distance and WHO class, has also been linked to decreased survival in PH associated with SSc (14, 35), although we were unable to demonstrate an association with functional class in this small series. Other markers of right ventricular dysfunction, such as brain natriuretic peptide (35) and pericardial effusion (10), may also portend a worse prognosis, as has been demonstrated for idiopathic PAH (33).
A novel finding of this study is the markedly reduced survival of SSc patients with PH complicating advanced ILD compared with those with PAH. The presence of ILD remained a significant predictor of mortality after adjusting for the severity of PH (PVR index). This subset of patients is routinely excluded from clinical trials of PAH therapies (25, 32, 36, 37) as well as trials investigating therapies for ILD (38). Yet, this appears to be an important subset of patients. Among a cohort of 619 SSc patients, Chang and coworkers found evidence of concomitant ILD (TLC <80% predicted) and PH (estimated right ventricular systolic pressure >35 mm Hg, determined by Doppler echocardiography) in 18%, while isolated PH was present in 19% (6). Importantly, mortality with combined disease was comparable with that observed for PH alone and considerably higher than for uncomplicated ILD.
The difference in survival rates between the study by Chang et al (6) and the present study is likely due to the case definitions of PH and ILD. In the study by Chang et al (6), PH was based on echocardiographic estimates of pulmonary pressure (right ventricular systolic pressure of 35 mm Hg), and ILD was defined as TLC <80% predicted. Our cohort included patients who underwent right heart catheterization and fulfilled the current consensus definition of PH. Further, we used the criteria of PAH treatment trials that included SSc patients with PAH, which allowed patients with TLC as low as 60% predicted to enroll if only minimal fibrosis was noted on computed tomography of the chest (37, 39, 40). The classification scheme used by Chang et al (6) would bias the results to favor no difference in survival between the 2 groups due to misclassification, which is what the authors found.
Investigators at the Royal Free Hospital observed a 2-year survival rate of ∼48% among 40 patients with combined pulmonary fibrosis and hypertension (34), which was nearly identical to the 46% rate found for these patients in our series. The 2-year survival rate among 108 patients with isolated PAH in that study was ∼68%. However, the survival curves merged at 40 months, and there was no significant difference between the 2 groups (P = 0.3). One-third of the subjects in the entire cohort had a mean pulmonary artery pressure of <32 mm Hg, whereas only 3 of 20 patients with ILD-associated PH and 2 of 39 patients with PAH in our series had a mean pulmonary artery pressure in that range. As in the PAH subset, we observed a higher mortality with increasing PVR index in the ILD-associated PH subset.
There was also a highly significant association between decreasing DLCO and reduced survival in patients with ILD-associated PH. A reduction in single-breath DLCO can result from a loss in lung volume and/or maldistribution of inspired gas (41), in addition to thickening of the alveolar–capillary membrane and reduction in pulmonary capillary blood volume, both of which may be altered as a result of pulmonary vascular disease (42). Diffusing capacity is reduced in SSc patients with ILD-associated PH relative to patients with ILD without PH, despite similar lung volumes (12). Agusti and coworkers (43) have shown that in patients with ILD, reduced diffusing capacity correlates with the rise in PVR and hypoxemia that occur with exercise, which was attributed to the inability to divert blood away from poorly ventilated units as a result of vascular remodeling. This is consistent with the greater need for supplemental oxygen noted in the patients with ILD-associated PH compared with those with PAH and the dramatically greater proportion of deaths due to hypoxemic respiratory failure.
Previous epidemiologic studies have shown that African American patients with scleroderma are more likely to have diffuse disease and poorer survival than white patients (44, 45). Consistent with the findings of Laing and colleagues (44), African American patients in our cohort were more likely to have diffuse disease, were more likely to have ILD-associated PH, and tended to have poorer survival compared with white patients. Further, within the ILD-associated PH group, the risk of death was significantly greater among African Americans, with a nearly 3-fold increased risk of death that persisted when adjusted for age. These findings confirm the results of previous studies showing racial differences in clinical expression of disease and should be a target of future studies of SSc.
Great strides have been made in the development of effective medical therapy for PAH. However, short-term (12-week) clinical trials have generally demonstrated reduced efficacy in patients with PAH related to connective tissue disease (primarily SSc) compared with patients with idiopathic PAH (25, 36, 37). A recent systematic review of randomized, controlled trials of PAH therapy showed no significant treatment effect as assessed by change in 6-minute walk distance for SSc patients with PAH compared with other patients with PAH (46). Similarly, long-term survival has consistently been poorer (9, 10, 31, 47, 48). However, it has been suggested that treatment with the endothelin receptor antagonist bosentan may reduce mortality compared with historical controls. Williams and colleagues (14) reported a 2-year survival rate of 71% in selected SSc patients with PAH treated initially with bosentan, compared with 47% in historical controls treated with various prostanoid agents. The latter figure is comparable with the 50% 2-year survival rate in the absence of specific PAH treatment reported by Steen and Medsger (49). The type of initial PAH therapy was not associated with survival in our patients; however, our sample size was small, and 71% of the patients received an endothelin receptor antagonist as the initial agent.
There is little information on the efficacy of PAH therapies in SSc patients with PH associated with advanced ILD, consisting only of anecdotal reports (50). Acutely, intravenous epoprostenol reduces pulmonary and systemic vascular resistance to an equal extent and worsens ventilation/perfusion relationships, whereas inhaled iloprost and sildenafil appear to have more favorable immediate effects (30, 50). Bosentan was not beneficial in SSc patients with ILD who did not have PH (51). Our experience suggests that despite currently available PAH therapies, 2-year survival in this group is dismal. Such patients, in the absence of other significant comorbidity, should be considered early for lung transplantation (52). The current donor lung allocation scheme in the US places a high priority on elevated pulmonary artery pressure in the setting of ILD (53).
Anticoagulation with warfarin is recommended for patients with idiopathic PAH, but there are no data supporting its use in other types of PAH, including SSc-related PAH; hence its designation as a weak recommendation based only on expert opinion (54). We could not detect a beneficial impact of warfarin use in this study. The use of warfarin was not randomized, and patients with more advanced disease may have been more likely to receive anticoagulation therapy. However, there was no difference in PVR index between PAH patients treated with warfarin and those who were not (data not shown).
There are several limitations of this study that may influence the generalizability of the results. First, all subjects were referred to our PH program, and thus selection bias toward inclusion of those patients who were most ill is likely. However, since both groups were as likely to be subject to referral bias, the relationship observed between the 2 groups is probably valid. Further, although we attempted to control for duration of disease using self-reported onset of symptoms consistent with scleroderma (first non–Raynaud's phenomenon symptom), it is possible that true duration of disease prior to the development of PH varied and thus may have skewed the survival analysis. However, the WHO recommends yearly screening for PH in patients with scleroderma; if applied in clinical practice (as is routine in our Scleroderma Center), this may attenuate potential lead-time bias.
Additionally, factors for which we have not accounted may confound the association between disease type and survival. For instance, although functional capacity has been shown to predict survival in cohorts of patients with various forms of PAH, we were unable to perform these analyses, since most patients with ILD-associated PH were not given 6-minute walk distance tests. Similarly, desaturation during exercise has been shown to predict survival in ILD (55), but these data were also not collected routinely on our patients. Further, although we have adjudicated the cause of death in most patients, it is possible that other factors, such as immunosuppressant toxicity, may have contributed to mortality in the ILD-associated PH group.
In summary, we have demonstrated significantly reduced survival in SSc patients with PH combined with advanced ILD compared with SSc patients with isolated PAH. This subset of patients has an extremely poor prognosis. The observed mortality in SSc patients with isolated PAH, treated with currently available agents, was lower than historical reports (49), but remains far from satisfactory. Earlier therapy may result in improved outcomes since mortality was associated with worsening hemodynamic factors. Clinical trials in PAH typically include these patients as well as patients with idiopathic PAH. Given the dramatically worse long-term outcomes in patients with SSc and PAH, studies focusing on this group of patients are warranted.
Dr. Girgis had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Mathai, Girgis.
Acquisition of data. Mathai, Hummers, Champion, Wigley, Zaiman, Girgis.
Analysis and interpretation of data. Mathai, Hummers, Hassoun, Girgis.
Manuscript preparation. Mathai, Hummers, Champion, Wigley, Zaiman, Hassoun, Girgis.
Statistical analysis. Mathai.
- 51Bosentan versus placebo in interstitial lung disease secondary to systemic sclerosis (SSc): the BUILD-2 Study [abstract]. Proc Am Thorac Soc 2006; 3: A243., , , , , , et al.