Predictors of mortality in connective tissue disease-associated pulmonary arterial hypertension: a cohort study

All patients had SSc or another underlying CTD, namely rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) or mixed connective tissue disease (MCTD). Definitions were based on the American College of Rheumatology diagnostic criteria for SSc [7], RA [8] and SLE [9] and the Alarcon-Segovia diagnostic criteria for MCTD [10]. Patients were defined as having limited or diffuse SSc according to the classification criteria of LeRoy et al. [11]. World Health Organization (WHO, Geneva, Switzerland) functional class (FC) [12] and 6MWD [13] at PAH diagnosis were recorded along with echocardiographic and RHC parameters. The presence of pericardial effusion on echocardiography was documented. We excluded patients with significant interstitial lung disease (ILD) defined based on extensive disease (> 20% lung involvement) on high-resolution CT lung (HRCT), or evidence of fibrosis on HRCT together with forced vital capacity (FVC) < 70% predicted and/or an FVC to diffusing capacity of carbon monoxide (DLCO) ratio < 1.6 in whom PAH was deemed secondary to lung disease [14]. Other covariates included autoantibody profile, anti-phospholipid antibody status (positive if either anti-cardiolipin or anti-beta2 glycoprotein IgG or IgM greater than upper limit of normal for assay) and the PAH study centre from which the patient was recruited.

The outcome variable was all-cause mortality. The date of death was recorded. Where data were available, the exact causes of death were also recorded and verified through chart review. The status (alive or dead) of patients at the time of censoring was confirmed by checking with the treating physician and through chart review. There were no losses to follow-up.

Specific PAH therapies, prescribed at physician discretion following RHC-confirmed diagnosis of PAH, were recorded at each visit. This included 'combination therapy', which was defined as treatment with more than one agent from any of the three classes (ERA, PDE5 inhibitors, prostacyclin analogues) at the same time. Indications for substitution or addition of a second agent were documented. We also recorded warfarin therapy of at least six months' duration, following PAH diagnosis. Indications and complications of warfarin therapy, and target INR were also recorded. In those who were not anticoagulated, we recorded contraindications for warfarin therapy as documented by the treating physician.

Patient characteristics at baseline are reported as mean ± standard deviation (SD) for continuous variables and proportions (percentages) for categorical variables. Differences in baseline hemodynamics between SSc patients and other CTD patients, and patients on monotherapy vs combination therapy were compared using the Student's t-test. Kaplan-Meier (K-M) curves were used to estimate survival in all patients and also in SSc patients compared with those who had other CTDs. Log-rank and Wilcoxon tests were used to determine univariate predictors of survival. After testing to ensure proportionality of hazard, Cox proportional hazards regression analyses were used to determine univariate and multivariable predictors of survival. These results were reported as hazard ratios (HR) with accompanying 95% confidence intervals (95% CI). Two-tailed P-values ≤ 0.05 were considered statistically significant.

All statistical analyses were performed using STATA 11.0 (Statacorp, College Station, TX, USA).

All patients received specific PAH therapy. Seventy patients (59.8%) received monotherapy, 12 patients (10.3%) sequential monotherapy and 34 patients (29.0%) combination therapy (Table 1). The main reasons for sequential and combination therapy were failure of initial PAH therapy or drug-related adverse effects significant enough to warrant a change in therapy.

Bosentan was the most commonly prescribed medication, used in 103 patients (88.0%). Sildenafil was the next most common medication used in 38 patients (32.5%), followed by sitaxentan in 18 (15.4%) and inhaled iloprost in 15 (13.4%). The most common combination therapies were bosentan and sildenafil (21 patients), followed by bosentan, sildenafil and inhaled iloprost (6 patients).

Mean ± SD mPAP at baseline was lower in patients with SSc than in patients with other CTDs (34.3 ± 11.7 mmHg vs 49.5 ± 10.3 mmHg, P < 0.001). SSc patients were also older at PAH diagnosis (62.6 ± 10.3 years vs 54.0 ± 17.1 years, P = 0.01) and had higher baseline 6MWD (334.6 ± 125.2 m vs 235.0 ± 110.3 m, P = 0.02) than patients with other CTDs.

Among all patients in the study, at baseline, those who subsequently received combination therapy had higher mPAP than those who received monotherapy or sequential monotherapy (40.0 ± 11.6 mmHg vs 34.0 ± 12.4 mmHg, P = 0.02). There were no differences in age at PAH diagnosis (59.6 ± 11.5 vs 62.5 ± 11.4 years, P = 0.21) or baseline 6MWD (303.3 ± 114.5 m vs 333.8 ± 131.5 m, P = 0.27) among the combination vs monotherapy or sequential monotherapy treatment groups.

Twenty-seven of the 32 (84.4%) deaths occurred in patients with SSc. Overall, one-year survival was 94%, two-year survival was 89% and three-year survival was 73% (Figure 1A). Median survival was approximately five years.

In univariate analysis (Table 2), factors associated with mortality were male sex, higher baseline mRAP on RHC, higher baseline WHO FC, lower baseline 6MWD, pericardial effusion, absence of warfarin therapy (Figure 1B) and lack of combination therapy (Figure 1C).

Cox regression analysis was performed to make adjustments for the effect of multiple covariates. In regression analysis, we took into consideration a desired ratio of independent-to-outcome variables of at most one to five, in order to ensure model stability (Table 3) [15]. Higher mRAP at diagnosis (HR = 1.13, 95% CI: 1.04 to 1.24, P = 0.007), lower baseline 6MWD (HR = 0.64, 95% CI: 0.43 to 0.97, P = 0.04), higher baseline WHO FC (HR = 3.42, 95% CI: 1.25 to 9.36, P = 0.04) and pericardial effusion (HR = 3.39, 95% CI: 1.07 to 10.68, P = 0.04) were predictive of mortality. Warfarin (HR = 0.20, 95% CI: 0.05 to 0.78, P = 0.02) and combination pulmonary vasodilator therapy (HR = 0.20, 95% CI: 0.05 to 0.83, P = 0.03) were protective. Male sex was not independently associated with mortality and, therefore, was removed from the final model. Overall, survival in patients with SSc (n = 104) was not significantly different compared to those with other CTDs (n = 13) (Figure 2). In further regression analyses of the SSc patients only, treatment with warfarin (HR = 0.33, 95% CI: 0.11 to 0.96, P = 0.04) and combination therapy (HR = 0.55, 95% CI: 0.33 to 0.92, P = 0.02) remained protective.

Twenty-eight of 36 (77.8%) patients on warfarin had SSc, 2 had MCTD, 2 SLE and 3 RA, and 1 had an undifferentiated CTD. Ten of 36 patients on warfarin died during follow-up. Among these, 7 had SSc, 1 had SLE, 1 MCTD and 1 RA. The main indication for anticoagulation in all 36 patients was PAH. However, three patients also had atrial fibrillation and three patients had a history of deep vein thrombosis (DVT). Of note, in the three patients with the history of DVT, PAH was deemed related to the CTD (WHO diagnostic group 1) [16] and not pulmonary thromboembolic disease. The target International Normalised Ratio (INR) in all patients was 2.0 to 3.0.

Univariate comparisons of characteristics of patients treated with warfarin, compared to those who were not treated with warfarin are summarised in Table 4. Patients on warfarin were significantly more likely to have a pericardial effusion (42.0% vs 33.4%, P = 0.001) and a higher baseline mPAP (42.0 vs 33.4 mm Hg, P = 0.001) than those not receiving warfarin. We found that there was no significant difference in baseline mRAP or baseline 6MWD between patients who were and those who were not anticoagulated (Table 4).

Among patients who were not treated with warfarin, a contraindication to anticoagulation was present in only 18 of 81 (22.2%) patients, and in the remainder the decision to not treat with warfarin was made at physician discretion. The most common contraindication in 16 patients was a history of gastrointestinal bleeding, with two other patients having a history of recurrent falls. On follow-up, only one patient with SSc, who was on warfarin for the treatment of PAH, had a complication of GI bleeding.