Pulmonary arterial hypertension (PAH) is a major cause of mortality in connective tissue disease (CTD), particularly in systemic sclerosis (SSc) 1. Prior to the introduction of advanced PAH therapies, such as endothelin receptor antagonists (ERA), prostacyclin analogues and phosphodiesterase type-5 inhibitors (PDE5), treatment options for PAH were limited. Current therapies increase exercise tolerance and improve hemodynamic parameters 2. A recent meta-analysis suggests that they also confer a survival benefit 3. In one contemporary cohort of patients with SSc-associated PAH (SSc-PAH), survival was 81% at one year and 71% at two years, compared to 68% and 47% respectively, in a historical cohort (P = 0.016) 4. Survival in CTD-associated PAH (CTD-PAH) is shorter than in idiopathic PAH (IPAH). This remains the case in the current treatment era, as demonstrated in a US registry of PAH patients wherein one-year survival was 86% among patients with CTD-PAH compared to 93% in patients with IPAH (P < 0.0001) 5.

Currently in Australia, prescription of specific PAH therapy is limited to government designated PAH treatment centres. All patients with right-heart catheter (RHC) proven PAH qualify for monotherapy with bosentan, ambrisentan, sildenafil or inhaled iloprost. For ongoing therapy, patients must demonstrate stability of both six-minute walk distance (6MWD) and echocardiographic parameters, although they are able to swap to an alternate agent if these criteria are not met. Combination PAH therapy is currently available at cost or on compassionate grounds. Sitaxentan was available in Australia until its worldwide withdrawal in early 2011. Intravenous prostacyclin analogues were not subsidised in the treatment of CTD-PAH in Australia at the time of this study.

In this study, our objective was to quantify survival and determine factors predictive of mortality among Australian patients with CTD-PAH, since the advent of specific PAH therapies.

All patients with CTD-PAH diagnosed on RHC 6 from November 2002 onwards were recruited (at the time of diagnosis of PAH) from six PAH treatment centres across Australia and followed prospectively, at three- to six-month intervals. Date of diagnosis of PAH was defined as the date of the RHC.

Clinical and hemodynamic variables were recorded during follow-up. Data were censored at 31 December 2009 for analysis. Ethics approval was obtained from the human research ethics committees of St. Vincent's Hospital Melbourne, Southern Health, Royal Adelaide Hospital, Royal Perth Hospital and Central Northern Adelaide Health Service. Patients in this study provided informed consent.

A total of 117 patients with incident CTD-PAH were recruited. Patient characteristics and baseline hemodynamics are summarised in Table 1. During a mean ± SD follow-up from PAH diagnosis of 2.6 ± 1.8 years, there were 32 (27.4%) deaths. All but three deaths were primarily due to PAH. In the remaining three deaths due to malignancy, PAH was a major contributor to death.

In this cohort study of patients with CTD-PAH confirmed by RHC, one- two- and three-year survivals were 94%, 89% and 73%, respectively. We found that independent predictors of mortality were more severe PAH at diagnosis, manifested by higher baseline mRAP, lower baseline 6MWD, higher baseline WHO FC and the presence of pericardial effusion. After adjustment for these covariates, anticoagulation and combination PAH therapy increased survival.

Our one-year survival of 94% is higher than the one-year survival of 86% reported by Chung et al. 5 in a cohort of CTD-PAH patients from a multi-centre US-based registry. Bearing in mind that the majority (95%) of patients in our study had SSc, our survival rates are also higher than those reported by Condliffe et al. 17 from the UK pulmonary hypertension service, where one-year survival was 78% and three-year survival was 47% in SSc-PAH, while one-year survival was 78% and three-year survival was 74% in SLE-associated PAH. There are several possible explanations for the differences in survival rates of our patients compared with other cohorts. These include differences in hemodynamic profile at PAH diagnosis and variations in the approach to treatment, with a significant proportion of our patients receiving combination therapy. In addition, it is possible that other cohorts may contain patients with overall more severe CTD, who are more likely to succumb to their disease.

Whilst elevated mRAP was associated with mortality in our study, elevated mPAP was not significantly associated with mortality in either univariate or multivariable analysis. This apparent discrepancy between mPAP and mRAP may be due to a proportion of patients having myocardial disease due to SSc, thereby compromising the ability of the right ventricle to generate an elevated mPAP. On the other hand, mRAP better reflects right ventricular dysfunction and, therefore, continues to increase with worsening disease. These relationships were borne out in a systematic review of predictors of survival in IPAH which found that whilst 10 publications evaluating this issue supported the association between increased mPAP and decreased survival, 19 did not 18. In the same systematic review, mRAP was the variable most commonly associated with mortality, with an independent predictive value for mortality reported in 17 of 28 studies.

6MWD is predictive of mortality in IPAH 19, and has been used as a primary endpoint in clinical trials of PAH therapy 2021222324. As 6MWD may be reduced due to impaired mobility and joint or muscle pain, especially in patients with CTD, some experts have considered it to be a less useful measure of exercise capacity in those with CTD-PAH than in those with IPAH 25. Our findings suggest that, despite these limitations in certain individuals, 6MWD may still be a useful predictor of mortality in CTD-PAH.

In our study, higher baseline WHO FC was also independently predictive of mortality. The prognostic value of WHO or New York Heart Association FC is supported by multiple studies in PAH, several of which have included patients with CTD-PAH 18. Lack of improvement in FC after specific PAH therapy has also been suggested as an indication to escalate monotherapy to combination therapy 2. However, WHO FC is a subjective measure and does not always correlate with other markers of PAH severity, such as mRAP and 6MWD 26.

In IPAH, the presence of pericardial effusion has been reported to be predictive of mortality in two retrospective studies, and the severity of pericardial effusion has been predictive of mortality in five retrospective studies 18. In our study, the presence of a pericardial effusion was predictive of mortality in CTD-PAH. This is consistent with the findings of Fisher et al. 27. Although the physiology of the accumulation of fluid in the pericardial space is incompletely understood, pericardial effusion in this context is likely due to increased venous pressure and right heart failure 28. Importantly, in our patients, there were no echocardiographic or other clinical features of serositis or pericarditis to indicate active inflammatory disease as the cause for pericardial effusion.

In multiple regression analysis, after adjustment for covariates, including use of specific PAH therapies, we found that anticoagulation with warfarin conferred a substantial survival benefit in CTD-PAH, which has not previously been reported. Correlates of warfarin therapy included more severe PAH, with anticoagulated patients more likely to have a higher mPAP and a pericardial effusion. This is consistent with an international survey of PAH experts, which found that clinicians were more likely to use warfarin in more severe CTD-PAH 29. This survey also revealed variation in physician beliefs and prescribing habits in relation to warfarin in CTD-PAH. Whilst it is accepted that thrombotic arteriopathy plays a role in the pathophysiology of PAH 30, the evidence for a survival benefit with anticoagulation remains limited. A survival benefit was suggested by a systematic review of warfarin therapy in IPAH in which five of seven retrospective analyses reported a survival benefit with warfarin use 31; no CTD-PAH patients were included in this analysis. A recent observational study of 275 patients with SSc-PAH showed a low probability of survival with warfarin; however, in this study, over a third of the patients on warfarin were not receiving specific PAH therapies 32.

In our study, the most common documented contraindication to anticoagulation in 16 patients was a history of gastrointestinal bleeding. This is a major consideration in patients with SSc-PAH as the prevalence of gastrointestinal vascular ectasia, a possible source of blood loss, is reported to be as high as 5.7% 33.

It is possible that our finding of a survival benefit with anticoagulation was confounded by sicker and frailer patients with severe multi-organ CTD being less likely to receive warfarin therapy. While data on PAH severity and other SSc manifestations were collected, we did not systematically collect data on non-SSc-related comorbidities which may have impacted the patients' overall health status and suitability for anticoagulation. Despite our efforts to adjust for all clinically important covariates, our observational study cannot fully overcome the effect of all potential confounders. Evaluation of the therapeutic efficacy of anticoagulation in CTD-PAH warrants a more definitive study, with random assignment of treatment in patients with similar baseline characteristics.

Combination specific PAH therapy in CTD-PAH targets more than one of the multiple biologic pathways involved in disease pathogenesis 2. Although a number of small, short-term trials of dual agent therapy have been conducted, results have been disparate and there are no data available beyond 12 weeks of therapy 2. In a cohort of 112 patients with PAH, among whom 40 had CTD (29 had SSc), Keogh et al. recently reported survival of 72% at 12 months and 48% at 24 months in patients on dual specific PAH therapy 34.

In this study, we have sought to determine survival and predictors of mortality in a cohort of CTD-PAH patients sourced from multiple centres, typical of those encountered in clinical practice. Although we included all CTD-PAH patients at each site, there was potential for selection bias in that the sites included in our study were large academic centres. However, in practice, as prescription of CTD-PAH therapy in Australia is limited to large government designated centres, patients with all grades of PAH severity are referred to these sites for treatment.

Overall, the findings of this study suggest that treatment with warfarin in addition to combination specific PAH therapy may improve survival in CTD-PAH. The efficacy of anticoagulation and combination therapy in CTD-PAH merits further evaluation.

6MWD: six minute walk distance; 95% CI: 95% confidence interval; CI: cardiac index; CTD: connective tissue disease; CTD-PAH: connective tissue disease-associated pulmonary arterial hypertension; DLCO: diffusing capacity of carbon monoxide; DVT: deep vein thrombosis; ERA: endothelin receptor antagonists; FVC: forced vital capacity; GI: gastrointestinal; HR: hazard ratio; HRCT: high-resolution computed tomography; ILD: interstitial lung disease; INR: International Normalised Ratio; IPAH: idiopathic pulmonary arterial hypertension; K-M curves: Kaplan-Meier curves; MCTD: mixed connective tissue disease; mPAP: mean pulmonary arterial pressure; mRAP: mean right atrial pressure; PAH: pulmonary arterial hypertension; PDE5: phosphodiesterase type-5 inhibitors; PVR: pulmonary vascular resistance; RA: rheumatoid arthritis; RHC: right-heart catheter; SD: standard deviation; SLE: systemic lupus erythematosus; SSc: systemic sclerosis; WHO FC: World Health Organization Functional Class.

Authors WS, EG and SP have received consultancies and speaking fees from Actelion Australia, GalxoSmithKline and Pfizer (amounts less than $10,000). Other authors have no disclosures.

GSN and MN contributed to study design, data collection and analysis, interpretation of findings and preparation of the manuscript. WS and SP contributed to study design, data collection, interpretation of findings and preparation of the manuscript. DP, EG, JR, AT, RM, CH, KC and JS contributed to data collection, interpretation of findings and preparation of the manuscript. All authors have approved the manuscript for publication.