Intravenous Prostanoids in Systemic Sclerosis-Associated Pulmonary Arterial Hypertension

A Single-Centre Experience

Maria Vlachou; Hossam Fayed; Adele Dawson; Sally Reddecliffe; Alexander Stevenson; Ross J Thomson; Benjamin Emmanuel Schreiber; J. Gerry Coghlan


Rheumatology. 2022;61(3):1106-1114. 

In This Article


Study Population

All consecutive patients from Royal Free Hospital, Pulmonary Hypertension Service Registry with SSc-PAH, initiated on i.v. prostacyclin analogues between January 2006 and May 2020, were included in the current analysis. In our study, we included patients where i.v. prostanoids were initiated after January 2006, this cut point was chosen to ensure that endothelin receptor antagonists and phosphodiesterase type 5 inhibitor therapies were available as background treatment options. Intravenous prostanoid therapy was considered in all patients as per European Society of Cardiology (ESC) guidelines [World Health Organization Functional Class (WHO FC) i.v., or WHO FC 3 where deteriorating or otherwise high risk].

The data were prospectively collected from January 2006 until November 2020 and retrospectively analysed in December 2020. Patients were diagnosed as having systemic sclerosis (SSc) if they fulfilled the American College of Rheumatology criteria and/or the LeRoy and Medsger criteria.[14–16] Patients who were diagnosed with systemic lupus erythematosus according to the American College of Rheumatology criteria as updated in 1997 and patients who were classified as having mixed CTD when they fulfilled at least one of the three most commonly used sets of criteria for mixed CTD: those of Sharp, those of Alarcon-Segovia and Villarreal, and those of Kasukawa and co-workers, were excluded from the study.[16–21]

The study was approved by the Royal Free Hospital Institutional Review Board, as an audit of current practice. As a retrospective analysis, including many patients who had died, consenting was not deemed practical, patients that refused consent for use of their data were identified using the message exchange for social care and health service.

Study Protocol

Patient survival was calculated from the date of initiation of i.v. prostanoids, following insertion of a Groshong or Hickman line for the purposes of administration of prostanoids. Death date was identified using the National Health Service Spine (demographics database) for the non-survivors. Patients were censored if they remained alive after 5 years or if they remained alive until the census date of 30 November 2020.

Baseline characteristics, before the initiation of i.v. prostanoids, and at first formal reassessment following initiation of i.v. prostanoids were prospectively collected, including age, gender, type of PAH treatment (no therapy, monotherapy, oral combination therapy), WHO FC, 6MWD (6-min walking distance), NT-proBNP (N-terminal pro-brain natriuretic peptide), haemodynamic characteristics and dose of i.v. prostanoids achieved at the time of the first follow-up catheterization, and the maximum dose achieved, as well. Patients with a total interstitial lung disease extent of ≥20%, or with moderate or severe emphysema were categorized as having lung disease. PVOD was identified by the development of falling oxygen saturation in association with new or progressive interseptal lobular thickening and centrilobular ground glass shadowing.[22]

Prior to 2011, combination therapy was rare as an upfront option, and commissioning delayed initiation of i.v. therapy after the clinical decision had been made. Therefore, we divided the SSc-PAH patients into two groups based on when they were commenced on i.v. prostanoids, (first group: January 2006 to December 2010; second group: January 2011 to November 2020), to investigate if a limitation comes from a calendar effect.

Risk Stratification

Risk assessment was based on the ESC guidelines, and the original REVEAL score.[1,23] Both scores were calculated at baseline, and at first formal reassessment. To determine the average ESC score, Kylhammar's methodology was followed.[24] The available parameters in our population were WHO FC, syncope presence and frequency, 6MWD, NT- proBNP, right atrial area, pericardial effusion, right atrial pressure, cardiac index, mixed venous oxygen saturation (SvO2).

The REVEAL score was calculated based on the original 1.0 score calculator, parameters included were: SSc-PAH diagnosis, presence/absence of renal insufficiency, presence/absence of male gender with an age >60 years, WHO FC, blood pressure and heart rate, 6MWD, NT-proBNP, presence/absence of pericardial effusion, diffusion capacity of carbon monoxide, mean right atrial pressure and pulmonary vascular resistance (PVR).[23] To allocate patients into meaningfully sized groups, they were classified as low risk or average risk if REVEAL points were ≤8, as moderate risk if REVEAL points were equal to 9, as high or very high risk if REVEAL points were ≥10.[23]

Survival was evaluated for all patients in whom a tunnelled line was inserted, irrespective of the actual treatment received. The clinical and haemodynamic efficacy of i.v. prostanoids was assessed in those alive and continuing on i.v. therapy at the time of the first full reassessment (median 6.5 months). Recalculated risk profile at first full reassessment was determined in survivors continuing on i.v. therapy and updated risk associated survival evaluated from the date of reassessment.

Statistical Analysis

Statistical analysis was performed with the SPSS software version 25.0 (SPSS, Chicago, IL, USA). Statistical significance was considered for P-values <0.05. Baseline parameters before the initiation of i.v. treatment and parameters obtained at the first formal reassessment of the patients after the initiation of i.v. prostanoids were recorded and analysed. Normally distributed continuous variables were expressed as mean ± standard deviation, whereas skewed variables were expressed as median with interquartile range (IQR). The Kolmogorov–Smirnov test was used to assess normality. Categorical variables were expressed as frequency (percentage). Changes between baseline and follow-up parameters were compared using paired t test for normally distributed variables, or Wilcoxon signed-rank test for non-normally distributed variables.

Univariable Cox proportional hazards analysis was performed to investigate the association of baseline parameters with 5-year estimated survival. Results were expressed as hazard ratios with 95% CI. Variables with a P-value <0.05 in the univariable analysis were entered in the multivariable analysis.

ESC score and REVEAL score were only in the Kaplan–Meier analyses, in order to avoid duplication, as their calculation were based on parameters that were included in the univariable analysis. The survival curves for all patients were calculated using Kaplan–Meier product-limit estimates, while comparisons among groups were assessed with log-rank test.