Pulmonary hypertension (PH) in sarcoidosis is associated with bad outcomes. Although there is interest in using pulmonary vasodilators (PVs) for PH in sarcoidosis, there are few data to support their use. In this study, a retrospective review of a cohort of patients with PH and sarcoidosis was conducted, focusing on those treated with PVs, and a meta-analysis of published reports indexed in MEDLINE was performed. Twenty-four patients were found. The rate of mortality or transplantation rate was 41.2%. Median survival without transplantation was 5.3 years. More patients who died or underwent transplantation during follow-up had moderate or severe lung fibrosis (66.7% vs 15.4%), had right ventricular dysfunction (80% vs 7.7%), and were in World Health Organization class IV (66.7% vs 30.8%). Body surface areas were lower in patients with events, as was cardiac output. Mortality was not different between patients treated with PVs and those not treated (54.5% vs 38.5%, p = 0.44) despite the treated patients’ having more right ventricular dysfunction and worse hemodynamics. In a Cox regression survival model, lower body surface area, right ventricular dysfunction, and the presence of moderate or severe lung fibrosis were predictors of worse outcomes, but not treatment with PVs. PV-treated patients (n = 11) showed improved 6-minute walk distances and decreased N-terminal pro–B-type natriuretic peptide levels during follow-up. There was a trend toward improvement in hemodynamic profile. Four studies plus the data from this study were included in the meta-analysis. Six-minute walk distance improved by 30.64 m after treatment. Hemodynamics improved, with a reduction in mean pulmonary arterial pressure of 8.03 mm Hg and a decrease in pulmonary vascular resistance of 4.23 Wood units. In conclusion, PH in sarcoidosis is associated with adverse outcomes, particularly when accompanied by right ventricular dysfunction and/or moderate or severe lung fibrosis. Treating selected patients can improve hemodynamics and functional parameters.
Pulmonary hypertension (PH) is common in patients with advanced sarcoidosis, being present in >70% of patients referred for lung transplantation and associated with a poor prognosis. PH associated with sarcoidosis is included in group 5 of the Dana Point classification of PH, because its pathophysiology is complex and not completely resolved. The rationale for including this condition in group 5 and not in group 3 (related to lung diseases) is the occurrence of PH in patients with sarcoidosis even without parenchymal lung disease. Compression of the pulmonary vessel by lymphadenopathy, vasospasm, vasculopathy, and fibrous destruction of the pulmonary vascular bed have all been proposed as causes of PH in sarcoidosis. Treatments used in patients with this condition have included immunosuppressive therapy, domiciliary oxygen, and in some cases pulmonary vasodilator therapies, as well as other supportive treatments. Although there is increasing interest in the use of pulmonary vasodilators (PV) for patients with sarcoidosis who have PH (endothelin receptor antagonists, phosphodiesterase-5 inhibitors, or prostanoids), there are few published data to support their use, and there have been no randomized placebo-controlled trials. In this work, we have analyzed a cohort of patients with PH and sarcoidosis demonstrated on right-sided catheterization, focusing particularly on those treated with specific PH therapy.
Methods
We performed a retrospective review of patients with sarcoidosis referred to the National Pulmonary Hypertension Unit of the Royal Free Hospital in London from 1999 to 2011. We identified all patients who underwent right-sided cardiac catheterization and then cross-checked individual medical records. We included patients with sarcoidosis who had PH, defined as a mean pulmonary artery pressure ≥25 mm Hg, confirmed by right-sided cardiac catheterization. We recorded the use of any PV therapy, such as sildenafil, bosentan, ambrisentan, iloprost, treprostinil, or epoprostenol. The decision to treat these patients with PVs was based on clinical judgment and funding decisions by patients’ local commissioners. Funding in the United Kingdom is generally approved only if the degree of PH is considered to be due to pulmonary arterial hypertension (Dana Point group 1). The severity of lung disease on computed tomography was determined by a review of the studies that were performed or reviewed by the radiologists at our center. Assessment of right ventricular function was based on tricuspid annular plane systolic excursion measured on echocardiography and “eyeball” estimation by the echocardiographer of the right ventricular ejection fraction. Survival was ascertained using a National Health Service database search performed in March 2012.
Data are expressed as mean ± SEM or as medians, depending on whether the data were normally distributed. Chi-square tests were used for comparisons of categorical values. Mann-Whitney U tests were used to compare variables across groups and Wilcoxon’s tests to assess changes in paired data. For survival analysis, we used Kaplan-Meier analysis with the log-rank test. These factors were analyzed in a Cox regression model to check if they were related to outcomes in this analysis. In this retrospective analysis, we first analyzed the data of the whole cohort of patients with sarcoidosis and PH and then focused on patients treated with PV therapy, especially those with hemodynamic data during follow-up after starting PV treatment. Statistical significance was set to a p value <0.05. SPSS version 15 (SPSS, Inc., Chicago, Illinois) for Microsoft Windows (Microsoft Corporation, Redmond, Washington) was used.
We also performed a meta-analysis of published data on the effect of licensed PVs, expanding this by including our data. We performed a review of published studies indexed in MEDLINE with a combination of the terms “pulmonary hypertension” and “sarcoidosis” with “bosentan,” “ambrisentan,” “epoprostenol,” “treprostinil,” “beraprost,” “sildenafil,” “tadalafil,” “iloprost,” and “treatment.” There was no language or year restriction. We subsequently reviewed the reference lists of narrative reviews, guidelines, and other retrieved documents to identify any publications not identified in the database search. The criteria for including studies in the meta-analysis were (1) Diagnosis of PH by right-sided cardiac catheterization, (2) diagnosis of sarcoidosis, (3) treatment with licensed therapies for PH, including ambrisentan, bosentan, sildenafil, tadalafil, epoprostenol, treprostinil, beraprost, and iloprost, (4) sample size ≥5 patients, and (5) a change in 6-minute walk distance or pulmonary hemodynamics (including at least mean pulmonary artery pressure, pulmonary vascular resistance, and cardiac output). We used the DerSimonian-Laird continuous 1-arm random-effects model to perform the analysis. Pooled effects on 6-minute walk distance and hemodynamic parameters are presented as weighted mean differences with corresponding 95% confidence intervals (CIs). Forest plots were created for each outcome. Statistical heterogeneity was assessed using the Cochrane Q statistic (with p values <0.10 considered significant). We also calculated I 2 statistics to estimate the proportion of variation attributable to between-study heterogeneity. In the studies in which the standard deviation of the mean change of the parameters studied was not provided, we imputed a change-from-baseline standard deviation using a correlation coefficient. We assumed as correlation coefficients in the formula, to estimate standard deviation of the mean change, the values calculated for our sample. We used the freeware MetaAnalyst for Microsoft Windows.
Results
From 1999 to December 2011, 24 patients ( Table 1 ) with sarcoidosis and evidence on right-sided cardiac catheterization of PH were seen in the National Pulmonary Hypertension Unit of the Royal Free Hospital in London. The median follow-up duration was 22.6 months. The mean age of patients at diagnosis of PH was 57.83 ± 13.05 years. There was a small predominance of women (54.2%), and almost 90% of patients were in functional class III or IV, with 42% in functional class IV. During follow-up, 9 patients (37.5%) died and 2 patients underwent lung transplantation (41.2% combined mortality and transplantation). Median survival without transplantation ( Figure 1 ) was 5.3 years (range 0.5 to 10.01), with a rate of mortality or transplantation at 2 years from diagnosis of 35% and a mortality rate of 33%. The severity of the disease of this cohort was confirmed by the presence of at least moderate pulmonary fibrosis on computed tomography in 63.2% of patients (although mean forced vital capacity was 70.24 ± 22.95%), right ventricular impairment in 27.8%, and poor baseline 6-minute walk distances (mean 212.8 ± 127.3 m).
| Patient | FU (mo) | Age at Diagnosis (yrs) | Gender | Death | Transplantation | Yr | WHO FC | 6MWD (m) | BNP (pmol/L) | BSA (m 2 ) | RAP (mm Hg) | WP (mm Hg) | MPAP (mm Hg) | PVR (WU) | CO (L/min) | FVC (%) | RV Impairment | PV | MSLF |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 153.27 | 43.24 | F | No | No | 1999 | — | — | — | 2.2 | 8 | 9 | 25 | 2.16 | 7.4 | 47 | No | No | No |
| 4 | 21.23 | 56.81 | F | Yes | No | 2002 | — | — | — | 1.57 | 7 | 6 | 40 | 8.5 | 4 | — | — | No | — |
| 7 | 97.7 | 45.13 | M | No | No | 2003 | II | 386 | — | 1.96 | 3 | 9 | 36 | 4.29 | 6.3 | — | — | No | — |
| 9 | 91.5 | 63.68 | F | No | No | 2004 | III | — | — | — | 13 | 17 | 24 | — | — | 86.5 | No | No | No |
| 2 | 25.27 | 61.65 | F | Yes | No | 2004 | — | — | — | 2.05 | 7 | 12 | 36 | 8.57 | 2.8 | — | — | Yes | — |
| 3 | 95.6 | 55.26 | F | No | No | 2004 | III | 132 | 1.9 | 5 | 11 | 26 | 3.85 | 3.9 | 57.7 | No | No | No | |
| 5 | 22.6 | 35.14 | M | No | Yes | 2004 | — | — | — | 1.9 | 9 | 8 | 51 | 11.94 | 3.6 | — | — | No | — |
| 6 | 64.3 | 68 | F | Yes | No | 2005 | IV | 78 | 243 | 1.8 | 8 | 6 | 56 | 20 | 2.5 | 60.3 | Yes | Yes | No |
| 8 | 4.2 | 75.45 | F | Yes | No | 2005 | — | — | — | 1.61 | 17 | 21 | 34 | 3.82 | 3.4 | — | — | No | — |
| 10 | 15.23 | 56.94 | M | Yes | No | 2006 | III | 160 | — | — | 8 | 13 | 53 | 5.63 | 7.1 | — | — | Yes | Yes |
| 11 | 29.77 | 79.85 | M | Yes | No | 2007 | III | 178 | 21 | 2.1 | 4 | 12 | 42 | 6.25 | 4.8 | 119 | Yes | Yes | No |
| 12 | 45.7 | 58.44 | M | No | No | 2008 | III | — | — | 1.9 | 5 | 9 | 28 | 3.39 | 5.6 | 49 | No | Yes | Yes |
| 13 | 42.7 | 40.5 | M | No | No | 2008 | IV | 231 | 300 | 1.65 | 16 | 10 | 58 | 22.86 | 2.1 | 73.7 | No | Yes | No |
| 14 | 9 | 39.04 | F | No | Yes | 2008 | IV | 396 | 94 | 1.72 | 3 | 7 | 30 | 4.04 | 5.7 | 108 | No | No | Yes |
| 15 | 12.2 | 65.34 | M | Yes | No | 2009 | IV | — | 261 | 1.71 | 3 | 10 | 31 | 5.53 | 3.8 | 35 | Yes | No | Yes |
| 16 | 27.43 | 76.06 | M | No | No | 2009 | IV | 132 | 39 | 1.98 | 15 | 5 | 25 | 3.13 | 6.4 | — | No | No | No |
| 18 | 24.33 | 54.01 | F | No | No | 2009 | IV | 69 | 6 | 2.1 | 14 | 18 | 45 | 4.29 | 6.3 | 64.2 | No | No | No |
| 17 | 6.67 | 69 | F | Yes | No | 2010 | IV | — | 165 | 1.56 | 17 | 10 | 37 | 7.94 | 3.4 | — | Yes | Yes | Yes |
| 19 | 22.6 | 65.26 | F | No | No | 2010 | III | 135 | 409 | 1.79 | 7 | 12 | 50 | 8.09 | 4.7 | 65.1 | No | Yes | No |
| 20 | 18.83 | 44.53 | M | No | No | 2010 | III | 333 | 41 | 1.79 | 8 | 13 | 47 | 5.31 | 6.4 | 50 | No | No | Yes |
| 21 | 18.17 | 70.22 | M | No | No | 2010 | II | 466 | 16 | 2 | 8 | 13 | 29 | 2.35 | 6.8 | — | No | No | No |
| 22 | 17.27 | 39.71 | M | No | No | 2010 | III | 220 | 1,099 | 2.56 | 22 | 15 | 63 | 9.6 | 5 | 75.1 | No | Yes | No |
| 23 | 8.4 | 54.05 | F | Yes | No | 2011 | IV | 63 | 252 | 2.02 | 16 | 28 | 49 | 8.75 | 2.4 | 76 | Yes | Yes | No |
| 24 | 9.07 | 70.62 | F | No | No | 2011 | III | 213 | 79 | 2.12 | 12 | 14 | 43 | 3.54 | 8.2 | 87 | No | Yes | No |
We compared baseline data to determine factors potentially related to the combined end point of mortality and transplantation. At baseline, predictors of death or transplantation were moderate or severe lung fibrosis (66.7% vs 15.4%, p = 0.02), right ventricular dysfunction (80% vs 7.7%, p = 0.002), and functional class IV (66.7% vs 30.8%, p = 0.14). Baseline body surface areas were lower in patients with events (1.80 ± 0.21 vs 1.98 ± 0.23 m 2 , p = 0.05), as were cardiac output (4.10 ± 1.41 vs 5.50 ± 1.81 L/min, p = 0.06) and cardiac index (2.04 ± 0.63 vs 2.87 ± 0.76 L/min/m 2 , p = 0.02). Survival free of transplantation ( Figure 1 ) was worse in patients in functional class IV (3.04 ± 0.91 vs 6.43 ± 1.06 years, p = 0.03), moderate or severe lung fibrosis on computed tomography (1.86 ± 0.57 vs 8.88 ± 1.83 years, p = 0.02), and right ventricular dysfunction (2.02 ± 0.90 vs 11.84 ± 0.88 years, p = 0.001). Although more patients died or underwent transplantation during follow-up in the group treated with PVs (54.5% vs 38.5%, p = 0.44), mean survival free of transplantation was not different between the 2 groups (3.20 ± 0.72 vs 7.88 ± 1.68 years, p = 0.45), mainly driven by the fact that the survival curves overlapped in the first 2.5 years of follow-up ( Figure 1 ).Variables (qualitative and quantitative) that resulted in significant decreases in mean event-free survival in previous analysis were reassessed using a Cox regression survival model adjusted for age and gender. Using this analysis, predictors of adverse outcome were lower body surface area (odds ratio [OR] 0.15, 95% CI 0 to 0.723, p = 0.03), right ventricular dysfunction (OR 83.1, 95% CI 2.2 to 3,102, p = 0.017), and the presence of moderate or severe lung fibrosis in computed tomography (OR 64, 95% CI 2.3 to 1,768, p = 0.04). Right ventricular dysfunction on echocardiography was the most powerful indicator of death or transplantation during follow-up. In the same analysis, functional class IV (OR 4.9, 95% CI 0.92 to 26.15, p = 0.062), lower cardiac output (OR 0.72, 95% CI 0.49 to 1.04), and lower cardiac index (OR 0.50, 95% CI 0.21 to 1.15) showed a tendency toward worse outcomes that did not reach statistical significance. Treatment with PVs was not related to a worse rate of death or transplantation (OR 1.45, 95% CI 0.40 to 5.22, p = 0.56) or crude mortality (OR 1.82, 95% CI 0.41 to 8.08, p = 0.45).
After initial evaluation including chest computed tomography, lung function tests, echocardiography, and right-sided cardiac catheterization, 14 patients (58%) were considered potential candidates for PV treatment ( Figure 2 ), all of them in functional class III or IV, but only 11 received therapy (45.8%), 2 with bosentan and 9 with sildenafil, because in 3 patients, funding for PV therapy was not approved by the commissioners. Three patients had wedge pressures >15 mm Hg at the baseline evaluation, 1 of whom was treated with PVs after normalization of filling pressures with diuretics. Patients treated with PVs had worse hemodynamic parameters at baseline, including higher mean pulmonary artery pressure (46.81 ± 10.62 vs 34.07 ± 9.1 mm Hg, p = 0.006), higher pulmonary vascular resistance (9.51 ± 6.26 vs 4.93 ± 2.76 Wood units, p = 0.02), and lower cardiac indexes (2.10 ± 0.86 vs 2.82 ± 0.61 L/min/m 2 , p = 0.04). There was a tendency that did not reach statistical significance to treat patients with higher levels of N-terminal pro–B-type natriuretic peptide (321.00 ± 337.47 vs 76.16 ± 95.54 pmol/L, p = 0.05) and more impaired right ventricular function (44.1% vs 11.1%, p = 0.14). No difference was found regarding the value of 6-minute walk distance or degree of lung involvement on computed tomography or spirometry. Of the 11 patients who received PVs, 1 died of sepsis before planned reevaluation, and 2 declined to undergo follow-up catheterization ( Figure 2 ), so we had 8 patients with follow-up hemodynamic and functional data. The median time to repeat right-sided cardiac catheterization was 5.3 months. Compared to baseline ( Table 2 ),patients receiving treatment showed significant improvement in 6-minute walk distances (+59.85 ± 77.43 m, p = 0.04) and reduced N-terminal pro–B-type natriuretic peptide (−240 ± 269 pmol/L, p = 0.04); functional class was unchanged (from 6 patients in class III and 2 patients in class IV at baseline to 7 patients in class III and 1 patient in class IV at follow-up). There was a nonsignificant trend toward improvement in the hemodynamic profile with decreasing pulmonary vascular resistance and increasing cardiac output and cardiac index; wedge pressure increased but mean pulmonary artery pressure was unchanged.
| Parameter | Baseline | Follow-Up | p Value |
|---|---|---|---|
| 6-minute walk distance (m) | 169.57 ± 56.66 | 229.42 ± 110.36 | 0.04 |
| N-terminal pro–B-type natriuretic peptide (pmol/L) | 381.60 ± 431.18 | 141.67 ± 167.91 | 0.04 |
| Heart rate (beats/min) | 87.00 ± 26.62 | 83.33 ± 28.00 | 0.18 |
| Right atrial pressure (mm Hg) | 9.87 ± 6.35 | 9.25 ± 4.06 | 0.73 |
| Wedge pressure (mm Hg) | 11.12 ± 2.85 | 12.50 ± 1.30 | 0.27 |
| Mean pulmonary artery pressure (mm Hg) | 45.75 ± 13.61 | 48.00 ± 17.41 | 0.88 |
| Pulmonary vascular resistance (Wood units) | 8.75 ± 6.02 | 7.45 ± 5.37 | 0.06 |
| Cardiac output (L/min) | 4.60 ± 1.90 | 5.20 ± 1.29 | 0.06 |
| Cardiac index (L/min/m 2 ) | 2.29 ± 0.84 | 2.63 ± 0.57 | 0.06 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
