Computed tomography findings of pulmonary venoocclusive disease in scleroderma patients presenting with precapillary pulmonary hypertension

Authors

  • S. Günther,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
  • X. Jaïs,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
  • S. Maitre,

    1. Université Paris-Sud, Kremlin-Bicêtre, France, and Hôpital Antoine Béclère, AP-HP, Clamart, France
    Search for more papers by this author
  • A. Bérezné,

    1. Université Paris-Descartes and Hôpital Cochin, AP-HP, Paris, France
    Search for more papers by this author
  • P. Dorfmüller,

    1. Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
    • Dr. Dorfmüller has received speaking fees from Actelion (less than $10,000).

  • A. Seferian,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
  • L. Savale,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
    • Dr. Savale has received consulting fees, speaking fees, and/or honoraria from Pfizer, Actelion, Eli Lilly, and GlaxoSmithKline (less than $10,000 each).

  • O. Mercier,

    1. Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
  • E. Fadel,

    1. Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
  • O. Sitbon,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
    • Dr. Sitbon has received consulting fees, speaking fees, and/or honoraria from Actelion, Bayer HealthCare, GlaxoSmithKline, Eli Lilly, Pfizer, and United Therapeutics (less than $10,000 each).

  • L. Mouthon,

    1. Université Paris-Descartes and Hôpital Cochin, AP-HP, Paris, France
    Search for more papers by this author
  • G. Simonneau,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
    • Drs. Simonneau and Humbert have received consulting fees, speaking fees, and/or honoraria from Actelion, GlaxoSmithKline, Eli Lilly, Novartis, Pfizer, and United Therapeutics (less than $10,000 each).

  • M. Humbert,

    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    Search for more papers by this author
    • Drs. Simonneau and Humbert have received consulting fees, speaking fees, and/or honoraria from Actelion, GlaxoSmithKline, Eli Lilly, Novartis, Pfizer, and United Therapeutics (less than $10,000 each).

  • D. Montani

    Corresponding author
    1. Université Paris-Sud, Hôpital de Bicêtre, AP-HP, Le Kremlin-Bicêtre, France, and INSERM U999, Centre Chirurgical Marie Lannelongue, Le Plessis-Robinson, France
    • Service de Pneumologie et Soins Intensifs Thoraciques, Centre de Référence de l'Hypertension Pulmonaire Sévère, Hôpital de Bicêtre, Assistance Publique Hôpitaux de Paris, 78, rue du Général Leclerc, 94270 Le Kremlin-Bicêtre, France
    Search for more papers by this author
    • Dr. Montani has received consulting fees, speaking fees, and/or honoraria from Actelion, Bayer, GlaxoSmithKline, Eli Lilly, Novartis, Pfizer, and United Therapeutics (less than $10,000 each).


Abstract

Objective

Pulmonary venoocclusive disease (PVOD) is an uncommon form of pulmonary hypertension (PH) characterized by obstruction of small pulmonary veins. Pulmonary venous involvement has been reported in pathologic assessment of patients with systemic sclerosis (SSc) presenting with precapillary PH. High-resolution computed tomography (HRCT) of the chest is a noninvasive diagnostic tool used to screen for PVOD. No HRCT data are available on SSc patients with precapillary PH. We undertook this study to evaluate the frequency and effect on prognosis of HRCT signs of PVOD in SSc patients with precapillary PH.

Methods

We reviewed chest HRCT data from 26 SSc patients with precapillary PH and 28 SSc patients without pulmonary arterial hypertension (PAH) or interstitial lung disease (ILD).

Results

The radiographic triad of HRCT signs of PVOD (lymph node enlargement [57.7% versus 3.6%], centrilobular ground-glass opacities [46.2% versus 10.7%], and septal lines [88.5% versus 7.1%]) was significantly more frequent in SSc patients with precapillary PH than in SSc patients without PAH or ILD (all P < 0.005). Indeed, 61.5% of SSc patients with precapillary PH had ≥2 of these signs. Cardiomegaly (P < 0.0001), pulmonary artery enlargement (P < 0.0001), and pericardial effusion (P < 0.0005) were also significantly more frequent in SSc patients with precapillary PH. Pulmonary venous involvement was histologically confirmed in 2 patients with radiographic signs of PVOD. The presence of ≥2 radiographic signs of PVOD was associated with the occurrence of pulmonary edema after initiation of PAH-specific therapy (in 8 of 16 patients) and with more rapid progression from diagnosis of PH to death.

Conclusion

HRCT signs of PVOD are frequently observed in SSc patients with precapillary PH, correlated with histologic assessment, and were associated with a high risk of pulmonary edema.

Pulmonary arterial hypertension (PAH) is a severe and progressive disease, characterized by elevated pulmonary artery pressure leading to right-sided heart failure and death (1, 2). A definite diagnosis of PAH requires right-sided heart catheterization (RHC) showing a mean pulmonary artery pressure (PAP) of ≥25 mm Hg at rest and a pulmonary capillary wedge pressure (PCWP) of ≤15 mm Hg (3). PAH is divided into several subcategories according to the classification of pulmonary hypertension (PH) proposed in European Respiratory Society/European Society of Cardiology guidelines: idiopathic PAH, heritable PAH, and PAH associated with different conditions including congenital heart disease, human immunodeficiency virus, portal hypertension, exposure to drugs and/or toxins, and connective tissue diseases (CTDs) (4).

In CTDs, PAH is a devastating and often fatal complication, especially in patients with systemic sclerosis (SSc), which has the highest mortality of all rheumatic disorders (5–8). Prospective studies have shown that PAH complicates SSc with a prevalence of 8–12% (8, 9) and is the major cause of death in these patients (5). Disease duration >10 years (9), late age at onset of SSc (10), and the severity and duration of existing Raynaud's phenomenon (11) are markers that entail an increased risk of developing PAH in this condition. As emphasized in recent European Respiratory Society/European Society of Cardiology guidelines (4), the diagnosis of precapillary PH is confirmed by RHC. Furthermore, high-resolution computed tomography (HRCT) may also reveal radiographic abnormalities including central pulmonary artery dilatation, right ventricular hypertrophy, right ventricular and arterial enlargement, dilated bronchial arteries, and mosaic pattern of attenuation due to variable lung perfusion (12, 13).

HRCT may also suggest pulmonary venoocclusive disease (PVOD) including centrilobular ground-glass opacities, septal lines, and lymph node enlargement (14, 15). Low PaO2 at rest, reduced diffusing capacity for carbon monoxide (DLCO), and occult alveolar hemorrhage reinforce the hypothesis of underlying PVOD (15). PVOD is characterized by involvement of pulmonary veins, which may also explain the risk of development of pulmonary edema after starting treatment for PAH (16, 17). Even though definite diagnosis of PVOD requires histologic proof by lung biopsy (18), a noninvasive approach including HRCT, pulmonary function testing, and bronchoalveolar lavage is preferred because surgical lung biopsy is a high-risk procedure in these critically ill patients (15). A recent study by Dorfmüller and colleagues of postmortem lung histology of patients with CTD-associated PAH suggested a frequent PVOD-like involvement of the postcapillary vascular bed in assessed lung samples (19). The main objectives of our study were to assess the frequency of radiographic signs of PVOD using chest HRCT and to evaluate the effect of this radiographic pattern on the risk of pulmonary edema after initiation of specific therapy for PAH and on the survival of SSc patients with precapillary PH.

PATIENTS AND METHODS

Patients.

We reviewed data from chest HRCT of 64 patients with SSc including 34 SSc patients with precapillary PH admitted to the French Referral Centre for Pulmonary Hypertension and 30 SSc patients without PAH or interstitial lung disease (ILD) referred to the French Referral Centre for Systemic Sclerosis. We selected consecutive patients referred to our center for suspected PH. Exclusion criteria were the absence of PH, blurred chest HRCT scan, and lung fibrosis radiographically characterized by honeycombing, coarseness, and extensive reticular pattern. In the group of SSc patients with precapillary PH, PH was confirmed in all 27 female and 7 male patients by RHC.

We reviewed data from HRCT of the chest in 30 consecutive SSc patients without PAH or ILD. Patients had to fulfill the American College of Rheumatology preliminary criteria for SSc (20) and/or the LeRoy and Medsger criteria for early SSc (21). PAH was ruled out in these patients based on normal echocardiography findings. CT scan data of patients with associated pulmonary fibrosis (8 SSc patients with precapillary PH and 2 SSc patients without PAH or ILD) resulted in the exclusion of these patients from the study. Complete HRCT data of the chest were recorded and analyzed in 26 SSc patients with precapillary PH and 28 SSc patients without PAH or ILD.

Clinical assessment.

Clinical features were recorded from patients referred to the French Referral Centre for Pulmonary Hypertension and the French Referral Centre for Systemic Sclerosis. Records of all clinical characteristics at the time of diagnosis of SSc with precapillary PH were stored in the Registry of the French Network of Pulmonary Hypertension. This registry was set up in agreement with French bioethics laws (French Commission Nationale de l'informatique et des Libertés), and all patients gave their informed consent (22).

Functional and hemodynamic characteristics incorporated functional class defined by the New York Heart Association (NYHA) (23), 6-minute walk test, arterial blood gas determination, and pulmonary function testing including DLCO, performed in accordance with international recommendations of the American Thoracic Society (24) and the European Respiratory Society. Results were expressed as percentages of the normal predicted values (% predicted). Baseline hemodynamic measurements included mean right atrial pressure, mean PAP, and mean PCWP. Cardiac output was determined by the thermodilution method, and SvO2 was recorded. Pulmonary vascular resistances (PVRs) were calculated as (mean PAP − mean PCWP)/cardiac output and were expressed in Wood Units. All patients underwent acute vasodilator testing with inhaled nitric oxide (25, 26). Occurrence of pulmonary edema after initiation of specific therapy for PAH and survival status of patients were obtained from medical records and/or telephone interviews.

Diagnostic evaluation was completed by chest HRCT, which is recommended in the diagnosis of PAH and also plays a major role in the diagnostic evaluation of SSc (4). All patients admitted to both centers underwent chest HRCT to rule out pulmonary fibrosis (which is often encountered in the setting of SSc), chronic thromboembolic PH, and specific radiographic abnormalities in the setting of PH.

All patients underwent a 6-minute walk test, determination of arterial blood gases, pulmonary function tests, and hemodynamic testing. When patients were not able to perform the 6-minute walk test because of severe dyspnea at rest (NYHA functional class IV), we used a value of 0 meters for the distance walked. Furthermore, we compared characteristics of SSc patients with precapillary PH with radiographic indications of venous involvement on chest HRCT to characteristics of patients with idiopathic PVOD confirmed by histology as previously reported (15).

Grading HRCT of the chest.

HRCT of the chest was performed using a HiSpeed scanner (General Electric Medical Systems) at end-inspiration in patients in a supine position. Thin-section CT was done with 1-mm section thickness at 10-mm intervals. CT scans of the chests of all patients were reviewed in a blinded manner by 1 radiologist (SM) and 2 pulmonologists (SG and DM) for the presence of various pathologic features (i.e., mediastinal lymphadenopathy, ground-glass opacities, septal lines, cardiomegaly, and pleural or pericardial effusion). Final decisions on the radiographic findings were reached by consensus. Exclusion criteria for pulmonary fibrosis were based on chest HRCT findings as by the presence of honeycombing, coarseness, and an extensive reticular pattern. SSc patients with previously known severe pulmonary fibrosis were excluded from analysis. Ten patients with underlying lung fibrosis were excluded (see above).

Different regions for lymph node enlargement included the tracheobronchial, subcarinal, and mediastinal regions on the right or left side. Ground-glass opacity was defined as increased opacity of the lung parenchyma not sufficient to obscure pulmonary vessels. Patterns of ground-glass opacity were divided into 2 categories according to the lobular distribution. Panlobular distribution corresponds to geographic regions of lung attenuation with relatively well-defined borders that were either homogeneous or heterogeneous. Centrilobular distribution corresponded to poorly defined nodular opacities <10 mm in diameter, predominantly in superior or inferior position or diffuse. Septal lines were related to thickened interlobular septa (fine linear areas of attenuation or patterns of multiple polygonal lines) with smooth, irregular, or nodular borders. Further pathologic features assessed by chest HRCT included pericardial and/or pleural effusion as well as pulmonary artery enlargement, defined by a ratio of the diameter of the main pulmonary artery to that of the thoracic aorta >1.

Histologic analysis.

In 2 SSc patients with precapillary PH, the histologic specimens were obtained at autopsy from explanted lungs. The pathologic hallmark of pulmonary arteriopathy observed in PAH was defined as medial hypertrophy, intimal thickening, and plexiform lesions. The pathologic hallmark of venoocclusive disease was defined as an extensive and diffuse obstruction of pulmonary veins and venules by intimal fibrosis, and interstitial capillary multiplication. Histologic signs of PVOD were compared to HRCT signs of PVOD.

Statistical analysis.

Statistical analysis was performed using StatView software, version 5.0 (Abacus Concepts). Data are presented as the mean ± SD unless stated otherwise. Comparisons between SSc patients with precapillary PH and SSc patients without PAH or ILD were assessed by Student's t-test and chi-square test. The same statistical analyses were used for comparisons in patients with ≥2 radiographic signs of PVOD and ≤1 radiographic sign. Survival status was evaluated using Kaplan-Meier survival curves and the log rank test. P values less than or equal to 0.05 were considered significant.

RESULTS

Demographic and clinical characteristics of the patients.

Demographic and clinical characteristics of the SSc patients with precapillary PH (n = 26) and the SSc patients without PAH or ILD (n = 28) are shown in Table 1. The mean ± SD age of SSc patients with precapillary PH was 64.7 ± 11.5 years, while that of SSc patients without PAH or ILD was 55.9 ± 16.7 years. Both groups were predominantly female (84.6% in SSc patients with precapillary PH versus 85.7% in SSc patients without PAH or ILD). Limited SSc was found in 80.7% of SSc patients with precapillary PH and in 89.3% of SSc patients without PAH or ILD.

Table 1. Demographic and clinical characteristics of the SSc patients with precapillary PH and the SSc patients without PAH or ILD*
 SSc patients with precapillary PH (n = 26)SSc patients without PAH or ILD (n = 28)P
  • *

    Except where indicated otherwise, values are the mean ± SD. SSc = systemic sclerosis; PH = pulmonary hypertension; PAH = pulmonary arterial hypertension; ILD = interstitial lung disease; NYHA = New York Heart Association; FEV1 = forced expiratory volume in 1 second; TLC = total lung capacity; DLCO = diffusing capacity for carbon monoxide; mean PAP = mean pulmonary artery pressure; PCWP = pulmonary capillary wedge pressure; CI = cardiac index; CO = cardiac output; PVR = pulmonary vascular resistance.

No. of women/no. of men (sex ratio)22/4 (5.5)24/4 (6)0.91
Type of SSc, no. (%)  0.62
 Limited21 (80.7)25 (89.3) 
 Diffuse5 (19.2)3 (10.7) 
NYHA functional class, no. (%)  <0.0001
 I0 (0)18 (64.3) 
 II4 (15.4)8 (28.6) 
 III19 (73.1)2 (7.1) 
 IV3 (11.5)0 (0) 
Six-minute walk test, meters329 ± 89443 ± 121<0.0006
Arterial blood gases, mm Hg   
 PaO267 ± 1589 ± 10<0.0001
 PaCO231 ± 439 ± 4<0.0001
Pulmonary function testing, % predicted   
 FEV190 ± 21102 ± 210.0443
 TLC94 ± 17107 ± 170.0083
 DLCO38 ± 1168 ± 15<0.0001
Hemodynamic parameters   
 Mean PAP, mm Hg43 ± 11
 Right atrial pressure, mm Hg6 ± 4
 PCWP, mm Hg7 ± 3
 CI, liters/minute/m22.92 ± 0.75
 CO, liters/minute4.93 ± 1.44
 PVR, Wood Units8.3 ± 4.2
 SvO2, %62 ± 10

Functional and hemodynamic characteristics of the patients.

Results of NYHA functional class, 6-minute walk test, and pulmonary function tests including forced expiratory volume in 1 second, total lung capacity, and DLCO are shown in Table 1. The majority of SSc patients with precapillary PH (73.1%) were in NYHA functional class III as compared to only 2 SSc patients without PAH or ILD (7.1%). SSc patients with precapillary PH walked 329 ± 89 meters in 6 minutes, while SSc patients without PAH or ILD walked 443 ± 121 meters (P < 0.006). Results of arterial blood gases at room air were as follows: PaO2 67 ± 15 mm Hg and PaCO2 31 ± 4 mm Hg in SSc patients with precapillary PH versus PaO2 89 ± 10 mm Hg and PaCO2 39 ± 4 mm Hg in SSc patients without PAH or ILD (both P < 0.0001). DLCO was significantly lower in SSc patients with precapillary PH (% predicted 38 ± 11 versus 68 ± 15; P < 0.0001). In SSc patients with precapillary PH, RHC confirmed severe precapillary PH (mean PAP 43 ± 11 mm Hg, cardiac index 2.92 ± 0.75 liters/minute/m2, PVR 8.3 ± 4.2 Wood Units).

Characteristics of SSc patients presenting with PH compared to patients with idiopathic PVOD.

Clinical and functional characteristics of SSc patients presenting with ≥2 radiographic signs of PVOD were broadly similar to those reported in our cohort of patients with histologically confirmed idiopathic PVOD. Patients with idiopathic PVOD have slightly worse hemodynamic parameters but are in slightly better NYHA functional classes as compared to SSc patients presenting with precapillary PH and suspected PVOD.

Analysis of HRCT of the chest in SSc patients with precapillary PH and SSc patients without PAH or ILD.

Chest HRCT findings in these 2 groups were compared. Lymph node enlargement was observed significantly more frequently in SSc patients with precapillary PH (P < 0.0001) (Table 2). Lymphadenopathy was mostly observed in the subcarinal region (50%) or tracheobronchial region (26.9%). Parenchymal abnormalities with a pattern of centrilobular ground-glass opacities were mostly diffuse (23.1%) or localized in the superior region (19.2%). Centrilobular ground-glass opacities were more frequently observed in SSc patients with precapillary PH (46.2%) than in SSc patients without PAH or ILD (10.7%) (P < 0.001). Septal lines were observed in 88.5% of SSc patients with precapillary PH and in 7.1% of SSc patients without PAH or ILD (P < 0.0001). In addition, some radiographic signs of PAH, including cardiomegaly (P < 0.0001), pulmonary artery enlargement (P < 0.0001), and pericardial effusion (P < 0.0005), were found significantly more frequently in SSc patients with precapillary PH. More than half of all SSc patients with precapillary PH (53.8%) had pericardial effusions compared with only 2 SSc patients without PAH or ILD (7.1%). A pleural effusion was observed in 1 SSc patient with precapillary PH (3.8%), while no SSc patients without PAH or ILD had evidence of pleural effusion.

Table 2. Radiographic characteristics on HRCT in SSc patients with precapillary PH and SSc patients without PAH or ILD*
 SSc patients with precapillary PH (n = 26)SSc patients without PAH or ILD (n = 28)P
  • *

    Values are the number (%). HRCT = high-resolution computed tomography; PA = pulmonary artery; PV = pulmonary vein (see Table 1 for other definitions).

Lymph node enlargement15 (57.7)1 (3.6)<0.0001
 Tracheobronchial lymphadenopathy7 (26.9)0 (0)<0.0112
 Subcarinal13 (50)1 (3.6)<0.0003
 Right hilar side2 (7.7)0 (0)0.4389
 Left hilar side3 (11.5)0 (0)0.2095
 Hilar bilateral4 (15.4)0 (0)0.1018
 Other4 (15.4)0 (0)0.1018
Parenchymal abnormalities   
 Panlobular ground-glass opacities4 (15.4)2 (7.1)0.5968
  Homogeneous3 (11.5)0 (0)0.2095
  Heterogeneous1 (3.8)2 (7.1)0.9466
 Centrilobular ground-glass opacities12 (46.2)3 (10.7)<0.001
  Superior5 (19.2)3 (10.7)0.9697
  Inferior1 (3.8)0 (0)0.9697
  Diffuse6 (23.1)0 (0)<0.0236
 Mosaic attenuation pattern0 (0)0 (0)
Septal lines23 (88.5)2 (7.1)<0.0001
Nodes11 (42.3)14 (50)0.7695
Other abnormalities   
 Cardiomegaly23 (88.5)2 (7.1)<0.0001
 PA enlargement22 (84.6)6 (21.4)<0.0001
 PV enlargement3 (11.5)0 (0)0.2095
 Pericardial effusion14 (53.8)2 (7.1)<0.0005
 Pleural effusion1 (3.8)0 (0)0.9697

In 16 SSc patients with precapillary PH (61.5%), we found 2 or 3 radiographic signs of PVOD on chest HRCT (i.e., septal lines, centrilobular ground-glass opacities, and lymph node enlargement). No differences were observed between the 2 groups regarding the type of SSc. However, we noted that patients with ≥2 radiographic signs of PVOD were exclusively in NYHA functional class III or IV and therefore had more severe disease than patients with ≤1 radiographic sign of PVOD. Patients with ≥2 radiographic signs of PVOD did significantly worse on the 6-minute walk test (294 ± 85 meters) than did patients with ≤1 radiographic sign of PVOD (381 ± 70 meters) (P = 0.013). In addition, these patients had more severe hypoxemia (PaO2 62 ± 12 mm Hg versus 76 ± 16 mm Hg; P = 0.016). DLCO was lower in patients with ≥2 radiographic signs of PVOD (% predicted 34 ± 9) than in patients with ≤1 radiographic sign of PVOD (% predicted 44 ± 11) (P = 0.019). Hemodynamic parameters in patients with ≥2 radiographic signs of PVOD were more severe in regard to mean PAP (48 ± 10 mm Hg versus 37 ± 11 mm Hg; P = 0.014) and PVR (9.8 ± 4.2 Wood Units versus 5.9 ± 4.2 Wood Units; P = 0.020) (Table 3). Of note, 2 patients presenting ≥2 radiographic signs of PVOD underwent bronchoscopy, and analysis of bronchoalveolar lavage fluid showed occult alveolar hemorrhage, defined by a Golde score >100.

Table 3. Comparison of characteristics of SSc patients with precapillary PH suspected of having PVOD*
 SSc patients with precapillary PH with ≥2 signs of PVOD on HRCT (n = 16)SSc patients with precapillary PH with ≤1 sign of PVOD on HRCT (n = 10)P
  • *

    Except where indicated otherwise, values are the mean ± SD. PVOD = pulmonary venoocclusive disease; HRCT = high-resolution computed tomography (see Table 1 for other definitions).

Sex  0.97
 No. of women139 
 No. of men31 
Type of SSc, no. (%)  0.94
 Limited13 (81.3)8 (80) 
 Diffuse3 (18.7)2 (20) 
NYHA functional class, no. (%)  0.013
 II0 (0)4 (40) 
 III13 (81.3)6 (60) 
 IV3 (18.7)0 (0) 
Six-minute walk test, meters294 ± 85381 ± 700.013
Arterial blood gases, mm Hg   
 PaO262 ± 1276 ± 160.016
 PaCO231 ± 432 ± 40.373
Pulmonary function testing, % predicted   
 FEV187 ± 2395 ± 190.359
 TLC91 ± 1998 ± 150.331
 DLCO34 ± 944 ± 110.019
Hemodynamic parameters   
 Mean PAP, mm Hg48 ± 1037 ± 110.014
 Right atrial pressure, mm Hg7 ± 55 ± 40.372
 PCWP, mm Hg7 ± 38 ± 40.559
 CI, liters/minute/m22.8 ± 0.83.1 ± 0.70.256
 CO, liters/minute4.6 ± 1.35.5 ± 1.60.138
 PVR, Wood Units9.8 ± 4.25.9 ± 4.20.020
 SvO2, %60 ± 1166 ± 60.129
Pulmonary edema with PAH-specific therapy, no. (%)8 (50)0 (0)<0.05

Comparisons between histologic analyses and HRCT signs of PVOD.

Lung samples from 2 SSc patients with precapillary PH were available for histologic analysis after lung transplantation. Both cases displayed similar vascular and interstitial remodeling. In addition to arterial medial thickening and concentric intimal fibrosis, we found impressive lesions within the capillary and the postcapillary vasculature. Septal veins and smaller preseptal venules frequently presented intimal fibrosis with considerable reduction of the native lumen (Figures 1B and C and 2B and C) and fibrous broadening of the septa. Moreover, patchy foci of interstitial capillary multiplication with consequently thickened alveolar septa were observed and were associated with small, strongly remodeled arterioles and venules (Figures 1D and 2D). Intraalveolar accumulation of siderin-laden macrophages was noted (Figure 2D).

Figure 1.

Radiographic and histologic description of venous involvement in a lung sample from a female systemic sclerosis patient with precapillary pulmonary hypertension. A, High-resolution computed tomography of the chest showing the presence of septal lines and ground-glass opacities. B, Septal vein displaying intimal occlusive fibrosis. C, Small preseptal venule with loose, lumen-narrowing concentric fibrosis. D, Interstitial capillary multiplication (hemangiomatosis-like) associated with muscularized arterioles. Original magnification × 100 in B and D; × 200 in C.

Figure 2.

Radiographic and histologic description of venous involvement in a lung sample from a male systemic sclerosis patient with precapillary pulmonary hypertension. A, High-resolution computed tomography of the chest showing marked radiographic signs of pulmonary venoocclusive disease including septal lines and ground-glass opacities. B, Septal vein with typical fibrotic intimal remodeling. C, Nearly occlusive preseptal venule. D, Hemangiomatosis-like pattern with thickening of alveolar septa and hypertrophic arterioles. Note numerous intraalveolar macrophages. Original magnification × 100 in B and D; × 200 in C.

Occurrence and management of pulmonary edema after initiation of specific therapy for PAH.

Diagnosis of pulmonary edema was based on physical examination and chest radiography. Patients with ≥2 radiographic signs of PVOD on chest HRCT are highly susceptible to developing pulmonary edema after starting PAH-specific drug therapy. Eight of 16 patients (50%) presenting ≥2 signs of PVOD developed pulmonary edema. Four patients were treated with endothelin receptor antagonists, 3 patients were treated with prostacyclin or its analogs, and 1 patient was treated with a phosphodiesterase V inhibitor. The median time between starting PAH-specific treatment and pulmonary edema was 70 days (range 31–394 days). Management of pulmonary edema included high-dose intravenous diuretics in all patients, withdrawal of PAH-specific treatment in 3 patients (37.5%), and dose reduction of intravenous prostacyclin in 1 patient. Management in 2 patients required dobutamine in an intensive care unit. After acute episodes, PAH-specific therapy was continued but dosage was changed in all patients. Four patients received inhaled prostacyclin, 3 patients received intravenous prostacyclin (2 received epoprostenol and 1 received treprostinil), and 2 patients were listed for urgent heart/lung transplantation. One of these patients underwent transplantation immediately after referral to our center, and the other patient's illness was managed successfully for 135 days by bridge therapy with intravenous prostacyclin until heart/double-lung transplantation.

To avoid pulmonary edema, intravenous epoprostenol is initiated in the intensive care unit, with slowly increasing dosage and high-dose diuretics. Before discharge of patients, the 6-minute walk test, blood gas analyses, and chest radiography are performed to rule out pulmonary edema.

Survival in SSc patients with precapillary PH according to HRCT signs of PVOD.

Patients presenting with ≥2 radiographic signs of PVOD had more rapid progression to death or lung transplantation. After 18 months of followup, 7 patients with high radiographic suspicion of venous involvement (≥2 radiographic signs of PVOD) had died, while all patients with ≤1 radiographic sign of PVOD were still alive. At 3 years of followup, 2 and 3 patients, respectively, were still alive in each group. Results of the Kaplan-Meier survival curve are shown in Figure 3.

Figure 3.

Comparison of followup in 16 systemic sclerosis (SSc) patients with precapillary pulmonary hypertension (PH) strongly suspected of having pulmonary venoocclusive disease (PVOD) (≥2 radiographic signs of PVOD on high-resolution computed tomography [HRCT]) versus followup in 10 SSc patients with precapillary PH having ≤1 radiographic sign of PVOD on HRCT. The x-axis shows the duration of followup in months. The box below the graph shows the number of surviving patients in the 2 groups at each followup time point. Cum. = cumulative.

DISCUSSION

In this study, chest HRCT showed that radiographic characteristics of PVOD such as lymph node enlargement, centrilobular ground-glass opacities, and septal lines are frequently encountered in SSc patients with precapillary PH. To our knowledge, there are no larger series specifically reporting radiographic signs of PVOD in the course of CTD. In a recently conducted study (15), we showed that the presence of radiographic abnormalities has a sensitivity of 75% and a specificity of 84.6% for the detection of PVOD. Of 26 evaluated SSc patients with precapillary PH in our study, 16 (61.5%) presented ≥2 radiographic signs of PVOD.

There have been some case reports of venous involvement in SSc patients with precapillary PH (27–30). Johnson and colleagues (29) assessed signs of PVOD in 4 women with SSc and precapillary PH. All these women presented the radiographic triad of PVOD (lymph node enlargement, centrilobular ground-glass opacities, and septal lines); 3 of them also had pleural and pericardial effusions. In 1 case, PVOD was confirmed by autopsy. Case reports from Morassut and colleagues (27) as well as from Saito and colleagues (28) confirmed existing venous involvement in SSc patients with precapillary PH by autopsy. In all reported cases, female sex was predominant, as in our study cohort. Furthermore, we found that radiographic signs of PAH (i.e., pulmonary artery enlargement and cardiomegaly, predominantly in the right ventricle) are observed significantly more often in SSc patients with precapillary PH than in SSc patients without PAH or ILD. Our findings therefore highlight the importance of chest HRCT as a noninvasive approach in the setting of PAH (13, 31). Nevertheless, clinicians should be aware of the possibility of underlying PAH without existing radiographic abnormalities on HRCT.

To our knowledge, there are no reported data of survival in SSc patients with precapillary PH with suspected pulmonary venous involvement. Recently presented data from 24 patients with histologically confirmed idiopathic PVOD suggest that these patients had a worse prognosis than patients with idiopathic PAH (15). Indeed, the survival rate in SSc patients with precapillary PH is poor (7), estimated to be 47% 3 years after diagnosis (32). However, the role of venous involvement as an underlying complication in SSc patients with precapillary PH was not systematically investigated. If venous involvement is suspected in SSc patients with precapillary PH, prognosis remains dismal. Survival rates were especially poor in our patients with ≥2 signs of PVOD on HRCT. Before the availability of PAH-specific therapy, the 1-year survival rate of SSc patients with precapillary PH was estimated to be 45% (33). The availability of new therapies including prostanoids, endothelin receptor antagonists, and phosphodiesterase V inhibitors and the established national screening program in patients with CTD have not modified overall survival (34). Hachulla and colleagues (35) recently showed an estimated 3-year survival rate of ∼50% in SSc patients with precapillary PH. Survival in these patients depends also on the severity of hemodynamic parameters. Unfortunately, in all reported studies, the role of venous involvement as an underlying complication in SSc with precapillary PH was not systematically analyzed. However, survival data in our study were based only on a small cohort of patients and should therefore be considered with caution. Nevertheless, our results are similar, when compared to results in patients with idiopathic PVOD, regarding poor prognosis and moderate response to PAH-specific therapy (15).

Dorfmüller and colleagues (19) highlighted a high rate of postcapillary involvement in CTD-associated PAH. They identified a significant obstructive vasculopathy predominantly involving the veins and preseptal venules in 6 of 8 CTD patients (4 with SSc, 2 with systemic lupus erythematosus [SLE], 1 with mixed CTD [MCTD] with PAH, and 1 with rheumatoid arthritis). This supports the hypothesis that the pulmonary venous system is very frequently involved in CTD-associated PH and explains the occurrence of pulmonary edema after initiation of therapy for PAH in this subtype of PH (16, 36, 37).

We reported that half of the SSc patients with associated PH who presented with ≥2 radiographic signs of PVOD developed pulmonary edema after initiation of PAH-specific treatment. As previously reported in idiopathic PVOD, this complication was not related to one class of drugs but may be observed with all PAH-specific therapies. The mechanism is thought to be the greater vasodilatation of the precapillary resistance vessels relative to the pulmonary capillaries and veins, associated with an increase in blood flow, leading to an increase in transcapillary hydrostatic pressure and transudation of fluid into the pulmonary interstitium and alveoli (38). Interestingly, none of the patients with ≤1 radiographic sign of PVOD developed pulmonary edema in the course of the disease. Thus, HRCT should be considered a useful diagnostic tool in the setting of SSc with precapillary PH to predict the risk of pulmonary edema after initiation of PAH-specific therapy, thereby enabling cautious use of these drugs in patients considered at high risk of pulmonary edema (presenting ≥2 radiographic signs of PVOD).

Recently reported data (39) showed a frequency of pleural effusions of 39% in patients with CTD-related PAH. In contrast to these results, we found that only 1 SSc patient with precapillary PH (3.8%) had a medical history of pleural effusion. Our data are consistent with results obtained by Wiener-Kronish and colleagues (40) showing the absence of pleural effusions in 27 patients with PAH. Nonetheless, we observed cardiac involvement with the presence of pericardial effusions in more than half of all evaluated SSc patients with precapillary PH (53.8%). These results are in accordance with previous findings in patients with idiopathic PVOD and in SSc patients with precapillary PH (32). Regarding the underlying type of SSc, 5 SSc patients with precapillary PH had a medical history of diffuse SSc. PAH is more often associated with the limited cutaneous form of SSc (9, 10, 41). Pulmonary fibrosis is encountered in almost 75% of patients with the diffuse cutaneous form of SSc (42, 43). The high prevalence of the limited cutaneous form of SSc in our study cohort is probably due to our excluding patients with pulmonary fibrosis.

Regarding the clinical performance of SSc patients with precapillary PH with ≥2 radiographic signs of PVOD, we noted reduced DLCO on pulmonary function tests and hypoxia on arterial blood gas determinations. These patients were exclusively in NYHA functional class III or IV. In addition, they had more severe hemodynamic parameters than patients with ≤1 radiographic sign of PVOD. These results are in accordance with our previous findings in histologically confirmed idiopathic PVOD (15) and with findings of Chung and coauthors, who confirmed lower DLCOand 6-minute walk test values in patients with CTD-related PAH as compared to patients with idiopathic PAH (32). Those authors found that severely reduced DLCO is specific to SSc patients with precapillary PH in comparison to other forms of CTD like SLE with PAH or MCTD with PAH. Authors of previous studies (9, 10) proposed that decreasing DLCO predicts the development of PAH in limited and diffuse SSc. Unfortunately, the role of alteration of the pulmonary venous system assessed by chest HRCT was not systematically reviewed in those studies. The comparison of patients with idiopathic PVOD to SSc patients presenting with precapillary PH showed similar clinical and hemodynamic parameters, suggesting similar pathophysiologic mechanisms. These observations reinforce suspicion of underlying venous involvement in SSc patients presenting with PH.

In our study, patients with ILD were excluded from analysis because of the difficulty of being confident in assessing radiographic signs of PVOD in this context. Johnson and coauthors reported that it was difficult to distinguish with certainty between CTD-associated PVOD and early interstitial lung fibrosis (29). Even if the data in the present study are limited, ours is the largest cohort of SSc patients with precapillary PH or SSc patients without PAH or ILD in whom radiographic signs of PH and PVOD were systematically reviewed. Indeed, we chose consecutive patients in a determined period to evaluate prognosis and limit selection bias. Of note, pathologic examination of the lung in 2 patients with lymph node enlargement, centrilobular ground-glass opacities, and septal lines (Figures 1A and 2A) formally confirmed the diagnosis of PVOD.

In conclusion, the presence of precapillary PH in SSc patients was significantly associated with radiographic abnormalities of chest HRCT including cardiomegaly, enlargement of pulmonary arteries, pericardial effusion, and signs of PVOD. Indeed, radiographic signs of PVOD, including the triad of lymph node enlargement, centrilobular ground-glass opacities, and septal lines, were frequently observed in SSc patients with precapillary PH and were highly suspicious for underlying venous involvement in this setting. Furthermore, in SSc with precapillary PH, the presence of ≥2 radiographic signs of PVOD identified a subgroup of patients with high risk of pulmonary edema after initiation of PAH-specific therapy and with poor outcome.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Montani 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 conception and design. Günther, Jaïs, Maitre, Dorfmüller, Humbert, Montani.

Acquisition of data. Günther, Jaïs, Maitre, Bérezné, Dorfmüller, Seferian, Savale, Mercier, Fadel, Simonneau, Humbert, Montani.

Analysis and interpretation of data. Günther, Jaïs, Maitre, Dorfmüller, Seferian, Savale, Mercier, Fadel, Sitbon, Mouthon, Simonneau, Humbert, Montani.

Acknowledgements

We thank Dr. Robert L. Owen (Professor of Medicine, University of California, San Francisco) for providing editorial assistance in preparing the manuscript.

Ancillary