Clinical phenotypes and outcomes of heritable and sporadic pulmonary veno-occlusive disease: a population-based study

Articles

Clinical phenotypes and outcomes of heritable and sporadic pulmonary veno-occlusive disease: a population-based study David Montani*, Barbara Girerd*, Xavier Jaïs, Marilyne Levy, David Amar, Laurent Savale, Peter Dorfmüller, Andrei Seferian, Edmund M Lau, Mélanie Eyries, Jérôme Le Pavec, Florence Parent, Damien Bonnet, Florent Soubrier, Elie Fadel, Olivier Sitbon, Gérald Simonneau, Marc Humbert

Summary Background Bi-allelic mutations of the EIF2AK4 gene cause heritable pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis (PVOD/PCH). We aimed to assess the effect of EIF2AK4 mutations on the clinical phenotypes and outcomes of PVOD/PCH. Methods We did a population-based study using clinical, functional, and haemodynamic data from the registry of the French Pulmonary Hypertension Network. We reviewed the clinical data and outcomes from all patients referred to the French Referral Centre (Pulmonary Department, Hospital Kremlin-Bicêtre, University Paris-Sud) with either confirmed or highly probable PVOD/PCH with DNA available for mutation screening (excluding patients with other risk factors of pulmonary hypertension, such as chronic respiratory diseases). We sequenced the coding sequence and intronic junctions of the EIF2AK4 gene, and compared clinical characteristics and outcomes between EIF2AK4 mutation carriers and non-carriers. Medical therapies approved for pulmonary arterial hypertension (prostacyclin derivatives, endothelin receptor antagonists and phosphodiesterase type-5 inhibitors) were given to patients according to the clinical judgment and discretion of treating physicians. The primary outcome was the event-free survival (death or transplantation). Secondary outcomes included response to therapies for pulmonary arterial hypertension and survival after lung transplantation. A satisfactory clinical response to specific therapy for pulmonary arterial hypertension was defined by achieving New York Heart Association functional class I or II, a 6-min walk distance of more than 440 m, and a cardiac index greater than 2·5 L/min per m² at the first reassessment after initiation of specific therapy for pulmonary arterial hypertension. Findings We obtained data from Jan 1, 2003, to June 1, 2016, and identified 94 patients with sporadic or heritable PVOD/PCH (confirmed or highly probable). 27 (29%) of these patients had bi-allelic EIF2AK4 mutations. PVOD/ PCH due to EIF2AK4 mutations occurred from birth to age 50 years, and these patients were younger at presentation than non-carriers (median 26·0 years [range 0–50.3] vs 60·0 years [6·7–81·4] years; p<0·0001). At diagnosis, both mutations carriers and non-carriers had similarly severe precapillary pulmonary hypertension and functional impairment. 22 (81%) of mutations carriers and 63 (94%) of non-carriers received therapy approved for pulmonary arterial hypertension. Drug-induced pulmonary oedema occurred in five (23%) of treated EIF2AK4 mutations carriers and 13 (21%) of treated non-carriers. Follow-up assessment after initiation of treatment showed that only three (4%) patients with PVOD/PCH reached the predefined criteria for satisfactory clinical response. The probabilities of event-free survival (death or transplantation) at 1 and 3 years were 63% and 32% in EIF2AK4 mutations carriers, and 75% and 34% in non-carriers. No significant differences occurred in event-free survival between the 2 groups (p=0·38). Among the 33 patients who had lung transplantation, estimated posttransplantation survival rates at 1, 2, and 5 years were 84%, 81%, and 73%, respectively. Interpretation Heritable PVOD/PCH due to bi-allelic EIF2AK4 mutations is characterised by a younger age at diagnosis but these patients display similar disease severity compared with mutation non-carriers. Response to therapy approved for pulmonary arterial hypertension in PVOD/PCH is rare. PVOD/PCH is a devastating condition and lung transplantation should be considered for eligible patients. Funding None.

Introduction Precapillary pulmonary hypertension is defined at rightheart catheterisation by a sustained increase in mean pulmonary artery pressure of at least 25 mm Hg and a normal pulmonary artery wedge pressure of no more than 15 mm Hg.1 Precapillary pulmonary hypertension can be heritable in the context of pulmonary arterial hypertension, pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis (PVOD/PCH).1

Pulmonary arterial hypertension is an uncommon cause of pulmonary hypertension characterised by proliferative remodelling and fibrosis of the small pulmonary arteries.1,2 Heritable pulmonary arterial hypertension can develop in patients carrying heterozygous mutations of the BMPR2 gene3 and other less common heterozygous gene mutations (such as ACVRL1, ENG, CAV-1, and KCNK3).4–6 Patients with pulmonary arterial hypertension and BMPR2 mutations present at a younger age with more severe

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

Lancet Respir Med 2017 Published Online January 10, 2017 http://dx.doi.org/10.1016/ S2213-2600(16)30438-6 See Online/Comment http://dx.doi.org/10.1016/ S2213-2600(16)30466-0 *Contributed equally University Paris–Sud, Faculté de Médecine, Paris, F-94270, France (D Montani MD, B Girerd PhD, X Jaïs MD, D Amar MD, L Savale MD, P Dorfmüller MD, A Seferian MD, E M Lau MD, F Parent MD, O Sitbon MD, G Simonneau MD, Prof M Humbert MD); AP-HP, Centre de Référence de l’Hypertension Pulmonaire Sévère, Département Hospitalo-Universitaire (DHU) Thorax Innovation (TORINO), Service de Pneumologie, Hôpital de Bicêtre, Le Kremlin Bicêtre, Paris, France (D Montani, B Girerd, X Jaïs, D Amar, L Savale, P Dorfmüller, A Seferian, E M Lau, F Parent, O Sitbon, G Simonneau, Prof M Humbert); UMR_S 999, Univ. Paris–Sud, INSERM, Laboratoire d’Excellence (LabEx) en Recherche sur le Médicament et l’Innovation Thérapeutique (LERMIT), Centre Chirurgical Marie Lannelongue, Le Plessis Robinson, Paris, France (D Montani, B Girerd, X Jaïs, D Amar, L Savale, P Dorfmüller, A Seferian, F Parent, G Simonneau, Prof M Humbert); M3C-Necker, Reference Centre for Complex Congenital Heart Diseases, Hôpital Universitaire Necker-Enfants malades, AP-HP, Université Paris Descartes, Paris, France (M Levy MD, D Bonnet MD); Centre Chirurgical Marie Lannelongue, Service de Chirurgie Thoracique, Vasculaire et Transplantation Cardiopulmonaire, Le Plessis Robinson, Paris, France (P Dorfmüller, J Le Pavec MD, Elie Fadel MD); Sydney Medical School, University of Sydney; Royal Prince Alfred Hospital,

1

Articles

Camperdown, VIC, Australia (E M Lau); and Département de Génétique, Hôpital Pitié Salpétrière, AP-HP, UMR_ S1166-ICAN; INSERM and UPMC Sorbonne Universités, Paris, France (M Eyries PhD, F Soubrier MD) Correspondence to: Prof Marc Humbert, Université Paris-Sud, Centre de Référence de l’Hypertension Pulmonaire Sévère, Service de Pneumologie, Hôpital Bicêtre, 94270 Le Kremlin-Bicêtre, Paris, France. [email protected]

Research in context Evidence before this study Pulmonary veno-occlusive disease (PVOD) and pulmonary capillary haemangiomatosis (PCH) are rare causes of pulmonary hypertension. Currently, PVOD and PCH are believed to represent spectrums of a common disease entity (PVOD/PCH). In 2014, bi-allelic mutations of the EIF2AK4 gene were shown to be a cause of heritable PVOD/PCH, but no information was available about whether the clinical phenotype and outcomes of heritable PVOD/PCH due to bi-allelic EIF2AK4 mutations are different to cases of sporadic disease.

non-carrier group. Both groups displayed similar pulmonary haemodynamic characteristics indicative of severe precapillary pulmonary hypertension, functional class impairment, and high-resolution CT findings. In the overall population with PVOD/ PCH, treatment with drugs for pulmonary arterial hypertension led to mild haemodynamic and functional improvement but this was associated with a significant risk of pulmonary oedema (21%). Transplantation-free survival for both EIF2AK4 mutations carriers and non-carriers was equally poor, with 32% and 34% of patients, respectively, who were event free at 3 years.

Added value of this study Our study provides the first systematic assessment of the effect of bi-allelic EIF2AK4 mutations in a large cohort of patients with PVOD/PCH at the French Referral Centre. We identified 27 patients with PVOD/PCH with EIF2AK4 mutations and 67 patients with sporadic PVOD/PCH. Bi-allelic EIF2AK4 mutation carriers were characterised by younger age at diagnosis with an equal sex ratio, whereas more male patients were affected in the

Implications of all the available evidence Patients with PVOD/PCH with bi-allelic EIF2AK4 mutations are substantially younger at diagnosis but have similarly severe disease in terms of haemodynamic characteristics and functional impairment, compared with patients without EIF2AK4 mutations. Sustained response to drugs for pulmonary arterial hypertension was not observed, resulting in the need for early lung transplantation in eligible patients.

disease, and are at increased risk of death or transplantation, compared with those without BMPR2 mutations.7–9 An autosomal recessive form of heritable precapillary pulmonary hypertension due to mutations of the EIF2AK4 gene (coding for eukaryotic translation initiation factor 2 α kinase 4) has been identified.10–12 Histological examination of the lungs from carriers of bi-allelic EIF2AK4 mutations has revealed extensive occlusion of pulmonary veins by fibrous tissue, intimal thickening of venules and small veins in the lobular septa, and localised capillary proliferation.10,13 These histological features correspond to the clinical entities of PVOD and PCH, which are believed to be manifestations of the same rare underlying condition (lowest estimate of prevalence is <1 case per million).1,10,12,13 The EIF2AK4 gene is the only gene identified in heritable PVOD/PCH. Beside heritable forms, environmental risk factors such as occupational organic solvent exposure and chemotherapy have also been associated with the development of PVOD/PCH.14–16 Distinguishing PVOD/PCH from pulmonary arterial hypertension on clinical grounds can be challenging, since the physical and haemodynamic findings for all three diseases are broadly similar, and PVOD/PCH might represent 5%–10% of cases initially thought to be idiopathic pulmonary arterial hypertension.13,17–19 An important clinical hallmark of PVOD/PCH is the possible occurrence of pulmonary oedema induced by medical therapies approved for pulmonary arterial hypertension.13,17,20 On the basis of these findings, PVOD/PCH are classified as a distinct subgroup in the updated clinical classification of pulmonary hypertension.1,21 Since PVOD/PCH can result from bi-allelic EIF2AK4 mutations in its heritable form or from environmental factors, the possibility is raised that EIF2AK4 mutation 2

status might be associated with a distinct phenotype of PVOD/PCH, as shown in patients with pulmonary arterial hypertension and BMPR2 mutations. To test this hypothesis, we obtained data from all adult and paediatric patients with PVOD/PCH in whom EIF2AK4 mutations were screened for and referred to the French Referral Centre for Severe Pulmonary Hypertension. Clinical, functional, and haemodynamic characteristics and outcomes were compared between patients with heritable PVOD/PCH with bi-allelic EIF2AK4 mutations and patients with PVOD/PCH without identifiable EIF2AK4 mutations.

Methods Study design and participants We did a population-based study using clinical, functional, and haemodynamic data from the registry of the French Pulmonary Hypertension Network,2 which includes all patients with pulmonary hypertension at the French Referral Centre for Severe Pulmonary Hypertension (Hôpital Bicêtre, Université Paris-Sud, Le Kremlin-Bicêtre, France) and the 22 associated centres across France. The French Referral Centre offers genetic counselling and mutation screening to all patients diagnosed with idiopathic, familial, or drug-induced pulmonary arterial hypertension, and to all patients with familial or sporadic PVOD/PCH. All patients or their legal representatives signed written informed consent, and all patients underwent genetic counselling before mutation screening. We reviewed the clinical data and outcomes from all patients with either confirmed or highly probable PVOD/ PCH who had DNA available and provided consent for mutation testing. PVOD/PCH was considered as confirmed in the presence of bi-allelic mutations of the

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

Articles

EIF2AK4 gene or extensive and diffuse obstruction of pulmonary veins and venules on lung biopsies, postmortem lung samples, or lung explants obtained after lung transplantation. PVOD/PCH was considered highly probable when patients fulfilled the following criteria: precapillary pulmonary hypertension confirmed by right-heart catheterisation, presence of two or more radiological abnormalities characteristic of PVOD/PCH on high-resolution CT of the chest (centrilobular groundglass opacities, interlobular septal lines, or lymph node enlargement), and low diffusing capacity for carbon monoxide (DLCO <60% predicted). We have previously validated that these clinical criteria provide high accuracy for the diagnosis of PVOD/PCH, since lung biopsy is contraindicated in the setting of severe pulmonary hypertension.13 Clinical characteristics at diagnosis and follow-up were stored in the Registry of the French Pulmonary Hypertension Network.2 This registry was set up in agreement with French bioethics laws (Commission Nationale de l’Informatique et des Libertés). Patients with PVOD/PCH associated with connective tissue disease or HIV infection, patients with spirometric impairment (FEV1/FVC <0·7 and FEV1 <60% predicted), and patients with lung fibrosis or other interstitial lung diseases on high-resolution CT were excluded.1

Screening of BMPR2 and EIF2AK4 mutations Since January, 2003, DNA from peripheral blood of consenting patients with heritable and sporadic PVOD/ PCH were collected and stored. Genetic variations of BMPR2 gene sequences were screened in all patients, as previously described.7 Since 2014, all patients with confirmed and highly probable PVOD/PCH have been retrospectively and prospectively screened for EIF2AK4 mutations. The entire coding sequence and intronic junctions of the EIF2AK4 gene were PCR-amplified using specific oligonucleotide primer pairs and subjected to bidirectional Sanger sequencing. We analysed the resulting sequence data with SeqScape software, version 2.5 (Applied Biosystems) in comparison with the EIF2AK4 reference sequence (NM_001013703). Nucleotide numbering reflects cDNA numbering with +1 corresponding to the A of the ATG translation initiation codon in the reference sequence according to the Human Genome Variation Society recommendations.

Follow-up and clinical outcomes Medical therapies approved for pulmonary arterial hypertension (prostacyclin derivatives, endothelin receptor antagonists and phosphodiesterase type-5 inhibitors) were given to patients according to the clinical judgment and discretion of individual treating physicians.1,22,23 Clinical, functional, and haemodynamic follow-up data were collected in the registry of the French Pulmonary Hypertension Network. Based on risk assessment for

pulmonary arterial hypertension in the European Society of Cardiology (ESC)/European Respiratory Society (ERS) 2015 guidelines,1 a satisfactory clinical response to specific therapy for pulmonary arterial hypertension was defined by achieving New York Heart Association (NYHA) functional class I or II, 6-min walk distance of more than 440 m,24 and a cardiac index greater than 2·5 L/min per m² at the first reassessment after initiation of specific therapy for pulmonary arterial hypertension.1 Occurrence of pulmonary oedema induced by pulmonary arterial hypertension therapy was prospectively reported in the data files of these patients and collected in the French registry. Pulmonary oedema was defined by the association of clinical deterioration, worsening hypoxaemia, and radiological changes. Time to death or lung transplantation and outcomes and survival after lung transplantation were recorded.

Statistical analysis Continuous data with normal distribution are presented as mean (SD). We used unpaired t tests to compare data between mutation carriers and non-carriers using pooled variance or Satherwaith’s method depending on equality of variances between groups. We used paired t tests to compare responses to specific pulmonary arterial hypertension therapy. Continuous data with nonparametric distribution were presented as median (range). We also used Mann-Whitney tests to compare data between mutation carriers and non-carriers. Categorical data were presented as number (%) and comparisons were based on χ² tests. We used the Friedman’s test to compare NYHA functional class before

109 patients with sporadic or heritable PVOD/PCH (with genetic testing)

47 confirmed PVOD/PCH

62 highly probable PVOD/PCH

15 excluded (chronic respiratory diseases) For the Human Genome Variation Society recommendations see http://www.hgvs.org/content/ guidelines

94 PVOD/PCH

27 EIF2AK4 mutation carriers (n=5 <18 years, n=22 ≥18 years)

67 EIF2AK4 mutation non-carriers (n=2 <18 years, n=65 ≥18 years)

Figure 1: Flow chart of patients included in the study PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis.

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

3

Articles

EIF2AK4 mutations carriers (n=27)

EIF2AK4 mutation non-carriers (n=67)

Age at diagnosis (years)

26 (0–50·3)

60 (6·7–81·4)

Sex, female/male (ratio)

14/13 (1·1)

19/48 (0·4)

NYHA functional class

p value

<0·0001 0·031 0·2

II

4 (15%)

3 (5%)

··

III

17 (65%)

50 (75%)

··

IV 6-min walk distance (m)

5 (19%)

14 (21%)

371 (167)

244 (152)

·· 0·0013

6-min walk distance (% pred)

49·3% (22·6)

44·1% (27·5)

0·42

mPAP (mm Hg)

49 (14)

46 (11)

0·29

PAWP (mm Hg)

7 (4)

8 (3)

0·42

CI (L/min per m²)

2·56 (0·90)

2·47 (0·70)

0·58

CO (L/min)

4·37 (1·47)

4·51 (1·50)

0·69

PVR (wood units)

11·4 (6·7)

SvO2 (%)

62% (13)

9·4 (4·4) 60% (9)

0·10 0·46

DLCO (% pred)

30% (7)

31% (10)

0·65

FEV1 (% pred)

94% (18)

85% (15)

0·032

TLC (% pred)

96% (18)

89% (14)

PaO2 (kPa)

9·1 (1·7)

7·2 (1·3)

<0·0001

PaCO2 (kPa)

3·7 (0·4)

4·0 (0·8)

0·16

Yes

1 (4%)

6 (11%)

··

No

23 (96%)

51 (90%)

··

Acute vasodilator response

0·080

0·35

Exposure to organic solvents

0

28 (42%)

<0·0001

Previous chemotherapy

0

7 (10%)

0·081

Data are median (range), mean (SD), or n (%). Comparisons between bi-allelic EIF2AK4 mutations carriers and non-carriers were analysed by the χ² test, and unpaired t-test as appropriate. ··=not applicable. NYHA=New York Heart Association. % pred=percentage of predicted value. mPAP=mean pulmonary artery pressure. PAWP=pulmonary artery wedge pressure. CI=cardiac index. CO=cardiac output. PVR= pulmonary vascular resistance. SvO2=mixed venous oxygen saturation. DLCO=diffusing capacity for carbon monoxide corrected for haemoglobin concentration. TLC=total lung capacity. PaO2=partial pressure of oxygen in arterial blood. PaCO2=partial pressure of carbon dioxide in arterial blood.

Table 1: Clinical, functional, and haemodynamic characteristics of patients at diagnosis of PVOD/PCH with and without bi-allelic EIF2AK4 mutations

See Online for appendix

4

and after treatment. This test is well suited for paired comparison of ordinal categorical data. Survival analyses were made from the time of the first diagnosis of pulmonary hypertension confirmed by right-heart catheterisation. Overall survival, event-free survival (time to death or lung transplantation), and survival after transplantation were estimated by the Kaplan-Meier method. Event-free survivals between EIF2AK4 mutation carriers and non-carriers were compared using a log-rank test. We assessed the association between mutation status and mortality using a multivariate cause-specific proportional hazard model. Lung transplantation was considered as a competing risk because patients were censored after transplantation and were not at risk of death anymore. Since we were more interested in aetiological research

than in prognostic research, a multivariate causespecific proportional hazard model was chosen rather than a subdistribution hazard approach.25 Aetiological research aims to investigate the causal relationship between risk factors or determinants and a given outcome. To this end, it uses hazard ratios (HRs) to estimate an effect size. By contrast, prognostic research aims to predict the probability of a given outcome at a given time for an individual patient.26 Because there is a direct relationship between the covariates and the cumulative incidence function, the subdistribution hazards model directly provides individual prediction based on covariates, or estimated probabilities of an event, given a patient’s characteristics. An important feature of this method is that patients who had a competing event remained in the risk set (instead of being censored), although they were no longer at risk of the event of interest. On the other hand, the proportional cause-specific hazards model directly quantifies the HRs among individuals who are actually at risk of developing the event of interest. Competing events were treated as censored observations.27,28 The HR was interpreted as: at any time after diagnosis of pulmonary arterial hypertension, mutation carriers had a hazard of dying higher or lower than that of non-carriers, and among patients who were alive and did not receive a lung transplant at that time. The hypothesis of proportional hazard was checked by plotting the Schoenfeld residuals against time. We considered a p value less than 0·05 as significant. We did all statistical analysis with Statview 5.0 (SAS Institute, Cary, NC, USA), except the multivariate cause-specific proportional hazard model computed using SAS 9.4 package (SAS Institute, Cary, NC, USA).

Role of the funding source There was no funding source for this study. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results Between Jan 1, 2003, and June 1, 2016, 109 patients with PVOD/PCH (47 confirmed and 62 highly probable) underwent genetic counselling and testing (figure 1). 15 patients with highly probable PVOD/PCH were excluded from the final analysis because of chronic respiratory disease. Thus, 94 patients were included in this study, of which 27 were carriers of bi-allelic EIF2AK4 mutations (mutations listed in appendix) and 67 were non-carriers of these mutations. Non-carriers corresponded to 47 highly probable and 20 confirmed cases of PVOD/PCH with histological assessment (13 explanted lungs, six lung biopsies, and one post-mortem). Of the 27 bi-allelic EIF2AK4 mutations carriers, 19 were familial cases (13 families) and eight were sporadic cases. Genetic analysis of 39 patients with PVOD/PCH (including

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

Articles

p<0·0001

100 p<0·0001 80

Age (years)

24 EIF2AK4 mutations carriers) have been previously reported (appendix).10 Seven cases were paediatric (aged <18 years), including five patients with bi-allelic EIF2AK4 mutations and two with histological confirmation and no identified mutation. Eight families (62%) of 13 were characterised by homozygous EIF2AK4 mutations (consanguinity was known in five families). 28 mutation non-carriers (42%) of 67 had a history of exposure to organic solvents and seven (10%) of 67 had previously received systemic chemotherapy (table 1). No history of solvent or chemotherapy exposure was noted in EIF2AK4 mutations carriers. Prevalence of EIF2AK4 mutations in specific subgroups was 100% (19/19) in patients with a family history of PVOD/PCH and 9% (7/75) in sporadic PVOD/PCH cases. Among patients with sporadic PVOD/PCH, prevalence of EIF2AK4 mutations varied widely depending on history of exposure to solvents or chemotherapy (0/35 in exposed patients and 25% [8/32]) in non-exposed patients. PVOD/PCH due to EIF2AK4 mutations occurred from birth to age 50 years, and patients were younger at presentation compared with non-carriers (median 26·0 years [range 0–50·3] vs median 60·0 years [6·7–81·4]; p<0·0001; figure 2). There was a male predominance in non-carriers. All patients with PVOD/PCH had advanced NYHA functional class at presentation; 6-min walk distance confirmed severe functional impairment with no difference between the two groups when expressed as percentage predicted adjusted for age, height, and sex (table 1). Low DLCO and hypoxaemia were identified in patients with PVOD/PCH, irrespective of the EIF2AK4 mutation status. Right-heart catheterisation at diagnosis showed severe precapillary pulmonary hypertension in both groups. No differences in haemodynamic characteristics were noted between mutation carriers and non-carriers (table 1). Acute vasodilator testing with inhaled nitric oxide was done in 81 (86%) of 94 patients with PVOD or PCH. No acute pulmonary oedema complicated vasodilator testing. Partial response to nitric oxide was observed with a mean pulmonary artery pressure (mPAP) decrease of 9·6%, and a median pulmonary vascular resistance (PVR) decrease of 16·2% (data not shown). One (4%) of 24 carriers of EIF2AK4 bi-allelic mutations and six (11%) of 57 non-carriers fulfilled the haemodynamic criteria for significant acute vasoreactivity (table 1).1 High-resolution CT of the chest was available in 25 EIF2AK4 bi-allelic mutations carriers. The scan showed septal lines in 25 (100%), nodular ground-glass opacities in 24 (96%), and mediastinal lymph node enlargement in 20 (80%) patients (figure 3). All mutation carriers presented with at least two of these radiological signs at diagnosis and 20 (80%) displayed all three radiological abnormalities. Other abnormalities included pulmonary artery dilatation (23 [92%]), right ventricular dilatation (22 [88%]), pericardial effusion (5 [20%]), and

60

40

20

0 Diagnosis

Death or lung transplantation

EIF2AK4 mutation carriers (n=27)

Diagnosis

Death or lung transplantation

EIF2AK4 mutation non-carriers (n=67)

Figure 2: Age at PVOD/PCH diagnosis and age at death or lung transplantation Data from EIF2AK4 bi-allelic mutation carriers and mutation non-carriers. Bottom and top of the box-plot represent the 25th and 75th percentiles. The band inside the box is the median. The ends of the whiskers represent 1·5 IQR above 75th percentile and below 25th percentile. Outliers are represented as circles. PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis.

pleural effusion (2 [8%]). By definition, high-resolution CT of mutation non-carriers with highly probable PVOD/ PCH had at least two radiological signs of the diseases. Medical therapies for pulmonary arterial hypertension (endothelin receptor antagonists, phosphodiesterase type 5 inhibitors, or prostacyclin derivatives) were initiated in 85 patients (22 EIF2AK4 bi-allelic mutation carriers and 63 non-carriers). Clinical and haemodynamic assessments following the initiation of therapy were available for 64 patients (17 EIF2AK4 mutation carriers and 47 non-carriers). Treatments at time of follow-up assessment are in table 2. There were no significant differences in the interval between pulmonary arterial hypertension-targeted therapy initiation and reassessment between EIF2AK4 mutation carriers and non-carriers (median time 4·3 months [range 1·5–22·1] in the entire cohort; p=0·38). Follow-up assessment after initiation of medical therapy for pulmonary arterial hypertension showed that only three patients with PVOD/PCH reached the predefined criteria for satisfactory clinical response (NYHA functional class I or II, 6-min walk distance >440 m and cardiac index >2·5 L/min per m²). However, significant improvements occurred in 6-min walk distance and pulmonary haemodynamics with a higher cardiac index and lower PVR on pulmonary arterial hypertension therapy (table 2). Mild clinical and haemodynamic improvements were only and rarely observed in a subgroup of mutation non-carriers (appendix). Regarding the 22 EIF2AK4 bi-allelic mutation carriers treated with pulmonary arterial hypertension therapy, five received transplantation before reassessment. Effects of pulmonary arterial hypertension therapy in the 17 EIF2AK4 bi-allelic mutation carriers are in the appendix.

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

5

Articles

A

B

Baseline

Second assessment

EIF2AK4 bi-allelic mutations (yes/no)

17/47*

17/47*

Sex, female/male

20/44

20/44

NYHA functional class

D

·· ·· 0·10

II

6 (10%)

15 (24%)

III

46 (72%)

38 (60%)

··

IV

12 (19%)

11 (17%)

··

289 (160)

321 (157)

46 (10)

46 (12)

6-min walk distance (m)

C

p value

mPAP (mm Hg) Cardiac index (L/min per m²)

2·43 (0·76)

2·84 (0·74)

PVR (wood units)

9·7 (4·7)

8·1 (5·0)

··

0·0142 0·87 <0·0001 0·0002

Medical therapy for pulmonary arterial hypertension

E

F

Figure 3: High-resolution CT scan of the chest in patients with PVOD/PCH (A) Ground-glass opacities with centrilobular pattern, poorly defined nodular opacities, and septal lines in an adult carrier of EIF2AK4 bi-allelic mutations. (B) Mediastinal lymph node enlargement in an adult carrier of EIF2AK4 bi-allelic mutations. (C) A paediatric case carrying EIF2AK4 bi-allelic mutations (patient 1 in table 3). (D) A paediatric case not carrying a EIF2AK4 bi-allelic mutation (patient 3 in table 3). (E and F) The chest of an EIF2AK4 mutation carrier before initiation of specific therapy for pulmonary arterial hypertension showed mild abnormalities with septal lines and ground-glass opacities (E); high resolution tomography of the chest was done 2 months after initiation of endothelin receptor antagonists for rapid worsening dyspnoea and showed a substantial increase of radiological abnormalities evocative of pulmonary oedema (F). PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis.

Drug-induced pulmonary oedema occurred in five (23%) of 22 EIF2AK4 mutation carriers and 13 (21%) of 63 non-carriers after a median delay of 1 month (0–38; 11 with endothelin receptor antagonists, two with prostacyclin derivatives, two with combination therapy, and three with calcium channel blockers). Among them, three patients developed severe pulmonary oedema requiring management in intensive care unit and listing for urgent lung transplantation. No significant clinical, functional, or haemodynamic differences were noted between patients with or without the development of drug-induced pulmonary oedema (appendix). Histological examinations of the lungs were available in 20 EIF2AK4 bi-allelic mutation carriers and 20 noncarriers. Regardless of the mutation status, all patients had pronounced venous or capillary involvement corresponding to PVOD/PCH (extensive and diffuse occlusion of pulmonary veins by fibrous tissue, intimal thickening involving venules and small veins in 6

ERA monotherapy

··

43 (67%)

··

PDE5 inhibitor monotherapy

··

8 (13%)

··

Prostacyclin derivative monotherapy

··

4 (6%)

··

ERA plus PDE5

··

4 (6%)

··

ERA plus prostacyclin derivative

··

4 (6%)

··

ERA plus PDE5 inhibitor plus prostacyclin derivative

··

1 (2%)

··

Data are n (%) or mean (SD). Response to specific therapy for pulmonary arterial hypertension was analysed using Friedman’s test, and paired t test as appropriate. ··=not applicable. NYHA=New York Heart Association. mPAP=mean pulmonary artery pressure. PVR=pulmonary vascular resistance. ERA=endothelin receptor antagonist. PDE5=phosphodiesterase type 5. PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis. *11 patients received specific therapies for pulmonary arterial hypertension but had no reassessment because of death (n=3) or lung transplantation (n=8).

Table 2: Clinical, functional, and haemodynamic characteristics at diagnosis and second assessment after initiation of medical therapy for pulmonary arterial hypertension in adult patients with PVOD/PCH

lobular septa, and localised capillary proliferation; figure 4). Eccentric intimal fibrosis of pulmonary arteries without plexiform lesions was also commonly observed. Five (19%) of 27 EIF2AK4 bi-allelic mutation carriers died after a median follow-up time of 27 months from diagnosis (range 0–63). 20 (74%) of 27 EIF2AK4 bi-allelic mutation carriers were transplanted after a median delay of 19 months (range 1–89), including six urgent lung transplantations through the high-priority allocation programme. Median age at transplantation was 27·2 years (range 13–51). Two (8%) of 27 mutations carriers were still alive at 34 and 35 months after diagnosis without transplantation (one patient was listed for lung transplantation). PVOD/PCH due to EIF2AK4 bi-allelic mutations had estimated 1-year, 2-year, and 3-year eventfree survival probabilities of 63%, 52%, and 32%, respectively (figure 5). In non-carriers of EIF2AK4 mutations, 40 (60%) of 67 patients died after a median follow-up of 19 months (range 5–122). 14 (21%) of 67 received a transplantation after a median delay of 9 months (range 2–40). Median age at the time of transplantation was 54·2 years (range 7·0–63·0). 13 (19%)

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

Articles

of 67 patients were still alive without transplantation after a median time of 20 months after diagnosis (range 3–131). Non-carriers had similar estimated 1-year, 2-year, and 3-year event-free survival probabilities of 75%, 44%, and 34%, respectively (figure 5). No significant differences occurred in event-free survival between the two groups (p=0·38 for log-rank test). However, bi-allelic EIF2AK4 mutation carriers were significantly younger at death or lung transplantation compared with mutation non-carriers (mean 26·8 years [10·9] vs 60·2 [16·5], p<0·0001; figure 2). The younger age of EIF2AK4 bi-allelic mutation carriers resulted in a higher likelihood of transplantation as compared to non-carriers, and by inference a better survival when time to death was assessed (patients with transplantation were censored at the time of transplantation; figure 5). Thereby, to include the competing risk of lung transplantation, a multivariate cause-specific proportional hazards model adjusted for age was done. No difference in survival was noted in the two groups (HR=0·37, 95% CI 0·10–1·44; p=0·15) for the whole population or when the analysis was restricted to adult patients (HR=0·59, 0·14–2·57; p=0·49). Among the 33 patients with PVOD/PCH who had lung transplantation, median follow-up after transplantation was 44 months (range 0·4–295). Estimated post-transplantation survival rates at 1 year, 2 years, and 5 years were 84%, 81%, and 73%, respectively. Two EIF2AK4 biallelic mutation carriers received a second lung transplantation after 24 and 34 months because of bronchiolitis obliterans syndrome. To date, no clinical or histological recurrence of PVOD or PCH has been identified. Clinical, haemodynamic, and radiological characteristics of the seven paediatric PVOD/PCH cases are given in table 3. Right-heart catheterisation and high-resolution CT of the chest were not available for the patient who presented with severe PVOD/PCH at birth and died aged 10 days. In this patient, pulmonary hypertension was detected by means of Doppler echocardiography, and PVOD/PCH was confirmed by the presence of bi-allelic EIF2AK4 mutations.

A

B

500 µm 100 µm

C

D

100 µm

E

200 µm

F

100 µm

1000 000 µm

Discussion

Figure 4: Pathological examination of the lungs in an adult patient with bi-allelic EIF2AK4 mutations (A–D) and a paediatric patient with PVOD/PCH without EIF2AK4 mutation (E–F) Examination was done with haematoxylin-eosin-saffron staining. (A) Severely remodelled septal vein (top) with fibrous intimal occlusion, leaving behind the image of a pauci-cellular fibrous band. Note another pulmonary vein (bottom) with remodelling and adjacent thickening of alveolar septa. (B) Preseptal muscularised venule with constrictive loose intimal fibrosis, identified by green dye within its lumen, previously injected into the hilar veins. (C) Magnification of patchy foci with alveolar septal thickening (see A): so-called pulmonary capillary haemangiomatosis-like foci correspond to a local multiplication of alveolar capillaries, comprising scattered inflammatory cells. (D) Pulmonary artery (centre) with its adjacent bronchiole (top) showing substantial eccentric intimal fibrosis and pulmonary capillary haemangiomatosis-like focus adjacent to the pulmonary artery (bottom). (E–F) Pathological examination of the lungs of a paediatric patient with PVOD/PCH without EIF2AK4 mutation (patient 6, table 3) showing a similar pattern with venous remodelling (E) and patchy capillary proliferation (F). PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis

PVOD/PCH occurring in patients with bi-allelic EIF2AK4 mutations is characterised by a younger age at diagnosis compared with non-carriers. Regardless of mutation status, all patients with PVOD/PCH presented with severe clinical, functional, and haemodynamic impairment. Despite the possible occurrence of drug-induced pulmonary oedema, a mild improvement with pulmonary arterial hypertension therapy occurred in the entire cohort, as reported previously.13 However, clinical and haemodynamic improvements with treatment were only noted in a subset of patients with PVOD/PCH without biallelic EIF2AK4 mutations, but not in mutation carriers. The poor response to specific pulmonary arterial hypertension therapies in these patients was confirmed by the small proportion of patients (3/64) reaching the criteria

of low-risk of mortality proposed for pulmonary arterial hypertension in the European pulmonary hypertension guidelines.1 Overall, this translated to a dismal prognosis with death or lung transplantation occurring in most cases within the first months or years of diagnosis. Pathological examination of the lungs from EIF2AK4 bi-allelic mutation carriers identified major pulmonary venous or capillary involvement, confirming that PVOD and PCH represent varied expressions of the same disease. PVOD/PCH belong to a rare and difficult-to-diagnose subgroup of pulmonary hypertension. Early identification of these conditions in patients with pulmonary hypertension is essential because of their progressive

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

7

Articles

Overall survival

A

B

100 Cumulative survival (%)

EIF2AK4 mutation non-carriers

EIF2AK4 mutation carriers

80 60 40 20 0 0

EIF2AK4 27

12 24 Time to death (months) 17 14

36

0

7

Non-carriers 67

12 24 Time to death (months) 48 25

36 19

Time to death or lung transplantation

C Cumulative survival (%)

100

D EIF2AK4 mutation non-carriers

EIF2AK4 mutation carriers

80 60 40 20 0 0 12 24 36 Time to death or lung transplantation (months)

EIF2AK4 27

17

14

7

0 12 24 36 Time to death or lung transplantation (months) Non-carriers 67

48

25

19

Figure 5: Overall survival and time to death or lung transplantation of EIF2AK4 bi-allelic mutation carriers and non-carriers Overall survival of EIF2AK4 bi-allelic mutations carriers (A) and non-carriers (B). For analysis of overall survival, patients receiving transplantation were censored at the time of transplantation. Time to death or lung transplantation of EIF2AK4 bi-allelic mutations carriers (C) and non-carriers (D).

EIF2AK4 bi-allelic mutations

clinical course and poor response to medical therapy. In our study, right-heart catheterisation revealed severe precapillary pulmonary hypertension at diagnosis. All patients had a normal pulmonary artery wedge pressure, confirming that conventional pulmonary haemodynamics cannot distinguish PVOD/PCH from pulmonary arterial hypertension. Conversely, it has been previously proposed that high-resolution CT of the chest could suggest abnormal findings in patients with PVOD/PCH, including septal lines, nodular ground glass opacities, and mediastinal lymph node enlargement.19,29 Our study confirms the high incidence of radiological abnormalities in patients with heritable PVOD/PCH carrying EIF2AK4 mutations. Additionally, a marked reduction in DLCO was noted in all patients, contrasting with a more preserved DLCO in heritable pulmonary arterial hypertension due to BMPR2 mutations.30 The low DLCO observed in patients with PVOD/PCH could be explained by a greater reduction in capillary blood volume from a compromised pulmonary vascular bed and poorer membrane diffusion as a result of capillary proliferation and interstitial oedema. High-resolution CT of the chest, and DLCO, thus represent major tools to detect PVOD/ PCH in patients with precapillary pulmonary hypertension. Importantly, the availability of genetic testing allows confirmation of a diagnosis of heritable PVOD/ PCH if bi-allelic EIF2AK4 mutations are found. This underscores the importance of offering genetic counselling and EIF2AK4 mutation screening to patients with sporadic and familial PVOD/PCH.31

Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Patient 6

Patient 7

Yes

Yes

Yes

Yes

No

No

Yes

Age at diagnosis (years)

17

16

15

12

10

Sex

Male

Female

Female

Male

Male

7 Female

Birth Male

NYHA functional class

III

III

IV

I–II

III

IV

··

mPAP (mm Hg)

75

39

54

24

49

73

··

PAWP (mm Hg)

8

8

7

11

8

5

··

Cardiac Index (L/min per m²)

1·94

4·49

1·68

1·54

3·2

··

PVR (wood units)

24·7

4·4

18·8

3·8 2·3

26·3

21

··

High-resolution CT of the chest Lymph node enlargement

No

Yes

Yes

Yes

Yes

Yes

··

Centrilobular ground-glass opacities

Yes

Yes

Yes

Yes

Yes

Yes

··

Interlobular septal lines

Yes

Yes

Yes

Yes

Yes

Yes

Medical therapy for pulmonary arterial hypertension

ERA

ERA plus PDE5 inhibitor

Prostacyclin derivatives

ERA plus PDE5 inhibitor

ERA

PDE5 inhibitor

··

Outcome

Lung transplantation at 4 months

Lung Lung transplantation transplantation at 5 months at 21 months

Lung transplantation at 17 months

Death at 17 months

Death at Lung transplantation 10 days at 7 months

No

PVOD/PCH=pulmonary veno-occlusive disease and/or pulmonary capillary haemangiomatosis. NYHA=New York Heart Association. mPAP=mean pulmonary artery pressure. PAWP=pulmonary artery wedge pressure. PVR=pulmonary vascular resistance. ERA=endothelin receptor antagonist. PDE5=phosphodiesterase type 5. ··=data not available.

Table 3: Clinical, functional, and haemodynamic characteristics at diagnosis, and outcomes of paediatric cases of PVOD/PCH

8

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

Articles

A major concern with pulmonary arterial hypertension therapy in PVOD/PCH is the risk of drug-induced pulmonary oedema.17,19 However, findings from small series and isolated case reports have suggested a potential benefit of these drugs as a bridge to lung transplantation in PVOD/PCH.13,19,32 Because of severe haemodynamic derangement at diagnosis, such medical therapies were initiated in most patients. Of note, 21% of non-carriers and 23% of mutation carriers developed drug-induced pulmonary oedema. As previously described, mild clinical, functional, or haemodynamic improvements occurred in non-carriers of mutations at first follow-up assessment.13,19,32 However, long-term outcomes remained poor in these patients (60% death and 21% lung transplantation after a median delay of 19 and 9 months, respectively). In 17 patients with heritable PVOD/PCH carrying EIF2AK4 mutations, no significant clinical or haemodynamic improvements were noted after medical therapy, although the small sample size could have resulted in insufficient power to detect a small treatment effect. This apparent refractoriness to treatment translated to five deaths at a median time of 19 months and the necessity to perform lung transplantation in 20 patients including six urgent procedures, at a median time of 27 months after diagnosis. Our results suggest that prognosis is dismal in heritable PVOD/PCH due to EIF2AK4 mutations because of the rapid evolution of the disease, absence of response to medical therapy, and the possible occurrence of severe pulmonary oedema after initiation of pulmonary arterial hypertension therapy. Thus, lung transplantation is the treatment of choice for PVOD/ PCH in eligible bi-allelic EIF2AK4 mutation carriers. Some results have suggested that patients with PVOD/ PCH might be at higher risk for death while on the transplantation waiting list.33 However, in our study survival after lung transplantation in PVOD/PCH appeared satisfactory, further emphasising the necessity to consider early referral for lung transplantation in eligible patients. Finally, no recurrence of PVOD/PCH has been observed to date after lung transplantation in EIF2AK4 mutation carriers as well as non-carriers. We also described seven paediatric cases of heritable and sporadic PVOD/PCH (five carriers of bi-allelic EIF2AK4 mutations). Pedigrees of families of patients carrying bi-allelic EIF2AK4 mutations allowed us to detect six more paediatric cases (brothers and sisters of index cases) deceased with a poorly documented diagnosis of pulmonary hypertension with insufficient information to be included in this report. Thus PVOD/ PCH could present in very early childhood, as previously reported,34 with the youngest patient with heritable PVOD/PCH diagnosed at birth. This information indicates for the first time that heritable PVOD/PCH can be a cause of pulmonary hypertension in newborn children with major consequences for genetic counselling (because of the risk of other cases in siblings).

Our study had several limitations. The first was the small number of patients included in this study. However, PVOD/PCH are orphan diseases and the present study represents the largest cohort of patients with PVOD/PCH published to date. The diagnosis of PVOD/PCH was confirmed in the mutation carriers group by the presence of bi-allelic mutations in the EIF2AK4 gene, but diagnosis was only definitively confirmed in 20 (30%) of noncarriers. Definitive diagnosis of PVOD/PCH in mutation non-carriers required pathological examination of the lung, in the context of either lung transplantation, postmortem examination, or lung biopsy. Because of the older age of mutation non-carriers, many patients were not eligible for transplantation and lung biopsy was contraindicated in the setting of severe pulmonary hypertension. Nevertheless, we only included patients with characteristic features of PVOD/PCH using strict clinical criteria, with the acknowledgement that these patients might not be representative of all phenotypes of PVOD/PCH without EIF2AK4 mutations. Finally, because of the rarity of PVOD/PCH, data collection was done over many years. Currently, no recommendations exist for the management of PVOD/PCH and these patients were treated according to clinical judgment. However, all patients included in this study were managed at the French Referral Centre with a broadly similar therapeutic approach. In conclusion, bi-allelic EIF2AK4 mutation carriers are significantly younger at diagnosis of PVOD/PCH compared with non-carriers. Our data suggest that they display no long-term response to medical therapy approved for pulmonary arterial hypertension and have an equally dismal prognosis compared to patients with sporadic PVOD/PCH. Heritable PVOD/PCH occurred across a broad age range from birth to middle adulthood. Overall, PVOD/PCH are devastating conditions and lung transplantation remains the treatment of choice for eligible patients. More research in the understanding of the role of the EIF2AK4 pathway in the maintenance of pulmonary vascular homeostasis should open new avenues for the identification of innovative therapeutic strategies in PVOD/PCH. Contributors DM, BG, and MH conceived the study. DM, BG, OS, GS, and MH designed the study. DM, BG, XJ, ML, DA, LS, PD, AS, EML, ME, JLP, FP, DB, FS, EF, GS, and MH obtained the data. DM and MH coordinated the study. DM, BG, EML, and MH analysed the data. All authors interpreted the results. DM, BG and MH drafted the manuscript with critical revisions for important intellectual content from all authors. All authors approved the final version. Declaration of interests DM, BG, XJ, LS, GS, and MH report grants, personal fees, and nonfinancial support from Actelion, Pfizer, Bayer, and GlaxoSmithKline, MSD, outside of the submitted work. All other authors declare no competing interests. Acknowledgments We thank the patients, their families, health-care providers from the French Pulmonary Hypertension Network for agreeing to collaborate, the patient association HTaP France, Pierre Clerson (Soladis Clinical

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6

9

Articles

Studies Roubaix) for the statistical analysis, the Fondation Eliane et Gérard Pauthier, under the auspices of Fondation de France, and the Fondation Maladies Rares. MH was awarded the Eliane and Gérard Pauthier 2016 Prize. References 1 Galiè N, Humbert M, Vachiery J-L, et al. 2015 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension: The Joint Task Force for the Diagnosis and Treatment of Pulmonary Hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS): Endorsed by: Association for European Paediatric and Congenital Cardiology (AEPC), International Society for Heart and Lung Transplantation (ISHLT). Eur Respir J 2015; 46: 903–75. 2 Humbert M, Sitbon O, Chaouat A, et al. Pulmonary arterial hypertension in France: results from a national registry. Am J Respir Crit Care Med 2006; 173: 1023–30. 3 Newman JH, Wheeler L, Lane KB, et al. Mutation in the gene for bone morphogenetic protein receptor II as a cause of primary pulmonary hypertension in a large kindred. N Engl J Med 2001; 345: 319–24. 4 Trembath RC, Thomson JR, Machado RD, et al. Clinical and molecular genetic features of pulmonary hypertension in patients with hereditary hemorrhagic telangiectasia. N Engl J Med 2001; 345: 325–34. 5 Ma L, Roman-Campos D, Austin ED, et al. A novel channelopathy in pulmonary arterial hypertension. N Engl J Med 2013; 369: 351–61. 6 Austin ED, Ma L, LeDuc C, et al. Whole exome sequencing to identify a novel gene (caveolin-1) associated with human pulmonary arterial hypertension. Circ Cardiovasc Genet 2012; 5: 336–43. 7 Sztrymf B, Coulet F, Girerd B, et al. Clinical outcomes of pulmonary arterial hypertension in carriers of BMPR2 mutation. Am J Respir Crit Care Med 2008; 177: 1377–83. 8 Girerd B, Montani D, Coulet F, et al. Clinical outcomes of pulmonary arterial hypertension in patients carrying an ACVRL1 (ALK1) mutation. Am J Respir Crit Care Med 2010; 181: 851–61. 9 Evans JDW, Girerd B, Montani D, et al. BMPR2 mutations and survival in pulmonary arterial hypertension: an individual participant data meta-analysis. Lancet Respir Med 2016; 4: 129–37. 10 Eyries M, Montani D, Girerd B, et al. EIF2AK4 mutations cause pulmonary veno-occlusive disease, a recessive form of pulmonary hypertension. Nat Genet 2014; 46: 65–69. 11 Tenorio J, Navas P, Barrios E, et al. A founder EIF2AK4 mutation causes an aggressive form of pulmonary arterial hypertension in Iberian Gypsies. Clin Genet 2015; 88: 579–83. 12 Best DH, Sumner KL, Austin ED, et al. EIF2AK4 mutations in pulmonary capillary hemangiomatosis. Chest 2014; 145: 231–36. 13 Montani D, Lau EM, Dorfmüller P, et al. Pulmonary veno-occlusive disease. Eur Respir J 2016; 47: 1518–34. 14 Montani D, Descatha A, Lau E, et al. Occupational exposure: a risk factor for pulmonary veno-occlusive disease. Eur Respir J 2015; 46: 1721–31. 15 Ranchoux B, Günther S, Quarck R, et al. Chemotherapy-induced pulmonary hypertension: role of alkylating agents. Am J Pathol 2015; 185: 356–71.

10

16

17

18 19

20

21

22 23

24 25

26

27

28 29

30

31

32

33

34

Perros F, Günther S, Ranchoux B, et al. Mitomycin-induced pulmonary veno-occlusive disease: evidence from human disease and animal models. Circulation 2015; 132: 834–47. Montani D, Achouh L, Dorfmüller P, et al. Pulmonary veno-occlusive disease: clinical, functional, radiologic, and hemodynamic characteristics and outcome of 24 cases confirmed by histology. Medicine (Baltimore) 2008; 87: 220–33. Mandel J, Mark EJ, Hales CA. Pulmonary veno-occlusive disease. Am J Respir Crit Care Med 2000; 162: 1964–73. Holcomb BW Jr, Loyd JE, Ely EW, Johnson J, Robbins IM. Pulmonary veno-occlusive disease: a case series and new observations. Chest 2000; 118: 1671–79. Palmer SM, Robinson LJ, Wang A, Gossage JR, Bashore T, Tapson VF. Massive pulmonary edema and death after prostacyclin infusion in a patient with pulmonary veno-occlusive disease. Chest 1998; 113: 237–40. Simonneau G, Gatzoulis MA, Adatia I, et al. Updated clinical classification of pulmonary hypertension. J Am Coll Cardiol 2013; 62: D34–41. Humbert M, Sitbon O, Simonneau G. Treatment of pulmonary arterial hypertension. N Engl J Med 2004; 351: 1425–36. Humbert M, Lau EMT, Montani D, Jaïs X, Sitbon O, Simonneau G. Advances in therapeutic interventions for patients with pulmonary arterial hypertension. Circulation 2014; 130: 2189–208. Enright PL, Sherrill DL. Reference equations for the six-minute walk in healthy adults. Am J Respir Crit Care Med 1998; 158: 1384–87. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509. Noordzij M, Leffondré K, van Stralen KJ, Zoccali C, Dekker FW, Jager KJ. When do we need competing risks methods for survival analysis in nephrology? Nephrol Dial Transplant 2013; 28: 2670–77. Andersen PK, Geskus RB, de Witte T, Putter H. Competing risks in epidemiology: possibilities and pitfalls. Int J Epidemiol 2012; 41: 861–70. Lau B, Cole SR, Gange SJ. Competing risk regression models for epidemiologic data. Am J Epidemiol 2009; 170: 244–56. Resten A, Maitre S, Humbert M, et al. Pulmonary hypertension: CT of the chest in pulmonary venoocclusive disease. AJR Am J Roentgenol 2004; 183: 65–70. Trip P, Girerd B, Bogaard H-J, et al. Diffusion capacity and BMPR2 mutations in pulmonary arterial hypertension. Eur Respir J 2014; 43: 1195–98. Girerd B, Montani D, Jaïs X, et al. Genetic counselling in a national referral centre for pulmonary hypertension. Eur Respir J 2015; 47: 541–52. Ogawa A, Miyaji K, Yamadori I, et al. Safety and efficacy of epoprostenol therapy in pulmonary veno-occlusive disease and pulmonary capillary hemangiomatosis. Circ J 2012; 76: 1729–36. Wille KM, Sharma NS, Kulkarni T, et al. Characteristics of patients with pulmonary venoocclusive disease awaiting transplantation. Ann Am Thorac Soc 2014; 11: 1411–18. Levy M, Eyries M, Szezepanski I, et al. Genetic analyses in a cohort of children with pulmonary hypertension. Eur Respir J 2016; 48: 1118–26.

www.thelancet.com/respiratory Published online January 10, 2017 http://dx.doi.org/10.1016/S2213-2600(16)30438-6