Abstract
This study aims to describe the haemodynamic and survival characteristics of patients with pulmonary hypertension in the recently individualised syndrome of combined pulmonary fibrosis and emphysema.
A retrospective multicentre study was conducted in 40 patients (38 males; age 68±9 yrs; 39 smokers) with combined pulmonary fibrosis and emphysema, and pulmonary hypertension at right heart catheterisation.
Dyspnoea was functional class II in 15%, III in 55% and IV in 30%. 6-min walk distance was 244±126 m. Forced vital capacity was 86±18%, forced expiratory volume in 1 s 78±19%, and carbon monoxide diffusion transfer coefficient 28±16% of predicted. Room air arterial oxygen tension was 7.5±1.6 kPa (56±12 mmHg). Mean pulmonary artery pressure was 40±9 mmHg, cardiac index 2.5±0.7 L·min−1·m−2 and pulmonary vascular resistance 521±205 dyn·s·cm−5. 1-yr survival was 60%. Higher pulmonary vascular resistance, higher heart rate, lower cardiac index and lower carbon monoxide diffusion transfer were associated with shorter survival.
Patients with combined pulmonary fibrosis and emphysema syndrome and pulmonary hypertension confirmed by right heart catheterisation have a dismal prognosis despite moderately altered lung volumes and flows and moderately severe haemodynamic parameters.
- Chronic obstructive pulmonary disease
- disproportionate
- emphysema
- pulmonary fibrosis
- pulmonary hypertension
- tobacco smoking
Idiopathic pulmonary fibrosis (IPF) is a severe chronic disease of unknown aetiology, with a median survival of 3 yrs. In smokers, some emphysema may be associated with IPF 1–3. We recently individualised the syndrome of combined pulmonary fibrosis and emphysema (CPFE) 4 based on high-resolution computed tomography (HRCT) of the chest in a homogeneous group of 61 patients, further characterised by severe dyspnoea on exertion, subnormal spirometry, severe impairment of gas exchange and a median survival of 6.1 yrs 5. CPFE is probably related to tobacco smoking, a common risk factor for both emphysema and fibrosis (with odds ratios of up to 3.6 in familial fibrosis) 6, 7.
Patients with advanced IPF have a high prevalence of pulmonary hypertension 8, 9, with 31–46% of patients with mean pulmonary arterial pressure (Ppa) >25 mmHg at right-sided heart catheterisation (RHC) at evaluation for lung transplantation 9–12 and 86% at the time of transplantation 13. Similarly, the prevalence of pulmonary hypertension (defined by mean Ppa >20 mmHg) in patients hospitalised for chronic obstructive pulmonary disease (COPD) is ∼50% 14, and may be as high as 50–90% in COPD patients evaluated for lung volume reduction surgery or lung transplantation 15, 16. The pejorative prognostic significance of pulmonary hypertension has been demonstrated in both IPF 9, 10, 17 and COPD 18.
Two studies have reported that pulmonary hypertension is frequent in patients with the CPFE syndrome 5, 19, with 47% of patients with estimated systolic right ventricular pressure ≥45 mmHg at echocardiography 5. The risk of developing pulmonary hypertension is much higher in CPFE than in IPF without emphysema (OR 19, 95% CI 5.1–68.7) 19. The prognosis of CPFE is worse than that of IPF without emphysema, an outcome determined by severe pulmonary hypertension and not only by the presence of associated emphysema 19. Indeed, pulmonary hypertension is associated with an increased risk of death in CPFE (hazard ratio 4.03), with 5-yr probability of survival of 25% in patients with pulmonary hypertension at echocardiography compared with 75% in those without pulmonary hypertension at diagnosis 5. CPFE syndrome has been included in the updated clinical classification of the aetiology of pulmonary hypertension under the category (3.3) of lung disease characterised by a mixed obstructive and restrictive pattern 20. However, pulmonary hypertension was evaluated by echocardiography only in both studies 5, 19 as haemodynamic analysis is not yet available for CPFE. Thus, the objective of the present study was to describe the haemodynamic characteristics and their relationship to survival in patients with CPFE and pre-capillary pulmonary hypertension demonstrated by RHC.
PATIENTS AND METHODS
Study design
This retrospective multicentre study was conducted by the French reference centre for rare pulmonary diseases (coordinator, J-F. Cordier), the French reference centre for pulmonary hypertension (coordinator, G. Simonneau) and the Groupe d’Etudes et de Recherche sur les Maladies “Orphelines” Pulmonaires (GERM“O”P), a collaborative group dedicated to the study of rare (so-called “orphan”) pulmonary diseases. Following the previous study 5, all participating physicians of the group were asked to prospectively report all cases of CPFE to the GERM“O”P registry, and were advised to perform at least yearly screening for pulmonary hypertension using echocardiography. RHC was performed at the discretion of the physician in cases of suspected severe pulmonary hypertension. Only cases with pre-capillary pulmonary hypertension confirmed at RHC were included, and clinical data were then collected retrospectively. Data collection ended in December 2008. HRCT scans were reviewed by two of us (V. Cottin and J-F. Cordier) to validate the imaging diagnostic criteria.
According to French legislation, the agreement of an ethics committee and informed consent are not required for retrospective collection of data corresponding to current practice. However, the database was anonymous and complied with the restrictive requirements of the Commission Nationale Informatique et Liberté, the organisation dedicated to privacy, information technology and civil rights in France. This study was approved by our institutional review board.
Inclusion criteria
The following criteria were required for inclusion.
1) Modified American Thoracic Society/European Respiratory Society criteria for the diagnosis of IPF 21, with exclusion of other known causes of interstitial lung disease, such as certain drug toxicities, environmental exposures and connective tissue diseases; impaired gas exchange (increased alveolar–arterial oxygen tension difference, decreased arterial oxygen tension with rest or exercise or decreased diffusing capacity of the lung for carbon monoxide (DL,CO)); bibasilar reticular abnormalities with basal and subpleural predominance, traction bronchiectasis and/or honeycombing, and with minimal ground-glass opacities on HRCT scan; transbronchial lung biopsy or bronchoalveolar lavage showing no features to support an alternative diagnosis; and at least three of the following: age >50 yrs, insidious onset of otherwise unexplained dyspnoea on exertion, duration of illness >3 months, bibasilar inspiratory crackles (dry or “velcro”-type in quality). As opposed to IPF criteria 21, evidence of restriction (reduced vital capacity) may or may not be present.
2) Presence of conspicuous emphysema (centrilobular and/or paraseptal) on HRCT scan, defined as well-demarcated areas of low attenuation delimitated by a very thin wall (>1 mm) or no wall.
3) Pre-capillary pulmonary hypertension defined by mean Ppa >25 mmHg at rest, with pulmonary arterial wedge pressure <15 mmHg and pulmonary vascular resistance (PVR) >240 dyn·s·cm−5 at RHC 22, 23.
Patients with connective tissue disease, hypersensitivity pneumonitis, drug-induced lung disease and pneumoconiosis were excluded from this study. Patients with pulmonary arterial hypertension related to portal hypertension, congenital heart disease, HIV infection, anorexigen exposure, and patients with pulmonary hypertension due to left heart disease and chronic thromboembolic pulmonary hypertension were excluded.
Investigations
We reviewed the medical records to collect information using a standardised form. The 6-min walk test (6MWT) and pulmonary function tests were performed according to recommendations 24, 25. RHC was performed as described 26. Date of diagnosis was defined as the date of RHC, and all data (symptoms, 6MWT, pulmonary function and echocardiography) were obtained at the time of the RHC. In the absence of guidelines on the treatment of pulmonary hypertension associated with parenchymal pulmonary diseases, treatment was left at the discretion of the physician, and included oral anticoagulation, diuretics, oxygen as needed and possible pulmonary arterial hypertension-specific therapy initiated after the RHC.
Statistical analysis
Microsoft Excel 2003 and SPSS 16.0 (SPSS, Chicago, IL, USA) were used for data analysis. All values were expressed as mean±sd. Two-tailed p-values <0.05 were considered statistically significant. Estimation of the probability of survival at each time point was performed using the Kaplan–Meier method, from the date of the first haemodynamic evaluation demonstrating pulmonary hypertension, to the end-points of death or censoring. All-cause mortality was used in survival statistics. Transplanted subjects were censored at the time of transplantation. Surviving patients were censored at the date of the last visit. Comparisons of survival were performed using the log rank test. The relationship between survival and selected baseline variables was examined for each variable using univariate analysis of hazard ratios based on the proportional hazards model.
RESULTS
Patient population
The study population included 40 patients (38 males and two females), with a mean±sd age of 68.2±8.9 yrs. Three patients had been included in a previous study 5. The mean delay between the first respiratory symptoms and the diagnosis of CPFE was 37±66 months. All patients except one were current or ex-smokers. One patient was exposed to agrochemical compounds. 13 patients (32%) had a history of atherosclerotic coronary artery disease and three (7%) of peripheral artery disease.
Baseline demographic and clinical data are shown in table 1⇓. Six patients (15%) had chronic bronchitis. 11 patients (27%) had clinical signs of right heart failure, two (5%) had a history of syncope. Finger clubbing was reported in 23 patients (57%) and basal crackles were present in 34 patients (85%).
Histopathology of the lungs available in six cases demonstrated a pattern of usual interstitial pneumonia and emphysema in all cases. The serum level of α1-antitrypsin measured in 14 cases was normal.
Functional evaluation and haemodynamics
New York Heart Association (NYHA) functional class was III or IV in 85% of the patients. 6-min walk distance (6MWD) was 244±126 m.
The mean delay between the diagnosis of CPFE and the RHC demonstrating pulmonary hypertension was 16±25 months, and the mean delay between the first respiratory symptoms and the diagnosis of pulmonary hypertension was 53±66 months. Results of RHC are presented in table 2⇓. 6MWD was 314±153 m in patients with NYHA functional class II, 255±126 m in class III and 180±91 m in class IV. Haemodynamic measurements showed a mean Ppa of 40±9 mmHg, and PVR of 521±205 dyn.s.cm−5. Mean Ppa, cardiac index and PVR did not significantly correlate with forced vital capacity (FVC), DL,CO or transfer factor of the lung for carbon monoxide (KCO).
In 27 patients (68%), the mean Ppa was >35 mmHg (and >40 mmHg in 48%), with a mean Ppa of 45±6 mmHg, PVR of 603±181 dyn·s·cm−5, cardiac index of 2.5±0.7 L·min−1·m−2 and 6MWD of 231±105 m.
Echocardiography showed dilated right cardiac cavities in 77% of cases, with paradoxical movement of the interventricular septum in 32% of the cases. Estimated systolic Ppa was 67±15 mmHg (range 20–100; n = 36) and was 35 mmHg or higher in 97% of patients. Pericardial effusion was not reported. The B-type natriuretic peptide level available for 14 patients was 340±298 pg·mL−1 (normal <100 pg·mL−1).
Mean values of lung volumes were within normal limits, contrasting with severely impaired gas exchange (mean KCO was 28±16% of predicted) (table 1⇑).
Outcome and survival analysis
Patients were followed for a median of 8±8 months (range 1–34 months). Treatment of pulmonary hypertension, pulmonary fibrosis and emphysema is shown on table 3⇓. 92% of the patients received long-term oxygen therapy. 24 patients (60%) received first-line therapy for pulmonary hypertension with bosentan, sildenafil or inhaled iloprost after RHC and were evaluated after 3–6 months. No statistically significant effect of treatment was observed regarding NHYA class, 6MWD or estimation of systolic Ppa at echocardiography.
At the end of the follow-up period, six patients (15%) had developed acute right heart failure, 14 (35%) had died, none had been lost to follow-up and four (10%) had been transplanted. The overall survival rate at 1 yr was 60±10% (fig. 1⇓). Death was due to hypoxaemia and chronic respiratory failure as a result of pulmonary hypertension and CPFE in 10 cases, cancer in three patients (lung, n = 2; throat, n = 1) and septic shock in one patient. A higher survival rate was observed in patients with DL,CO higher than the median value of 22% pred than in those with lower DL,CO (estimated 1-yr survival of 79.5±13.1% versus 43.5±18%; p = 0.046); in patients with PVR lower than the median value of 485 dyn·s·cm−5 (1-yr estimate of survival of 100% versus 47.6±15.1%; p = 0.008); and in patients with a cardiac index higher than the median value of 2.4 L·min−1·m−2 (1-yr survival of 79.1±13.8% versus 45.8±14.2%; p = 0.044). 10 of 20 patients with cardiac index <2.4 L·min−1·m−2 died, compared with two of the 20 patients with a cardiac index >2.4 L·min−1·m−2 (p = 0.01). The median survival in patients with cardiac index <2.4 L·min−1·m−2 was only 7.5 months (95% CI 1.2–13.9 months), and the median survival in patients with PVR >485 dyn·s·cm−5 was 6.6 months (95% CI 5.2–8.0 months). Nonsignificant trends for higher survival rate were observed in patients with higher transfer coefficient, lower mean Ppa, higher 6MWD and NYHA class II or III.
The results of the univariate analysis relating survival time to clinical, functional and haemodynamic characteristics measured at baseline in the overall population are shown in table 4⇓. High mean Ppa, high PVR, high heart rate and low DL,CO were significantly associated with a poor outcome. In addition, there was a trend for a poor outcome in patients with NYHA functional class IV, lower KCO and lower cardiac index. Similar results were obtained when selected numerical variables (mean Ppa, cardiac index and PVR) were treated as nonlinear and separated into two categories or into four quartiles (data not shown). No significant effect of therapy was observed in patients who received medical treatment for pulmonary hypertension compared with conventional therapy alone.
DISCUSSION
This is the first study of pulmonary hypertension confirmed by RHC in patients with CPFE.
CPFE is a distinct syndrome contrasting with both solitary IPF and emphysema by demonstrating relatively preserved lung volumes and airflow measurements, respectively 4. It is associated with a poor outcome related to the high prevalence of pulmonary hypertension, a characteristic feature in the natural history of CPFE syndrome 5, 19. In this study, we showed that: 1) pulmonary hypertension was demonstrated at RHC with a mean delay of only 16 months after the diagnosis of CPFE at HRCT scan; 2) patients had severe dyspnoea (with 85% in functional class III or IV) and severe exercise limitation (with a mean 6MWD of 244±126 m), despite subnormal spirometry and moderately severe haemodynamic parameters; 3) pulmonary hypertension in CPFE was associated with a dismal prognosis, with a 1-yr survival of only 60%; 4) a lower cardiac index, a lower transfer factor and increased PVR were associated with a shorter survival.
The pulmonary function characteristics of the patients included in the present study were strikingly similar to that of our previous report 5, although only three patients were included in both studies. This reproducibility underscores the clinical relevance of defining the syndrome of CPFE with simple diagnostic criteria, thus justifying our pragmatic approach based on the presence of conspicuous features of both emphysema and fibrosis on HRCT (e.g. noticeable without quantification of imaging features). Mean values of FVC, total lung capacity and forced expiratory volume in 1 s (FEV1) were normal, contrasting with severely impaired diffusion capacity of the lung, with mean values of DL,CO and KCO of only 24% and 28% pred, respectively, and severe hypoxaemia in 92% of the patients. The severe impairment of diffusion capacity probably represents the additive or synergistic effects of emphysema, fibrosis and pulmonary vascular disease, and is one of the hallmarks of CPFE syndrome 27. Most patients with CPFE syndrome were males, as previously shown 5, 19. Pulmonary hypertension is present in 47–90% of patients with CPFE, based on echocardiographic measurement of right ventricular systolic pressure, and is associated with an increased risk of death 5, 19. Since only patients with pre-capillary pulmonary hypertension confirmed at RHC were included in this study, we could not further evaluate the proportion of CPFE patients with pulmonary hypertension or the proportion of patients with post-capillary pulmonary hypertension. Selection bias towards the most severe cases cannot be excluded. However, echocardiography especially lacks specificity and accuracy in patients with advanced lung disease, including COPD 28 and IPF 29, 30, frequently leading to overdiagnosis of pulmonary hypertension 29, 30. The present study is the first to report on haemodynamic evaluation in CPFE patients with pre-capillary pulmonary hypertension confirmed by RHC, the gold standard for the diagnosis of pulmonary hypertension, thus allowing prognostic analysis according to haemodynamic parameters. The delay in diagnosing pulmonary hypertension in patients with CPFE seemed to be mostly related to the natural history of disease, with pulmonary hypertension detected at echocardiography during follow-up; therefore, we perform echocardiography at least once a year in any patient with CPFE.
Although no formal comparison can be made from this retrospective analysis, it is noteworthy that patients with pulmonary hypertension and CPFE had a dismal prognosis, with a 60% probability of survival at 1 yr from the diagnosis of pulmonary hypertension, similar to the probability of survival at 1 yr of 72% in patients with IPF and associated pulmonary hypertension at RHC 10. The reported median survival from the diagnosis is ∼3 yrs in IPF 31 and 6 yrs in one series of CPFE 5, 19; however, survival from the diagnosis of fibrosis was lower in patients with CPFE than in those with IPF without emphysema when compared within a single institution, mostly due to a higher incidence of pulmonary hypertension at echocardiography in CPFE 5, 19. In contrast, the survival was 36% at 5 yrs in patients with COPD and pulmonary hypertension (mean Ppa >25 mmHg) at onset of long-term oxygen therapy (Global Initiative for Chronic Obstructive Lung Disease IV) 18. The 1-yr survival of incident cases of pulmonary arterial hypertension in the national French registry was 88%, although with a worse haemodynamic profile than that of the present cohort (with a higher PVR index of 1,640 dyn·s·cm−5·m−2 and similar cardiac index of 2.5 L·min−1·m−2) 32. Thus, CPFE with associated pulmonary hypertension is a most severe condition with especially poor prognosis, worse than that of solitary COPD with associated pre-capillary pulmonary hypertension, and somewhat similar to that of IPF with pre-capillary pulmonary hypertension.
Pulmonary hypertension occurring in the context of chronic parenchymal lung disease is usually mild or moderate (i.e. with mean Ppa <35–40 mmHg). Recently, attention has focused on a subgroup of COPD patients, with severe “out-of-proportion” pre-capillary pulmonary hypertension despite long-term oxygen therapy 14, 15, 33, arbitrarily defined by mean Ppa >35–40 mmHg 14. These patients are prone to right heart failure and may share similarities with those with idiopathic pulmonary arterial hypertension 16. Interestingly, 68% of the patients included in the present study had pulmonary hypertension that was disproportionate to the underlying parenchymal lung disease, with mean Ppa >35 mmHg. The mean value of FVC was 86% pred in CPFE (compared with 49% pred in patients with IPF and associated pulmonary hypertension 10), and the mean value of FEV1 was 78% (compared with 55% in patients with COPD and disproportionate pulmonary hypertension 14).
Although the efficacy of drugs specifically indicated in pulmonary arterial hypertension has not been demonstrated in patients with pulmonary parenchymal disorders and associated out-of-proportion pulmonary hypertension, a large number of patients from the present study were treated off-label on an individual basis, thereby providing some preliminary information on the efficacy and safety of pulmonary hypertension therapy in this condition. No significant effect of treatment was found on survival. However, this result must be interpreted with caution, owing to the retrospective design, heterogeneity of treatment and lack of systematic haemodynamic assessment of the effect of treatment. Whether patients with CPFE and out-of-proportion pre-capillary pulmonary hypertension may benefit from treatment could be best evaluated in randomised controlled trials, although the feasibility of trials is challenged in such a rare and severe condition. Anyway, careful individual evaluation of patients under treatment should be obtained prospectively. Younger patients should be evaluated early for lung transplantation.
Prognostic factors in pulmonary hypertension associated with parenchymal pulmonary disease have not been extensively evaluated. The present study identified high PVR, low cardiac index and low DL,CO as significant predictors of a worse prognosis. Several haemodynamic factors associated with a shorter survival in pulmonary arterial hypertension were thus also associated with worse prognosis in this group of patients with CPFE and associated pulmonary hypertension. Other factors associated with a shorter survival in pulmonary arterial hypertension, such as NYHA functional class, 6MWD, pericardial effusion, elevated B-type natriuretic peptide levels and elevated right atrial pressure 34 were not significantly associated with survival in the present study, possibly due to insufficient statistical power.
Our study has several limitations, in particular its observational and uncontrolled design, with retrospective collection of data. Our results are subject to selection and treatment bias. Indication for therapy and choice of drug were not uniform among patients, limiting evaluation of the effect of treatment. Data presented here should not be interpreted as a proper evaluation of efficacy of treatment, which will require specific studies. However, cases of CPFE syndrome were prospectively identified by participating centres; data regarding haemodynamic parameters and survival were unlikely to be affected by the study design. Multivariate analysis could not be performed due to the sample size, and possible confounding effects of various variables related to survival time could not be evaluated. Long-term follow-up was not available in all patients owing to recent diagnosis; patients who were censored for short follow-up were not significantly different at baseline than the rest of the group; given the high mortality rate, it is unlikely that different results would have been found had the whole cohort been followed up for a longer period of time.
In conclusion, pulmonary hypertension may appear within a mean of only 16 months after the diagnosis of CPFE syndrome, mostly in patients requiring long-term oxygen therapy. Prognosis is poor, despite moderately severe haemodynamic parameters, with a 1-yr survival of 60% from the diagnosis of pulmonary hypertension.
Support statement
Financial support was provided by Hospices Civils de Lyon, Lyon, France ("PHRC regional 2005"; PHRC 2009).
Statement of interest
A statement of interest for M. Humbert can be found at www.erj.ersjournals.com/misc/statements.dtl
Acknowledgments
Contributors from the GERM”O”P group: A Berezné (Paris, France), D. Coëtmeur (St-Brieuc, France), I. Danner-Boucher (Nantes, France), D. Funke (Bern, Switzerland), D. Israel-Biet (Paris, France), E. Marchand (Yvoir, Belgium) and L. Mouthon (Paris, France).
The authors are indebted to all physicians who took care of the patients. We thank S. Zeghmar and A.C. Cadoré (Lyon, France) for data extraction and data entry.
Footnotes
-
For editorial comments see page 9.
- Received March 7, 2009.
- Accepted July 18, 2009.
- © ERS Journals Ltd