Abstract
Rationale: Incident pulmonary arterial hypertension was underrepresented in most pulmonary hypertension registries and may have a different disease profile to prevalent disease.
Objectives: To determine the characteristics and outcome of a purely incident, treatment-naive cohort of idiopathic, heritable, and anorexigen-associated pulmonary arterial hypertension and to determine the changes in presentations and survival over the past decade in the United Kingdom and Ireland.
Methods: All consecutive newly diagnosed patients from 2001 to 2009 were identified prospectively.
Measurements and Main Results: A total of 482 patients (93% idiopathic, 5% heritable, and 2% anorexigen-associated pulmonary arterial hypertension) were diagnosed, giving rise to an estimated incidence of 1.1 cases per million per year and prevalence of 6.6 cases per million in 2009. Younger patients (age ≤ 50 yrs) had shorter duration of symptoms, fewer comorbidities, better functional and exercise capacity, higher percent diffusing capacity of carbon monoxide, more severe hemodynamic impairment, but better survival compared with older patients. In comparison with the earlier cohorts, patients diagnosed in 2007–2009 were older, more obese, had lower percent diffusing capacity of carbon monoxide,, and more comorbidities, but better survival. Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL) equation, REVEAL risk score, and Pulmonary Hypertension Connection Registry survival equation accurately predicted survival of our incident cohort at 1 year.
Conclusions: This study highlights the influence of age on phenotypes of incident pulmonary arterial hypertension and has shown the changes in demographics and epidemiology over the past decade in a national setting. The results suggest that there may be two subtypes of patients: the younger subtype with more severe hemodynamic impairment but better survival, compared with the older subtype who has more comorbidities.
Most clinical trials and registries of idiopathic pulmonary arterial hypertension (IPAH) have included a mixture of incident and prevalent cases. Incident IPAH is reported to have worse outcome compared with prevalent cases. Demographics of IPAH seem to have changed compared with the NIH registry.
This is the first major national study of incident, treatment-naive patients with IPAH. The demographics and epidemiology of incident IPAH have continued to change over the past decade in a national setting. Survival has improved over the past decade but only after adjusting for the changing demographics of the more recent cohorts.
In rare diseases, such as pulmonary arterial hypertension (PAH), registries provide valuable information on the baseline characteristics and outcomes of the disease. In some of the earlier registries, definition and assessment of PAH were not standardized, numbers were small, and a significant number of patients did not have right heart catheterization to confirm the diagnosis (1–4). In addition, most pulmonary hypertension (PH) registries were composed of a mixture of incident and prevalent patients (5–9). However, prevalent patients have better prognosis compared with incident patients (10). We are concerned that current knowledge of PAH based on data obtained from a mixed population may be biased. Patients with severe or rapidly progressive disease may die early and never live long enough to be enrolled in registries. Conversely, patients with stable disease for a number of years have better survival and may be overrepresented in registries. In addition, patients outside registry participating centers were not included, potentially introducing further selection bias.
Recent reports from contemporary registries (8, 9, 11, 12) suggested that the typical idiopathic patients with PAH are now older and have better survival compared with patients from the NIH registry (5, 6). However, it is unclear whether younger and older patients in the current treatment era differ in their baseline characteristics and long-term outcomes. We are also interested to know whether the demographics, baseline characteristics, and survival of incident idiopathic PAH have changed over the past decade. Since 2001, the diagnosis, initiation of treatment, and long-term management of all adult patients with PH in the United Kingdom and Ireland has been centralized to eight designated PH centers (13). This setup also provides an opportunity to study the demographic and survival trends of virtually all newly diagnosed, treatment-naive patients within a single region with a similar healthcare system.
Five survival prediction equations are currently available to predict prognosis in PAH: NIH equation (6), French equation (10), Pulmonary Hypertension Connection Registry (PHC) equation (12), Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL) equation (14), and REVEAL risk score (15). It is unclear whether any of these models, which were all derived from a mixed incident-prevalent cohort, are applicable to a purely incident PAH population.
The objectives of this study are (1) to determine the epidemiology, baseline characteristics, and outcomes of an incident cohort of idiopathic, heritable, and anorexigen-associated PAH; (2) to determine whether younger and older patients differ in baseline characteristics and outcomes; and (3) to determine whether the demographics and epidemiology have changed over the past decade. Some of the results of this study have been previously reported in the form of abstracts (16–18).
This was a longitudinal observational study. Patients were identified prospectively and entered into local databases at the time of diagnosis. Data were collected on all consecutive treatment-naive, incident cases of idiopathic, heritable, and anorexigen-associated PAH diagnosed between January 1, 2001 and December 31, 2009 in all eight PH centers in the United Kingdom and Ireland. Only anonymized and nonidentifiable data were collected and the UK National Information Governance Board was informed of the data collection. This study was designed and conducted to define current care and hence formal ethics approval was deemed unnecessary by the West of Scotland Research Ethics Committee.
Idiopathic PAH was defined as PAH with no identifiable causes of PH. Patients with heritable and anorexigen-associated PAH have similar baseline characteristics and outcomes to idiopathic PAH (19) and were therefore analyzed together. PAH was defined as mean pulmonary arterial pressure greater than or equal to 25 mm Hg at rest and pulmonary artery occlusion pressure less than or equal to 15 mm Hg with a normal or reduced cardiac output. The date of diagnostic right heart catheterization was taken as the date of diagnosis. Patients with significantly abnormal pulmonary function test (FEV1, FVC, or TLC <60% predicted) (7, 20) or with evidence of parenchymal lung disease (of any severity) on chest computed tomography were excluded.
Analysis was performed using SPSS 18 (SPSS Inc, Chicago, IL) and SAS v9.2 (SAS Institute, Cary, NC). Mean ± SD or median (interquartile ranges [IQR]) were used to describe quantitative variables. Comparisons between two independent groups were performed using Student t test or Mann-Whitney U test. For multiple comparisons of quantitative data, one-way analysis of variance or Kruskal-Wallis test were used. Categorical variables were described by frequencies and percentages and comparisons between groups performed using chi-square or Fisher exact tests. Patients were divided into younger and older subgroups by the median age. Patients were divided into three cohorts based on the year of diagnosis (2001–2003, 2004–2006, and 2007–2009) to identify any changing trends over the course of the study period.
Survival endpoint was taken as either date of death or censoring. Patients were censored if they were transplanted, lost to follow-up, or alive on December 31, 2009. Survival was estimated using the Kaplan-Meier method and survival curves compared using the log-rank test.
Multivariate Cox regression was used to evaluate the prognostic significance of year of diagnosis after adjusting for variables known to have prognostic significance (age, functional class, 6-min walk distance, % predicted diffusion capacity of carbon monoxide [DlCO], mean right atrial pressure, and cardiac index) (6, 7, 11, 12, 14).
Expected survival was calculated for each patient for each of the survival prediction equations and compared with observed survival of the same patient estimated using the Kaplan-Meier method. The Brier score was used to assess the overall performance of each prediction model and can range from 0 to 1 (21–23). Lower score represents higher accuracy. Additional details on statistical analysis including handling of missing data are provided in the online supplement. P value less than 0.05 was considered significant throughout.
The combined population of the United Kingdom (http://www.statistics.gov.uk) and Ireland (http://www.cso.ie) in 2001–2003, 2004–2006, and 2007–2009 was estimated to be 61, 62, and 64 million, respectively. This gave rise to estimated incidence (per million per year) of idiopathic, heritable, and anorexigen-associated PAH in the United Kingdom and Ireland of 0.7, 0.9, and 1.1 in 2001–2003, 2004–2006, and 2007–2009, respectively. The estimated prevalence in 2009 was 6.6 cases per million population.
A total of 500 patients satisfied the inclusion and exclusion criteria. Five patients were too ill for baseline diagnostic right heart catheterization and were started on empirical PH treatment. Their diagnoses were subsequently confirmed on right heart catheterization on treatment. All other patients with missing baseline right heart catheterization data (n = 18) were excluded leaving a final total of 482 patients in this study cohort. Out of the 482 patients, 92.9% (n = 448) were idiopathic, 5.4% (n = 26) heritable, and 1.7% (n = 8) anorexigen-associated PAH. Baseline characteristics of the whole cohort are shown in Table 1 (additional results on comparisons of selective baseline characteristics between age quartile subgroups are available in the online supplement).
| Baseline Characteristics | All Cases | Age ≤ 50 | Age > 50 | P Value* |
| Age, yr (n = 482) | 50.1 (17.1) | 36.5 (9.3) | 65.1 (8.3) | <0.001 |
| Sex, % female (n = 482) | 69.9% (337) | 73.2% (180) | 66.5% (157) | 0.112 |
| Ethnicity, % nonwhite (n = 424) | 12.3% (52) | 15.4% (34) | 8.9% (18) | 0.041 |
| Symptoms (n = 437) | ||||
| Peripheral edema | 30.9% (135) | 23.5% (52) | 38.4% (83) | 0.001 |
| Syncope | 22.7% (99) | 33.5% (74) | 11.6% (25) | <0.001 |
| Presyncope | 11.9% (52) | 14.9% (33) | 8.8% (19) | 0.048 |
| Duration of symptoms, mo, median (IQR) (n = 404) | 18 (9–36) | 12 (6–24) | 24 (12–36) | <0.001 |
| Functional class (n = 456) | ||||
| I and II | 15.5% (72) | 19.7% (45) | 11.9% (27) | 0.031 |
| III | 66.7% (304) | 65.9% (151) | 67.4% (153) | |
| IV | 17.5% (80) | 14.4% (33) | 20.7% (47) | |
| Smoking history (n = 361) | ||||
| Current smoker | 14.4% (52) | 16% (28) | 12.9% (24) | <0.001 |
| Ex-smoker | 41.3% (149) | 30.9% (54) | 51.1% (95) | |
| Never smoker | 44.3% (160) | 53.1% (93) | 36% (67) | |
| Comorbidities (n = 455) | ||||
| IHD | 12.1% (58) | 1.3% (3) | 24.2% (55) | <0.001 |
| Hypertension | 26.6% (121) | 11% (25) | 42.3% (96) | <0.001 |
| Atrial fibrillation | 5.3% (24) | 0% (0) | 10.6% (24) | <0.001 |
| Diabetes | 14.3% (65) | 5.3% (12) | 23.3% (53) | <0.001 |
| Hypothyroidism | 11.6% (53) | 7.5% (17) | 15.9% (36) | 0.005 |
| 6MWD, m (n = 260) | 292.4 (123) | 330 (118.7) | 246.1 (112.2) | <0.001 |
| Pulmonary function | ||||
| % FEV1 (n = 350) | 85.3 (14.7) | 85.5 (13.6) | 85.1 (15.8) | 0.785 |
| % FVC (n = 346) | 94 (16.4) | 91.2 (14.6) | 96.7 (17.6) | 0.002 |
| % TLC (n = 204) | 95.4 (13.9) | 96.3 (13.4) | 94.4 (14.3) | 0.324 |
| % DlCO (n = 331) | 62 (20.9) | 67 (17.6) | 56.7 (22.8) | <0.001 |
| Hemodynamics | ||||
| mRAP, mm Hg (n = 439) | 10.1 (6) | 10 (5.8) | 10.2 (6.1) | 0.738 |
| PASP, mm Hg (n = 427) | 86.2 (21.1) | 88.7 (22.6) | 83.8 (19.2) | 0.016 |
| PADP, mm Hg (n = 424) | 34.7 (12.3) | 38.6 (12.9) | 31.1 (10.5) | <0.001 |
| mPAP, mm Hg (n = 457) | 54.1 (13.9) | 57.2 (14.9) | 51 (12.2) | <0.001 |
| PAOP, mm Hg (n = 408) | 9.2 (3.5) | 8.8 (3.5) | 9.5 (3.5) | 0.035 |
| TPG, mm Hg (n = 404) | 44.4 (13.7) | 47.8 (14.5) | 41.3 (12.2) | <0.001 |
| SvO2, % (n = 395) | 61.5 (9.5) | 62.5 (10.3) | 60 (8.6) | 0.053 |
| CO, L⋅min−1 (n = 410) | 4 (1.5) | 4 (1.5) | 4 (1.5) | 0.830 |
| CI, L⋅min−1⋅m−2 (n = 366) | 2.1 (0.7) | 2.1 (0.7) | 2.1 (0.7) | 0.683 |
| PVR, WU (n = 395) | 12.8 (6.3) | 13.9 (6.7) | 11.8 (5.8) | 0.001 |
| PVRI, WU⋅m2 (n = 355) | 23.1 (10.3) | 24.8 (11.2) | 21.5 (9.1) | 0.003 |
The median age was 50 years (IQR, 36–65) and was similar between males and females. There was no statistically significant difference in sex ratio across the four age quartiles. A total of 13.5% (n = 65) of patients were older than 70 years of age at the time of diagnosis.
Ethnicity data were available in 424 patients of which 12.3% were of nonwhite ethnic origin. Compared with whites, nonwhites were younger (39.5 [IQR, 32–59.8] vs. 50 yr [IQR, 36–65]; P = 0.015); had greater female predominance (84.6% vs. 68.3%; P = 0.016); and lower proportion of current or ex-smoker (34.2% vs. 58.6%; P = 0.007). There was no statistically significant difference in body mass index, functional class, exercise capacity, or hemodynamics between whites and nonwhites. Patients with missing ethnicity data had similar baseline characteristics to the white ethnic group.
Mean body mass index was 28.3 ± 6.3 kg/m2 and was higher in females (females, 28.8 ± 6.6 kg/m2 vs. males, 27.3 ± 5.5 kg/m2; P = 0.02). A total of 37.9% of females and 24.8% of males were obese.
Presenting symptoms were available in 437 patients. The median duration of symptoms (n = 404) was 18 months (IQR, 9–36) and was similar in males and females. Dyspnea was the most common symptom (98.9%; n = 432) at the time of diagnosis. Other less common symptoms included peripheral edema (30.9%; n = 135); chest pain (29.1%; n = 127); syncope (22.7%; n = 99); light-headedness (19.7%; n = 86); fatigue (13.5%; n = 59); cough (13.1%; n = 57); presyncope (11.9%; n = 52); palpitation (8.9%; n = 39); weight loss (4.8%; n = 21); and hemoptysis (3.9%; n = 17).
Older (age > 50 yr) patients were more likely to present with peripheral edema, whereas younger (age ≤ 50 yr) patients were more likely to present with syncope and presyncope (Table 1).
Baseline exercise capacity was evaluated with 6-minute walk test in 260 patients and incremental shuttle walk test in 120 patients. Mean 6-minute walk distance and median shuttle distance at the time of diagnosis were 292.4 ± 123 m and 190 m (IQR, 82.3–340), respectively.
Most patients had severe symptoms by the time of diagnosis, with 66.7% and 17.5% in functional class III and IV, respectively. Distribution of functional class was similar between males and females. A greater proportion of younger patients were in functional class I and II compared with older patients (Table 1).
A total of 14.4% (n = 52) and 41.3% (n = 149) of patients were current and ex-smokers, respectively. History of smoking was more prevalent in older (Table 1) and male patients (11.4% of male patients were current and 57.9% ex-smoker compared with 15.8% and 33.6% of female patients were current and ex-smoker, respectively; P < 0.001).
Patients had preserved spirometry and lung volumes but disproportionately reduced % DlCO (Table 1). Older patients had lower % DlCO compared with their younger counterparts (Table 1).
When compared across smoking history subgroups, ex-smokers had lower % DlCO (55.9% ± 20.8%) compared with never (65.9% ± 20.2%; P = 0.001) and current smokers (64.6% ± 21.3%; one-way analysis of variance; P = 0.001).
Patients had severe hemodynamic compromise (Table 1) and this was similar for males and females. Younger patients had significantly higher mean pulmonary artery pressures and transpulmonary gradient, lower pulmonary artery occlusion pressures, similar cardiac indices, and higher pulmonary vascular resistance indices compared with older patients (Table 1; see Table E1 in the online supplement).
First-line treatment data were available in 479 patients. A total of 44.3% (n = 212) of patients were started on an endothelin receptor antagonist; 29.2% (n = 140) on a phosphodiesterase type V inhibitor; 18.8% (n = 90) on prostaglandins; 5% (n = 24) on a calcium channel blocker; 2.1% (n = 10) on combination therapy; and 0.6% (n = 3) of patients received no treatment.
A total of 46.3% (n = 222) and 6% (n = 29) of patients received combination and triple therapy, respectively, at some point in their disease. A higher proportion of younger patients received sequential combination therapy (57.6% vs. 34.5%; P < 0.001), triple therapy (10.6% vs. 1.3%; P < 0.001), prostaglandins (51.8% vs. 28.1%; P < 0.001), calcium channel blocker (10.2% vs. 3%; P = 0.002), and transplantation (5.3% vs. 0.4%; P = 0.002) compared with their older counterparts. There was no statistically significant difference in treatment between males and females.
Fourteen patients received heart-lung or bilateral lung transplantation and 12 patients had atrial septostomy over the study period.
In the later cohort (2007–2009), patients were older, more obese, and had more comorbidities and lower % DlCO (Table 2). There was no statistically significant difference in sex, distribution of functional class, exercise capacity, or pulmonary hemodynamics over the three time periods.
| Baseline Characteristics and Treatment | 2001–2003 | 2004–2006 | 2007–2009 | P Value* |
| Age, yr, median (IQR) (n = 482) | 45 (34.5–59) | 52 (36–65) | 52 (37.8–68) | 0.003 |
| Duration, mo, median (IQR) (n = 404) | 18 (8–39) | 18 (8.5–30) | 18 (10–30) | 0.883 |
| BMI, kg/m2 (n = 398) | 28.1 (6.4) | 26.9 (5.8) | 29.8 (6.5) | <0.001 |
| % Obese (n = 398) | 31.2% (29) | 25.5% (37) | 43.1% (69) | 0.004 |
| Comorbidities (n = 455) | ||||
| IHD | 5.8% (7) | 11.4% (18) | 16.3% (33) | 0.018 |
| Diabetes | 5.7% (6) | 15% (23) | 18.3% (36) | 0.012 |
| % DlCO (n = 331) | 68.6 (21.8) | 61.8 (19) | 59.2 (21.3) | 0.009 |
| First-line treatment (n = 479) | ||||
| PD5i | 7.6% (9) | 20.8% (33) | 48.8% (98) | <0.001 |
| ERA | 51.3% (61) | 57.9% (92) | 29.4% (59) | |
| Prostaglandins | 31.1% (37) | 17% (27) | 12.9% (26) | |
| CCB | 7.6% (9) | 3.1% (5) | 5% (10) | |
| Combination | 2.5% (3) | 0.6% (1) | 3% (6) | |
| None | 0% (0) | 0.6% (1) | 1% (2) |
A total of 129 patients died over the study period with observed 1-, 2-, 3-, and 5-year survival of 92.7%, 84%, 73.3%, and 61.1%, respectively. Younger patients (aged ≤50 yr) had better survival with 1-, 2-, 3-, and 5-year survival of 94.7%, 91%, 87.2%, and 74.7% compared with 1-, 2-, 3-, and 5-year survival of 90%, 75.5%, 57.1%, and 43.7% in patients aged more than 50 years (Figure 1) (log-rank; P < 0.001). Patients aged greater than 50 years were more likely to die compared with patients aged less than or equal to 50 years (unadjusted hazard ratio, 3; P < 0.001).
Figure 1. Kaplan-Meier survival curves according to age. Patients less than or equal to 50 years old (corresponding to median age) had better survival compared with patients greater than 50 years old.
[More] [Minimize]There was no difference in overall survival for the three time periods. However, after adjusting for age, functional class, 6-minute walk distance, % DlCO, mean right atrial pressure, and cardiac index, patients diagnosed in 2001–2003 were more likely to have a shorter survival time compared with patients diagnosed in 2007–2009 (hazard ratio, 1.96; P = 0.019).
Some survival prediction equations performed better (i.e, lower Brier score) than others (Table 3). At 1 year, there was no difference in accuracy among the two REVEAL models and the PHC equation (see Table E2). At 2 and 3 years, the PHC and French equations performed better than the NIH equation (see Tables E3 and E4). Accuracy of all survival equations (NIH, French, and PHC) decreased with time from diagnosis. Figure 2 shows the survival curves of actual observed survival of the UK and Ireland cohort versus predicted survival of the same patients calculated using the NIH, French, PHC, and REVEAL equations.
| Brier Scores By Bootstrap (95% Confidence Interval) | |||
| Survival Prediction Models | 1-Year Prediction | 2-Year Prediction | 3-Year Prediction |
| NIH equation | 0.1231 (0.1034–0.1448) | 0.2103 (0.1836–0.2384) | 0.2799 (0.2437–0.3137) |
| French equation | 0.0642 (0.0413–0.0919) | 0.1206 (0.0918–0.1539) | 0.1831 (0.1458–0.2251) |
| PHC equation | 0.0499 (0.0214–0.0851) | 0.1015 (0.0651–0.1439) | 0.1691 (0.1263–0.2181) |
| REVEAL equation | 0.0462 (0.0177–0.0815) | ||
| REVEAL risk score | 0.0468 (0.0193–0.0809) | ||
Figure 2. Observed versus predicted survival using the NIH equation, French equation, Pulmonary Hypertension Connection Registry (PHC) equation, and Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL) equation. Only 1-year predicted survival was calculated for the REVEAL equation. Predicted survival from the REVEAL and PHC equations were comparable with the observed survival of this United Kingdom and Ireland incident idiopathic, heritable, and anorexigen-associated pulmonary arterial hypertension cohort. The NIH and French equations underestimated survival of this study cohort.
[More] [Minimize]The present study describes the largest series so far reported of incident, treatment-naive, idiopathic, heritable, and anorexigen-associated patients with PAH. It is a national study encompassing all patients diagnosed with idiopathic, heritable, and anorexigen-associated PAH in the United Kingdom and Ireland. First, we found age influenced the phenotypic characteristics of patients. Second, we found that despite the aging demographics over the past decade, survival of patients in the United Kingdom and Ireland has improved. However, there was no change in duration of symptoms before diagnosis. Patients were still diagnosed late with severe functional and hemodynamic impairment. Finally, we validated five survival prediction equations with our incident cohort and demonstrated superiority of some prediction equations over others.
Our study showed age influenced the phenotypic characteristics of incident idiopathic, heritable, and anorexigen-associated PAH. Although we dichotomized our study cohort into younger and older subgroups and showed significant differences between the two, in real life these are likely to represent the two ends of a continuous spectrum of the same disease. When the study cohort was divided into four groups by age quartile (see Table E1), similar relationship was observed between these variables and aging. Comorbidities in older patients may partly explain their worse functional capacity despite less severe hemodynamic impairment. Although Shapiro and coworkers (24) also reported similar age-related findings with lower 6-minute walk distance, transpulmonary gradient, and pulmonary vascular resistance in elderly patients with PH, 28% of all patients and 56% of the elderly group in that study had pulmonary artery occlusion pressure greater than or equal to 15 mm Hg and thus did not meet hemodynamic criteria for idiopathic PAH. Although a normal pulmonary artery occlusion pressure does not exclude pulmonary venous hypertension (25), especially in older patients with comorbidities, we believe it is unlikely many patients in this study were misclassified. All patients were diagnosed by multidisciplinary teams of PH experts at each center in accordance with contemporary national and international PH guidelines (13, 26, 27). Routine work-up of new patients includes ECG, echocardiogram, and cardiac magnetic resonance imaging. Results of these assessments were taken into consideration in a multidisciplinary setting before a final diagnosis was made in any new patient.
Younger patients in this study had a significantly higher frequency of syncope and presyncope, whereas older patients were more likely to present with peripheral edema. Syncope and presyncope were also common in childhood idiopathic PAH, whereas edema was rare (28). Younger patients with their higher functional class and exercise capacity may exert themselves further leading to presyncope or syncope. Onset of peripheral edema may reflect right ventricular failure. Interestingly, even in patients with similar mean right atrial pressure, older patients are more likely to present with edema compared with younger patients (results not shown). Older patients, despite their less severe hemodynamic impairment, seem to have less physiologic reserve than their younger counterparts to cope with the progressive PH and subsequent failing right ventricle. This needs to be taken into account when deciding choice of treatment for individual patients.
Older patients in this study have lower % DlCO compared with their younger counterparts. Most reference equations for DlCO use age, height, and sex in a linear equation to predict DlCO (29). However, the rate of decline of DlCO accelerates with aging (30). This may potentially result in overprediction of predicted DlCO in older patients. A higher proportion of current and ex-smokers observed in older patients in this study may also explain their lower % DlCO because of the presence of coexisting smoking-related lung disease. However, we have excluded patients with any coexisting parenchymal lung disease from this study. Because parenchymal lung abnormalities on chest computed tomography are difficult to quantify objectively, we therefore excluded abnormalities of all severity from this study.
Demographics, epidemiology, and survival of idiopathic, heritable, and anorexigen-associated PAH have changed in the United Kingdom and Ireland between 2001 and 2009. We observed that over the 9-year study period, patients were older, more obese, and have more comorbidities in the later years. We speculate that these changing demographics are largely caused by changes in referral pattern rather than an actual change in the disease. Alternatively, this may indicate that we are dealing with two possible subtypes of patients: younger patients similar to the primary PH cohort of the NIH registry (5), versus the older patients characterized by the presence of comorbidities. The incidence also seemed to increase year by year, largely caused by a rising number of referrals to PH centers. Although there was no difference in overall survival among the three time periods, when survival was adjusted for age and other known prognostic factors, it is encouraging to observe that the survival of incident idiopathic, heritable, and anorexigen-associated PAH in the United Kingdom and Ireland has improved over the past decade. However, delays in diagnosis persist. Not only was the delay unchanged within the United Kingdom and Ireland over the past decade, but there was no improvement when compared with patients in the NIH registry from the 1980s (5). We believe the persistent delay in diagnosis over time is largely caused by the increasing proportion of elderly patients with PAH. We found older patients have longer duration of symptoms before diagnosis. It is possible that the nonspecific PH symptoms were more commonly attributed to their old age or other more common illnesses before PAH was suspected. More education is needed outside the PH community to recognize that idiopathic PAH is no longer a disease affecting only young patients.
The ability to predict prognosis helps to guide treatment decisions. The NIH equation (6) was derived from a cohort from the predisease targeted therapy era and was used in many clinical trials to suggest survival benefit of the study drug. Our results confirmed that the NIH equation underestimated survival of idiopathic PAH and should no longer be used in the current age of modern PAH therapies. The French equation (10) also underestimated survival of our study cohort. We believe there are three possible explanations. First, the French equation may simply reflect outcomes of patients in France in 2002–2006, at a time when several currently approved PAH therapies were not yet available (10). Second, patients with less severe disease or those who remained stable on oral therapies may not be referred to one of the French University Hospitals. In contrast, our study is a truly national study inclusive of all idiopathic patients with PAH of all severity in the United Kingdom and Ireland. Finally, in addition to using the same pulmonary function criteria as the French registry, we have also excluded patients with any radiologic evidence of coexisting parenchymal lung disease from our study. These patients with coexisting lung disease may have worse survival and some may be included in the French registry. Although REVEAL equation (14) and risk score (15) have lower Brier scores than the PHC (12) equation, the differences were not statistically significant. In addition, analysis of REVEAL equation and risk scores was limited to only 1-year prediction. Because there was no difference in accuracy among the REVEAL equation (14), REVEAL risk score (15), and the PHC equation (12), user-friendliness of an equation may also influence its applicability in daily clinical practice. Finally, all equations were better at predicting survival at 1 year than at longer terms. This is not unexpected because the course of disease of a patient can be highly variable. Any deteriorations or changes to treatment during follow-up may have an impact on their subsequent outcome.
The main limitation of our study is the unavoidable incompleteness in our dataset caused by its observational nature, with a small amount of missing data and uneven frequency of follow-up and investigations performed at each visit. Future studies looking into the relationship between age and right heart function at baseline and on treatment may help to understand the phenotypic and survival differences observed between the younger and older patients.
The current UK and Ireland study is the largest incident series of treatment-naive idiopathic, heritable, and anorexigen-associated patients with PAH to date. We have shown younger patients have a different phenotype and survival compared with their older counterparts. The demographics and epidemiology have changed and survival has improved from 2001 to 2009. However, patients still present in advanced stages. Awareness of the changing demographics and epidemiology of idiopathic, heritable, and anorexigen-associated PAH as reported in this study is imperative to the early diagnosis and treatment of this otherwise rare and fatal disease. Finally, we have demonstrated that some survival prediction equations more accurately predict survival of incident PAH in the United Kingdom and Ireland than others.
The authors thank Jonathan Alsop (Numerus Ltd.) for providing statistical advice, and Dr. Benji Schreiber and Dr. Konstantinos Dimopoulos for comments on the manuscript.
| 1. | Tueller C, Stricker H, Soccal P, Tamm M, Aubert JD, Maggiorini M, Zwahlen M, Nicod L. Epidemiology of pulmonary hypertension: new data from the Swiss registry. Swiss Med Wkly 2008;138:379–384. |
| 2. | Stricker H, Domenighetti G, Popov W, Speich R, Nicod L, Aubert JD, Soler M. Severe pulmonary hypertension: data from the Swiss Registry. Swiss Med Wkly 2001;131:346–350. |
| 3. | Jing ZC, Xu XQ, Han ZY, Wu Y, Deng KW, Wang H, Wang ZW, Cheng XS, Xu B, Hu SS, et al.. Registry and survival study in Chinese patients with idiopathic and familial pulmonary arterial hypertension. Chest 2007;132:373–379. |
| 4. | Appelbaum L, Yigla M, Bendayan D, Reichart N, Fink G, Priel S, Schwartz Y, Richman P, Picard E, Goldman S, et al.. Primary pulmonary hypertension in Israel: a national survey. Chest 2001;119:1801–1806. |
| 5. | Rich S, Dantzker D, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, Fishman AP, Goldring RM, Groves BM, Koerner SK, et al.. Primary pulmonary hypertension: a national prospective study. Ann Intern Med 1987;107:216–223. |
| 6. | D'Alonzo GE, Barst R, Ayres SM, Bergofsky EH, Brundage BH, Detre KM, Fishman AP, Goldring RM, Groves BM, Kernis JT, et al.. Survival in patients with primary pulmonary-hypertension: results from a national prospective registry. Ann Intern Med 1991;115:343–349. |
| 7. | Humbert M, Sitbon O, Chaouat A, Bertocchi M, Habib G, Gressin V, Yaici A, Weitzenblum E, Cordier JF, Chabot F, et al.. Pulmonary arterial hypertension in France: results from a national registry. Am J Respir Crit Care Med 2006;173:1023–1030. |
| 8. | Badesch DB, Raskob G, Elliott CG, Krichman AM, Farber HW, Frost AE, Barst RJ, Benza RL, Liou TG, Turner M, et al.. Pulmonary arterial hypertension baseline characteristics from the REVEAL registry. Chest 2010;137:376–387. |
| 9. | Thenappan T, Shah SJ, Rich S, Gomberg-Maitland M. A USA-based registry for pulmonary arterial hypertension: 1982–2006. Eur Respir J 2007;30:1103–1110. |
| 10. | Humbert M, Sitbon O, Chaouat A, Bertocchi M, Habib G, Gressin V, Yaici A, Weitzenblum E, Cordier JF, Chabot F, et al.. Survival in incident and prevalent cohorts of patients with pulmonary arterial hypertension. Eur Respir J 2010;36:549–555. |
| 11. | Humbert M, Sitbon O, Chaouat A, Bertocchi M, Habib G, Gressin V, Yaici A, Weitzenblum E, Cordier JF, Chabot F, et al.. Survival in patients with idiopathic, familial, and anorexigen-associated pulmonary arterial hypertension in the modern management era. Circulation 2010;122:156–163. |
| 12. | Thenappan T, Shah SJ, Rich S, Tian L, Archer SL, Gomberg-Maitland M. Survival in pulmonary arterial hypertension: a reappraisal of the NIH risk stratification equation. Eur Respir J 2010;35:1079–1087. |
| 13. | National Pulmonary Hypertension Centres of the UK and Ireland. Consensus statement on the management of pulmonary hypertension in clinical practice in the UK and Ireland. Heart 2008;94:i1–i41. |
| 14. | Benza RL, Miller DP, Gomberg-Maitland M, Frantz RP, Foreman AJ, Coffey CS, Frost A, Barst RJ, Badesch DB, Elliott CG, et al.. Predicting survival in pulmonary arterial hypertension insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL). Circulation 2010;122:164–172. |
| 15. | Benza RL, Gomberg-Maitland M, Miller DP, Frost A, Frantz RP, Foreman AJ, Badesch DB, McGoon MD. The REVEAL risk score calculator in patients newly diagnosed with pulmonary arterial hypertension. Chest 2012;141:354–362. |
| 16. | Ling Y, Johnson M, Kiely D, Condliffe R, Elliot C, Gibbs S, Howard L, Pepke-Zaba J, Sheares K, Corris P, et al.. Demographic trends and changes in long term outcome of incident idiopathic, heritable and anorexigen-associated pulmonary arterial hypertension between 2001 to 2009: results from the Pulmonary Hypertension Registry of the UK and Ireland [abstract]. Presented at the European Respiratory Society Annual Congress, September 24–28, 2011, Amsterdam, The Netherlands. |
| 17. | Ling Y, Johnson M, Kiely D, Condliffe R, Elliot C, Gibbs S, Howard L, Pepke-Zaba J, Sheares K, Corris P, et al.. Influence of age on clinical phenotypes of incident idiopathic pulmonary arterial hypertension: results from the Pulmonary Hypertension Registry of the UK and Ireland [abstract]. Thorax 2011;66:A34–A5. |
| 18. | Ling Y, Johnson M, Kiely D, Condliffe R, Elliot C, Gibbs S, Howard L, Pepke-Zaba J, Sheares K, Corris P, et al.. Prediction of survival in pulmonary arterial hypertension using survival equations: results from the Pulmonary Hypertension Registry of the UK and Ireland [abstract]. Thorax 2011;66:A35–A36. |
| 19. | Souza R, Humbert M, Sztrymf B, Jais X, Yaici A, Le Pavec J, Parent F, Herve P, Soubrier F, Sitbon O, et al.. Pulmonary arterial hypertension associated with fenfluramine exposure: report of 109 cases. Eur Respir J 2008;31:343–348. |
| 20. | Peacock AJ, Murphy NF, McMurray JJV, Caballero L, Stewart S. An epidemiological study of pulmonary arterial hypertension. Eur Respir J 2007;30:104–109. |
| 21. | Steyerberg EW, Vickers AJ, Cook NR, Gerds T, Gonen M, Obuchowski N, Pencina MJ, Kattan MW. Assessing the performance of prediction models: a framework for traditional and novel measures. Epidemiology 2010;21:128–138. |
| 22. | Brier G. Verification of forecasts expressed in terms of probability. Mon Weather Rev 1950;78:1–3. |
| 23. | Gerds TA, Cai T, Schumacher M. The performance of risk prediction models. Biom J 2008;50:457–479. |
| 24. | Shapiro BP, McGoon MD, Redfield MM. Unexplained pulmonary hypertension in elderly patients. Chest 2007;131:94–100. |
| 25. | Halpern SD, Taichman DB. Misclassification of pulmonary hypertension due to reliance on pulmonary capillary wedge pressure rather than left ventricular end-diastolic pressure. Chest 2009;136:37–43. |
| 26. | National Pulmonary Hypertension Centres of the UK and Ireland. Consensus statement on the management of pulmonary hypertension in clinical practice in the UK and Ireland. Thorax 2008;63:ii1–ii41. |
| 27. | Galiè N, Hoeper MM, Humbert M, Torbicki A, Vachiery JL, Barbera JA, Beghetti M, Corris P, Gaine S, Gibbs JS, et al..; 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 the International Society of Heart and Lung Transplantation (ISHLT). Guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Respir J 2009;34:1219–1263. |
| 28. | Moledina S, Hislop AA, Foster H, Schulze-Neick I, Haworth SG. Childhood idiopathic pulmonary arterial hypertension: a national cohort study. Heart 2010;96:1401–1406. |
| 29. | Crapo RO, Hankinson JL, Irvin C, MacIntyre NR, Zoter KZ, Rise RA. Single-breath carbon monoxide diffusion capacity (transfer factor). Recommendations for a standard technique: 1995 update. Am J Respir Crit Care Med 1995;152:2185–2198. |
| 30. | Sherrill DL, Enright PL, Kaltenborn Walter T, Lebowitz Michael D. Predictors of longitudinal change in diffusion capacity over 8 years. Am J Respir Crit Care Med 1999;160:1883–1887. |
Supported by unrestricted educational grants from Pfizer, GSK, Actelion, and Bayer.
Author Contributions: Y.L. was involved in the conception and design of the study, acquisition, analysis, and interpretation of the data, and writing the manuscript, and had substantial involvement in its revision before submission. All other authors were involved in the conception and design of the study, analysis and interpretation of the data, and writing the manuscript, and had substantial involvement in its revision before submission.
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.201203-0383OC on July 12, 2012
Author disclosures
