Characteristics and survival of adult Swedish PAH and CTEPH patients 2000-2014.

Abstract Objectives: The Swedish Pulmonary Arterial Hypertension Register (SPAHR) is an open continuous register, including pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension (CTEPH) patients from 2000 and onwards. We hereby launch the first data from SPAHR, defining baseline characteristics and survival of Swedish PAH and CTEPH patients. Design: Incident PAH and CTEPH patients 2008–2014 from all seven Swedish PAH-centres were specifically reviewed. Results: There were 457 PAH (median age: 67 years, 64% female) and 183 CTEPH (median age: 70 years, 50% female) patients, whereof 77 and 81%, respectively, were in functional class III–IV at diagnosis. Systemic hypertension, diabetes, ischaemic heart disease and atrial fibrillation were common comorbidities, particularly in those >65 years. One-, 3- and 5-year survival was 85%, 71% and 59% for PAH patients. Corresponding numbers for CTEPH patients with versus without pulmonary endarterectomy were 96%, 89% and 86% versus 91%, 75% and 69%, respectively. In 2014, the incidence of IPAH/HPAH, associated PAH and CTEPH was 5, 3 and 2 per million inhabitants and year, and the prevalence was 25, 24 and 19 per million inhabitants. Conclusion: The majority of the PAH and CTEPH patients were diagnosed at age >65 years, in functional class III–IV, and exhibiting several comorbidities. PAH survival in SPAHR was similar to other registers.


Introduction
Clinical research registries have been fundamental in defining the natural history and outcome of diseases related to underlying mechanisms, presentation at diagnosis and response to treatment strategies. Registers have also been important in characterising patient populations and developing risk stratification tools. [1] This is evident in pulmonary arterial hypertension (PAH), where an early register showed a median survival of only 1-2.8 years for untreated PAH patients from diagnosis, depending on the underlying cause. [2] PAH is characterised by pulmonary vasoconstriction and remodelling, the latter including medial hypertrophy, intimal proliferation and fibrosis, adventitial thickening, in situ thrombosis and complex lesions. These changes lead to elevated pulmonary vascular resistance, right ventricular failure and eventually death. [3] Pulmonary hypertension (PH) includes a broad spectrum of diseases, diagnosed by right heart catheterisation and categorised into five groups and a number of subgroups, [3] where treatment and outcome vary. [1] In 2013, an extensive review including 11 major PH/PAH registers [1,[4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21] from six countries was published. [1] Ten registers presented PAH populations, as a whole or as selected subgroups, and one register included chronic thromboembolic pulmonary hypertension (CTEPH) besides PAH. [19] The number of participating centres in each register ranged from 1 to 55 and the number of patients from 72 to 3515. [1] The registers varied with respect to prospective or retrospective design, the inclusion of prevalent and/or incident cases and the time period of inclusion. Furthermore, most register reports did not include patient comorbidities, a factor of growing importance considering the increased age at the time of diagnosis.
The Swedish Pulmonary Arterial Hypertension Register (SPAHR) constitutes an open continuous register initiated in 2008 that includes patients with PAH or CTEPH from 2000 and onwards (Appendix). For the first time, in an era of changing demographic patterns and the initiation of new PAH therapies, we report baseline characteristics, comorbidities and survival of prevalent and incident Swedish PAH and CTEPH patients from 2000 to 2014, but with a focus on the incident cases from 2008 to 2014.

Study subjects and ethical consideration
The patients were locally informed about their participation in SPAHR, including clinical follow-up, and had the right to refuse participation. SPAHR was approved by the National Board of Health and Welfare and the Swedish Data Inspection. The study was approved by the local ethics board in Lund (Dnr 2014/92) and adhered to the Declaration of Helsinki.

Study design and methods
The Swedish Pulmonary Arterial Hypertension Register and recruiting PAH centres SPAHR constitutes an open continuous register with the intention to register all Swedish patients with PAH (Group 1), including idiopathic/hereditary (IPAH/HPAH) and associated PAH (APAH), or with CTEPH (Group 4). Registration of PAH and CTEPH patients in SPAHR began January 2008, with the aim to retrospectively include those that were alive in the year of 2000, or diagnosed afterwards. The main objective of the present report was to describe baseline characteristics, comorbidities and survival of the adult Swedish PAH and CTEPH patients, with a focus on the incident population from 2008 to 2014. Data from the whole population, e.g. prevalent and incident cases from 2000 to 2014, are reported in the online supplement (Supplementary Tables I and II).
All Swedish PAH-centres located at the University hospitals in Gothenburg, Link€ oping, Lund, Stockholm, Umeå, Uppsala and € Orebro have been including patients in SPAHR since the start. Swedish PAH-centres, in addition, to evaluate and treat patients with PAH, are also responsible for the investigation and treatment of CTEPH patients. The PAH-centres therefore refer the completed evaluation of their CTEPH patients, including pulmonary angiography, for operability assessment, to specialised centres in pulmonary endarterectomy (PEA), such as Karolinska University hospital in Stockholm, Sweden, and Skejby hospital in Århus, Denmark.
Each PAH centre was responsible for allocating their patients to the correct PH group. Pulmonary function test and computer tomography were clinically used to rule out PH due to lung disease (PH group 3). Echocardiography along with invasive pulmonary artery wedge pressure (PAWP) from right heart catheterisation was additionally clinically used to rule out PH due to left heart disease (PH group 2). Validation of the data entered in the register was performed by on-site monitoring by an independent monitor where missing data and erroneous values were checked against original medical journal source data. The monitoring met all predefined criteria.
The national coverage was calculated with the support of the National Board of Health and Welfare. It was estimated that in 2014 approximately 85% of all patients in Sweden with a PAH diagnosis on PAH-targeted therapy had been included in SPAHR. The lack of full completeness may partly be due to registration in the Swedish national register for patients with grown-up congenital heart disease.
Diagnosis and demographic data in SPAHR Baseline in the register was defined as the date of the right heart catheterisation that established the PH diagnosis. Precapillary PH was defined at rest by a mean pulmonary arterial pressure (MPAP) 25 mmHg, a PAWP 15 mmHg, at a normal or reduced cardiac output, according to the Dana Point Classification. [3] Vasoreactivity test by NO-inhalation at baseline had been registered in 89% of the incident IPAH/ HPAH patients.
Data from clinical investigations at the time of diagnosis and follow-up visits have been entered at each PAH-centre. Beside haemodynamic evaluations, these investigations include demographics, comorbidities, World Health Organization Functional Class (FC), 6-min walk test and blood biochemistry (e.g. NT-proBNP). Electrocardiography, echocardiography, pulmonary function test, pulmonary ventilation/perfusion scintigraphy, pulmonary angiography and high-resolution computer tomography may also have been entered when locally feasible. In 2010, the patient reported outcome measurements (Cambridge Pulmonary Hypertension Outcome Review, CAMPHOR) were added.

Analysis
To reflect the adult incident population, aged 18 or older, patients diagnosed with PAH (PH group 1) or CTEPH (PH group 4), from 1 January 2008, through 31 December 2014, were included in the treatment and survival analysis. Moreover, adult patients alive anytime between 1 January 2000 and 31 December 2007, were included in a separate cohort survival analysis. The combined prevalent and incident population 2000-2014 is described in the Supplementary Tables I and II. Baseline characteristics, comorbidities and treatment allocations are presented as medians with interquartile ranges (IQR) and as proportions with 95% confidence intervals (CI). Decimals were rounded off to the nearest whole numbers. National population data from Statistics Sweden (in Swedish, Statistiska Centralbyrån, SCB, http://www.statistikdatabasen.scb.se) were used for calculation of prevalence and incidence rates. Data in the survival analysis were censored at death, transplantation, change of baseline diagnosis or loss to follow-up, or 31 December 2014, whatever came first. Survival stratified by diagnosis and by cohort years are presented as Kaplan-Meier survival plots with Log-Rank tests for overall comparisons. For comparison of outcomes, univariate and multivariable Cox regression analyses, adjusted for age and sex were used. Hazard ratios with 95% CI are presented as (HR [95% CI]). A p value <0.05 was considered statistically significant. SPSS version 23.0 (Armonk, NY) was used for statistical computation.

Incident population 2008 through 2014
The number of incident patients 2008-2014 (Table 1) with PAH or CTEPH, respectively, were 457 and 183, respectively, whereof 301 and 130, respectively, were alive, 31 December 2014. Baseline diagnosis for the incident population was divided even between IPAH/HPAH of 35% and APAH of 36%, whereas 29% were diagnosed with CTEPH ( Figure 1A). In 2014, the proportions had changed such that 50%, 29% and 21% were diagnosed with IPAH/HPAH, APAH and CTEPH ( Figure 1B), respectively. The age distribution for the incident PAH and CTEPH patients, divided by gender are shown in Figure 2.
In 2014, the prevalence of IPAH/HPAH, APAH and CTEPH were 25, 24 and 19 per million inhabitants, respectively, whereas the corresponding incidence were 5, 3 and 2 per million inhabitants and year, respectively.

Combined prevalent and incident population 2000-2014
The median age at baseline, for the combined population of PAH patients 2000-2014, was 63 years, 66% were females and 80% in functional class III-IV (Supplementary Tables I).

Incident population 2008-2014
The median age at baseline, for the incident population 2008-2014, was 67 years, 64% were females and 77% were in functional class III-IV (Table 1). The median age of the CTEPH patients at baseline was 70 years, 50% were females and 81% in functional class III-IV (Table 1). Both the PAH and CTEPH patients showed exercise limitations as measured by 6-min walk distance and increased levels of NT-proBNP. Baseline haemodynamic profiles were similar in patients with IPAH/HPAH and CTEPH.

Incident population 2008-2014
The first treatments registered after baseline, distributed for PAH subgroups, gender and age are shown in Table 3. There were no clear differences in treatment allocation between men and women or between age groups.

Discussion
The Swedish PAH Register SPAHR include incident and prevalent patients with PAH or CTEPH from 2000 and onwards, and is here presented for the first time, with a focus on the incident population from 2008 to 2014. To the best of our knowledge, this represents the first multicentre PAH and CTEPH register study in Scandinavia. All Swedish PAH centres, located to the seven Swedish University hospitals, report to SPAHR. Thus, the vast majority of patients with PAH-specific therapy in Sweden have been included in the analysis. This allows a description of a contemporary real-life Swedish patient population, including baseline characteristics, comorbidities and initial treatments, as well as survival.
In line with other registers [1], the present report shows an improved survival compared to historical data of untreated IPAH patients, [2] with a mean survival for the incident IPAH/HPAH and APAH-CTD populations of 4.7 and 4.3 years, respectively. This improvement is attributed to new treatments and treatment strategies as well as an increased awareness of PAH and its phenotypes. Survival is, however, still severely impaired, which may be explained by diagnosis late in the disease and patients being old at the time of diagnosis, thus exhibiting more comorbidities. In addition, the response to PAH therapy among older patients may differ compared to the younger population. [1,14] Among the PAH subgroups, APAH-CHD showed the best and APAH-CTD the worst survival. Similar results have previously been found in a local report from Southern Sweden [22] and in studies emphasising the devastating outcome in patients with connective tissue disease. [15] Our data furthermore support the concept of survival bias with regards to prevalent and incident cases of PAH [9], however, the difference disappeared when the cohorts were adjusted for age and sex.
PAH has for long been considered a disease that predominantly affect young to middle age women, but recent reports suggest a different scenario. The majority of the PAH patients in the present report were 60 years or older at the time of diagnosis and it was only below the age of 70 that more women than men were diagnosed. This was reflected both in those living with the disease and those newly diagnosed. The present report complement findings from registries in the United Kingdom and Ireland [20], suggesting two subtypes of patients, older with worse functional capacity and more comorbidities and younger with more severe haemodynamic impairment, but a better response to PAH treatment and better survival. The COMPERA register indeed also showed IPAH more frequently diagnosed in elderly patients with a more balanced gender ratio, exhibiting other clinical features and lower response to medical therapy and a higher age-adjusted mortality. [23] The REVEAL register from the USA also reported IPAH patients as older than earlier believed, however, the mean age of 49 was lower than in SPAHR and the female predominance was maintained. [12,17,18] Our SPAHR evaluation represents an upto-date, multicentre Swedish register study that additionally also complement the Swiss PH register, reporting a mean age at diagnosis of 57 and 63 years for PAH and CTEPH, respectively [24], as well as the Danish single centre study where incident PAH patients exhibited a mean age of 50 years at diagnosis between January 2000 and March 2012. [26] These studies [24,25] did, however, not describe patient comorbidities to the same extent as our SPAHR investigation. The older age at PAH diagnosis in SPAHR, with a median age at diagnosis of 67 years is also supported by the study of incident cases in Germany 2014, exhibiting a mean age at diagnosis of 64 years. [26] Finally, the CTEPH population in SPAHR showed a predominance of diagnosis at 60 years or older and those who underwent PEA exhibited a better survival than those who were not operated. Even though these patient cohorts differ, depending on the location of the vascular disease, or with regards to comorbidities and age, it emphasises the importance of referring CTEPH patients for operability assessment.
The high prevalence of IPAH/HPAH patients in Sweden of 25 per million inhabitants exceeded those reported from the Scottish (9 per million), French (5.9 per million) and Spanish (4.6 per million) registers. [1,6,[8][9][10]19] In addition, the incidence of 5 per million inhabitants and year was also higher than in most other reports of 1-2.6 per million inhabitants and year. [1,6,[8][9][10]19] However, a recent report from Germany in 2014, shows similar numbers as in SPAHR, with a prevalence of 25.9 per million inhabitants and an incidence of 3.9 per million inhabitants and year. [26] Interestingly, both SPAHR and the German cohort share the high age at diagnosis. [26] This might reflect an increased awareness of PAH as well as an improved screening process of patients with dyspnoea. However, it might also indicate that present haemodynamic criteria may not be sufficient discriminators for PAH in the elderly population. New criteria for PAH diagnosis that take into account the numbers and combination of comorbidities could therefore potentially be of importance in future definitions of PAH, even when a diastolic dysfunction is not found. Systemic hypertension, diabetes, atrial fibrillation and ischaemic heart disease were common among the PAH patients >65 years in the present report. While it is not surprising that an elderly population have more comorbidities than younger individuals, it possibly highlights the importance of accounting for the relation between comorbidities and age at the time of diagnosis. Thus, it is possible, that even though the elderly fulfilled the Dana Point haemodynamic criteria for PAH, the comorbidities may have had an additional influence, even though not detected by echocardiography, or by an elevated PAWP.
In conclusion, the present evaluation of the Swedish incident PAH and CTEPH patients from 2008 to 2014, as well as the combined population of prevalent and incident cases from 2000 to 2014, represents an up-to-date uniform national coverage of these patient cohorts, including baseline characteristics, comorbidities and survival, in a Scandinavian country in the modern treatment era. The diversity and increase of comorbidities with older de novo PAH patients at diagnosis, predominantly diagnosed late in FC III-IV, are highlighted. Though survival has improved, as new PAH-targeted treatments and treatment strategies have evolved, survival remains poor and differ between PAH subgroups.
Identifying individual treatment strategies, as well as new markers for earlier diagnosis and detection of treatment response, are therefore essential. In addition, the influence of comorbidities on present criteria for the diagnosis of PAH and response to PAH-targeted treatment need further attention in future studies, especially in the elderly group.