Abstract
We aimed to characterize and estimate survival rates in patients diagnosed with pulmonary arterial hypertension (PAH) in western Denmark in the modern management era. All incident cases of PAH were consecutively enrolled in our single-center prospective cohort study between January 2000 and March 2012. A total of 134 patients fulfilling the inclusion criteria were followed up from first diagnostic right heart catheterization to either death or the end of the study. Kaplan-Meier survival analysis was used to estimate 1-, 3-, and 5-year survival rates with 95% confidence intervals (CIs). Survival in the total cohort was 86.4% (95% CI, 79.3%–91.2%) after 1 year, 72.9% (95% CI, 64.1%–79.9%) after 3 years, and 65.4% (95% CI, 55.8%–73.4%) after 5 years. Significantly better survival was seen in the group of patients with PAH associated with congenital heart disease than in the group of patients with idiopathic PAH, heritable PAH, connective tissue disease, HIV infection, and portal hypertension. In conclusion, survival rates in the Danish PAH population were similar to or slightly better than survival rates estimated in other modern registries. However, PAH remains a fatal disease, despite modern targeted therapies.
Pulmonary arterial hypertension (PAH) is a progressive disease that despite optimal treatment often results in early death. The disease is characterized by proliferation and remodeling of the vessel wall in the small arteries of the pulmonary circulation. This causes a progressive increase in pulmonary vascular resistance and, hence, an increase in pulmonary arterial pressure. The elevated afterload increases the workload of the right ventricle and ultimately reduces right ventricle function, leading to right ventricle failure and death.1–3 PAH is a rare disease, with an approximated prevalence reported between 15 and 50 cases per million.4–6 According to the latest World Health Organization (WHO) classification, 7 PAH is clinically grouped into idiopathic PAH (IPAH), familial/heritable PAH (FPAH), and PAH with associated conditions (APAH), such as connective tissue disease (APAH-CTD), congenital heart disease, HIV infection, and portal hypertension. Each of these subgroups shares the same pathophysiology.
The development of specific therapies for PAH (endothelin receptor antagonists, prostanoids, and phosphodiesterase 5 inhibitors) has improved the prognosis. In the pretreatment era, the National Institutes of Health (NIH) registry reported very low survival rates, with 1- and 3-year survival rates of 68% and 48%, respectively, in a population of 194 patients with primary pulmonary hypertension (PH). 8 In the era of modern treatment, several registry-based studies have been published and have indicated improved survival. Two US studies have reported survival in PAH: the Pulmonary Hypertension Connection (PHC) registry and the REVEAL (Registry to Evaluate Early and Long-Term PAH Disease Management) registry. The PHC registry evaluated 576 incident and prevalent cases among patients with IPAH, FPAH, and APAH and found 1- and 3-year survival rates of 86% and 69%, respectively. 9 The REVEAL registry studied 2,635 incident and prevalent cases of PAH in patients enrolled from 2006 through 2009. With estimated 1- and 3-year survival rates of 85% and 68%, 10 respectively, the results were similar to those of the PHC registry. Comparably, a French multicenter registry included 674 incident and prevalent cases of PAH from 2002 to 2003 and found estimated 1-and 3-year survival rates of 87% and 67%, respectively. 11
No analysis of the Danish population of patients with PAH has been published. Therefore, our aim was to characterize and estimate survival rates for all patients from western Denmark diagnosed with PAH (WHO group 1 PH) during the modern management era.
METHODS
The Danish PH registry
The registry was initiated in January 2000 at Aarhus University Hospital, Aarhus, Denmark. Patients from hospitals in western Denmark, a background population of 3 million, are referred to our center, where diagnostic workup and treatment of PH is centralized. The total population of Denmark is 5.6 million. Between January 2000 and March 2012, patients satisfying the definition of PH were consecutively enrolled in the Danish PH registry. The registry collected information for demographic profiles and data from diagnostic workups. During follow-up visits, parameters from clinical examinations were collected and registered. These parameters included data on hemodynamic right heart catheterization, 6-minute walk distance (6MWD), WHO functional class, echocardiography, and medication.
Study design
We designed a single-center prospective cohort study based on incident cases of PAH in the population of western Denmark. PAH was defined as precapillary PH with the absence of other reasons for PH, such as lung disease or chronic thromboembolic PH. A hemodynamic profile with mean pulmonary arterial pressure of ≥25 mmHg and pulmonary arterial wedge pressure of ≤15 mmHg was used to define precapillary PH. 12 PAH was further classified according to the Nice classification into subgroups based on etiology. 7 Patients who met the defined hemodynamic profile were included in the current study. According to the latest classification, 7 patients with pulmonary capillary hemangiomatosis (PCH) and veno-occlusive pulmonary disease (PVOD) were classified as a separate group (group 1) and excluded from the study. For purposes of analysis, the five PAH subgroups were aggregated into two diagnosis groups: congenital heart diseases (Eisenmenger syndrome and systemic-to-pulmonary shunts) and noncongenital diseases (IPAH, FPAH, APAH-CTD, HIV infection, and portal hypertension).
Follow-up
All follow-up visits were conducted at our institution and managed by a small group of clinical PH specialists. Follow-up interval and initiation of relevant therapy were determined at the physician's discretion, and treatment was based on to current international recommendations.12–15 Treatment (both targeted and conventional therapy) was registered at each visit. Because of changes in therapy, availability, and recommendations through the time period, treatment is registered as administration on last visit.
Outcome and exposure variables
Primary outcome in the study was all-cause mortality. Survival time was estimated from the date of diagnostic right heart catheterization. At end of the study on February 25, 2013, vital status was obtained by query in the Danish Civil Registration System, which is the national registry of all Danish citizens. 16 At baseline and during follow-up, demographics, invasive hemodynamics, WHO functional class, and 6MWD were registered. Age at the time of diagnosis was categorized as follows: 0–39, 40–59, and ≥60 years. WHO functional class was categorized into three groups: WHO functional class 1/2, 3, and 4.
Statistical analysis
Continuous variables were reported as means with SD or as medians with interquartile range (IQR) where appropriate. Categorical variables were reported as absolute frequencies and percentages. Kaplan-Meier curves were generated with time from diagnosis as the timescale and all-cause mortality as the outcome. Patients lost to follow-up were censored at the time of their last visit. Patients who underwent lung transplantation were censored at the time of transplantation. Otherwise, patients were censored at the end of the study. When examining individual Kaplan-Meier survival curves for sex, diagnosis, age, and WHO functional class, an unadjusted log-rank analysis was used. A P value (two-sided) below 0.05 was considered statistically significant. STATA software (STATA IC for Windows, ver. 11.2) was used for statistical analyses.
This study was approved by and compliant with the requirements of the Danish Data Protection Agency and the Danish Health and Medicines Authority.
RESULTS
A total of 1,142 incident patients with PH were successively enrolled in the Danish PH registry. Of these, 137 met the definitions of PAH; 3 patients were excluded because of PVOD or PCH. Thus, the study population consisted of 134 patients: IPAH (n = 43), FPAH (n = 1), collagen vascular disease associated (n = 31), congenital systemic-to-pulmonary shunt (n = 28), Eisenmenger syndrome (n = 20), portal hypertension (n = 8), and HIV related (n = 3). During the study period, 1 patient was lost to follow-up due to emigration, and 1 patient underwent lung transplantation. For the total cohort and the two diagnosis groups (congenital heart diseases and noncongenital diseases), a selection of demographic, hemodynamic, and other baseline characteristics are presented in Table 1.
Baseline characteristics
Note: Data are mean (SD), unless otherwise indicated. PAH: pulmonary arterial hypertension; IPAH: idiopathic PAH; FPAH: familial/heritable PAH; APAH-CTD: connective tissue disease–associated PAH; APAH-CHD: congenital heart disease–associated PAH; WHO: World Health Organization; FC: functional class; 6MWD: 6-minute walk distance; mPAP: mean pulmonary arterial pressure; RAP: right atrial pressure; PAWP: pulmonary arterial wedge pressure; CI: cardiac index; PVR: pulmonary vascular resistance; NO: nitric oxide.
Data were incomplete for 6MWD for the group with congenital heart diseases.
The mean age at diagnosis was 50 ± 20.9 years. The majority of patients were female. Female sex was especially predominant in the noncongenital group. The majority of the cohort was in WHO functional class 3 on diagnosis. On the last follow-up visit, 71% of the population was receiving therapy with at least one targeted medication, and 33% were receiving combination therapy (Table 2).
Medical therapy on last follow-up visit
Note: Data are percentage of the total cohort. PDE-5I: phosphodiesterase 5 inhibitors; ETA: endothelin receptor antagonists.
Survival
The median follow-up period for the cohort was 3.6 years (IQR, 1.5–7.4). At the end of the study, 49 patients were dead, 1 was lost to follow-up, and 84 were alive. Median survival time from diagnostic right heart catheterization for the 49 deaths was 1.6 years (IQR, 0.5–3.9). Table 3 presents survival estimates for the total cohort and the two diagnosis groups. The Kaplan-Meier curve for the total cohort is illustrated in Figure 1. Survival was significantly better in patients with congenital heart disease than in patients with IPAH, FPAH, APAH-CTD, HIV infection, and portal hypertension (Fig. 2). In addition, survival was better in patients with WHO functional class 1/2 at baseline than in patients with WHO functional class 3/4. Furthermore, older age at diagnosis was associated with increased mortality, while no difference in survival was observed between male and female patients. No significant difference was observed in survival times for patients diagnosed before and after 2005. This was also the case for the comparison of patients diagnosed before and after 2008.
Kaplan-Meier survival estimates
Note: Data are mean (95% confidence interval).
Kaplan-Meier curve for the total cohort of 134 patients with pulmonary arterial hypertension.
Kaplan-Meier curves according to sex, age at diagnosis, World Health Organization (WHO) functional class, and diagnosis group.
DISCUSSION
For the first time, characterization and survival estimation in a Danish population-based cohort of PAH patients treated in the modern management era have been presented. Comparison of our survival data with those of the historic pre-modern-management NIH registry indicates improvement in survival. We found a 1-year survival of 86% and a 3-year survival of 73%, while the NIH registry found 68% survival after 1 year and 48% survival after 3 years (Table 4). This correlates with the French and US studies, which indicated similar improvement since the introduction of modern therapy. Baseline characteristics in our cohort showed that the majority of patients were female, mean age was approximately 50 years, and the majority were in WHO functional class 3 on diagnosis. Similar trends were found in the French registry and the two US registries.4,9,10 Higher WHO functional class and older age at diagnosis were associated with worse outcome (Fig. 2). No difference was found between the sexes. Survival for patients with a congenital diagnosis was significantly better than that for patients with a noncongenital diagnosis, thus indicating a correlation between etiology of the disease and survival. The difference in survival between groups could also be partially explained by a higher representation of younger patients in the congenital group (Table 1), which likely results in lower noncardiac mortality. An association between survival and a congenital diagnosis has also been reported in both the US REVEAL registry and the French registry.
Comparison of study populations and survival among registries
Note: NIH: National Institutes of Health; PHC: Pulmonary Hypertension Connection; PH: pulmonary hypertension; NA: not available; WHO: World Health Organization; FC: functional class.
The ratio of incident to prevalent cases was not specified; however, both were included.
There was no difference in survival with regard to time of diagnosis during the enrollment period.
Survival in an incident cohort
The survival estimates presented in our population represent survival in an incident cohort. Data from the French registry indicate a significantly worse survival for incident cases than for prevalent cases. 11 This is most likely explained by the fact that patients with a good prognosis (i.e., longer-term survivors) are included in the prevalent cohort to a greater degree than patients with a poorer prognosis who have died before inclusion in a prevalent cohort. The PHC and REVEAL studies also had a large percentage of prevalent cases.9,10 The difference in 1-year survival between the studies is minimal (Table 4). Our cohort had a 1-year survival of 86%, similar to the French and US studies, in which survival ranged from 85% to 87% after 1 year. Regarding longer-term survival, our cohort revealed a 3-year survival rate of 72.9%, compared with 67%–69% in the previous studies. After 5 years, our cohort presented a 65.4% survival rate, compared with 57%–61% in the two US cohorts (follow-up in the French study was limited to 3 years).9–11 Given that incident cases are expected to show worse survival than prevalent cases, our long-term survival data are intriguing and might suggest slightly better survival in the Danish population than in the US and French populations.
Limitations
Our small study population is a major limitation. It prevented us from identifying independent predictors of mortality, and caution is needed when making comparisons between studies; thus, larger studies are warranted. Our center performs all diagnostic workups for and treatment of PH for all of western Denmark. Although this ensures that all patients diagnosed with PAH are included in the Danish PH registry, patients suspected of PH may never be referred from local hospitals because of comorbidities or high age. This referral bias might result in overly optimistic survival estimates in our study.
More focus on PH and the establishment of a national registry would yield more detailed knowledge on the prognosis of this fatal disease.
Conclusion
In conclusion, our results indicate that survival in the population of PAH patients has improved since targeted therapy was introduced. The survival rates estimated in our study indicate that survival within the Danish population is similar to or possibly slightly better than survival estimates presented in other PAH registries.
Footnotes
ACKNOWLEDGMENTS
We give special thanks to Dorte Frydendahl Bjerre, Aarhus University Hospital, for supporting and helping in the collection of data.