Author + information
- Received December 18, 2001
- Revision received April 11, 2002
- Accepted May 15, 2002
- Published online August 21, 2002.
- Olivier Sitbon, MD*,* (, )
- Marc Humbert, MD*,
- Hilario Nunes, MD*,
- Florence Parent, MD*,
- Gilles Garcia, MD*,
- Philippe Hervé, MD*,
- Maurizio Rainisio, PhD* and
- G.érald Simonneau, MD*
- ↵*Reprint requests and correspondence:
Dr. Olivier Sitbon, Service de Pneumologie et Réanimation, Hôpital Antoine Béclère, 157 rue de la Porte de Trivaux, 92141 Clamart, France.
Objectives We sought to determine the factors associated with long-term survival in patients with primary pulmonary hypertension (PPH) treated with continuous epoprostenol infusion.
Background Epoprostenol improves survival in patients with PPH in New York Heart Association (NYHA) functional class III or IV. However, some patients do not benefit from epoprostenol and must be considered for lung transplantation. The best timing for listing these patients on a lung transplantation program is currently unknown.
Methods Between December 1992 and January 2001, 178 patients with PPH in NYHA functional class III or IV were treated with epoprostenol. The 6-min walk test (WT) and right-sided heart catheterization were performed at baseline, after three months on epoprostenol and thereafter once a year.
Results Overall survival rates at one, two, three, and five years were 85%, 70%, 63%, and 55%, respectively. On univariate analysis, the baseline variables associated with a poor outcome were a history of right-sided heart failure, NYHA functional class IV, 6-min WT ≤250 m (median value), right atrial pressure ≥12 mm Hg, and mean pulmonary artery pressure <65 mm Hg. On multivariate analysis, including both baseline variables and those measured after three months on epoprostenol, a history of right-sided heart failure, persistence of NYHA functional class III or IV at three months, and the absence of a fall in total pulmonary resistance of >30%, relative to baseline, were associated with poor survival.
Conclusions Survival of patients with PPH treated with epoprostenol depends on the severity at baseline, as well as the three-month response to therapy. These findings suggest that lung transplantation should be considered in a subset of patients who remain in NYHA functional class III or IV or in those who cannot achieve a significant hemodynamic improvement after three months of epoprostenol therapy, or both.
Primary pulmonary hypertension (PPH) is an uncommon disorder of unknown etiology, characterized by a progressive increase in pulmonary artery pressure (PAP) and pulmonary vascular resistance, often leading to right ventricular failure and death, especially in patients in New York Heart Association (NYHA) functional class III or IV (1–3).
In the national North American Registry of 194 patients with PPH, the median survival was 2.8 years (2). Most patients included in the registry received conventional therapy with an oral anticoagulant, diuretic, oxygen, digoxin, and calcium channel blocker. Among the baseline hemodynamic variables, high levels of right atrial pressure (RAP) (≥20 mm Hg) and mean PAP (≥85 mm Hg) and a low cardiac index (CI) (<2 l/min/m2) were associated with poor survival.
Epoprostenol (i.e., prostaglandin I2 or prostacyclin [PGI2]) is a potent, short-acting vasodilator and inhibitor of platelet aggregation that is produced by the vascular endothelium (4). Long-term treatment with continuous intravenous infusion of PGI2 (Flolan by Glaxo Smith Kline, Brentford, UK) has been shown to improve exercise capacity, hemodynamics, and quality of life in most patients with severe PPH refractory to conventional therapy (5,6). A significant improvement in survival with PGI2 treatment was demonstrated in an unblinded, randomized, 12-week study of patients with severe PPH (5). In contrast, there are few data on the long-term survival of patients with PPH treated with continuous intravenous PGI2, although Shapiro et al. (7) described significantly improved survival in a retrospective analysis of 69 patients with PPH treated with PGI2 over a five-year period, as compared with an historic group of patients from the U.S. National Institutes of Health Registry.
Initially proposed as a bridge to lung transplantation, intravenous PGI2 is now considered as a first line therapy and an alternative to lung transplantation in severe patients with PPH. However, some patients with PPH do not benefit from this therapy and in these cases the only therapeutic option remaining is lung transplantation. There is a need to identify the factors that enable the early prediction of a failure of PGI2 therapy so that these patients may be placed on a waiting list for lung transplantation at the earliest possible opportunity. The major objective of the present study was to analyze the usefulness of clinical and hemodynamic parameters that predict the outcome in PGI2-treated patients with PPH.
From December 1992 to January 2001, 178 patients who were above 15 years old with severe PPH were treated with long-term intravenous infusion of PGI2. The diagnosis of PPH was established by standard diagnostic criteria (1,8). All patients were classified as being in NYHA functional class III or IV, despite optimal medical therapy, including oral anticoagulants, diuretics, long-term oxygen, and oral vasodilators. Exclusion criteria were as: 1) pulmonary hypertension associated with connective tissue disease, congenital heart disease, portal hypertension, or human immunodeficiency viral infection; 2) distal chronic thromboembolic pulmonary hypertension; 3) chronic pulmonary disease; and 4) an acute pulmonary vasodilator response that predicted a clinical response to oral calcium channel blockers (9). An historic group of 135 patients with PPH matched for NYHA functional class and who never received PGI2 was used for comparative survival analysis. Data from this control group were extracted from our personal data base. Patients from this historic group were comparable to those from the PGI2-treated group in terms of functional class, exercise capacity and hemodynamics (10). This retrospective study was approved by our Institutional Review Board.
Hemodynamics and exercise capacity
The baseline evaluation included a medical history and physical examination, routine blood tests, a nonencouraged 6-min walk test (WT), as previously described (11), and right-sided heart catheterization using standard techniques. Acute vasodilator responsiveness was evaluated using a short-term vasodilator challenge (inhaled nitric oxide or intravenous PGI2), as previously described (12).
A clinical evaluation, including NYHA functional class and 6-min WT, was done every three months. Repeated right-sided heart catheterization was performed after three months and one year of treatment with PGI2 and thereafter once a year.
For each patient, venous access for the infusion of PGI2 (Flolan) was obtained by insertion of a permanent, tunneled Groshong catheter (Bard Access Systems, Salt Lake City, Utah) into a subclavian vein. Epoprostenol was infused continuously with the use of a portable infusion pump (CADD-Plus by Sims Deltec Inc., St. Paul, Minnesota). During the initial hospital period, patients were trained in pump programming, drug preparation, sterile technique, and catheter care. Epoprostenol therapy was initiated at a dose of 1 ng/kg/min, and the dose was increased by 1 ng/kg/min every 12 h up to 10 ng/kg/min. After this dose was achieved, patients were discharged from the hospital. Dose adjustments were made systematically to reach a mean level of 14 ± 4 ng/kg/min at three months. Thereafter, dose adjustments were based on clinical symptoms consistent with clinical deterioration or the occurrence of adverse events, distance walked during exercise testing, and hemodynamic measurements.
Data were stored on a PC-based data spreadsheet. Analysis was performed using the Statview version 5.0 statistical package (SAS Institute, Cary, North Carolina). All data are expressed as the mean value ± SD. One-way analysis of variance with repeated measures was performed for functional and hemodynamic values obtained at baseline, after three months and after one year of treatment with intravenous PGI2. Multiple comparisons were made when the F test value was statistically significant. A p value <0.05 was considered as statistically significant. The Student t test and chi-square test for independence were used to compare differences between mean values and between rates measured at baseline for patients in NYHA functional class III and those in class IV, as appropriate.
For the survival analysis, we used the commencement of PPH therapy (conventional or PGI2) as the start point to determine the survival duration. The Kaplan-Meier method was used to estimate the proportion of patients surviving at each time point. Univariate analysis based on the proportional hazards model was used to examine the relationship between survival and selected demographic, medical history and hemodynamic variables measured at the initial catheterization and after three months of PGI2 therapy. The results are expressed as hazard ratios with 95% confidence intervals. For continuous variables, we chose to separate patients into two groups on both sides of the median value. Patients who underwent lung transplantation were considered as censored at the date of transplantation.
Multivariate analysis based on the Cox proportional hazards regression model was used to examine the independent effect of each variable on survival, controlling for possible confounding variables. Two separate regressions were performed: the first one included only baseline variables and the second one included both baseline variables and those measured after three months of PGI2 therapy.
Baseline clinical and hemodynamic variables
The patients’ clinical characteristics, exercise capacity, and hemodynamic measurements at baseline are shown in Table 1.
The frequencies of symptoms at the initiation of PGI2 therapy were as follows: 100% for dyspnea on exertion, 42% for syncope or near syncope, and 35% for lower limb edema. Raynaud’s phenomenon was reported by 27 patients (15%).
At baseline, the mean distance achieved in the 6-min WT was 240 ± 146 m (range 0 to 550 m), and the median value was 250 m.
Long-term follow-up of patients
Functional class and exercise capacity
Patients were followed for a mean period of 26 ± 21 months (range 0.5 to 98 months). At three months, 10 patients were dead before evaluation and 2 had undergone lung transplantation. The remaining 166 patients were evaluated according to their NYHA functional class and 6-min WT. Their mean dose of PGI2 was 14 ± 4 ng/kg/min (range 7 to 34 ng/kg/min). The NYHA functional class improved in 125 patients (75%), was unchanged in 40 patients (24%) and worsened in only 1 patient. Ninety-two patients (55%) were classified in NYHA functional class I (n = 2) or II (n = 90), and 74 (45%) were in class III (n = 67) or IV (n = 7). The mean increase in the 6-min walk distance was 125 m from baseline (251 ± 144 m at baseline to 376 ± 114 m at 3 months; p < 0.001). The 6-min WT improved in 149 patients (90%), with a mean increase of 147 m. The mean increase in the distance walked was significantly greater for patients who were initially in NYHA functional class IV (222 ± 142 m) than for patients in class III (88 ± 99 m; p < 0.0001), but the absolute distance walked in 6 min remained lower (326 ± 106 m vs. 398 ± 108 m; p = 0.0002).
At one year, 14 additional patients were dead and 11 had undergone lung transplantation. In the 130 patients who were evaluated, the mean dose of PGI2 was 21 ± 7 ng/kg/min (range 8 to 40 ng/kg/min). Eighty-two patients (63%) were classified in NYHA functional class I (n = 5) or II (n = 77), and 48 patients (37%) were still in class III (n = 42) or IV (n = 6). In terms of NYHA functional class, 115 patients remained stable, 5 improved, and 10 worsened between 3 and 12 months of treatment. In these 130 patients evaluated at one year, the mean increase in the 6-min WT was 142 m after three months (253 ± 144 m to 395 ± 107 m; p < 0.001), with stabilization after one year of treatment with PGI2 (395 ± 112 m).
One hundred and fifty-six patients underwent repeat right-sided heart catheterization after three months of continuous PGI2 therapy, and 107 patients had it after one year. The results of the hemodynamic variables are summarized in Table 2. In 77 patients, a long-term evaluation was performed after 41 ± 17 months (16 to 96 months) of PGI2 therapy (mean dose 33 ± 11 ng/kg/min [range 9 to 63 ng/kg/min]). In these patients, hemodynamic improvement was sustained, with a slight increase in the CI (2.5 ± 0.6 l/min/m2 at one year to 2.8 ± 0.6 liters/min per m2 at follow-up; p = 0.002), resulting in a lower level of total pulmonary resistance (TPR) (25.3 ± 6.9 U/m2 to 22.8 ± 6.9 U/m2; p = 0.001).
Minor complications attributable to PGI2 were frequent and included jaw pain, headache, diarrhea, flushing, leg pain, and nausea/vomiting. Seventy-six episodes of catheter-related sepsis occurred in 53 patients (0.19 catheter-related infections per patient-year). In four patients, a severe catheter-related infection was directly responsible for death, including three nosocomial infections that were acquired in the intensive care unit where the patients were hospitalized for severe right-sided heart failure. Seven patients developed severe pulmonary edema leading to death. In these patients, a postmortem examination of the lungs demonstrated pulmonary veno-occlusive disease or pulmonary capillary hemangiomatosis.
At the time of follow-up (28 ± 22 months), 93 patients remained on long-term PGI2 treatment (mean follow-up 38 ± 22 months [range 3.9 months to 8.6 years]); 27 patients had undergone lung or heart/lung transplantation after a mean period of 14 months of PGI2 therapy (1 month to 5 years); and 58 patients had died after a mean period of 18 months of PGI2 therapy (5 days to 5.5 years). The overall survival rates at one, two, three, and five years were 85%, 70%, 63%, and 55%, respectively (Fig. 1). Patients who were classified in NYHA functional class IV at baseline had a poorer prognosis than those in class III (Fig. 2A). The estimated rates of patients surviving at one, two, and three years were 76%, 60%, and 47% for patients in NYHA functional class IV, as compared with 90%, 76%, and 71% for patients in class III (p = 0.001 by the log-rank test). After three months of treatment with PGI2, patients whose symptoms had improved such that they could be reclassified in functional class I or II had an increased probability of survival. In these patients, the survival rates at one, two, and three years were 100%, 93%, and 88%, respectively, as compared with 77%, 46%, and 33% in patients persisting in class III or IV (Fig. 2B).
The majority of deaths (n = 44; 76%) occurred during the first two years of PGI2 therapy. The main cause of death was progressive right-sided heart failure in 39 patients. Other causes of death included sepsis (n = 8), pulmonary edema (n = 7), hemorrhage (n = 3), and stroke (n = 1). Postmortem verification was performed in 17 patients, revealing typical plexiform lesions (n = 10), pulmonary veno-occlusive disease (n = 5), and pulmonary capillary hemangiomatosis (n = 2) (13).
Factors associated with survival
The results of univariate analysis of the relationship between survival and variables measured at baseline are shown in Table 3. Mortality was not associated with patient age or gender, the presence of Raynaud’s phenomenon, anorexigenic exposure, or a history of syncope. Only a history of right-sided heart failure, baseline NYHA functional class IV, and baseline 6-min walk distance ≤250 m (corresponding to the median value) were associated with poor survival. Two additional hemodynamic variables were significantly related to an increased risk of death: mean RAP ≥12 mm Hg and, paradoxically, mean PAP <65 mm Hg. The CI, mixed venous oxygen saturation (Svo2), and calculated TPR were not related to survival in this patient group.
The results of univariate analysis relating survival time to selected variables measured after three months of treatment with PGI2 are shown in Table 4. Persistent NYHA functional class III or IV, despite PGI2 therapy, was strongly associated with an increased risk of death (Fig. 2B), as well as the absolute distance walked in 6 min (Fig. 3A). Conversely, the change in the distance walked from baseline was not a predictor of survival in this study (Fig. 3B). Hemodynamic variables associated with better survival were mean RAP <10 mm Hg, a paradoxical high level of mean PAP ≥59 mm Hg, Svo2 ≥62%, an increase in CI of >0.5 l/min/m2 from baseline and a decrease in TPR of >30% relative from baseline.
Two separate multivariate analyses were performed to determine, first, the effects of selected baseline values on survival and, second, the effects of selected values measured both at baseline and after three months of PGI2 therapy. In the univariate analysis of baseline variables, three clinical variables (NYHA functional class, history of right-sided heart failure, and 6-min walk distance) and two hemodynamic variables (RAP and mean PAP) were significantly related to mortality. In the multivariate analysis, we examined the independent effect of each independent variable in the presence of others on mortality. In the multivariate Cox proportional hazards regression analysis, including baseline variables, only a history of right-sided heart failure and RAP ≥12 mm Hg were independently related to poor survival. When both baseline and three-month variables were included in the multivariate regression analysis, three variables remained independently associated with mortality: a history of right-sided heart failure, the persistence of NYHA functional class III or IV after three months of PGI2 therapy, and the absence of a significant fall in TPR (30% relative to baseline) after three months of PGI2 therapy.
Our study provides new insights on the long-term outcome, survival and prognostic factors in patients in NYHA functional class III or IV who have PPH treated with continuous intravenous PGI2. This study shows that survival is mainly related to simple pretherapeutic clinical variables, such as NYHA functional class and distance achieved in the 6-min WT. In addition, both the clinical and hemodynamic responses to PGI2 therapy in the first three months appeared to be major predictors of survival in this patient group. Indeed, patients in NYHA functional class I or II after three months of PGI2 therapy had a markedly better survival than did patients in class III or IV. Moreover, a sustained hemodynamic response to PGI2, which can be measured by the fall in pulmonary vascular resistance after a three-month treatment period, seems to be a major prognostic factor. Although there was no formal control group in this study, validated historic data on patients with PPH were available to compare the long-term survival between patients treated with PGI2 and those not treated (2,10,14).
The 6-min WT is a reliable tool for the assessment of exercise capacity in patients with PPH (5,15,16). Furthermore, it is an independent predictor of mortality (17). In the present study, we demonstrated that the distance achieved during the 6-min WT before starting PGI2 therapy correlates with survival, as previously suggested by Barst et al. (5). Furthermore, although the absolute value of the 6-min WT performed after three months of PGI2 therapy correlated with survival, its increase from baseline did not.
Long-term treatment with intravenous PGI2 improves exercise capacity, hemodynamics, and survival in most patients with PPH in NYHA functional class III or IV. Despite these favorable outcomes, continuous intravenous infusion of PGI2 is a far from ideal treatment of severe pulmonary hypertension, as it is complicated, uncomfortable for patients and very costly. Side effects such as flushing, headache, jaw pain, and diarrhea are common. More important complications can occur, mainly related to the delivery system: the incidence of catheter-related infections was 0.19 per patient-year in our study, similar to incidences previously reported (0.1 to 0.4 per patient-year) (5,6). Despite significant clinical improvement, hemodynamic parameters usually remained severely compromised in many patients on long-term PGI2 therapy. This suggests that PGI2 given over the long term may have other effects in addition to pulmonary vasodilation at rest, including myocardial protection from ischemic injury, as well as effects on pulmonary vascular growth and remodeling (6,18,19). Furthermore, PGI2 may improve hemodynamics at exercise, even in the absence of a significant hemodynamic change at rest (20).
In our study, a relatively low mean PAP value (i.e., below the median value at baseline and after three months of PGI2 therapy) was a predictor of poor survival. This finding is counterintuitive and may be a function of random analysis of a wide variety of variables, so that one inevitably may produce an apparent level of significance that is not truly justified. Moreover, mean PAP reflects only one aspect of pulmonary hemodynamics, and one should discuss pulmonary resistance (i.e., mean PAP/CI) rather than mean PAP alone. This is clearly demonstrated in our present study, where the reduction in TPR after three months of PGI2 therapy (>30% relative to baseline) is a clear indicator of better survival, whereas the reduction in mean PAP does not modify the patients’ overall survival (Table 4). These invasive criteria appear to provide important prognostic information in addition to simple clinical measurements, such as NYHA functional class and the 6-min WT. On the basis of this information, we recommend performing a repeat invasive hemodynamic procedure after a three-month period of PGI2 treatment in these patients.
Timing for lung transplant listing
Our study provides the basis for a rational approach to the management of patients with severe PPH receiving long-term PGI2 infusion. The current shortage of organ donors, coupled with a five-year survival rate of only 50%, precludes systematic listing of patients with PPH for lung or heart/lung transplantation. Therefore, it is widely accepted that lung transplantation should be reserved for those patients with PPH who fail to respond to optimal medical therapy. Our current results suggest that such a subset of patients may be identified as those who remain in NYHA functional class III and/or IV or who do not achieved a 30% fall, relative to baseline, in TPR after three months of PGI2 therapy.
Survival of patients with PPH treated with epoprostenol depends on the severity at baseline, as well as the three-month response to therapy, which is evaluated by both clinical measures and hemodynamics. These findings suggest that lung transplantation should be considered in a subset of patients who remain in NYHA functional class III or IV or who do not achieve a significant hemodynamic improvement after three months of epoprostenol therapy.
☆ This study was supported by Université Paris-Sud.
- cardiac index
- pulmonary artery pressure
- New York Heart Association
- epoprostenol (prostacyclin)
- primary pulmonary hypertension
- right atrial pressure
- mixed venous oxygen saturation
- total pulmonary resistance
- walk test
- Received December 18, 2001.
- Revision received April 11, 2002.
- Accepted May 15, 2002.
- American College of Cardiology Foundation
- Rubin L.J.,
- Groves B.M.,
- Reeves J.T.,
- Frosolono M.,
- Handel F.,
- Cato A.E.
- Shapiro S.M.,
- Oudiz R.J.,
- Cao T.,
- et al.
- Rich S, editor. Primary Pulmonary Hypertension: Executive Summary from the World Symposium—Primary Pulmonary Hypertension 1998. Available from the World Health Organization via the Internet at: http://www.who.int/ncd/cvd/pph.html
- Sitbon O.,
- Humbert M.,
- Jagot J.L.,
- et al.
- Guyatt G.H.,
- Sullivan M.J.,
- Thompson P.J.,
- et al.
- Sandoval J.,
- Bauerle O.,
- Palomar A.,
- et al.
- Jones D.,
- Benjamin C.,
- Linseman D.