Survival in Primary Pulmonary Hypertension with Long-Term Continuous Intravenous Prostacyclin

Methods

In a previous 8-week randomized study [19], 11 of 25 patients were randomly assigned to receive prostacyclin plus conventional therapy, and 14 of the 25 patients were randomly assigned to receive conventional therapy alone. At the completion of the 8-week study, all survivors were eligible to enter this open, multicenter, uncontrolled trial of long-term intravenous infusion of prostacyclin regardless of their baseline hemodynamic response to prostacyclin or the treatment group to which they were initially assigned. The clinical diagnosis of primary pulmonary hypertension was established in all patients before entry on the basis of the criteria of the NIH Registry [3]. Thromboembolic disease was excluded on clinical grounds by perfusion lung scanning, or, when this was inconclusive, by pulmonary angiography. Patients with associated conditions such as portal hypertension, human immunodeficiency virus infection, collagen vascular diseases, and pulmonary vasculitides were excluded from this study. Sterile, lyophilized prostacyclin sodium powder (Flolan, epoprostenol sodium), synthesized by the Upjohn Co. (Kalamazoo, Michigan) and formulated by the Wellcome Research Laboratories (Beckenham, Kent, United Kingdom), was refrigerated until use. Immediately before administration, prostacyclin was reconstituted with a sterile buffer (pH, 10.5) at a concentration of 5 mg/mL and was filtered. All patients were treated with warfarin.

Hemodynamic measurements were obtained by catheterization of the right side of the heart using standard techniques. After baseline hemodynamic measurements were obtained, an intravenous infusion of prostacyclin was begun at a rate of 2 ng/kg of body weight per minute and increased by increments of 2 ng/kg per minute every 10 to 15 minutes. Hemodynamic measurements were repeated at the end of the infusion of each dose. The acute infusion dose was not further increased when one or more of the following occurred: a greater than 40% decrease in systemic arterial pressure, a greater than 40% increase in heart rate, or symptoms such as nausea, vomiting, or severe headache. After the acute dose-ranging study, the dose of prostacyclin was decreased to a dose that did not result in any adverse effects. The long-term infusion dose was adjusted based on results of the most recently completed dose-ranging study or clinical needs of the patient. Patients were allowed to continue prostacyclin therapy until transplantation or death.

Six-minute walk tests were done to assess exercise capacity [19]. Tests were carried out before long-term prostacyclin therapy was initiated and after 6, 12, and 18 months of long-term therapy.

Venous access for the infusion of prostacyclin was obtained by inserting a permanent intravenous catheter into a jugular or subclavian vein and tunneling subcutaneously. Prostacyclin was infused continuously through a portable pump (Autosyringe AS2F, Travenol Inc., Hooksett, New Hampshire or CADD-1 Model 5100 HF, Pharmacia Deltec Inc., St. Paul, Minnesota). Before hospital discharge, patients were thoroughly trained in catheter care, sterile technique, and drug preparation and administration.

Safety was monitored by physical examination, routine hematologic and chemical profiles, electrocardiograms, and adverse experience assessments.

The prostacyclin doses and hemodynamic and exercise data are presented as the mean ± SD. Changes in exercise capacity and hemodynamics were analyzed using the Wilcoxon signed-rank test for paired data [20]. Survival analysis was done using a proportional hazards regression model [21] that models survival time against group (the 17 NYHA class III and IV patients treated with prostacyclin compared with the 31 NYHA class III and IV patients [historical controls] who were treated with anticoagulant agents). The model stratifies patients according to transplantation status and NYHA class and controls through a confounder score for baseline mean pulmonary artery pressure, mean right atrial pressure, and cardiac index. These three hemodynamic variables have been shown to be associated with survival in patients with primary pulmonary hypertension [4], and the particular covariate used here as a confounder score is the predicted survival at 1 year, which is a function of the three hemodynamic variables mentioned above. Patients were censored at the time of transplantation. We constructed Kaplan-Meier curves for the patients treated with prostacyclin and for the historical controls [22]. A two-sided P value of ≤ 0.05 was considered statistically significant. The Equation developed from the NIH Primary Pulmonary Hypertension Registry [4] was used to predict survival for the patients treated with prostacyclin as well as for the historical controls. The formula

P(t) = H(t) A^(x,y,z)

H(t) = 0.88 −0.14 t + 0.01t²

A(x,y,x) = e^{(0.007325x + 0.0526y −0.3275 z)}

where x = mean pulmonary artery pressure; y = mean right-atrial pressure; z = cardiac index; and t = 1, 2, or 3 years, estimates a patient's chances for survival (P[t]) at 1, 2, and 3 years given the values of the patient's three hemodynamic variables: mean pulmonary artery pressure, mean right-atrial pressure, and cardiac index.

Results

Patients

Eighteen patients with primary pulmonary hypertension (17 adults and 1 child) were entered into this long-term study after informed consent was obtained. From the 25 patients enrolled in the preceding 8-week randomized study [19], the 10 surviving patients already receiving prostacyclin plus conventional therapy and 8 of the 11 survivors from the group receiving conventional therapy alone elected to enter this long-term study. The clinical and demographic characteristics are shown in Table 1. The mean age was 35.9 ± 13.4 years. Twelve patients were female and 6 patients were male. Seventeen of the 18 patients were NYHA class III or IV despite conventional therapy, which consisted of vasodilators, oxygen, diuretic agents, cardiac glycosides, and anticoagulant agents as deemed necessary. Six patients were receiving oral vasodilator agents before entering this study, and 5 of these patients continued this therapy.

Treatment Regimen

Before long-term treatment with continuous intravenous prostacyclin was started and during the baseline acute dose-ranging study, the range of maximal tolerated doses of prostacyclin for the 18 patients was 4 to 22 ng/kg per minute. After the acute dose-ranging study, the dose of prostacyclin was decreased until it did not result in any adverse effects. The initial dose of the long-term prostacyclin infusion ranged from 2 to 8 ng/kg per minute (mean, 6.9 ± 3.0 ng/kg per minute). Sixteen patients had repeat hemodynamic evaluation and dose-ranging study after 6 months of therapy, and the hemodynamics of 14 patients were re-evaluated at 12 months. Three patients had transplantation at 6, 10, and 12 months, respectively, before repeat hemodynamic measurements were obtained. One patient declined catheterization at 12 months. One patient had catheterization at 12 months but not at 6 months. The mean dose of long-term prostacyclin was 17.6 ± 11.2 ng/kg per minute (n = 14) at 1 year, 36.7 ± 21.2 ng/kg per minute (n = 11) at 2 years, and 52.9 ± 30.2 ng/kg per minute (n = 7) at 3 years.

Exercise Capacity

Exercise endurance evaluated on the basis of the 6-minute walk test is shown in Figure 1. At 6 and 18 months, patients could walk, on average, more than 100 meters farther than they could before prostacyclin therapy was started. The length of the 6-minute walk increased from 264 ± 160 meters at baseline to 370 ± 119 meters at 6 months, 348 ± 142 meters at 12 months, and 408 ± 138 meters at 18 months (P < 0.001 at 6 months and P = 0.02 at 18 months compared with baseline).

Hemodynamics

Mean hemodynamic measurements at baseline (n = 18), 6 months (n = 16), and 12 months (n = 14) are shown in Table 2. In the patients who had follow-up cardiac catheterization at 6 months, the cardiac index increased 18% (CI, 0.1% to 36.7%), mean pulmonary artery pressure decreased 9% (CI, 1.4% to 15.7%), and total pulmonary resistance improved 26% (CI, 6.1% to 46.3%). Among patients re-evaluated at 12 months, a 27% increase remained in the cardiac index (CI, 1.3% to 51.9%) and a 32% decrease remained in total pulmonary resistance over baseline (CI, 9.7% to 53.6%). Although total systemic resistance decreased, systemic arterial pressure was unchanged. There was a significant inverse correlation between the change in the 6-minute walk time from baseline to 12 months and the corresponding changes in both mean pulmonary artery pressure (r = −0.733;CI, −0.319 to −0.923)and mean right atrial pressure (r = −0.754;CI, −0.178 to −0.945).Changes in no other hemodynamic variable were correlated with exercise capacity. The baseline values and effects of long-term continuous intravenous prostacyclin on hemodynamic variables for each of the 18 individual patients are presented in Tables 3 and 4. The patients were divided into two groups: patients who did not have transplantation (n = 10) and patients who had transplantation (n = 8). Although no significant differences at baseline or at follow-up were noted between the patients who had transplantation and those who did not, the overall hemodynamic profile in the group of patients who had transplantation was slightly worse than that of the patients who did not. The baseline mean right atrial pressure, mean pulmonary artery pressure, cardiac index, and 6-minute walk time in the 8 patients who had transplantation were 14.0 ± 7.7 mm Hg, 64.9 ± 14.4 mm Hg, 1.64 ± 0.54 L/min per m², and 240 ± 177 meters, respectively, compared with 7.3 ± 5.2 mm Hg, 58.5 ± 14.9 mm Hg, 2.08 ± 0.60 L/min per m², and 284 ± 151 meters, respectively, in the 10 patients who did not have transplantation. Moreover, the relative changes with prostacyclin treatment were greater in the patients who had transplantation. The last follow-up cardiac catheterization and repeat 6-minute walk in the 8 patients who had transplantation was 13.7 ± 8.3 months after starting prostacyclin therapy. The changes from baseline in the mean right atrial pressure, mean pulmonary artery pressure, cardiac index, and 6-minute walk at last follow-up in these 8 patients were −6.1± 9.0 mm Hg, −5.3± 13.6 mm Hg, 0.72 ± 0.82 L/min per m², and 75 ± 154 meters, respectively, compared with −0.3± 2.5 mm Hg, −7.7± 8.1 mm Hg, 0.36 ± 0.82 L/min per m², and 53 ± 105 meters, respectively, in the 10 patients who did not have transplantation measured at a similar follow-up period (13.2 ± 2.5 months).

Survival

Of the 18 patients who were treated with continuous prostacyclin in this study, 6 patients still receive prostacyclin (range of therapy, 50 to 69 months), 8 patients have had transplantation (5 are still alive and 3 have died), and 4 patients died while receiving a continuous infusion of prostacyclin (after 12 to 40 months of treatment). The mean duration of prostacyclin treatment before transplantation was 17 ± 9 months. Long-term follow-up data for the 18 patients are shown in Figure 2.

The baseline hemodynamic data and predicted survival [4] of the 17 NYHA class III and IV patients treated with continuous prostacyclin were compared with that of the 31 NYHA class III and IV patients treated with standard therapy, including anticoagulant agents, from the NIH Primary Pulmonary Hypertension Registry (Table 5). Patient 5 (NYHA class II) was excluded from the survival analysis. Thirteen NYHA class III (76%) and 4 class IV (24%) patients were in the prostacyclin group, and 28 NYHA class III (90%) and 3 NYHA class IV (10%) patients were in the historical control group. No significant differences were present between baseline hemodynamics or predicted survival. Long-term survival was, however, significantly improved in patients who received continuous prostacyclin compared with those in the NIH Registry who received standard therapy (Figure 3). The 1-, 3-, and 5-year survival rates for the patients treated with prostacyclin were 87%, 63%, and 54%, respectively, compared with 77%, 41%, and 27% for the patients from the NIH Registry treated with standard therapy (historical controls relative to patients treated with prostacyclin were stratified according to transplantation status and NYHA class; hazard ratio, 2.9 [CI, 1.0 to 8.0; P = 0.045]). The probability of survival of the NYHA class III and IV patients treated with prostacyclin was also compared with their predicted survival estimated by the NIH Primary Pulmonary Hypertension Registry equation [4]. Long-term treatment with prostacyclin showed improved survival for these 17 patients compared with their predicted survival.

Six patients still receive continuous intravenous prostacyclin: Five patients (patients 7, 10, 14, 17, and 18) continue to improve clinically (range of therapy, 50 to 69 months), and one patient (patient 13) remains clinically stable while receiving long-term prostacyclin infusion (52 months of therapy; current dose, 25 ng/kg per minute). Substantial increases in doses may explain the sustained beneficial responses of some of the patients who receive long-term therapy. The current prostacyclin doses for the five patients who remain clinically improved are 51, 105, 35, 68, and 91 ng/kg per minute, respectively.

Complications

Minor complications related to the medication were common and seen in varying degrees in almost all patients; these included loose stools, jaw pain, flushing, warmth, photosensitivity, and headaches. Although patients were receiving anticoagulant therapy during the study, clotting of the delivery system occurred nine times in five patients. Thrombolytic agents were commonly used to dissolve the clot. Of the four deaths that occurred in patients who died while receiving continuous prostacyclin, two were attributable to the drug and delivery system. One patient experienced a temporary interruption of the infusion, resulting in an abrupt deterioration. The other patient developed sepsis after a thrombus was removed from the indwelling catheter. The two other deaths were attributable to disease progression. Seven episodes of nonfatal sepsis occurred in three patients. Three patients had catheter replacement (two because of documented sepsis and one because of venous thrombosis). Mechanical problems occurred in five patients, including pump malfunctions in two patients. Additional mechanical problems occurred with the syringe, tubing, and catheter system in these five patients, which caused the temporary interruption of prostacyclin infusion. While therapy was interrupted, patients experienced an increase in the symptoms of primary pulmonary hypertension, and one patient experienced syncope.

Discussion

Primary pulmonary hypertension has been considered a progressive, fatal disease [4]. The median survival of patients with mild to moderate symptoms (NYHA classes I and II) was 6 years compared with 2.5 years for class III patients and only 6 months for class IV patients in the NIH Registry [4].

The effects of long-term vasodilator therapy in patients with primary pulmonary hypertension have been inconclusive; some studies have suggested beneficial effects [5, 7-9, 11, 12, 23-29], and others have shown deleterious effects [30-34]. Approximately 25% of patients with primary pulmonary hypertension have a reactive pulmonary vascular bed: With acute vasodilator drug testing, pulmonary arterial pressure decreases occur concomitantly with an increase in cardiac output and no significant systemic hypotension. Recently, Rich and colleagues [8] reported that long-term therapy with calcium channel blockers in patients with a reactive pulmonary vascular bed improved long-term survival. In approximately 50% of patients with primary pulmonary hypertension, vasodilator drugs acutely increase cardiac output without affecting pulmonary arterial pressure; in the other 25% of patients, acute vasodilator testing either increases pulmonary arterial pressure as cardiac output increases or decreases systemic arterial pressure without affecting pulmonary arterial pressure or cardiac output. The effects on survival of long-term vasodilator therapy in these patients remain unclear, although long-term vasodilator treatment is considered contraindicated in the latter group of patients with presumably “fixed” pulmonary vascular disease.

Our study evaluated the effects of long-term continuous infusion of prostacyclin in patients with severe primary pulmonary hypertension. Seventeen of the 18 patients treated with continuous prostacyclin were classified as NYHA III or IV when therapy with prostacyclin was started. Although these patients were severely symptomatic despite maximal conventional therapy and were among the least responsive to acute vasodilator testing with prostacyclin, long-term hemodynamic improvements were observed. Thus, the absence of an acute hemodynamic response to prostacyclin does not appear to preclude improvement with long-term therapy. Furthermore, survival significantly improved with continuous prostacyclin when compared with standard therapy.

In an uncontrolled study, Jones and colleagues [35] first observed improvement in exercise tolerance in 10 patients with primary pulmonary hypertension treated with continuous prostacyclin. In a previous randomized study [19], we showed that continuous infusion of prostacyclin improved exercise capacity for at least 8 weeks. Our current study shows that patients receiving long-term prostacyclin therapy maintain increased exercise endurance for at least 18 months and that this improvement is correlated with reductions in pressures of the right side of the heart.

Fuster and colleagues [13] reported a 3-year survival rate of less than 20% for patients with primary pulmonary hypertension (n = 115) who have a mixed venous oxygen saturation lower than 63%. The mean mixed venous oxygen saturation for the patients treated with prostacyclin before starting prostacyclin therapy was 59% ± 12%; it increased to 67% ± 7% after 6 months of prostacyclin treatment and remained improved after 12 months of therapy at 64% ± 12%. More importantly, the 3-year survival rate for these patients was 63% compared with the 20% reported by Fuster and coworkers.

Prostacyclin is delivered by a portable infusion pump that is connected to a catheter inserted into the jugular or subclavian vein; major complications have been attributable to the drug delivery system. Although the delivery of prostacyclin is more cumbersome and potentially more hazardous than oral vasodilator therapy, our results show that continuous long-term infusion of prostacyclin results in sustained clinical and hemodynamic improvement in patients with severe primary pulmonary hypertension who are refractory to conventional therapy. Early in our experience with continuous prostacyclin therapy, dose adjustments were dictated by increasing symptoms reported by our patients. We have learned to be more aggressive with dose increases over time and currently attempt to increase doses before symptoms recur. We have generally increased the dose in increments of 1 to 2 ng/kg per minute every 1 to 4 weeks. We believe that increased doses of prostacyclin over time may explain the sustained beneficial responses of some of our patients who receive long-term therapy.

Eight of our patients had thoracic transplantation. The decision to proceed to transplantation in these patients was based on the individual preferences of the patients or the persistence of hemodynamic abnormalities and symptoms that portended a poor prognosis with medical therapy. Our study was not a randomized trial comparing continuous prostacyclin therapy and transplantation. In addition, our preceding 8-week randomized trial [19] was small, and therefore the survival analysis in this study was based on comparisons with a historical control group derived from a national registry that preceded the widespread performance of lung transplantation. Accordingly, our results do not provide insight into the optimal selection or timing for either of these treatment alternatives.

Our study suggests that continuous infusions of prostacyclin may be particularly useful as a bridge to transplantation for severely ill patients. Approximately 1300 patients in the United States are awaiting lung or heart and lung transplantation [36]. The current estimated waiting time for lung transplantation is approximately 1 year; for heart and lung transplantation, the wait is longer than 18 months [36]. Our experience has been that 30% to 40% of patients with primary pulmonary hypertension awaiting transplantation die before a suitable donor organ becomes available. The most experienced centers report 1-and 2-year survival rates of approximately 75% [37-39]. The significantly improved survival of 80% at 18 months compared with 58% with standard therapy suggests that prostacyclin may be particularly useful in sustaining seriously ill patients awaiting transplantation. It should be emphasized, however, that this aggressive and complex treatment strategy poses additional risks and complications in a hemodynamically fragile patient population, and that serious adverse events occur despite careful monitoring. In addition, dose requirements tend to increase over time. Nevertheless, the sustained improvements in hemodynamics, symptoms, and survival in several patients treated for prolonged periods of time (longer than 4 years) suggest that prostacyclin therapy may also be a suitable alternative to transplantation in selected patients.

Presented at the American College of Cardiology 42nd Annual Scientific Session, 17 March 1993, Anaheim, California.