Intravenous iloprost for treatment failure of aerosolised iloprost in pulmonary arterial hypertension

Pulmonary arterial hypertension (PAH) is characterised by progressive obliteration of the pulmonary vascular bed that almost inevitably results in progressive right heart failure and death 1, 2. Treatment with continuous intravenous prostacyclin (epoprostenol) has been shown to improve exercise capacity, haemodynamics and survival in patients with pulmonary hypertension 3–5. Alternatively, continuous intravenous infusion of iloprost, a stable prostacyclin analogue, is used in some countries and seems to be as efficient as intravenous epoprostenol in these patients 6, 7.

A major drawback of continuous intravenous prostaglandin treatment is the requirement of a permanent central venous access that is prone to infectious complications 3, 4. To circumvent this problem, novel prostaglandins such as inhaled iloprost 8, oral beraprost 9, and subcutaneous treprostinil 10 have recently been introduced as alternatives. Furthermore, the orally available dual endothelin receptor antagonist bosentan is the first nonprostanoid substance for which efficacy has been demonstrated in randomised, controlled trials 11, 12. It is tempting to use these new substances as noninvasive first line treatment of patients with PAH and to reserve intravenous prostaglandins for the most severe cases and treatment failures. However, it is unknown if intravenous prostaglandins do provide clinical benefit when the novel, noninvasive prostaglandins are not sufficiently effective.

Aerosolised iloprost is among the novel therapies that have recently been introduced for PAH 8. In open studies, aerosolised iloprost had beneficial acute and long-term effects in this group of patients 13–16. The efficacy of inhaled iloprost in patients with PAH was recently confirmed by a European randomised, placebo-controlled, multicentre trial that demonstrated significant improvement in exercise capacity and haemodynamics in patients treated with aerosolised iloprost (unpublished data). The inhaled route of administering iloprost has less side-effects than the intravenous approach and obviates the need of a permanent central venous access. However, concern has been raised that this form of treatment may be less effective than treatment with intravenous prostacyclin 17. No study has yet addressed efficacy and safety of inhaled iloprost compared with intravenous epoprostenol or iloprost, respectively.

At Hannover Medical School, Germany, most of the patients with PAH have been receiving first line treatment with aerosolised iloprost since 1997. These patients have been followed closely, including repeated outpatient visits every 3 months and yearly follow-up right heart catheter examinations. If there was no improvement with aerosolised iloprost (primary treatment failure) or if a patient deteriorated after initial improvement (secondary treatment failure) switching to continuous intravenous iloprost was recommended, although it is unknown if patients whose clinical conditions deteriorate under treatment with aerosolised iloprost benefit from intravenous iloprost.

In this paper, the clinical results of 16 PAH patients who received continuous intravenous iloprost after primary or secondary treatment failure with inhaled iloprost are described.

Patients and methods

Of 140 PAH patients referred to the authors' centre between February 1997 and August 2001, 93 received aerosolised iloprost as first line treatment (some patients with severe haemodynamic impairment were immediately treated with intravenous iloprost, while some received calcium channel blockers, oral beraprost sodium or subcutaneous treprostinil as first line treatments).

Treatment with aerosolised iloprost was continued for as long as the patients showed clinical improvement or stabilisation in New York Heart Association (NYHA) class II or III. Right heart catheterisations were performed before initiation of inhaled iloprost and 3 months later, followed by annual catheter studies or when indicated by clinical requirements. In case of progression of pulmonary hypertension, as determined by clinical assessment, 6‐min walking distance and haemodynamic deterioration during repeated right heart catheterisations, the dose of inhaled iloprost was increased as described below. If there was further deterioration, treatment with continuous intravenous iloprost was initiated. Inhaled iloprost was tapered and discontinued within 3–7 days after intravenous therapy was started. Of the 93 patients treated initially with aerosolised iloprost, 16 were eventually switched to intravenous iloprost because of clinical and haemodynamic deterioration. Eligible patients were admitted on the waiting list for lung transplantation when intravenous treatment was begun.

Before treatment with aerosolised iloprost was started, all patients were informed about the investigational character of this therapy and the possibility that intravenous iloprost might be more effective. The authors' institutional review board approved this approach.

Results

Sixteen patients received intravenous iloprost for failure of treatment with inhaled iloprost. The patients' clinical characteristics, haemodynamic variables before initiation of intravenous iloprost, and exercise capacities are shown in table 1⇓. Two patients had porto-pulmonary hypertension (No. 6 and 11), while the others suffered from primary pulmonary hypertension. All patients had severe limitation of exercise capacity; three were in NYHA class III, and 13 in NYHA class IV. The walking distance in the 6‐min walk test was 179±118 m (range 0–350 m). Cardiac function was severely compromised as indicated by a cardiac index of 1.6±0.2 L·min⁻¹·m⁻² (range 0.9–2.0 L·min⁻¹·m⁻²) and a mixed venous oxygen saturation of 52±6% (range 40–60%). The patients had been treated with aerosolised iloprost for 13±10 months (range 1–33 months). Of the 16 patients who were switched to intravenous iloprost, six were treated for <6 months with inhaled iloprost. These patients had no substantial improvement with inhaled iloprost at any time. In contrast, the other 10 patients showed improvement in 6‐min walking distance and haemodynamics for up to 24 months with inhaled iloprost but eventually haemodynamic and clinical deterioration occurred.

Of the 16 patients who received intravenous iloprost for failure of treatment with inhaled iloprost, eight did not show improvement of exercise capacity or NYHA class. Five of these patients died after 3–16 (median 3) months of intravenous treatment. Two patients underwent lung transplantation while one patient is alive, receiving intravenous iloprost for 9 months and awaiting lung transplantation. This patient is confined to a wheelchair due to advanced right heart failure. In contrast, the remaining eight patients showed substantial clinical improvement. One of these patients received a lung transplant 3 months after intravenous iloprost was started while the remaining patients were in stable clinical condition after receiving intravenous iloprost for 9±2 months (range 7–11 months).

Before switching from inhaled to intravenous iloprost, three patients were in functional NYHA class III and 13 patients in NYHA class IV. After a mean treatment period of 9±2 months, all patients who responded to treatment were in NYHA class III (table 1⇑). In this subgroup of patients, the walking distance in the 6‐min walk test increased from 205±94 m to 329±59 m.

There were no significant differences between the patients who derived benefit from intravenous iloprost and those who did not in terms of age, duration of disease, duration of treatment with inhaled iloprost, haemodynamic variables, or vasoreactivity as determined by the acute haemodynamic response to inhaled iloprost (data not shown).

Discussion

There is substantial evidence from open studies that aerosolised iloprost is an effective treatment for PAH 14, 15. A recently concluded European placebo-controlled multicentre trial confirmed these observations (unpublished data). Because of these results it is expected that European regulatory agencies will grant approval for inhaled iloprost as treatment of PAH in the year 2002.

A major drawback of this therapy is the relatively short duration of action of aerosolised iloprost. The haemodynamic effects are detectable only for 30–60 min after each inhalation 13 so that even with 6–9 inhalations·day⁻¹, continuous pulmonary vasodilation is not achievable. Although vasodilation is probably not the only mechanism by which prostaglandins act in PAH, the short-lived haemodynamic action raised concerns that aerosolised iloprost might be less effective than continuous intravenous prostaglandins.

Prospective studies comparing inhaled iloprost with intravenous epoprostenol or iloprost, respectively, are not available. One way to find out whether intravenous prostaglandins are more effective than inhaled iloprost is to introduce intravenous treatment in those patients who deteriorate while receiving inhaled iloprost. In the current investigation, replacing inhaled iloprost by intravenous iloprost caused substantial increases in exercise capacity in eight of 16 patients whereas the remaining patients had no benefit. Seven of the patients in the latter group died or underwent lung transplantation, while seven patients in the former group were alive and substantially improved after 9±2 months of intravenous treatment (one patient who underwent lung transplantation died in the early postoperative period). Two aspects of these findings are remarkable: first, intravenous iloprost was clearly more effective than aerosolised iloprost in some patients with PAH; second, half of the patients had no benefit from switching to intravenous iloprost. The superior efficacy of intravenous iloprost was anticipated but the finding that intravenous treatment failed to provide clinical improvement in half of the patients was surprising and disturbing at the same time. Several reasons may have accounted for the overall unsatisfying efficacy of intravenous iloprost in this highly selected subgroup of patients.

When initiated as first-line treatment, intravenous epoprostenol and iloprost have been reported to be effective in the vast majority of patients with PAH 3, 4, 19. In the current authors' centre, intravenous treatment was given only to patients with no long-term response to aerosolised iloprost. Thus, it is possible that a subgroup of patients were selected who were unresponsive to treatment with prostaglandins. The authors were unable to identify clinical or haemodynamic variables that would discriminate these patients from those who remain responsive to intravenous iloprost despite failure of inhaled iloprost.

Treatment with inhaled iloprost was performed for 13±10 months before the patients received intravenous iloprost. Thus, the patients had more advanced disease than they would have had if intravenous iloprost was instituted as first-line treatment. The fact that the patients in this study had a cardiac index of 1.6±0.2 L·min⁻¹·m⁻² and a mixed venous oxygen saturation of 52±6% at the time when intravenous treatment was introduced, underscores the notion that this group of patients had very severe disease. This fact probably contributed to the high rate of nonresponders to intravenous iloprost. However, the possibility that some of the patients would have fared better with earlier introduction of intravenous treatment cannot be ruled out and it will be a major task of further studies to determine criteria that help to decide when a patient should be switched from noninvasive to invasive prostaglandin treatment.

The optimal dosing regimen of intravenous prostaglandins in patients with PAH is unclear. Continuous administration of prostacyclin or iloprost may cause tachyphylaxis. Therefore, it is common practice in most centres to gradually increase the dose over time 4, 20 although it is still uncertain to what extent this is really necessary. In several centres, the dose of intravenous epoprostenol or iloprost is increased whenever side-effects wane. At least some patients might be overtreated with this approach 21. A slightly more conservative protocol was used in this study by increasing the dose of intravenous iloprost only when the clinical efficacy was unsatisfactory. With this regimen, the mean dose of iloprost in the patients was 2.0±0.4 ng·kg⁻¹·min⁻¹, which is close to the mean dose of 2.1 ng·kg⁻¹·min⁻¹ that has been used by Higenbottam et al. 6 in another study. It is possible that a more aggressive dosing regimen might have been more efficient. However, the doses of intravenous iloprost were not different in patients who gained benefit and those who did not.

The fact that intravenous iloprost was used instead of intravenous epoprostenol may also be a matter of debate. The evidence for efficacy of epoprostenol in PAH is much stronger than for intravenous iloprost and only a few studies have been performed to compare both substances 6, 7. From these data, it seems likely that iloprost is as effective as epoprostenol in the treatment of pulmonary hypertension but this has not been scientifically proven. Iloprost has the advantage of being much more stable than epoprostenol 22 and is therefore easier to handle. In addition, the longer half-life makes iloprost potentially safer than epoprostenol, since interruption of drug supply does not result in immediate cessation of the drug's action. Furthermore, in Germany, iloprost is less expensive than epoprostenol. Nevertheless, the authors cannot exclude with certainty that these results might have been better if epoprostenol had been used.

This study has several limitations including the small number of patients. The criteria for switching from aerosol to intravenous treatment were not rigorously defined. The most important limitation comes from the retrospective analysis and the fact that there was no control group. It has to be kept in mind that the presented data were merely observational rather than from a clinical trial. In light of the strong evidence for the efficacy and potentially life-saving effects of intravenous prostaglandins in PAH, it would have been unethical to withhold these substances from severely ill patients who deteriorate while receiving inhaled iloprost.

This study is further limited by the lack of follow-up data from right heart catheterisation. However, since dosing of iloprost was adjusted according to clinical patterns in the majority of patients, no indication for repeated cardiac catheterizations was seen.

Despite these limitations, this data clearly demonstrates that continuous intravenous iloprost may have substantial favourable effects in patients with pulmonary arterial hypertension who deteriorate while being treated with aerosolised iloprost. However, not all patients derive benefit from this approach. Since there seems to be no means to identify patients who will respond to intravenous iloprost after treatment failure with inhaled iloprost, intravenous prostaglandin treatment should be considered in all pulmonary arterial hypertension patients who deteriorate while receiving iloprost aerosol.