Adverse effects occur with all PAH therapies. Clearly however, benefit outweighs the risk when initiating single-agent PAH therapy, regardless of the class of drug. Patients are generally followed closely, and PAH drug doses are decreased or the drugs are discontinued if adverse effects are intolerable. Severe adverse effects leading to hospitalization, and particularly to death are distinctly unusual. Even when intravenous therapy is undertaken, fatal sepsis from line infections, or other severe line complications are rare when patients are treated by experts and followed closely [1, 20–24].

In the 81 patient, randomized, epoprostenol trial, 41 patients received this drug [1]. Adverse effects included jaw pain, diarrhea, flushing, headaches, nausea, and vomiting. Serious complications were usually due to the delivery system and included four episodes of nonfatal, catheter-related sepsis. There were 26 episodes of malfunction of the drug-delivery system resulting in temporary interruption of the infusion. These included occlusions, perforations, and dislodgements of the catheter and pump malfunction. When epoprostenol therapy was interrupted, patients experienced an increase in PAH symptoms. Irritation or infection at the catheter site occurred in seven patients, bleeding at the catheter site in four, and catheter-site pain in four. More than 20 years have passed since this study was initiated. PH centers now have substantial experience in preventing and addressing adverse effects.

More recently, we studied the transition from epoprostenol to IV treprostinil and found no serious adverse events that could be attributed to treprostinil [20]. When we examined de novo therapy with IV treprostinil, side effects were mild and consistent with those reported with prostacyclin treatment [21]. Finally, we conducted a long-term (48-week), multicenter, prospective, open-label, uncontrolled, study of continuous IV treprostinil in 16 patients on no prior PAH specific therapy at baseline and 31 patients transitioned at baseline from IV epoprostenol [22]. During the study, 5 patients died of causes not considered related to the therapy, and 7 discontinued due to adverse events.

With regard to inhaled prostanoid therapy, common adverse effects following administration of inhaled treprostinil in the double-blind trial included cough (54 %), headache (41 %), nausea (19 %), flushing (15 %) and throat irritation (14 %) or pain (11 %) [23]. Over a 2-year treatment period in the TRIUMPH trial extension, adverse events resulting in drug discontinuation occurred in 19 % of patients; these included worsening PAH, cough, headache, throat discomfort and pneumonia [24].

A number of clinical trials have included a study drug and one or two background therapies, and while on multi-agent therapy an increase in adverse effects would be expected, these studies have shown that severe, life-threatening adverse events in patients on three PAH therapies including a parenteral or inhaled prostanoid would be appear to be unusual.

Ideally, in any setting, when a novel therapeutic approach is considered, particularly when it involves added patient inconvenience or potential discomfort and expense, as much preparation and preliminary data collection should be undertaken as possible. A randomized trial comparing the above-described protocol with a standard goal-oriented approach would be ideal. Such a trial could be undertaken. In an ideal world, an additional arm could substitute an inhaled prostanoid arm (perhaps combined with an oral prostanoid; i.e., treprostinil) for the IV prostanoid. A second option would be to randomize less severely ill patients to the aggressive three-drug protocol, while proceeding with the protocol in more advanced cases of PAH who would ordinarily undergo upfront parenteral prostanoid therapy. That way, patients with more severe PAH would be getting the most aggressive therapy possible. Finally, once the data for oral prostanoid receptor agonists are available, it could be judged if such therapy could substitute for parenteral/inhaled prostanoid therapy in an upfront, multi-agent protocol perhaps reducing adverse effects and offering a more convenient approach.

In any setting in which randomization is not done, prospective data should at least be collected, either in the form of a pilot study, or as a long-term single-arm study in which detailed data are collected. There is always a concern that triple-combination therapy could conceivably have antagonistic effects, but this does not appear to be evident in the limited PH literature thus far.

Several studies have suggested that patients on background PAH therapy, however, may have less robust responses than patients who had been untreated when a new therapy was initiated [25, 26]. In the FREEDOM C study evaluating oral treprostinil, more than 40 % of patients were on both a PDE-5I and an ERA, and most of the remainder of the patients were on one or the other [27]. In this study, the primary end point (6MWD at week 16) did not achieve significance. Other possible explanations for this were the premature discontinuation of the study drug due to adverse events associated with higher-dose tablets. A 1-mg twice-daily dose of oral treprostinil is approximately equivalent to 10 ng/kg/min of infused treprostinil and dose increases of 0.5 or 1 mg were poorly tolerated by most patients [28–30]. No patient with access to the 0.25-mg tablets discontinued because of adverse events. In the follow-up study, FREEDOM C2, again essentially all patients were on either one or two background therapies, and again the primary endpoint was not met [30]. It is notable, however, that patients receiving oral treprostinil combined with background PDE-5I therapy were found to have approximately twice the 6MWD treatment effect (15.0 m) as those receiving background ERA therapy (7.7 m). This potential effect was not, however, demonstrated in patients receiving both ERA and PDE-5I background therapy (4.0 m). In FREEDOM C2, the 6MWD treatment effect tended to be greater in patients receiving a more recent diagnosis (28 m for patients with PAH diagnosed within the first year) than in patients with a longer disease duration (−2 m if the diagnosis was made ≥3.6 years prior to therapy). It is possible that these data reflect a survivor effect, with favorable responders to their current PAH therapy having less room for response to additional therapies. The patients who received intensive PAH-specific treatment for a longer period could simply be less responsive to the addition of new therapy. While there is the possibility that earlier initiation of multi-agent PAH therapy could result in a patient “using up” their therapeutic options, there is no clear data clinical supporting this.