Inhaled Treprostinil in Pulmonary Hypertension in the Context of Interstitial Lung Disease: A Success, Finally

Pulmonary hypertension in the context of interstitial lung disease (PH-ILD) is one of the most fatal medical conditions patients and doctors are faced with. The vascular component of advanced ILD is difficult to tackle and obviously differs from pulmonary arterial hypertension (PAH), as multiple high-quality clinical trials failed to convincingly demonstrate a clinical benefit of pulmonary vasoactive drugs in various PH-ILD populations, whereas those drugs are effective and approved in PAH (1–5). Some drugs like ambrisentan and riociguat even showed harmful effects in PH-ILD populations and were consequently banned from treatment in this indication (2, 4). One potential cause for this differential effect of pulmonary vasoactive drugs in PAH and PH-ILD might be the induction or aggravation of $\dot{\text{V}}$/$\dot{\text{Q}}$ mismatch in ILD lungs if vasodilators are administered systemically. Another reason might simply be the fact that vasoreagibility is fundamentally different in vessels located in otherwise normal lung tissue in idiopathic PAH or cemented in fibrotic tissue in PH-ILD (6). At least the $\dot{\text{V}}$/$\dot{\text{Q}}$ mismatch can be avoided by inhalative administration of vasodilating drugs, which will be distributed preferentially to the well-ventilated areas of the lungs, thus even reducing intrapulmonary mismatch and shunting. This concept has been successfully explored in single dose studies more than 20 years ago (7, 8). The recently published INCREASE study could show for the first time a significant positive effect of inhaled treprostinil, a stable prostacyclin analogon with potent vasodilative effects on pulmonary vasculature, in a population of patients with fibrotic ILD and PH confirmed with right heart catheter (9). The primary endpoint of the pivotal study—change in 6-minute-walking distance (6MWD) from baseline to Week 16—was significantly positive, showing improvement of 6MWD of 31 m (P < 0.001) in the active treatment group for the very first time in this population (9). The 6MWD can be regarded as a clinically meaningful endpoint per se, as it addresses reduced exercise tolerance, which is the cardinal clinical limitation of this patient population. Nonetheless, the question arises: how does this effect translate into other clinically meaningful endpoints? Time to clinical worsening has successfully been used as another clinical meaningful study endpoint in multiple PAH trials (10). In the INCREASE pivotal study, therefore, time to clinical worsening defined by hospitalization for a cardiopulmonary indication, a decrease in 6MWD greater than 15% from baseline, death from any cause, or lung transplantation was already a composite secondary endpoint and significantly favored the treprostinil-treated group (9).

Beyond the primary analysis, there remained urging questions to answer: does the observed effect apply equally to the different ILD subgroups included in this study? What happens after clinical worsening has occurred, obviously excluding death and transplantation? Does this mean that treatment is not efficacious and should be stopped, or could continuing treatment be beneficial, nonetheless? In this issue of the Journal, the INCREASE authors (pp. 198–207) set out to answer exactly these important questions in a post hoc analysis of the INCREASE data set focusing on ILD subgroups and consecutive clinical worsening events (11). The subgroup analysis revealed a very consistent favorable trend for all subgroups and for all the event categories investigated, which, however, reached statistical significance only for the overall number of events in the largest group of idiopathic interstitial pneumonias (11). This is not to blame, as statistical power was adjusted to the total study population and not to the subgroups. The clinical worsening events included a ⩾15% decline in 6MWD, cardiopulmonary hospitalization, lung transplantation, or death during the study and a ⩾10% decline in the FVC and exacerbations of underlying lung disease; the latter two criteria were collected as safety endpoints and were not included as part of the prespecified composite of clinical worsening. This elegant analysis demonstrated a huge imbalance of worsening events in favor of treprostinil therapy: in the inhaled treprostinil group, 11/163 patients had >1 event versus 26/163 patients in the placebo group; this also included a clearly lower mortality rate with inhaled treprostinil, suggesting that continuing treatment is the way to go (11). However, there are limitations that apply to this study and to post hoc analyses in general. Most importantly, the two subgroup cohorts of patients with >1 worsening event can no longer be regarded as a randomized subset, as the selection of these groups occurred not in a randomized fashion but following clinical events. Although the baseline characteristics show good matching, we do not know whether the functional status of the patients was still comparable at the time when the first event occurred (11). Moreover, two of the worsening categories, ⩾10% decline in FVC and exacerbation of the pulmonary disease, were captured as safety variables and might therefore be less well standardized as variables that are captured for endpoint assessments. Because FVC decline and exacerbation accounted for almost 50% of all worsening events, this is a potential weakness of this analysis. Despite acknowledging these limitations, the current analysis provides a very consistent picture of the positive effects of continued inhaled treprostinil in PH-ILD and clearly further supports the primary analysis (9, 11). After a long and sometimes painful journey of negative and inconclusive clinical trials, inhaled treprostinil therapy is a success for the treatment of PH-ILD, finally. The question to be addressed now and in future trials will be is this impressive effect of inhaled treprostinil durable?