Command Palette
Search for a command to run...
Inhaled treprostinil in PH-COPD: PERFECT stops for harm and the extension closes
United Therapeutics closed the open-label extension of PERFECT after the Phase 3 parent trial of inhaled treprostinil in pulmonary hypertension due to COPD was halted for excess serious adverse events and suggestive excess mortality.
Mechanism Risk Score
| Component | Points |
|---|---|
| Phase-weighted failure burden | 15.7 / 40 |
| Archetype severity | 8.5 / 25 |
| Temporal recency | 4.3 / 15 |
| Genetic evidence deficit | 5.9 / 15 |
| Programmatic saturation | 2.5 / 5 |
For PTGIR in Pulmonary hypertension due to COPD (PH-COPD), the Mechanism Risk Score is 37/100 (yellow band). The score is a failure-burden index derived from Claidex post-mortems on this target–disease pair, not a probability of approval.
MRS 37/100 (YELLOW). 1 programs across PTGIR have been documented for PTGIR in Pulmonary hypertension due to COPD (PH-COPD): 1 Phase 3, 0 Phase 2, 0 Phase 1 — of which 0 were efficacy failures, 1 safety, 0 biomarker, and 0 operational (enrollment, sponsor, or funding). The most informative failure on file is Inhaled treprostinil in PH-COPD: PERFECT stops for harm and the extension closes. This score quantifies the documented failure burden; the Open Targets association score of 0.61 reflects moderate genetic support, neither rescuing nor compounding the failure record. The MRS is not a prediction of future trial outcomes — it is a structured summary of the empirical record, recomputed live from the Claidex claims table, and intended to flag mechanisms where any new program must explicitly resolve each prior failure mode before pursuit is justified.
This score does not predict whether the next trial will succeed. It flags how heavy the documented mechanistic failure record is before a new program is justified.
What was tried
NCT03794583 was the open-label extension of the PERFECT program, United Therapeutics' Phase 3 evaluation of inhaled treprostinil in pulmonary hypertension due to chronic obstructive pulmonary disease (PH-COPD). The extension was single-arm, gave inhaled treprostinil solution four times daily during waking hours, and was meant to carry responders forward from the randomized parent study. It enrolled 41 patients, recorded a primary completion date of November 2022, and was terminated. The registry reason reads "Sponsor's decision." That decision did not happen in isolation. It followed the parent PERFECT randomized trial, which an independent data and safety monitoring committee stopped early.
The biological hypothesis
Treprostinil is a prostacyclin analog and an agonist at the prostacyclin (IP) receptor, the product of PTGIR (ChEMBL CHEMBL1237119). IP receptor signaling raises cyclic AMP in pulmonary vascular smooth muscle, producing vasodilation and anti-proliferative effects. In Group 1 pulmonary arterial hypertension this is one of the oldest validated axes in the field. Open Targets scores PTGIR to pulmonary hypertension at 0.61, a figure driven almost entirely by clinical evidence rather than human genetics, which fits a target defined by pharmacology rather than Mendelian disease.
The PH-COPD bet rested on a real prior success in a neighboring indication. INCREASE tested inhaled treprostinil in pulmonary hypertension due to interstitial lung disease and met its primary endpoint, improving six-minute walk distance, and a prespecified analysis showed an improvement in forced vital capacity (Waxman 2021, Nathan 2021). ILD and COPD are both Group 3 pulmonary hypertension, both common, both short on therapies. Extending a positive Group 3 result from ILD to COPD looked like a reasonable next step.
What actually happened
PERFECT did not reach a positive readout. The randomized trial enrolled 76 patients, 66 of whom received inhaled treprostinil and 58 placebo across its crossover and contingent parallel designs. The DSMC halted it on the totality of evidence that treprostinil increased the risk of serious adverse events, with suggestive evidence of increased mortality. The incidence of an adverse response was 36.4 percent (24 of 66) on treprostinil versus 27.6 percent (16 of 58) on placebo. The most common treatment-emergent serious adverse events were acute respiratory failure and COPD exacerbations (Nathan 2024). The open-label extension recorded here was then closed by sponsor decision, the predictable consequence of stopping the parent trial for harm.
Aggregate post-marketing data show treprostinil is a physiologically active drug with a real adverse event profile. The openFDA FAERS database holds 10,858 treprostinil reports, 8,293 of them serious, including 1,146 deaths, with dyspnoea (1,881) and decreased oxygen saturation (532) among the frequent terms. Those signals come mostly from approved PAH use and are not specific to COPD, but they point at the same physiology that PERFECT exposed.
Failure mechanism, best guess
The most likely mechanism is worsened ventilation-perfusion matching. COPD lungs are heterogeneously diseased. Hypoxic pulmonary vasoconstriction is a protective reflex that shunts blood away from poorly ventilated alveoli toward better-ventilated units. An inhaled vasodilator that reaches the pulmonary circulation can blunt that reflex non-selectively, increasing perfusion to low-ventilation regions, widening the shunt fraction, and lowering arterial oxygen. Acute respiratory failure and COPD exacerbations, the dominant serious events in PERFECT, are exactly what that mechanism would produce.
The contrast with INCREASE is the diagnostic clue. The target was engaged in both trials. In ILD the lung is more uniformly restricted and the vasodilator did not worsen gas exchange enough to cause net harm. In COPD the same target engagement was harmful. This is a translational mismatch within Group 3 pulmonary hypertension, not a failure of IP receptor pharmacology. The drug worked as designed and the design was wrong for the lung.
How to prevent this next time
Two quantitative tools would have changed this program before it reached a Phase 3 mortality signal.
First, a physiology-based safety run-in with an explicit gas-exchange stopping rule. The known liability of pulmonary vasodilators in obstructive lung disease is hypoxemia from V/Q mismatch. A short crossover measuring arterial oxygen and shunt fraction on drug versus off, with a predefined number-needed-to-harm threshold, would have quantified the hazard in tens of patients rather than revealing it as serious adverse events across a full trial.
Second, Bayesian predictive-probability monitoring on a harm-weighted composite. Rather than a fixed interim look, the trial could have tracked
with the posterior updated continuously and an early stop triggered when the predictive probability of net benefit fell below a preset floor. With an adverse response rate of 36.4 percent versus 27.6 percent, that floor would have been crossed early, sparing later enrollees. Enrichment by baseline phenotype, separating precapillary-dominant COPD-PH from patients with severe airflow obstruction and gas-exchange limitation, would have further narrowed who, if anyone, could tolerate the drug.
The single highest leverage change would have been to gate the COPD program on a gas-exchange safety run-in with a prespecified oxygenation stopping rule before committing to a Phase 3 outcome trial.
What this means for similar programs
Group 3 pulmonary hypertension is not one disease, and a positive result in one subtype does not license expansion to another. The ILD-to-COPD jump treated a regulatory category as if it were a shared physiology. For inhaled or systemic pulmonary vasodilators in any obstructive lung disease, the V/Q mismatch hazard should be treated as the default expectation and measured first. The broader lesson holds for any program that extends a validated mechanism into an adjacent indication: the question is not whether the target is engaged but whether engagement helps or harms in the new tissue context.
Open questions
Is there a precapillary-dominant COPD-PH subgroup, identifiable by hemodynamics or imaging, in which controlled pulmonary vasodilation is safe or beneficial? Would a more pulmonary-selective delivery or a lower starting dose change the gas-exchange risk? Did the PERFECT mortality signal reflect the drug, the underlying severity of enrolled patients, or both? And does the failure in COPD change how regulators should treat the umbrella Group 3 label in future trial design?
Sources
ClinicalTrials.gov. Open-Label Extension Study of Inhaled Treprostinil in Patients With PH-COPD, NCT03794583. https://clinicaltrials.gov/study/NCT03794583.
ChEMBL. Treprostinil (CHEMBL1237119), prostanoid IP receptor agonist. https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL1237119/.
openFDA FAERS. Treprostinil adverse event reports. https://api.fda.gov/drug/event.json?search=patient.drug.medicinalproduct:treprostinil.
Open Targets Platform. PTGIR to pulmonary hypertension association (MONDO_0005149). https://platform.opentargets.org/evidence/ENSG00000160013/MONDO_0005149.
Related failure claims
Linked claims sharing target, indication, or failure mechanism.