This is a prospective, open-label, multicenter, single-arm cohort study conducted from September 2018 to February 2022 at eight centers in China. After a 4-week screening period, all eligible patients will receive 5 mg of ambisentan once daily for a 12-week initial treatment period. This is followed by a 12-week extension treatment period during which the 5 mg dose will be maintained or the dose will be increased up to 10 mg once daily. The domestic ambrisentan used in this study is provided by China Jiangsu Hansoh Pharmaceutical Group Co.

Chinese subjects aged 18–75 years, diagnosed with group 1 PAH of WHO-updated clinical classification (19) (WHO-FC II–III), who have not received PAH medication treatment or are currently receiving stable doses of non-endothelin receptor antagonist (non-ERA) PAH medications for at least 4 weeks, were enrolled in this study. All subjects must have a baseline 6-min walk distance of at least 150 m. Right heart catheterization (RHC) has been performed within 6 months prior to screening and meets the following hemodynamic criteria: mean pulmonary artery pressure (mPAP) ≥25 mmHg, pulmonary vascular resistance (PVR) ≥3 wood units, and pulmonary artery wedge pressure (PAWP) or left ventricular end-diastolic pressure ≤15 mmHg. Pulmonary function tests have been performed within 6 months prior to screening and meet the following criteria: total lung capacity ≥60% and forced expiratory volume in the first second (FEV1) ≥55% of predicted normal values.

Subjects with serum alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels >2 × ULN, serum bilirubin levels >1.5 × ULN, severe hepatic insufficiency (Child–Pugh grade C), severe renal insufficiency (creatinine clearance <30 ml/min), hemoglobin concentration <100 g/L or hematocrit <30%, severe hypotension (diastolic blood pressure <50 mmHg or systolic blood pressure <90 mmHg) were excluded. Pregnant and lactating women were excluded.

The primary endpoint for efficacy was Δ6MWD at week 12. The secondary endpoints included Δ6MWD at week 24, changes in the Borg scale, plasma BNP level, WHO-FC, and echocardiogram from baseline to weeks 12 and 24. The echocardiogram measurements include TAPSE (tricuspid annular plane systolic excursion) and TAPSE/PASP (pulmonary arterial systolic pressure). Clinical worsening is defined as the occurrence of all-cause death, hospitalization for PAH deterioration, and change or addition of PAH medications within 24 weeks of treatment duration. The exploratory endpoints were risk stratification improvement, TTCI, and TTCI-low within 24 weeks of treatment duration. TTCI was defined as the time from initiation of treatment to the first occurrence of improvement in at least one level of four-strata risk stratification without clinical worsening. TTCI-low was defined as the time from initiation of treatment to the first occurrence of patients achieving low risk of four-strata risk stratification without clinical worsening (20). All efficacy endpoints were assessed at baseline and every 6 weeks. The safety endpoints were adverse events and their severity, laboratory tests, and liver function.

The four-strata risk stratification was determined in accordance with the latest guideline protocol (1, 21). In brief, 6MWD >440, 320–440, 165–319, and <165 m were respectively scored as 1, 2, 3, and 4 points; WHO-FC I, II, III, and IV were respectively scored as 1, 1, 3, and 4 points; and BNP <50, 50–199, 200–800, and >800 ng/L were, respectively, scored as 1, 2, 3, and 4 points. At least two of the above three variables must be utilized in the risk stratification process, and the rounded average value was the risk point. The final calculated 1, 2, 3, and 4 risk points indicated low, intermediate-low, intermediate-high, and high risk, respectively.

Assuming a dropout rate of 10% and a study power of 90%, a sample size of 40 patients was required to detect a two-sided test significant difference of an increase in Δ6MWD of 41.8 m with a standard deviation of 83 m after 12 weeks of treatment (6, 18).

This study used the intent-to-treat (ITT) population to describe baseline/demographic data and follow-up efficacy endpoints, including those who received at least one dose of ambrisentan and had baseline measurements and at least one follow-up measurement. First, risk stratification was calculated by raw data of 6MWD, BNP, WHO_FC. Then, the last observation carry forward (LOCF) method of imputation for missing data was used for 6MWD, Borg scale, WHO-FC, risk stratification, TAPSE, and TAPSE/PASP. Subgroup analyses were conducted based on age, gender, PAH classification, WHO-FC, risk, and treatment groups at baseline. Normal variables were described by means (SDs), non-normal variables were described by medians (quartiles), and ordinal variables were described by composition ratios. For comparing baseline and follow-up efficacy endpoints, normal variables were compared by paired t-tests and non-normal variables and ordinal variables were compared by Wilcoxon signed-rank tests. TTCI, TTCI-low, and TTCW were analyzed by Kaplan–Meier analysis and univariate/multivariate COX model regression analysis. Two treatment group analyses were performed using two independent-sample t-tests to compare normal variables and Mann–Whitney U tests to compare non-normal and ordinal variables. All statistical analyses were conducted using SPSS version 26.0. Sample size calculations were performed using PASS version 15.0.

Safety was assessed by the safety population, including only those who received at least one dose of ambrisentan treatment.

A total of 83 subjects were successfully enrolled and completed initial treatment. Nine patients withdrew from the study primarily due to drug interruption caused by the COVID-19 pandemic, resulting in 74 subjects entering the extension phase of treatment (Figure 1). The first subject was enrolled in December 2018. The last subject follow-up time was in September 2021. The number of subjects enrolled by each center can be seen in Supplementary Table S1 in the supplementary materials. The average exposure time of all subjects was 160.1 (44.4) days. One subject was excluded from ITT population because of the lack of baseline efficacy endpoint data. Thus, 82 subjects were enrolled in the ITT population, including 48 subjects who had never received PAH medication therapy (naive group) and 34 subjects who had been receiving one sort of non-ERA PAH medication on a stable dose (add-on group).

The demographic/baseline characteristics are shown in Table 1. The study population was primarily women (81.7%), with a median age of 36 years. The baseline 6MWD of the patients was 403.5 (78.0) m. The baseline risk stratification was mostly intermediate risk, with 36 (43.9%) patients at intermediate-low risk, 45 (54.9%) patients at intermediate-high risk, and no subjects at low risk. Sildenafil was the most combined PAH medication in the add-on group (85.3%). Except for age (P = 0.034), there were no significant differences in other demographic/baseline characteristics between the naive group and add-on group subjects.

A significant improvement in Δ6MWD was observed since week 6, with further significant clinical improvement at week 12 (42.2 m, P < 0.0001) and week 24 (53.4 m, P < 0.0001). As illustrated in Figure 2 and Table 2, changes in the Borg scale, WHO-FC, BNP, TAPSE, and TAPSE/PASP all demonstrated significant improvements at weeks 12 and 24, which was consistent with Δ6MWD. The result of subgroup analyses of Δ6MWD was similar to the pattern noted in the overall population. Also, there were no statistical differences in Δ6MWD between subgroups (Supplementary Table S2 in the supplementary material). There were no statistical differences in Δ6MWD at week 24 between the naive group (50.9 m) and the add-on group (56.9 m, P = 0.609). Additionally, there were no significant differences in other traditional efficacy endpoints between the two groups (Table 2).

A total of 29 (35.4%) subjects achieved low-risk status within 24 weeks without clinical worsening (Figure 2D). Kaplan–Meier analysis of TTCI-low revealed that the cumulative improvement rate was 41.4% within 24 weeks (Figure 3D). Subjects in the intermediate-low-risk group reached low-risk status significantly earlier than those in the intermediate-high-risk group (log-rank P < 0.0001, Figure 3E).

Univariate COX model regression analysis showed TTCI-low was associated with Δ6MWD at week 24 and most of the baseline characteristics, including RHC (sRVP, mPAP, PVR, SvO₂), 6MWD, WHO-FC, BNP, echocardiogram (TAPSE, PASP, TAPSE/PASP), and risk stratification at baseline (Supplementary Table S3 in the supplementary material).

A total of 30 (36.6%) and 53 (64.6%) subjects improved in at least one level of four-strata risk stratification within 12 and 24 weeks without clinical worsening (Table 2). Kaplan–Meier analysis of TTCI revealed that the median improvement time was 131 (95% CI: 125.4–136.7) days, and the cumulative improvement rate was 75.1% (Figure 3A and Table 2). There was no significant difference in TTCI between intermediate-low-risk subjects and intermediate-high-risk subjects (median improvement time was 133 (95% CI: 73.9–178.1) days vs. 126 (95% CI: 128.6–137.4) days, log-rank P = 0.51, Figure 3B). TTCI of naive group patients is shorter than that of the add-on group patients [median improvement time and cumulative improvement rate were respectively 94.0 (95% CI: 55.6–132.4) days and 84.1% vs. 173.0 (95% CI: 121.4–224.6) days and 57.9%, log-rank P = 0.008, Figure 3C and Table 2].

Univariate COX model regression analysis showed that TTCI was associated with the treatment group, Δ6MWD at week 24, change in WHO-FC at week 24, and PASP and TAPSE/PASP at baseline. After adjustment for these above variables, the treatment group, Δ6MWD, and change in WHO-FC at week 24 were independent influencing factors of TTCI, with hazard ratios of 0.314 (95% CI: 0.164–0.603, P < 0.001), 1.009 (95% CI: 1.002–1.016, P = 0.017), and 2.974 (95% CI: 1.703–5.191, P < 0.001), respectively (Table 3).

Three (3.6%) subjects experienced clinical worsening events. One was hospitalized for PAH, two added other PAH medications within the study duration (one added sildenafil and one added tadalafil), and no patient experienced all-cause death or lung transplantation within the study duration (Table 4 and Figure 3F).

A total of 45 (54.2%) subjects experienced adverse events. The most common adverse events were flushing (14.5%), palpitation (9.6%), and peripheral edema (8.4%). No significant changes were observed in hematological parameters. One patient withdrew from the study due to a > 3 × ULN elevation in alanine aminotransferase (ALT) enzymes (Table 4), and all the other adverse events were mild to moderate in severity.