Abstract
Rationale
Inhibition of aromatase with anastrozole reduces pulmonary hypertension in experimental models.
Objectives
We aimed to determine whether anastrozole improved the 6-minute-walk distance (6MWD) at 6 months in pulmonary arterial hypertension (PAH).
Methods
We performed a randomized, double-blind, placebo-controlled phase II clinical trial of anastrozole in subjects with PAH at seven centers. Eighty-four postmenopausal women with PAH and men with PAH were randomized in a 1:1 ratio to receive anastrozole 1 mg or placebo by mouth daily, stratified by sex using permuted blocks of variable sizes. All subjects and study staff were masked. The primary outcome was the change from baseline in 6MWD at 6 months. By intention-to-treat analysis, we estimated the treatment effect of anastrozole using linear regression models adjusted for sex and baseline 6MWD. Assuming 10% loss to follow-up, we anticipated having 80% power to detect a difference in the change in 6MWD of 22 meters.
Measurements and Main Results
Forty-one subjects were randomized to placebo and 43 to anastrozole, and all received the allocated treatment. Three subjects in the placebo group and two in the anastrozole group discontinued the study drug. There was no significant difference in the change in 6MWD at 6 months (placebo-corrected treatment effect, −7.9 m; 95% confidence interval, −32.7 to 16.9; P = 0.53). There was no difference in adverse events between the groups.
Conclusions
Anastrozole did not show a significant effect on 6MWD compared with placebo in postmenopausal women with PAH and in men with PAH. Anastrozole was safe and did not have adverse effects.
Clinical trial registered with www.clincialtrials.gov (NCT03229499).
Keywords: pulmonary hypertension, clinical trial, anastrozole, sex hormones
At a Glance Commentary
Scientific Knowledge on the Subject
To our knowledge, this is the first multicenter phase II randomized clinical trial of aromatase inhibition in pulmonary arterial hypertension.
What This Study Adds to the Field
We have shown that anastrozole 1 mg daily did not affect the 6-minute-walk distance at 6 months. Anastrozole was safe and well tolerated and did not appear to have adverse effects on the right ventricle.
Pulmonary arterial hypertension (PAH) includes idiopathic pulmonary arterial hypertension (IPAH) and heritable forms as well as PAH associated with systemic conditions such as connective tissue disease and portal hypertension. In PAH, the small muscular pulmonary arteries show endothelial proliferation and smooth muscle hypertrophy, in situ thrombosis, and plexiform lesions. Right ventricular (RV) failure ensues, leading to exercise limitation and death. Although a number of therapies have been approved for PAH, most current treatments target the same three pathobiologic pathways (prostacyclin, nitric oxide, and endothelin-1).
Female sex is the best established risk factor for IPAH and heritable PAH; however, females with PAH have better outcomes than males (1–3), suggesting a potential role for sex hormones in the pathogenesis for PAH. A single nucleotide polymorphism in the promoter region of aromatase (which converts androgens to estrogen) not only was associated with a higher concentration of circulating 17β-estradiol (E2) but also increased the risk of PAH in patients with cirrhosis in two studies (4, 5). Pre- and postmenopausal females and males with PAH have higher circulating E2 concentrations, which are associated with worse disease burden, including shorter 6-minute-walk distance (6MWD), although premenopausal females showed variability in these relationships on the basis of menstrual phase (6–10).
Administration and/or deprivation of estrogens has mixed effects in experimental models of pulmonary hypertension (PH). Aromatase is produced in the smooth muscle cells of the small muscular pulmonary arteries of both female animal models of PH and women and men with PAH (11). Administration of the aromatase inhibitor anastrozole (AN) reduced pulmonary arterial pressures, pulmonary vascular changes, and indices of RV hypertrophy in two different experimental models of PH (11); metformin had similar effects via aromatase inhibition (12). We published a pilot two-center, randomized, double-blind, placebo-controlled trial which showed that AN significantly increased the 6MWD at 3 months in 12 subjects with PAH compared with 6 subjects with PAH treated with placebo (13).
We conducted a 12-month multicenter phase II clinical trial of AN in subjects with PAH. We hypothesized that AN would increase the 6MWD at 6 months compared with placebo. Other secondary outcomes were assessed through 12 months.
Methods
Trial Design
The PHANTOM (Pulmonary Hypertension and Anastrozole) study was a seven-center, randomized, double-blind, placebo-controlled parallel 12-month study of AN in patients with PAH. The protocol called for the recruitment of 84 subjects with PAH. The trial protocol was approved by the single institutional review board at the University of Pennsylvania and the data and safety monitoring board, which was appointed by the NHLBI. Details of the methods, the protocol, and the manual of procedures are provided online. The trial was registered with www.clinicaltrials.gov before recruitment was initiated (NCT03229499).
Participants
We included postmenopausal women and men >18 years of age with PAH who were able to perform 6-minute-walk testing. We excluded patients who were being treated with estrogen or antihormone therapy and those with a history of invasive breast cancer, World Health Organization (WHO) class IV functional status, or severe osteoporosis. Patients were allowed to receive concurrent treatment with any PAH-specific medications. Complete inclusion and exclusion criteria are provided in Table E1 of the online supplement. Participants were recruited from seven academic pulmonary vascular disease clinics (Stanford University, University of Pennsylvania, University of Colorado, Washington University, Vanderbilt University, Johns Hopkins University, and Rhode Island Hospital/Brown University). All participants provided written informed consent.
Study Interventions and Procedures
Potentially eligible patients were identified through prescreening of medical records. After informed consent and successful screening, subjects were randomly assigned in a 1:1 ratio by a Web-based computerized system to AN 1 mg (Zydus Pharmaceuticals) once daily or manufactured placebo (Temple University), which appeared identical. The randomization scheme was blocked and stratified by sex. All subjects were provided with open-label vitamin D3 2,000 IU daily (Major Pharmaceuticals and Rugby Laboratories) and received dietary counseling about maintaining 1,200-mg calcium intake each day.
Subjects were to be evaluated at the study centers at baseline, 3 months, 6 months, and 12 months with telephone calls at 9 months. Remote study visits were instituted instead of in-person visits after clinic shutdowns during the COVID-19 pandemic (see protocol).
Outcomes
The primary outcome was the difference in change from baseline in 6MWD at 6 months. Secondary outcomes included the changes from baseline in 6MWD at 3, 6, and 12 months; in RV function assessed by echocardiography at 6 and 12 months; in N-terminal prohormone brain natriuretic peptide (NT-proBNP) and other estrogen and biomarker concentrations; in functional class; in physical activity measured by actigraphy; in 36-item Short Form Health Survey scores; and in emPHasis-10 scores at 3, 6, and 12 months. Time to clinical worsening (TTCW) was assessed through 12 months. TTCW was defined as the time to the first event, including addition of new PAH therapies or dose increases in previously stable PAH therapy for increased symptoms, hospitalization for PAH progression and/or right-sided heart failure, lung transplant, atrial septostomy, and all-cause death. An independent blinded committee adjudicated all potential clinical worsening endpoints and deaths. Changes in bone mineral density from baseline to 12 months were assessed using dual-energy X-ray absorptiometry. Details of endpoint assessments, laboratory analyses, and monitoring plan are provided in the online supplement, protocol, and manual of procedures.
Randomization and Masking
The randomization assignment was generated by a biostatistician who remained blinded throughout the study, using permuted blocks of sizes 2 and 4 and stratified by sex. When an eligible patient received his or her allocation, research coordinators were provided with the kit number for that patient, which contained both the study intervention or placebo and vitamin D. All staff and subjects were masked in terms of the assignment to either the AN or placebo arm, and the tablets were identical in appearance. Only the research pharmacist and the analyst for the data and safety monitoring board closed session were unblinded.
Sample Size
With a 1:1 ratio of randomization to AN and placebo arms and an anticipated 10% loss to follow-up rate at 6 months, we estimated that a sample size of 84 subjects would provide 80% power to detect a mean difference of 0.64 SD units or 22 meters in change in 6MWD between AN and placebo using a two-sided test, with α = 0.05 representing the minimally important mean group difference in PAH (14). We anticipated 80% power to detect a difference of 31 meters for males or females only.
Statistical Methods
All analyses were conducted according to the intention-to-treat principle. Continuous variables were presented as mean ± SD or median (interquartile range [IQR]) and categorical variables as frequency (percent). The primary endpoint was compared between the two study arms using a linear regression model adjusted for sex and baseline 6MWD. Other continuous endpoints at 6 months were analyzed similarly, adjusted for baseline values. Longitudinal and continuous secondary outcomes were compared between the two arms using linear mixed-effects models adjusted for sex, baseline value of the corresponding variable, time since baseline visit, and within-subject correlation. Parametric estimates and significance testing were performed using the residual maximum likelihood method. Longitudinal and categorical secondary outcomes were compared between the two arms using logistic mixed-effects models, adjusted for sex, baseline status, time since baseline visit, and within-subject correlation. Covariates were included as fixed effects, and a subject-specific intercept was included as a random effect, in the regression models. The comparison of TTCW between the two treatment arms was performed using the Cox proportional hazards model, adjusting for sex and baseline 6MWD. The proportional hazards assumption was assessed using both survival graphs and formal statistical tests of zero slopes in the Schoenfeld residuals. Subset analyses of the primary outcome were also performed within females and males using the same approach. Potential effect modification by sex was assessed using a drug × sex interaction term in prespecified hypothesis-generating secondary analyses.
The primary analyses used complete data without imputation. To assess potential bias resulting from missing data, we conducted sensitivity analyses under the assumption of missing at random using the inverse probability weighting method (15). In the linear regression model for the change in 6MWD at 6 months, each subject was weighted by the inverse probability of being a complete case, estimated using arm assignment, sex, and baseline 6MWD.
Because of a batch with a technical error in the assay, there were some missing E2 concentrations. These were imputed from estrone, sex, age, and follow-up time using multiple imputation by chained equations. We did not plan for formal interim analyses or early stopping boundaries for the trial. All analyses were conducted with R version 4.2.2. No multiplicity adjustment was made for the secondary analyses; nominal P values are reported.
Results
We screened 2,144 patients with PAH during the enrollment period (Figure 1). The first patient was randomized on December 7, 2017, and the last on July 7, 2021, with final follow-up on July 19, 2022. Eighty-four participants were randomized, and 79 were assessed for the primary outcome at 6 months (Figure 1); 81 patients completed follow-up at 12 months, of whom 78 had 6-minute-walk data. The primary outcome was assessed at 6 months; however, the trial had a total duration of 12 months to collect data on long-term effects and safety.
Figure 1.
CONSORT (Consolidated Standards of Reporting Trials) diagram. 6MWD = 6-minute-walk distance; 6MWT = 6-minute-walk test.
Forty-three subjects were randomized to AN and 41 to placebo (Table 1). The two groups were similar in terms of age, sex at birth, race, ethnicity, and body mass index. Approximately half in each arm had IPAH or heritable PAH. Most patients in both arms were in WHO functional class II, and 80% were receiving dual or triple therapy for PAH. 6MWD was significantly reduced from normal in the study sample (median, 83% predicted normal 6MWD; median of the actual vs. predicted 6MWD, −89 m) (16).
Table 1.
Baseline Characteristics of Intention-to-Treat Population
| Characteristic | n | Anastrozole (n = 43) | Placebo (n = 41) |
|---|---|---|---|
| Age, yr | 84 | 58.6 ± 11.8 | 60.3 ± 10.2 |
| Sex at birth, female | 84 | 26 (60.5%) | 25 (61.0%) |
| Race | |||
| White | 84 | 31 (72.1%) | 33 (80.5%) |
| Black | 5 (11.6%) | 3 (7.3%) | |
| Asian | 4 (9.3%) | 1 (2.4%) | |
| Other | 3 (7.0%) | 4 (9.8%) | |
| Ethnicity | |||
| Not Hispanic/Latino | 84 | 41 (95.3%) | 36 (87.8%) |
| Hispanic/Latino | 2 (4.7%) | 5 (12.2%) | |
| Body mass index, kg/m2 | 84 | 30.7 ± 6.4 | 28.3 ± 5.9 |
| PAH etiology | |||
| Idiopathic | 84 | 18 (41.9%) | 15 (36.6%) |
| Associated with connective tissue disease | 8 (18.6%) | 11 (26.8%) | |
| Drug or toxin induced | 6 (14.0%) | 4 (9.8%) | |
| Associated with congenital heart disease | 5 (11.6%) | 2 (4.9%) | |
| Heritable | 3 (7.0%) | 4 (9.8%) | |
| Associated with HIV infection | 2 (4.7%) | 2 (4.9%) | |
| Portopulmonary hypertension | 1 (2.3%) | 3 (7.3%) | |
| Number of background therapy classes | |||
| Single | 84 | 8 (18.6%) | 5 (12.2%) |
| Dual | 22 (51.2%) | 25 (61.0%) | |
| Triple | 13 (30.2%) | 11 (26.8%) | |
| WHO functional class | |||
| I | 84 | 6 (14.0%) | 3 (7.3%) |
| II | 28 (65.1%) | 26 (63.4%) | |
| III | 9 (20.9%) | 12 (29.3%) | |
| Six minute walk distance, m | 84 | 442.8 ± 119.5 | 416.3 ± 115.3 |
| NT-proBNP, pg/ml | 84 | 187 [79, 424] | 256 [132, 556] |
Definition of abbreviations: NT-proBNP = N-terminal prohormone brain natriuretic peptide; PAH = pulmonary arterial hypertension; WHO = World Health Organization.
Values are mean ± SD, median [IQR], or n (%).
No patients were lost to follow-up over the 12 months, and no patient completely withdrew from the study (Figure 1). Four subjects in the placebo arm discontinued study drug at 89, 162, 178, and 359 days, and two subjects in the AN arm discontinued at 81 and 91 days because of adverse events but continued follow-up. Because only three subjects in the placebo arm and two in the AN arm discontinued the experimental intervention before 6 months, we did not conduct sensitivity analyses for nonadherence excluding these subjects.
Randomization to AN did not affect the difference in change in 6MWD at 6 months after adjustment for baseline 6MWD and sex (placebo-adjusted treatment effect, −7.9 m; 95% confidence interval [CI], −32.7 to 16.9 m; P = 0.53) (Figure 2A). The estimates were similar when including 3-, 6-, and 12-month assessments using linear mixed-effects models adjusted for sex, baseline 6MWD, time, and within-subject correlation (−8.8 m; 95% CI, −29.3 to 11.7 m; P = 0.39). Sensitivity analysis to account for missing data using inverse probability weighting of being a complete case yielded similar results (Table E2). Sex-specific analyses of the 6MWD showed similar results (data not shown), and we found no evidence of interaction between treatment assignment and sex in terms of 6MWD at 6 months (P for interaction = 0.93) or over 1 year (P for interaction = 0.71).
Figure 2.
Longitudinal changes in outcomes. Least squares mean (95% confidence interval) change from baseline in (A) 6-minute-walk distance and (B) log(N-terminal prohormone brain natriuretic peptide), adjusted for baseline value and sex.
The least squares mean difference in change in log(NT-proBNP [pg/ml]) over 6 months was 0.02 (95% CI, −0.12 to 0.15; P = 0.80) after adjustment for baseline log(NT-proBNP) and sex and was similar when including 3-, 6-, and 12-month assessments using linear mixed-effects models adjusted for sex, baseline log(NT-proBNP), time, and within-subject correlation (0.01; 95% CI, −0.10 to 0.12; P = 0.88) (Figure 2B).
The odds ratio for being in WHO functional class III/IV versus I/II for AN compared with the placebo arm at 6 months (adjusted for sex and baseline functional class) was 2.34 (95% CI, 0.63–10.2; P = 0.22), and over 12 months, it was 2.31 (95% CI, 0.64–8.36; P = 0.20) after adjustment for sex, baseline WHO functional class, time, and within-subject correlation (Figure E1). AN treatment was not associated with the mental component or physical component summary scores of the 36-item Short Form Health Survey or the emPHasis-10 score (Figure E2). There was no difference in any physical activity features between patients randomized to AN versus placebo (Figure E3).
Echocardiography was performed at baseline, 6 months, and 12 months. There were no differences in RV free wall strain, RV systolic or diastolic areas, fractional area change, tissue Doppler, RV systolic pressure, Tei index, tricuspid annular plane systolic excursion, tricuspid annular plane systolic excursion/RV systolic pressure, stroke volume, or cardiac index after adjustment for sex, baseline value, time, and within-subject correlation between the AN and placebo arms. There were no effects of AN on indices of left ventricular function (data not shown).
Randomization to AN significantly reduced concentrations of log(estrone [pg/ml]) at 6 months (−1.18; 95% CI, −1.43 to −0.93; P < 0.001; n = 80) after adjustment for sex and baseline estrone and over 12 months (0.86; 95% CI, −1.02 to −0.69; P < 0.001; n = 84) after adjustment for sex, baseline estrone, time, and within-subject correlation (Figure 3A). As anticipated, postmenopausal women had very low E2 concentrations, especially after treatment with AN, leading to a substantial number of E2 concentrations being below the level of detection or in a low range. For results determined to be between 0.25 and 1 pg/ml, we imputed a value of 0.625 pg/ml. For samples below the lower limit of detection (0.25 pg/ml), we imputed a value of 0.125 pg/ml. For participants who had E2 concentrations that were not missing because of the technical error, AN significantly reduced log(E2 [pg/ml]) concentrations at 6 months (−0.57; 95% CI, −0.91 to −0.23; P = 0.001; n = 61) with adjustment for sex and baseline level and over 12 months using a linear mixed-effects model with adjustment for sex, baseline E2, time, and within-subject correlation (−0.38; 95% CI, −0.60 to −0.16; P = 0.001; n = 64) (Figure 3B). Expressed another way, AN reduced estrone concentrations by (median [IQR]) 95% (99%, 70%) compared with placebo, which showed a small increase in estrone of 2% (−18%, 15%) at 12 months. AN reduced E2 concentrations by 48% (91%, 17%) compared with placebo, which showed no change (−30%, 14%) at 12 months.
Figure 3.
Longitudinal changes in estrogens. Least squares mean change (95% confidence interval) from baseline in (A) log(estrone) and (B) log(estradiol), adjusted for baseline value and sex.
Sensitivity analyses imputing E2 concentrations affected by the technical problem from estrone, sex, age, and time using multiple imputation by chained equations showed similar results (Figure E4). The treatment effect of AN on estrone and E2 did not differ by sex at 6 months (estrone P for interaction = 0.13; E2 P for interaction = 0.80) and over 12 months (P for interaction = 0.08 and 0.62, respectively). AN had no effects on dehydroepiandrosterone-sulfate, C-reactive protein, IL-6, insulin, monocyte chemoattractive protein-1, or oxidized low-density lipoprotein but may have decreased vascular endothelial growth factor (Table E3).
Median follow-up time was 364 days (IQR, 356, 376) and similar in both arms. An initial clinical worsening event occurred in eight placebo patients (21.3 per 100 person-years) and seven AN patients (17.3 per 100 person-years) (Table 2). Three cardiovascular deaths occurred in the placebo group, and none occurred in the AN group. The TTCW did not differ between the two groups (hazard ratio for AN vs. placebo adjusted for 6MWD and sex was 1.06; 95% CI, 0.37–3.06; P = 0.99) (Figure 4). The proportional hazards assumption as assessed using both survival graphs and formal statistical tests of zero slopes in the Schoenfeld residuals was not violated.
Table 2.
Clinical Worsening Events
| Anastrozole (n = 43) | Placebo (n = 41) | HR (95% CI)* | P Value | |
|---|---|---|---|---|
| Patients with first clinical worsening event, n (%) | 7 (16.2) | 8 (19.5) | 1.06 (0.37, 3.06) | 0.99 |
| Patients with a clinical worsening event, n (%) | ||||
| Death | 0 (0.0) | 3 (3.6) | — | — |
| Hospitalization for PAH progression or right-sided heart failure | 7 (16.2) | 3 (7.3) | 4.53 (0.95, 21.6) | 0.058 |
| Addition of new PAH therapy or dose increase for increased symptoms | 2 (4.7) | 5 (12.2) | 0.41 (0.08, 2.14) | 0.29 |
| Lung transplant | 0 (0.0) | 0 (0.0) | — | — |
| Atrial septostomy | 0 (0.0) | 0 (0.0) | — | — |
Definition of abbreviations: CI = confidence interval; HR = hazard ratio; PAH = pulmonary arterial hypertension.
Data are n (%).
Adjusted for 6-minute-walk distance and sex.
Figure 4.
Time to clinical worsening.
Adverse events occurred in 38 (93%) subjects randomized to placebo and 40 (93%) subjects randomized to AN (P = 0.98). Serious adverse events occurred in 10 (24%) and 16 (37%) subjects, respectively (P = 0.20). The most common adverse events are listed in Table 3. We specifically inquired about the side effects of AN, including hot flashes, vaginal dryness, and arthralgias (noted in the table). There was no significant effect of AN on bone mineral density at the lumbar spine, total hip, or femoral neck (Table E4).
Table 3.
Adverse Events
| Event | Anastrozole (n = 43) | Placebo (n = 41) |
|---|---|---|
| Upper respiratory infection | 9 (20.9%) | 6 (14.6%) |
| Dyspnea | 7 (16.3%) | 4 (9.8%) |
| Edema limbs | 5 (11.6%) | 5 (12.2%) |
| Hot flashes* | 6 (14.0%) | 4 (9.8%) |
| Arthralgia* | 6 (14.0%) | 4 (9.8%) |
| Headache | 5 (11.6%) | 4 (9.8%) |
| Diarrhea | 5 (11.6%) | 3 (7.3%) |
| Depression | 3 (7.0%) | 5 (12.2%) |
| Nausea | 3 (7.0%) | 5 (12.2%) |
| Pain in extremity | 3 (7.0%) | 5 (12.2%) |
| Hypotension | 3 (7.0%) | 4 (9.8%) |
| Back pain | 4 (9.3%) | 3 (7.3%) |
| Fatigue | 4 (9.3%) | 3 (7.3%) |
| Pruritus | 1 (2.3%) | 4 (9.8%) |
| Myalgia | 4 (9.3%) | 1 (2.4%) |
| Pneumonia | 3 (7.0%) | 2 (4.9%) |
| Vaginal dryness* | 2 (7.7%) | 3 (12.0%) |
| Anemia | 2 (4.7%) | 3 (7.3%) |
| Catheter placement | 2 (4.7%) | 3 (7.3%) |
| Cough | 3 (7.0%) | 2 (4.9%) |
| Joint range of motion decreased | 1 (2.3%) | 3 (7.3%) |
| Coronavirus infection | 3 (7.0%) | 1 (2.4%) |
| Vertigo | 3 (7.0%) | 0 (0.0%) |
| Anxiety | 0 (0.0%) | 3 (7.3%) |
| Presyncope | 0 (0.0%) | 3 (7.3%) |
| Palpitations | 0 (0.0%) | 3 (7.3%) |
| Heart failure | 0 (0.0%) | 3 (7.3%) |
Only includes events that occurred in at least 5.0% of either arm. Data shown as n (%).
Events specifically queried at each study visit (vaginal dryness queried only in those born female, n = 51).
Discussion
In this phase II, placebo-controlled, randomized clinical trial, AN did not have an effect on 6MWD at 6 months or 6MWD, NT-proBNP, RV function, or health-related quality of life over 1 year in PAH. TTCW was similar between the AN and placebo groups, although more deaths in the placebo group were balanced by more hospitalizations in the AN group. Retention in the trial was outstanding (especially considering that study activities occurred during the COVID-19 pandemic), as was adherence to the study drug. There were minimal missing data, and the trial fully recruited its initial target number of subjects in a timely fashion. The study had adequate power to detect the minimally important group mean difference in 6MWD (14). Although higher than in the prior study of AN, 6MWD was still significantly reduced when compared with that predicted by age, sex, height, and weight (13, 16). AN was safe and well tolerated in this population of postmenopausal women with PAH and men with PAH and was not associated with serious adverse events or worsened RV function by echocardiography or NT-proBNP at 1 year, addressing concerns that reduction in E2 could be detrimental to the right ventricle (RV).
Female sex is a well-established risk factor for PAH, specifically for IPAH and heritable PAH. Although one study found that female sex increased the risk of portopulmonary hypertension, a more recent study did not (5, 17). In certain animal models, female sex predisposes to PH and higher severity, whereas similar risks in male and female animals are seen in other models. Estrogen has potentially beneficial proangiogenic, antiapoptotic, and cardioprotective effects on the RV, such that some have argued that inhibiting estrogen may actually be harmful in patients with PAH (18–20). Studies have shown that women with PAH have better RV systolic function at baseline than men and that RV ejection fraction increases more in women than in men after PAH treatment, potentially explaining the longer survival in women with PAH (1, 21–25). This study showed that reduction in estrogen concentrations did not appear to be harmful to patients with PAH, and particularly RV size and function, within the time frame studied.
Administration of AN after Sugen-hypoxia administration reduced RV systolic pressure, pulmonary vascular remodeling, and RV mass in the female animals (11). Chen and colleagues demonstrated that fulvestrant and AN both prevented and treated PH in the BMPR2 transgenic mouse model (26). In a pilot, randomized, double-blind, placebo-controlled phase II study that allocated 12 patients with PAH to AN and 6 to placebo, patients receiving AN had a significantly longer 6MWD over 12 weeks (13). AN reduced circulating E2 concentrations by 40%, but to a lesser effect than normally seen in postmenopausal women with breast cancer treated with AN (>60–90%) (27, 28). Liquid chromatography/mass spectrometry is currently the standard for measuring sex hormones but may be challenged by the exceptionally low E2 concentrations seen in postmenopausal women, especially those receiving AN. We were able to show that AN lowered estrone and E2 to a degree similar to that seen in our pilot study, despite many concentrations being below detection and technical issues with a single batch of samples.
Unlike our prior study, this study showed a null finding despite having adequate power to detect the minimal clinically important group difference in 6MWD (14) observed in our prior study with similar estrogen suppression. Although this study was not powered to detect smaller differences in 6MWD or in time to clinical worsening or cardiovascular death, there was no signal of benefit in terms of biomarkers, physical activity, quality of life, or echocardiographic parameters, strengthening our conclusions. Explanations for the discrepant prior results include type I error, lack of treatment with vitamin D in the prior study (provided to all patients in this study) (29), differences in the study samples, or the short duration of the prior study. Although more patients in the present study were on combination therapy (reflecting the evolving standard of care) and had longer 6MWD, the 6MWD was still reduced, making a “ceiling effect” an unlikely explanation. Even if this trial enrolled a “healthier” population, we would still expect some signal of an effect on the many secondary endpoints, which we did not see. We examined certain groups for treatment heterogeneity (men and women) a priori; however, we have resisted trying to identify “responder” groups in the main analysis of this null trial, which can be fraught with problems and can lead to type I error. The composite TTCW was similar between the arms; however, patients receiving AN seemed to have more PAH hospitalizations, and patients receiving placebo seemed to have more deaths. However, these comprised very small numbers of events, making inferences difficult.
Limitations
Our study has several limitations. Although there were some missing data, the completeness of the data, considering the COVID-19 pandemic and clinic shutdowns, and the lower-than-anticipated withdrawal of study treatment and from the study overall are major strengths of the study. The study population was limited to males and postmenopausal females because aromatase inhibitors alone do not suppress estrogen concentrations in the setting of functioning ovaries. Premenopausal females with PAH may have a distinct estrogen-related endophenotype (7, 30), which could be responsive to treatment (NCT03528902). We did not include patients with severe PAH (e.g., WHO functional class IV patients), limiting generalizability to this group. There were technical issues in measuring E2 in a random sample of patients, and E2 concentrations below the level of detection for the liquid chromatography/mass spectrometry assay at baseline made change hard to detect. This was mitigated by the use of estrone (typically higher, making post–aromatase inhibitor measurement easier), which showed a significant decrease in men and women, confirming that patients did adhere to the study treatment, which acted in the expected fashion.
The hemodynamic response to AN would have been of interest; however, the requirement for multiple right heart catheterizations in otherwise stable, treated patients would have rendered the trial infeasible. In addition, we have shown that short-term hemodynamics are not adequate surrogate endpoints in PAH trials (31). Although a small number of participants discontinued the study intervention (which would bias to the null), the significant reductions in estrogen concentrations make this an unlikely explanation for the results.
In summary, we have shown that AN 1 mg each day lowered E2 and estrone but did not affect the 6MWD, echocardiographic parameters of RV function, physical activity, quality of life, and other biomarkers in PAH. Aromatase inhibition in postmenopausal women (and in men) appeared safe in PAH. Future randomized controlled trials of hormone signaling in premenopausal women are warranted.
Data Sharing
Data sharing is described in the Online Material.
Supplemental Materials
Acknowledgments
Acknowledgment
The authors thank the following individuals for their assistance with this study: Tiffany Sharkoski, M.S.; Mamta Patel, R.N.; Diane Pinder, B.S.; Lisa Wesby, M.S.; Marie Durborow, B.S.; Tess LaPatra, B.S. DXA Core: Ann Schwartz, Ph.D. Echocardiography Core: Barbara G. Manahan, R.D.C.S.; Elle McIntire, R.D.C.S.; Jennifer Warmsbecker, R.D.C.S. Adjudication Committee: Steven Nathan, M.D.; Joel Wirth, M.D.; Belinda Lebron, M.D., M.S. Data and Safety Monitoring Board: Scott M. Palmer, M.D., M.H.S., Chair; Karen A. Fagan, M.D.; Ann H. Partridge, M.D., M.P.H.; Harold Collard, M.D.; Michael Szarek, Ph.D., M.S.; Nancy M. P. King, J.D. Unblinded statistician: Scarlett Bellamy, Ph.D.
Footnotes
Supported by the National Institutes of Health (R01 HL134905, R01 HL134904, R01 HL141268, and 5UL1TR002243-03). The liquid chromatography/mass spectrometry studies were supported by the Medical Research Council (MR/T015713) and the British Heart Foundation (RG/F/21/110047). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Author Contributions: S.M.K., C.E.V., R.F., S.S.E., and A.D.: study concept and design, study conduct, interpretation of the data, and drafting and revision of the manuscript. R.Z., T.B., M.C., S.C.M., A.H., G.L., M.D., J. Minhas, and H.S.: study conduct and interpretation of the data. R.A., M.M., and I.S.: study conduct and interpretation of the data. E.A.: study concept and design and interpretation of the data. N.S.: study conduct, data analysis, and interpretation of the data. J. Moutchia: data analysis, interpretation of the data, and drafting and revision of the manuscript. All authors reviewed the work critically for important intellectual content and reviewed the final version before submission.
A data supplement for this article is available via the Supplements tab at the top of the online article.
Originally Published in Press as DOI: 10.1164/rccm.202402-0371OC on May 15, 2024
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1. Jacobs W, van de Veerdonk MC, Trip P, de Man F, Heymans MW, Marcus JT, et al. The right ventricle explains sex differences in survival in idiopathic pulmonary arterial hypertension. Chest . 2014;145:1230–1236. doi: 10.1378/chest.13-1291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Humbert M, Sitbon O, Chaouat A, Bertocchi M, Habib G, Gressin V, et al. Survival in patients with idiopathic, familial, and anorexigen-associated pulmonary arterial hypertension in the modern management era. Circulation . 2010;122:156–163. doi: 10.1161/CIRCULATIONAHA.109.911818. [DOI] [PubMed] [Google Scholar]
- 3. Evans JDW, Girerd B, Montani D, Wang X-J, Galiè N, Austin ED, et al. BMPR2 mutations and survival in pulmonary arterial hypertension: an individual participant data meta-analysis. Lancet Respir Med . 2016;4:129–137. doi: 10.1016/S2213-2600(15)00544-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Roberts KE, Fallon MB, Krowka MJ, Brown RS, Trotter JF, Peter I, et al. Pulmonary Vascular Complications of Liver Disease Study Group Genetic risk factors for portopulmonary hypertension in patients with advanced liver disease. Am J Respir Crit Care Med . 2009;179:835–842. doi: 10.1164/rccm.200809-1472OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Al-Naamani N, Krowka MJ, Forde KA, Krok KL, Feng R, Heresi GA, et al. Pulmonary Vascular Complications of Liver Disease Study Group Estrogen signaling and portopulmonary hypertension: the Pulmonary Vascular Complications of Liver Disease Study (PVCLD2) Hepatology . 2021;73:726–737. doi: 10.1002/hep.31314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Baird GL, Archer-Chicko C, Barr RG, Bluemke DA, Foderaro AE, Fritz JS, et al. Lower DHEA-S levels predict disease and worse outcomes in post-menopausal women with idiopathic, connective tissue disease- and congenital heart disease-associated pulmonary arterial hypertension. Eur Respir J . 2018;51:1800467. doi: 10.1183/13993003.00467-2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Baird GL, Walsh T, Aliotta J, Allahua M, Andrew R, Bourjeily G, et al. Insights from the menstrual cycle in pulmonary arterial hypertension. Ann Am Thorac Soc . 2021;18:218–228. doi: 10.1513/AnnalsATS.202006-671OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ventetuolo CE, Baird GL, Barr RG, Bluemke DA, Fritz JS, Hill NS, et al. Higher estradiol and lower dehydroepiandrosterone-sulfate Levels are associated with pulmonary arterial hypertension in men. Am J Respir Crit Care Med . 2016;193:1168–1175. doi: 10.1164/rccm.201509-1785OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jing Z-C, Xu X-Q, Han Z-Y, Wu Y, Deng K-W, Wang H, et al. Registry and survival study in Chinese patients with idiopathic and familial pulmonary arterial hypertension. Chest . 2007;132:373–379. doi: 10.1378/chest.06-2913. [DOI] [PubMed] [Google Scholar]
- 10. Denver N, Homer NZM, Andrew R, Harvey KY, Morrell N, Austin ED, et al. Estrogen metabolites in a small cohort of patients with idiopathic pulmonary arterial hypertension. Pulm Circ . 2020;10:2045894020908783. doi: 10.1177/2045894020908783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mair KM, Wright AF, Duggan N, Rowlands DJ, Hussey MJ, Roberts S, et al. Sex-dependent influence of endogenous estrogen in pulmonary hypertension. Am J Respir Crit Care Med . 2014;190:456–467. doi: 10.1164/rccm.201403-0483OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Dean A, Nilsen M, Loughlin L, Salt IP, MacLean MR. Metformin reverses development of pulmonary hypertension via aromatase inhibition. Hypertension . 2016;68:446–454. doi: 10.1161/HYPERTENSIONAHA.116.07353. [DOI] [PubMed] [Google Scholar]
- 13. Kawut SM, Archer-Chicko CL, DeMichele A, Fritz JS, Klinger JR, Ky B, et al. Anastrozole in pulmonary arterial hypertension. A randomized, double-blind, placebo-controlled trial. Am J Respir Crit Care Med . 2017;195:360–368. doi: 10.1164/rccm.201605-1024OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Moutchia J, McClelland RL, Al-Naamani N, Appleby DH, Blank K, Grinnan D, et al. Minimal clinically important difference in the 6-minute-walk distance for patients with pulmonary arterial hypertension. Am J Respir Crit Care Med . 2023;207:1070–1079. doi: 10.1164/rccm.202208-1547OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Hogan JW, Lancaster T. Instrumental variables and inverse probability weighting for causal inference from longitudinal observational studies. Stat Methods Med Res . 2004;13:17–48. doi: 10.1191/0962280204sm351ra. [DOI] [PubMed] [Google Scholar]
- 16. Enright PL, Sherrill DL. Reference equations for the six-minute walk in healthy adults. Am J Respir Crit Care Med . 1998;158:1384–1387. doi: 10.1164/ajrccm.158.5.9710086. [DOI] [PubMed] [Google Scholar]
- 17. Kawut SM, Krowka MJ, Trotter JF, Roberts KE, Benza RL, Badesch DB, et al. Pulmonary Vascular Complications of Liver Disease Study Group Clinical risk factors for portopulmonary hypertension. Hepatology . 2008;48:196–203. doi: 10.1002/hep.22275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Frump AL, Yakubov B, Walts A, Fisher A, Cook T, Chesler NC, et al. Estrogen receptor-α exerts endothelium-protective effects and attenuates pulmonary hypertension. Am J Respir Cell Mol Biol . 2023;68:341–344. doi: 10.1165/rcmb.2022-0224LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Frump AL, Albrecht M, Yakubov B, Breuils-Bonnet S, Nadeau V, Tremblay E, et al. 17β-Estradiol and estrogen receptor α protect right ventricular function in pulmonary hypertension via BMPR2 and apelin. J Clin Invest . 2021;131:e129433. doi: 10.1172/JCI129433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hester J, Ventetuolo C, Lahm T. Sex, gender, and sex hormones in pulmonary hypertension and right ventricular failure. Compr Physiol . 2019;10:125–170. doi: 10.1002/cphy.c190011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kawut SM, Al-Naamani N, Agerstrand C, Berman Rosenzweig E, Rowan C, Barst RJ, et al. Determinants of right ventricular ejection fraction in pulmonary arterial hypertension. Chest . 2009;135:752–759. doi: 10.1378/chest.08-1758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Ventetuolo CE, Mitra N, Wan F, Manichaikul A, Barr RG, Johnson C, et al. Oestradiol metabolism and androgen receptor genotypes are associated with right ventricular function. Eur Respir J . 2016;47:553–563. doi: 10.1183/13993003.01083-2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kawut SM, Lima JA, Barr RG, Chahal H, Jain A, Tandri H, et al. Sex and race differences in right ventricular structure and function: the Multi-Ethnic Study of Atherosclerosis–Right Ventricle study. Circulation . 2011;123:2542–2551. doi: 10.1161/CIRCULATIONAHA.110.985515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Ventetuolo CE, Ouyang P, Bluemke DA, Tandri H, Barr RG, Bagiella E, et al. Sex hormones are associated with right ventricular structure and function: the MESA-Right Ventricle study. Am J Respir Crit Care Med . 2011;183:659–667. doi: 10.1164/rccm.201007-1027OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Swift AJ, Capener D, Johns C, Hamilton N, Rothman A, Elliot C, et al. Magnetic resonance imaging in the prognostic evaluation of patients with pulmonary arterial hypertension. Am J Respir Crit Care Med . 2017;196:228–239. doi: 10.1164/rccm.201611-2365OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Chen X, Austin ED, Talati M, Fessel JP, Farber-Eger EH, Brittain EL, et al. Oestrogen inhibition reverses pulmonary arterial hypertension and associated metabolic defects. Eur Respir J . 2017;50:1602337. doi: 10.1183/13993003.02337-2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Kyvernitakis I, Albert US, Kalder M, Winarno AS, Hars O, Hadji P. Effect of anastrozole on hormone levels in postmenopausal women with early breast cancer. Climacteric . 2015;18:63–68. doi: 10.3109/13697137.2014.929105. [DOI] [PubMed] [Google Scholar]
- 28. Bajetta E, Martinetti A, Zilembo N, Pozzi P, La Torre I, Ferrari L, et al. Biological activity of anastrozole in postmenopausal patients with advanced breast cancer: effects on estrogens and bone metabolism. Ann Oncol . 2002;13:1059–1066. doi: 10.1093/annonc/mdf083. [DOI] [PubMed] [Google Scholar]
- 29. Atamañuk AN, Litewka DF, Baratta SJ, Seropian IM, Perez Prados G, Payaslian MO, et al. Vitamin D deficiency among patients with pulmonary hypertension. BMC Pulm Med . 2019;19:258. doi: 10.1186/s12890-019-1011-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. van Wezenbeek J, Groeneveldt JA, Llucià-Valldeperas A, van der Bruggen CE, Jansen SMA, Smits AJ, et al. Interplay of sex hormones and long-term right ventricular adaptation in a Dutch PAH-cohort. J Heart Lung Transplant . 2022;41:445–457. doi: 10.1016/j.healun.2021.11.004. [DOI] [PubMed] [Google Scholar]
- 31. Ventetuolo CE, Gabler NB, Fritz JS, Smith KA, Palevsky HI, Klinger JR, et al. Are hemodynamics surrogate end points in pulmonary arterial hypertension? Circulation . 2014;130:768–775. doi: 10.1161/CIRCULATIONAHA.114.009690. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.