Fulvestrant for the Treatment of Pulmonary Arterial Hypertension

To the Editor:

Postmenopausal women and men with pulmonary arterial hypertension (PAH) have higher circulating estrogen levels than respective control subjects, and higher levels of estrogen are associated with worse hemodynamics and lower 6-minute-walk distance (6MWD) in PAH (1, 2). We have shown increased expression of estrogen receptor α (ERα) in the small muscular pulmonary arteries of women with PAH (3). In PAH, hematopoietic progenitor cells (HPCs) are increased in the bone marrow, and estradiol increases HPC proliferation and circulation via ERα leading to lung microvessel growth (4, 5). The link between ERα signaling and HPCs has not been well studied in PAH.

Fulvestrant is a U.S. Food and Drug Administration (FDA)-approved intramuscularly administered ERα inhibitor for the treatment of metastatic breast cancer and for combination use. Fulvestrant blocks dimerization of ERα, decreases ERα expression, and limits nuclear translocation of transcriptional activating factors. We aimed to determine if fulvestrant would improve right ventricular function, reduce HPCs and other biomarkers, and increase 6MWD in PAH.

Methods

We performed a proof-of-concept open-label pilot study of fulvestrant in PAH. The trial protocol (see online supplement) was approved by the institutional review board and registered at clinicaltrials.gov before enrollment (NCT02911844). We enrolled five postmenopausal women with PAH. We excluded patients using hormone therapy or warfarin, with a history of breast cancer, or with World Health Organization class IV functional status or Child-Pugh C liver disease (see clinicaltrials.gov for complete criteria and online supplement). Endpoints included changes from baseline to follow-up (∼9 weeks) in 6MWD, plasma N-terminal pro-brain natriuretic peptide, HPCs, estradiol levels and other blood biomarkers, and echocardiographic measures (right ventricular systolic pressure, stroke volume, and tricuspid annular plane systolic excursion). The lung uptake of ¹⁸F-fluoroestradiol (¹⁸F-FES), an established probe for estradiol binding to ER, was measured using positron emission tomography–computed tomography imaging before and after treatment with fulvestrant under a distinct protocol and informed consent (NCT02899533) and under an investigational new drug application (FDA #126041) (6).

For flow cytometry, peripheral blood mononuclear cells (PBMCs) were isolated by density-gradient centrifugation and washed to remove platelet contamination. PBMCs (10⁶ cells) were labeled with PeCy5.5 anti-CD45 (Invitrogen; cat# MHCD4518), fluorescein isothiocyanate anti-CD34 (BD Biosciences; cat# 555821), and allophycocyanin anti-CD133 (Miltenyi Biotech; cat# 130-90-826) or with appropriate isotype controls and were analyzed on a MacsQuant 10 (Miltenyi Biotech). HPCs were defined as CD34⁺CD133⁺ cells, which were CD45dim (expressed as percentage of PBMCs) (7). Estradiol and other estradiol metabolites were measured in plasma using liquid chromatography–mass spectrometry (8).

Fulvestrant was administered in a 500-mg dose intramuscularly on Days 0, 14, 28, and 56 by study personnel. Research assessments were performed at baseline (Day 0) before the first dose of fulvestrant and at 9 weeks (Day 63).

Sign-rank tests and Spearman rho were used as appropriate. The analysis plan included reporting P values, but they were not used for inference. A data-sharing statement is provided online.

Results

The median age of the study subjects was 55 years (range, 52–66 yr), three were non-Hispanic white, two were non-Hispanic black, and the median body mass index was 31.7 kg/m² (range, 26.6–41.4 kg/m²). Two had idiopathic PAH, two had connective tissue disease–related PAH, and one had HIV PAH. Four were in World Health Organization functional class II and one in class III. Two were treated with phosphodiesterase 5 inhibitors, one with endothelin-receptor antagonists, and three with combination therapy. The median change in 6MWD after treatment with fulvestrant was +31 m (interquartile range, 10–32 m; P = 0.50) (Table 1), and there was no change in functional class (not shown). There was no change in tricuspid annular plane systolic excursion or right ventricular systolic pressure, but the change in stroke volume was 6.4 ml (IQR, 2.2–16.2 ml; P = 0.07). Figure 1 shows HPC levels before and after treatment with fulvestrant. The correlation (rho) between the changes in 16α-hydroxyestradiol (estriol) (16OHE2) and HPCs after treatment with fulvestrant was 0.90 (P = 0.04). There was no change in ¹⁸F-FES uptake in the lungs after the administration of fulvestrant.

Table 1.

Baseline, follow-up, and median (interquartile) change in assessments for patients

	N	Baseline	Follow-Up (∼9 wk)	Median Change	P Value
6-min-walk distance, m	5	347 (336 to 488)	365 (357 to 519)	31 (10 to 32)	0.50
Echocardiography
Tricuspid annular plane systolic excursion, mm	4	19.5 (14.5 to 28.5)	25 (18 to 27.5)	2 (−3 to 5.5)	0.47
Right ventricular systolic pressure, mm Hg	3	86 (35 to 100)	87 (40 to 89)	1 (−11 to 5)	1.0
Right ventricular index of myocardial performance	4	0.29 (0.26 to 0.66)	0.52 (0.37 to 0.82)	0.095 (−0.04 to 0.2)	0.47
Stroke volume, ml	4	56 (36 to 73.5)	62.2 (48 to 79.9)	6.4 (2.2 to 16.2)	0.07
Biomarkers
17-β Estradiol, pg/ml	5	3.24 (2.94 to 5.64)	2.68 (0.76 to 3.94)	−1.26 (−1.7 to −0.26)	0.14
17-α Estradiol, pg/ml	5	10.38 (3.5 to 17.3)	10.4 (9.8 to 31.9)	6.3 (−1.68 to 14.6)	0.50
Estrone, pg/ml	5	29.86 (22.24 to 31.02)	22.96 (19.84 to 47.82)	−6.26 (−8.06 to 13.24)	0.69
16α-hydroxyestrone (16OHE1), pg/ml	5	43.58 (2.64 to 58.7)	28.9 (9.38 to 34.26)	−9.32 (−23.22 to −1.46)	0.22
16α-hydroxyestradiol (16OHE2), pg/ml	5	16.34 (3.82 to 17.18)	6.86 (1.72 to 9.66)	−2.1 (−10.32 to −2.1)	0.06
N-terminal pro–brain natriuretic peptide, pg/ml	5	33.8 (28.1 to 1,542)	70.3 (26.9 to 2,150)	42.1 (6.7 to 114)	0.14
Homeostatic model assessment index	5	3.9 (1.4 to 6.1)	3.0 (1.7 to 4.6)	0.3 (−1.0 to 0.6)	0.89
Total cholesterol, mg/dl	5	187 (157 to 227)	185 (167 to 203)	−2 (−16 to 2)	0.42
High-density lipoprotein, mg/dl	5	45 (40 to 51)	47 (40 to 60)	9 (2 to 9)	0.13
Triglycerides, mg/dl	5	139 (107 to 196)	133 (123 to 137)	−16 (−30 to 3)	0.42
Low-density lipoprotein, mg/dl	5	112 (87 to 122)	108 (100 to 110)	−12 (−17 to −4)	0.22
Interleukin-6, pg/ml	5	2.8 (2.4 to 2.9)	1.8 (1.2 to 2.9)	−0.7 (−1.8 to 0.06)	0.35
Oxidized low-density lipoprotein, mU/L	5	51,828.7 (44,746.4 to 55,361.4)	49,983 (45,283.6 to 57,910)	−259.9 (−5,377.9 to 537.2)	0.69
18F-fluoroestradiol uptake, standardized uptake value	5	0.73 (0.62 to 1)	0.69 (0.64 to 1.03)	0.04 (−0.09 to 0.7)	0.69

Figure 1.

Hematopoietic progenitor cells (expressed as percentage of peripheral blood mononuclear cells [PBMCs]) before and after fulvestrant.

Two patients had diarrhea, one patient had injection site pain, one had headache, one had hot flashes and shoulder pain, and one had worsened dyspnea, joint pain, worsening hemodynamics, and increased N-terminal pro-brain natriuretic peptide. No adverse events were considered serious.

Discussion

In this pilot proof-of-concept study, fulvestrant may have been associated with numerically higher 6MWD, increasing stroke volume, and a decrease in 16OHE2 and circulating HPCs. There were no changes in biomarkers of insulin resistance, inflammation, or in other echocardiographic measures. There was no change in ¹⁸F-FES uptake in the lungs. The study population was well defined and available, and the intervention was successfully administered to all subjects. Fulvestrant was well tolerated in this small study of short duration, although one person had clinical worsening (20%; 95% confidence interval, 0–72%). We recently published a small double-blind randomized clinical trial, which showed that anastrozole significantly improved the 6MWD by 26 m without an increase in adverse events (9). There is a larger phase II trial of anastrozole ongoing (PHANTOM [Pulmonary Hypertension and Anastrozole Trial]; NCT03229499).

Estrogen stimulates HPCs to mobilize and proliferate via ERα (5), and bone marrow xenografting from patients with PAH to animals leads to HPCs, which cause pulmonary hypertension (4, 10). Fulvestrant may have decreased 16OHE2 (produced from estradiol and 16α-hydroxyestrone) which induces migration and proliferation of blood outgrowth endothelial cells from PAH patients (11) and circulating HPCs. 16-hydroxylated metabolites may also contribute to experimental PAH via miR-29, bone morphogenetic protein receptor type II signaling, and insulin response (12–16). Reduction of 16OHE2 and circulating HPCs could reverse existing vascular remodeling. Recently, administration of fulvestrant and anastrozole to the bone morphogenetic protein type II transgenic mouse model of PAH for 2 weeks showed improvement in morphologic changes and hemodynamics (12). In that study, estrogen inhibition also reversed metabolic abnormalities, including reducing oxidized lipids and insulin resistance, which was not seen in our study and might be specific to the experimental model.

Using positron emission tomography/computed tomography imaging, we did not find changes in ¹⁸F-FES uptake in the lungs of patients with PAH with fulvestrant. This might be because the signal is insufficient to detect ER blockade in the small muscular vessels of the lungs or because we did not achieve sufficient ER blockade with the dose used. The latter is unlikely, as this “high-dose” regimen is the standard for breast cancer and, in one patient, a focal area of increased uptake in the left breast disappeared after treatment with fulvestrant. Subsequent follow-up of that patient with mammogram and ultrasound did not show any clinically significant abnormality. The study is limited by the small size and open-label and uncontrolled design, making it difficult to directly attribute the findings to the study intervention. The short duration of the study may have affected the results. We administered the study drug to all patients, who would need to identify someone (or self-administer the drug) for chronic use. Because of the cost of fulvestrant, changes in the patent may affect feasibility of future investigation.

Future studies of fulvestrant and other modifiers of the estrogen pathway will benefit from being longer in duration, including sophisticated assessment of estrogen metabolites, and assessing patient-reported outcomes. Inclusion of biomarkers that measure angiogenic potential may shed light on the mechanism of the intervention. A larger phase II study with randomization and a placebo arm may be the next step in studying fulvestrant in PAH.