Sex and menopause differences in response to tadalafil: 6-minute walk distance and time to clinical worsening

Abstract Abstract

Pulmonary arterial hypertension (PAH) is a female-predominant disease, but there are little data on treatment response by sex and menopausal status. In this retrospective analysis of the Pulmonary Arterial Hypertension and Response to Tadalafil (PHIRST) randomized clinical trial, we assessed treatment response between the sexes by examining change in 6-minute walk distance (6MWD) and time to clinical worsening (TCW). We examined the effect of menopausal status on the same treatment measures. 6MWD was recorded before and after 16 weeks of treatment with tadalafil or placebo in the PHIRST study cohort of 340 subjects (264 females, 76 males). A univariate analysis was used to assess the effect of sex on change in 6MWD and TCW. Multivariate linear regression and Cox proportional hazards models were built for 6MWD and TCW, respectively. Women were subdivided by age as a surrogate for menopausal status. The linear trend test and the log-rank test were performed on change in 6MWD and TCW by age. For tadalafil-treated patients, a significant difference in change in 6MWD by sex (mean: 48.6 m for males vs. 34.7 m for females; P = 0.01) was found, but it was not significant in multivariate analysis (P = 0.08). There was a trend toward a female age-dependent effect in change in 6MWD; the premenopausal group showed the greatest improvement. A significant sex- or age-dependent effect on TCW was not present. In conclusion, this retrospective analysis of the PHIRST trial suggests that men and premenopausal women may experience greater functional improvement when treated with tadalafil than older women, but there was no consistent sex or menopausal effect on TCW.

Pulmonary arterial hypertension (PAH) is a group of disorders that are characterized by elevated pulmonary arterial resistance, leading to eventual right heart failure.^1,2 PAH is a female-predominate disease; according to the REVEAL (Registry to Evaluate Early and Long-Term PAH Disease Management) Registry, group 1 PAH is four times more likely in female patients than in male patients.^3-6 While the exact cause of the observed female predominance is unknown, it may potentially relate to a disruption in the protective effect of estrogen on the pulmonary vasculature and myocardium.^7-11 A recent study examining patient factors as a predictor of treatment response with tadalafil found male sex to be predictive of improved change in 6-minute walk distance (6MWD).¹² Despite this, there is a paucity of literature investigating the effect of menopause on response to PAH treatment. Providing more tailored therapy to patients may allow for more cost-conscious, targeted, and safer treatment for patients with PAH.

Two prior sex-specific treatment response studies found opposing results for response to tadalafil and endothelin receptor antagonists (ERAs). ERAs have demonstrated a more robust change in 6MWD in female patients, while tadalafil has demonstrated a significant increase in 6MWD in male patients.^12,13 The etiology of this sex difference between the treatment agents is unknown and could possibly involve differences in nitric oxide (NO) and endothelin signaling between the sexes.^14,15 Multiple animal studies have demonstrated that estrogen decreases pulmonary arterial vasoconstriction by decreasing expression of endothelin 1 and increasing release of NO and prostaglandin.^8,16-20 Neither of these studies addressed the effect of menopause on treatment response or evaluated time to clinical worsening (TCW) as a clinical end point.

In the general population, many vasoreactive conditions, such as migraine headache and Raynaud’s phenomenon, increase after menopause.¹⁷ This postmenopausal increase is also demonstrated in PAH and is postulated to be due to estrogen withdrawal, as it has been demonstrated in animal models that lower estrogen and progesterone states produce increased pulmonary vasoconstriction.^17,21-24 Despite the known effects of estrogen on the pulmonary vasculature, it remains unclear how menopause affects response to treatment of PAH in general, including with phosphodiesterase type 5 (PDE-5) inhibitors. A treatment agent with a postmenopausal-specific treatment benefit would be extremely valuable given the high prevalence of PAH in this population compared with that in men and younger women.

Tadalafil is a commonly prescribed and well-tolerated PDE-5 inhibitor. Multiple studies have indicated that treatment with PDE-5 inhibitors improves outcomes in PAH, but none have reported any sex-specific or menopausal status–specific treatment effects on TCW.^2,3 Our present study examines sex-specific treatment responses to tadalafil by assessing change in 6MWD and TCW using data from the pivotal randomized, placebo-controlled Pulmonary Arterial Hypertension and Response to Tadalafil (PHIRST) trial.³ We also report the effects of menopausal status on treatment response to tadalafil using age as a surrogate for menopausal status.

Methods

Selection of patients

Patient selection included those enrolled in the PHIRST study.³ For the purposes of this study, deidentified information was used, and therefore a separate institutional review board was not required. For the initial PHIRST study, the local institutional review boards or independent ethics committees approved the protocol, and written consent was obtained from all patients.³ Subjects were at least 12 years of age and had symptomatic PAH (the youngest male was 19 years old; the youngest female was 14 years old). Subjects exposed to anorexigens or given a diagnosis of connective tissue disease, HIV infection, congenital pulmonary shunts, or idiopathic PAH were included in this study. The hemodynamic criteria for PAH included mean pulmonary arterial pressure of ≥25 mmHg, pulmonary arterial wedge pressure of ≤15 mmHg, and pulmomary vascular resistance of ≥3 Wood units. Patients with a 6MWD of <150 or >450 m were excluded. Patients treated with intravenous epoprostenol, inhaled or intravenous iloprost, or subcutaneous or intravenous treprostinil were excluded. Patients taking a maximum dose of bosentan (125 mg) twice daily for a minimum of 12 weeks at the time of screening continued to receive bosentan in addition to the study medicine.

Study design

In total, 405 subjects (317 females, 88 males) were enrolled in a 16-week, double-blind, double-dummy, placebo-controlled multicenter study. Subjects were randomized into groups receiving placebo or 2.5, 10, 20, or 40 mg of tadalafil once daily. Baseline 6MWD and hemodynamics were recorded. 6MWD was recorded before and after 16 weeks of treatment with tadalafil or placebo in a cohort of 340 subjects (264 females, 76 males).³ Clinical worsening was defined as death, lung or heart-lung transplantation, atrial septostomy, hospitalization due to worsening PAH, initiation of new PAH-approved therapy, or worsening World Health Organization (WHO) functional class over the 16 weeks. 6MWD and WHO functional class were measured at baseline and at weeks 4, 8, 12, and 16.

Statistical analysis

When examining the effect of sex on change in 6MWD and TCW, only the tadalafil-treated patients were included (n = 323; males = 71, females = 252). For the univariate analyses, the Wilcoxon rank sum test was used to assess the effect of sex on change in 6MWD. A log-rank test was used to assess the effect of sex on TCW. Females were subdivided by age as a surrogate for menopausal status: 14–44 years, premenopausal; 45–54 years, perimenopausal; and ≥55 years, postmenopausal (based on the average age of menopause, 52.4 years).²⁵ The linear trend test with first degree of orthogonal polynomials was used to assess the effect of age on change in 6MWD among females and separately for males within the same age groups. A log-rank test was performed on TCW by age among females. Kaplan-Meier plots were constructed to investigate clinical worsening by sex and age. Multivariate linear regression and Cox proportion hazards models were constructed to assess the effect of sex on change in 6MWD and TCW, respectively, by adjusting for age, sex, race, etiology of PAH, WHO functional class, background therapy with bosentan, and baseline 6MWD. Goodness of model fit, colinearity, and numerical stability were also evaluated. All analyses were performed using SAS.

Results

Baseline characteristics by sex and menopausal status were collected and are reported in Tables 1 and 2. Between male and female patients, there was a significant difference in baseline weight, PAH etiology, and age. The female cohort was slightly younger (mean age: 53 vs. 57 years). For female patients, there was a significant difference between menopausal status groups in baseline 6MWD, with postmenopausal females having the lowest baseline 6MWD. In addition, there was a significant difference in PAH pathogenesis across the sex and menopausal states. The majority of each group reported having idiopathic PAH; however, atrial septal defect was predominant in premenopausal females, and cardiovascular disease was predominant in postmenopausal females.

Table 1.

Baseline characteristics of subjects by sex

	Male		Female
	n	Value	n	Value
Characteristic, mean (minimum, maximum)
Age, yearsa	88	57 (19, 90)	317	53 (14, 86)
Weight, kgb	88	82 (50, 155)	317	73 (40, 129)
Baseline 6MWD, m	88	337 (168, 450)	316	346 (150, 463)
Race
White	72	82.8	254	80.4
African American or black	5	5.7	29	9.2
Asian	10	11.5	24	7.6
American Indian or Alaska Native	0	0	5	1.6
Pacific Islander	0	0	2	0.6
Other	0	0	2	0.6
PAH etiologyb
Idiopathic/familial	68	77.3	179	56.5
Collagen vascular disease	9	10.2	86	27.1
Anorexigen use	2	2.3	14	4.4
Associated with an atrial septal defect	2	2.3	30	9.5
Surgical repair of VSD or PDA ≥1 year	7	8	8	2.5
Concomitant use of bosentan	41	46.6	175	55.2
Treatment dose
Placebo	17	19.3	65	20.5
Tadalafil 2.5 mg	18	20.5	64	20.2
Tadalafil 10 mg	13	14.8	67	21.1
Tadalafil 20 mg	20	22.7	62	19.6
Tadalafil 40 mg	20	22.7	59	18.6
WHO functional class
1	1	1.1	3	0.9
2	25	28.4	105	33.1
3	60	68.2	204	64.4
4	2	2.3	5	1.6

^NoteExcept where otherwise noted, data are no. (%) of patients. 6MWD: 6-minute walk distance; PAH: pulmonary arterial hypertension; VSD: ventricular septal defect; PDA: patent ductus arteriosus; WHO: World Health Organization.

^aP = 0.01.

^bP = 0.0001.

Table 2.

Baseline characteristics of subjects by menopause status

	Premenopausal		Perimenopausal		Postmenopausal
	n	Value	n	Value	n	Value
Characteristic, mean (minimum, maximum)
Age, yearsa	99	35 (15, 44)	71	49 (44, 55)	147	67 (56, 86)
Weight	99	70 (41, 128)	71	76 (40, 125)	147	74 (41, 129)
Baseline 6MWDa	99	371 (210, 450)	70	367 (180, 450)	147	318 (150, 463)
PAH etiologyb
Idiopathic/familial	50	50.5	47	66.2	82	55.8
Anorexigen use	4	4	5	7	5	3.4
Connective tissue disease	20	20.2	15	21.1	51	34.7
Associated with an atrial septal defect	21	21.2	4	5.6	5	3.4
Surgical repair of VSD or PDA ≥1 year	4	4	0	0	4	2.7
Concomitant use of bosentan	60	60.6	41	57.7	74	50.3
Treatment
Placebo	22	22.2	13	18.3	30	20.4
Tadalafil 2.5 mg	19	19.2	8	11.3	37	25.2
Tadalafil 10 mg	16	16.2	21	29.6	30	20.4
Tadalafil 20 mg	19	19.2	16	22.5	27	18.4
Tadalafil 40 mg	23	23.2	13	18.3	23	15.6

^NoteExcept where otherwise noted, data are no. (%) of patients. 6MWD: 6-minute walk distance; PAH: pulmonary arterial hypertension; VSD: ventricular septal defect; PDA: patent ductus arteriosus.

^aP < 0.0001.

^bP = 0.0001.

For tadalafil-treated patients, males were found to have a significantly greater improvement in 6MWD compared with females (mean [SD]: 48.6 [55] m for males vs. 34.7 [54] m for females; P = 0.01; Table 3). This difference was not statistically significant in the multivariate analysis (P = 0.08) after adjusting for covariates: age, sex, race, etiology of PAH, WHO functional class, background therapy with bosentan, and baseline 6MWD. A backward model selection was performed with 0.05 as the significance level for staying in the model to assess any covariate correlation. This model selection resulted in elimination of the variables background therapy with bosentan (P = 0.8854), race (P = 0.1935), and sex (P = 0.0647).

Table 3.

Change in 6-minute walk distance (6MWD) and time to clinical worsening (TCW) by sex

	Tadalafil		Placebo
Sex, parameter	n	Mean (SD)	n	Mean (SD)
Male
Change in 6MWD, m	62	48.6 (55)a	14	26.4 (63)
TCW, days	8	79 (36)b	3	62.3 (13)
Female
Change in 6MWD, m	210	34.7 (54)a	54	16.5 (52)
TCW, days	21	61 (29)b	10	39 (28)

^aP = 0.01 for the difference between male and female change in 6MWD.

^bP = 0.49 for the difference between male and female change in TCW.

Overall, tadalafil treatment demonstrated a beneficial effect on TCW for patients who experienced clinical worsening compared with placebo in both male and female patients (Table 3). Female patients randomized to receive tadalafil were found to have a shorter mean (SD) TCW compared with male patients randomized to receive tadalafil (61 [29] vs. 79 [36] days). A log-rank test did not show a consistent effect of sex on TCW for tadalafil-treated patients (P = 0.49). This difference in TCW between the sexes was found to be not statistically significant in the multivariate analysis (P = 0.41).

To assess the impact of menopausal status on treatment response, female subjects were subdivided by age as a surrogate for menopausal status (Table 4). Tadalafil treatment demonstrated a consistent beneficial effect on change in 6MWD from baseline compared with placebo in these subgroups. For tadalafil-treated patients, there was a trend among female subjects for an age-dependent effect on mean change in 6MWD: the premenopausal group showed the greatest improvement (mean [SD]: 45.7 [55] m), followed by the perimenopausal (mean [SD]: 41.1 [50] m) and the postmenopausal (mean [SD]: 24.5 [54] m) group. The linear trend of the effect of age on 6MWD in females was found to be statistically significant (P = 0.02). This age-related trend was not present in male subjects subdivided into the same age groups: the younger (mean [SD]: 73.2 [48] m) and older (mean [SD]: 42.8 [56] m) men showed greater improvement than middle-aged men (mean [SD]: 34.4 [52] m).

Table 4.

Change in 6-minute walk distance (6MWD) and time to clinical worsening (TCW) by menopausal status

	Tadalafil		Placebo
Age group, parameter	n	Mean (SD)	n	Mean (SD)
Premenopausal (14–44 years old)
Change in 6MWD, m	61	45.7 (55)	21	22.7 (45)
TCW, days	6	50.2 (31)	2	43 (20)
Perimenopausal (45–55 years old)
Change in 6MWD, m	52	41.1 (50)	12	33.4 (45)
TCW, days	4	84.5 (13)	1	29
Postmenopausal (>55 years old)
Change in 6MWD, m	97	24.5 (54)	21	0.8 (60)
TCW, days	11	58.3 (28)	7	39.3 (32.5)

Tadalafil was found to have a beneficial effect on TCW for patients who experienced clinical worsening in all female age groups compared with placebo. For tadalafil-treated patients, the log-rank test did not show a consistent age-dependent effect on TCW in females (P = 0.78).

Discussion

In this sex-based post hoc analysis of the PHIRST study that assessed the effect of tadalafil in patients with PAH, we found a significant difference between the sexes in 6MWD change but no consistent effect of sex on TCW. This difference in 6MWD was not significant in the multivariate analysis. In the female population, a trend was found for greater improvement in 6MWD for premenopausal compared with postmenopausal women. This menopausal status trend was not present for TCW.

The baseline characteristics were fairly similar between males and females, with a significant difference in PAH etiology, age, and weight. The female cohort reported more atrial septal defects and was slightly younger than the male cohort. Body mass index data were missing for most patients in this study; therefore, it is unclear whether the weight difference between the sexes is physiologically significant. Female patients were grouped according to age as a surrogate marker for menopausal status; significant differences between the age groups in the distribution of PAH etiology and in 6MWD at baseline were noted. It is unclear how significant the etiology distribution was to the overall results, since the majority of each age group consisted of women with idiopathic PAH. The difference in baseline 6MWD did reveal a significantly lower 6MWD for older women, which was considered when evaluating change in 6MWD.

Our data indicate that males treated with tadalafil may have a clinically significant functional improvement compared with females, supporting the findings of the recently published tadalafil sex-response study.¹² The authors of that study suggested that this was due to male patients having lower levels of NO and, therefore, a more robust response to the enhanced NO signaling resulting from PDE-5 inhibition.^12,26 Testosterone may also play an important role given its pulmonary vasodilator action via inhibition of the L-type voltage-gated calcium channel.²⁷ It seems likely that the observed difference in functional improvement between the sexes in this study is due to the multifactorial effects of testosterone and estrogen on the pulmonary vasculature.

The difference in 6MWD between the sexes for those treated with tadalafil was statistically significant by the nonparametric Wilcoxon rank sum test. After controlling for multiple variables, including age, sex, race, etiology of PAH, WHO functional class, baseline 6MWD, and background therapy with bosentan, the observed differences in 6MWD between the sexes was no longer statistically significant. This may be due to correlation of variables in the multivariate regression model. The correlated variables compete for explaining the change in 6MWD, which may result in variables losing significance compared with the univariate analysis.

Prior studies have demonstrated significant effects of cotreatment with bosentan and PDE-5 inhibitors on change in 6MWD.^3,28,29 To examine these correlations, a backward model selection was performed, which demonstrated that changes in 6MWD are best explained by age, etiology of PAH, WHO functional class, and baseline 6MWD. It can be noted that sex was marginally significant in this patient cohort but that background therapy with bosentan and race were not.

Our study demonstrated a trend toward more progressive worsening in postmenopausal females on the basis of change from baseline 6MWD. Interestingly, this trend was not seen for males in the same age groups, suggesting that this progressive female worsening is not necessarily due to age. Postmenopausal females may likely have more comorbidities than younger women, which may in part account for this trend. This trend may also be related to the low circulating levels of cardioprotective estrogen present after menopause and the estrogen-dependent mechanism of PDE-5 inhibitors.^20,30 Ovariectomized mice that were treated with sildenafil did not experience the antiremodeling properties that their ovary-retaining counterparts did and, interestingly, after exogenous estrogen replacement regained benefit from sildenafil.²⁰ On the basis of these animal data and the findings of our study, it appears that the treatment mechanism of PDE-5 inhibitors in women may require estrogen. This raises the question of whether PDE-5 inhibitors should be the optimal therapy for postmenopausal women.

Limitations of this study include inconsistencies in tadalafil dosing and bosentan background therapy. In addition, information on hormone replacement therapy and hormone-based contraceptive use in this patient cohort was not available. The assumption of menopausal status was based on age, and no actual information on patients’ menopausal status was available at the time of this study. The use of 6MWD as a primary end point and the short duration of the PHIRST study also present potential limitations. 6MWD should be viewed as an intermediate end point; further research into the impact of sex and menopausal status on the effects of PDE-5 inhibitors on mortality and morbidity is required.^31,32

In conclusion, while PAH is a disorder with a known female and postmenopausal predominance, it appears that males and premenopausal females may experience greater functional improvement when treated with tadalafil than older women. No consistent effect of sex or menopause on TCW was observed. All subjects regardless of sex or age demonstrated improvement with tadalafil compared with placebo. These additional findings with regard to sex and estimated menopausal status provide clinicians with further information that can be used to individualize treatment for patients.

Source of Support: HCC receives research support from National Heart, Lung, and Blood Institute grants P01 HL-103455, U01 HL-108642, R34 HL-117344, and RO1 HL-11475; the Vascular Medicine Institute; the Institute for Transfusion Medicine; and the Hemophilia Center of Western Pennsylvania.

Conflict of Interest: YR and JC are employed by United Therapeutics. All other authors: none declared.