Background
Pulmonary arterial hypertension (PAH) is a chronic progressive disorder of the pulmonary vasculature characterised by abnormal remodelling of small, peripheral resistance vessels in the lung. The occlusion of these pulmonary arterioles leads to a persistently elevated pulmonary vascular resistance (PVR), a raised pulmonary arterial pressure (PAP), and ultimately death due to right ventricular hypertrophy, decreased cardiac output and heart failure. Since the recent 2018 World Symposium on Pulmonary Hypertension, PAH is defined haemodynamically by the presence of mean PAP (mPAP) > 20 mmHg alongside an elevated PVR ≥ 3 Wood Units [
1]. The underlying pathological hallmarks of PAH include pulmonary arterial endothelial cell (EC) dysfunction, excessive vasoconstriction, pulmonary artery EC and smooth muscle cell (SMC) proliferation, inflammation & fibrosis, in situ thrombosis and right ventricular (RV) hypertrophy.
Currently, PAH is primarily treated with drugs from four major drug classes targeting the endothelin, prostacyclin and nitric oxide (NO) signalling pathways. However, these approved therapies mainly treat the excessive pulmonary vasoconstriction/reduced vasodilation but fail to alter the cardiac and pulmonary remodelling associated with the later phases of the disease. Therefore, there is an urgent unmet medical need for new drugs to new therapeutic targets that offer greater tolerability/compliance and overall efficacy.
An exciting alternative approach to achieving improved vasodilation is to reduce the excessive vasoconstriction by antagonizing/inhibiting the action of the prostanoid thromboxane (TX) A
2. TXA
2, synthesised from arachidonic acid (AA) by the sequential actions of cyclooxygenase (COX)-1/COX-2 and TXA
2 synthase (TXA
2S), mediates its actions in humans by binding to the TPα and TPβ isoforms of the T Prostanoid receptor (the TP). TXA
2 is a potent mediator of platelet aggregation and induces constriction of various types of smooth muscle (SM) including vascular, renal and pulmonary SM. In addition, TXA
2 acts as a potent
pro-inflammatory and
pro-mitogenic agent promoting inflammation, fibrosis, blood vessel remodelling and/or restenosis following endothelial injury and is the main COX-derived SM constrictory prostanoid produced within the lung [
2‐
5]. Hence, imbalances in the levels of TXA
2, or of TXA
2S or of the TP have been implicated in a range of cardiovascular, renal and pulmonary diseases, including in PAH [
2‐
5]. The TXA
2-TP axis also regulates key mitogenic/ERK and RhoA-mediated signalling cascades, explaining, at least in part, the role of TXA
2 in increasing cell proliferation and migration such as occurs in restenosis, vascular remodelling, and in a range of cancers in which the TXA
2-TP axis is increasingly implicated [
2‐
8].
Moreover, and also critically, the TP not only mediates the actions of TXA
2 but also those of the isoprostane 8-iso-prostaglandin (PG) F
2α (also termed 15-F2t-isoprostane), a non-enzymatic-, free-radical- derived product of arachidonic acid produced in abundance during oxidative injury/hypoxia, including in various cardiovascular and pulmonary diseases [
9]. Significantly in the context of PAH, 8-iso-PGF
2α is increased over 3-fold in PAH patients and correlates with PAH status and disease progression [
2,
9,
10]. A recent study has also demonstrated the pivotal role of 8-iso-PGF
2α signalling through the TP in the pathogenesis of pulmonary fibrosis [
11], a key hallmark of PAH disease progression. Thus, TP antagonists will have benefits over existing PAH therapies in that they will not only inhibit the adverse actions of TXA
2 itself but also those of 8-iso-PGF
2α which is present in high amounts in the hypoxic environ of the PAH lung [
10].
As further evidence of its role in PAH, TXA
2 or its stable analogues are also used widely to induce pulmonary hypertension (PH) in large preclinical animal models such as the pig [
12]. Infusion of the TXA
2 mimetic U46619 results in stable PH, with significant increases in mPAP, PVR and in cardiac output (CO) [
12‐
14]. Furthermore, within a hypoxia-induced PH model in pigs, disruption of TXA
2 production and TP signalling attenuates the negative effects on pulmonary and cardiac parameters [
15]. Following treatment with the TP antagonist Daltroban, the significant hypoxia-induced rises in mPAP, PVR and mean right atrial pressure were significantly reduced [
15].
NTP42 is a novel antagonist of the TP and is currently in development for the treatment of PAH. During its development, over 250 small chemical compounds were characterised in calcium mobilisation assays in human embryonic kidney (HEK) 293 cells over-expressing TPα and TPβ following stimulation with the TXA
2 mimetic U46619 or the isoprostane 8-iso-PGF
2α [
16,
17]. Following this primary screen, prioritised leads were then subject to secondary screening by examining their ability to inhibit TP (U46619)- mediated aggregation of human platelets ex vivo [
16,
17]. Key leads in this series, including the drug candidate
NTP42, were confirmed to display potent antagonist activity, excellent specificity, pharmacokinetic, pharmacodynamic, toxicology and efficacy profiles in a range of in vitro, ex vivo and in vivo pre-clinical models following oral delivery [
16,
17]. Mechanistically, TP antagonists such as
NTP42 may be promising therapeutic drugs for PAH, not only inhibiting the excessive vasoconstriction but also preventing the micro-vessel thrombosis and, potentially, limiting the pulmonary artery remodelling, as well as the inflammation and fibrosis found in PAH. In addition, as also stated, TP antagonists will inhibit signalling by 8-iso-PGF
2α, the free-radical derived isoprostane generated in abundance in the clinical setting of PAH [
2,
9‐
11]. Thus, the aim of this study was to investigate the efficacy of
NTP42 in the monocrotaline (MCT)-induced PAH rat model, alongside current standard-of-care compounds.
Discussion
Treatment of PAH has primarily focused on correcting imbalances between the multiple vasodilator and vasoconstrictor pathways in the pulmonary circulation that results in haemodynamic abnormalities including elevated pulmonary arterial pressure (PAP) and increasing pulmonary vascular resistance (PVR). Unfortunately, currently approved therapies are often short-acting, exhibit limited efficacy at modifying or preventing disease progression, and many have side effects. While the goals of current treatments are to achieve increased exercise capacity, improved quality of life, and to slow disease progression and lower mortality risk, none of the currently-approved therapies have been shown to slow pathological progression [
25]. PAH remains an incurable condition with a high mortality rate, underscoring the need for new drugs to new therapeutic targets that will offer greater overall efficacy and tolerability/compliance [
26].
TXA
2, as a potent vasoconstrictor, inducer of platelet aggregation and smooth muscle constrictor and mitogen, has drawn significant attention as a potential therapeutic target for PAH. Several lines of evidence implicate the TXA
2-TP axis as a contributor to PAH, and indeed in other classes of pulmonary hypertension. The importance of TXA
2 as a driver of disease pathology is indicated by findings in paediatric patients with PAH caused by congenital heart disease defects. These patients typically have elevated urinary and plasma levels of thromboxane (TX) B
2, the stable end-product metabolite of TXA
2 [
27]. In infants with persistent pulmonary hypertension of the newborn (PPHN), elevated levels of TXB
2 positively correlated with a significantly increased pulmonary artery pressures [
28]. Furthermore, in adult PAH patients, 24-h excretion levels of TXB
2 are increased, as compared with normal controls, whereas, in contrast, 24-h excretion of 2,3-dinor-6-keto-prostaglandin F
1α (a stable metabolite of prostacyclin) are significantly decreased [
29]. Results from a recent randomised clinical trial of Aspirin and Simvastatin (ASA-STAT) showed that PAH patients with higher levels of TXA
2 were associated with more advanced disease, and with worse clinical outcome/survival [
30]. With regards to the TP, employing radioligand binding studies, its expression was shown to be significantly elevated in the right ventricles (RVs) from PAH patients compared to non-diseased controls [
31]. Additionally, while expression in normal RV was found to be primarily perinuclear, the TP was strongly expressed throughout the cell surface in cardiomyocytes of patients with PAH [
32].
Expanding on the growing evidence for the role for the TP in PAH, this current study demonstrated abundant expression of both TPα and TPβ isoforms of the TP in the human lung, both in normal control and PAH disease tissues. Thereafter, this study demonstrated, for the first time, that TP antagonism by treatment with the novel TP antagonist NTP42 prevented the development of PAH and ameliorated its progression in the MCT-preclinical model of PAH in rodents. NTP42 reduced MCT-induced PAH as determined from the haemodynamic measurements, and at least to a similar extent as the standard-of-care drug Sildenafil or Selexipag. Moreover, detailed morphometric analysis of pulmonary vascular remodelling and histological analysis of inflammation and fibrosis indicate that the TP antagonist NTP42 was superior to Sildenafil and Selexipag in reducing the vascular remodelling, mast cell recruitment and pulmonary collagen deposition in the MCT-treated animals. From a histological assessment, lung tissue from NTP42-treated animals and ‘No MCT’ control animals displayed similar histochemistry across the various parameters analysed (i.e., alveolar wall thickening, gas exchange distances, inflammatory infiltrates, vessel remodelling) which, in turn, was radically different and more pronouncedly diseased in ‘MCT Only’-, Sildenafil- and Selexipag-treated animals. These findings illustrate a significant benefit of NTP42 compared with the approved drugs Sildenafil and Selexipag in lung pathology/histology outcomes. Notably, in the current study, NTP42 shows at least equivalent haemodynamic outcomes or significantly greater lung histology/pathology benefit than Sildenafil even when NTP42 (0.25 mg/kg BID, PO) was used at a 200-fold lower dosage than Sildenafil (50 mg/kg BID, PO).
While similar haemodynamic benefits were observed for
NTP42 in comparison to both Sildenafil and Selexipag in alleviating the MCT-induced increase in mPAP, and where
NTP42 and Sildenafil, but not Selexipag, treatment resulted in decreases in RVSP, it was highly notable that TP antagonism via
NTP42 was the only treatment that showed significant decreases in pulmonary mast cell infiltration and fibrosis. Whereas mast cells have traditionally been recognised as critical in allergic and nonallergic immune responses, a growing body of evidence implicates these cells in cardiovascular disease, including in PAH. Mast cell infiltration around small pulmonary vessels has been noted in models of pulmonary hypertension, and indeed within plexiform lesions of PAH patients [
33]. They also stimulate the proliferation of endothelial and smooth muscle cells [
34,
35]. Through their release of histamine, heparin and chymase as well as multiple other molecules, mast cells contribute strongly to pro-fibrotic activities, either directly through effects on fibroblasts and fibrocytes, or indirectly through the recruitment and activation of various inflammatory and structural cell types [
36]. Additionally, and of particular relevance to this study, activated mast cells produce significant quantities of TXA
2, alongside other prostanoids [
37]. Taken together, these previous findings consolidate the hypothesis that mast cells may contribute to the pathogenesis of PAH. The effect of
NTP42 in reducing mast cell recruitment within the lung, and the potential resulting pulmonary fibrosis, is of particular note and demonstrates a unique benefit of TP antagonism in this study, at least. While deemed beyond the scope of the current study, it will be of interest to explore how
NTP42 may impact on other pathways within the innate and/or adaptive immune systems. Furthermore, it must be noted that while toluidine blue is a routinely-used stain for mast cells within the tissues of animals from MCT-PAH investigations [
38‐
40], additional specific markers such as c-Kit/CD117 may be employed in future studies as a further validation of the effect of TP antagonism on mast cell recruitment.
Previous investigations have attempted to clinically assess the effects of disrupting the TXA
2-TP axis in the treatment of PAH. Inhibition of TXA
2 production using the TXA
2 synthase (TXA
2S) inhibitor CGS13080 resulted in a modest improvement in pulmonary haemodynamics in a small study of patients with PAH [
41]. Showing promise in preclinical investigations, the dual TXA
2S inhibitor/TP antagonist terbogrel prevented pulmonary hypertension and the development of pulmonary artery dysfunction in a chronic hypoxia-induced porcine pulmonary hypertension model [
42]. However, a multicentre, randomised, placebo-controlled Phase II trial of terbogrel for use in adults with PAH had to be terminated prematurely because of severe leg pain, which occurred almost exclusively in terbogrel-treated patients [
43]. As a TXA
2S inhibitor, while terbogrel was pharmacologically effective in reducing TXA
2 metabolites, it also led to a rise in levels of prostacyclin metabolites [
43]. In PAH patients, prostacyclin-associated leg pain is a recognised debilitating adverse effect in PAH patients on prostacyclin therapy [
44]. In follow-up pharmacokinetic and pharmacodynamic assessments, treatment with terbogrel across a range of doses resulted in up to 10-fold increases in prostacyclin metabolites [
45]. Thus, it is now widely accepted that inhibiting TXA
2 synthesis, either through the administration of selective TXA
2S inhibitor or dual TXA
2S inhibitor/TP antagonist compounds, shifts the enzymatic conversion of the common precursor endoperoxide substrates PGG
2/PGH
2 away from TXA
2 biosynthesis towards generation of the pain/nociceptive-inducing prostacyclin [
43].
Bearing in mind these side-effects associated with increases in prostacyclin levels as a result of TXA
2S inhibition, it is crucial that any proposed inhibitor of the TXA
2-TP signalling axis have a positive drug profile and yet retain the ability to ameliorate the development of PAH.
NTP42 is a highly potent and selective TP antagonist with efficacy in the low nanomolar range (Supplemental Figures
1 &
2). Specifically,
NTP42 antagonised intracellular calcium mobilization following stimulation of mammalian HEK 293 cells stably over-expressing the TPβ isoform of the human TP (HEK.TPβ line) with the TXA
2 mimetic U46619 (IC
50 8.86 ± 3.07 nM; Supplemental Figure
1A). Furthermore, and as previously discussed, as a highly specific antagonist of the TP,
NTP42 will directly inhibit the actions of the isoprostane 8-iso-PGF
2α, which can exert pulmonary hypertensive effects in several distinct ways [
46]. Herein,
NTP42 antagonised intracellular calcium mobilization in the HEK.TPβ line following stimulation with 8-iso-PGF
2α (IC
50 8.04 ± 3.74 nM; Supplemental Figure
1B).
NTP42 was also shown to antagonize U46619-induced aggregation of human platelets, where
NTP42 was confirmed to have an IC
50 of 10.6 ± 1.7 nM (
Supplemental Figure 2). In the context of TP specificity,
NTP42 was confirmed to be an entirely selective TP antagonist, with no agonist actions at the TP (
Supplemental Figure 3). Specifically,
NTP42 was assessed for activity at the 7 other prostanoid receptors, namely the prostaglandin (PG) D
2 (DP
1), PGE
2 (EP
1, EP
2, EP
3, EP
4), PGF
2α (FP) and PGI
2/prostacyclin (IP) prostanoid receptors, and was confirmed to exhibit no agonist or antagonist effects at these receptors (
Supplemental Figure 3 &
Supplemental Table 2). Furthermore, unlike the dual TXA
2S inhibitor/TP antagonist terbogrel,
NTP42 does not inhibit TXA
2S (
Supplemental Figure 4). Thus, in contrast to terbogrel which, as stated, has been clinically evaluated for efficacy in PAH [
43],
NTP42 is a selective TP antagonist and does not affect levels of the nociceptive agent prostacyclin.
Limitations of study
Based on the choice of the PAH preclinical model, several limitations with the current study must be acknowledged. In the case of the MCT-PAH model, animals respond to a single administration of the highly toxic pyrrolizidine alkaloid MCT and the disease develops quickly where even within 6 h post-MCT administration, changes associated with disease induction can be observed. Thereafter, over a period of 1–2 weeks, PAH develops and without any drug intervention, a high proportion of animals typically succumb to the disease through death or animals must be sacrificed due to morbidity concerns [
47]. Thus, owing to the relatively rapid development and progression of the disease, the MCT-induced PAH model is frequently best employed as a prophylactic or early intervention model where the test drug is administered either simultaneously with the MCT or within a short duration post-MCT induction, as was the case in this study. The therapeutic effects of
NTP42 should therefore be further investigated using conditions viewed as more reminiscent of a treatment model, rather than a prophylactic model. This could be achieved within the MCT-induced PAH model by delaying treatment with
NTP42 to strike a balance between undue loss of animals due to high mortality rates versus definitively demonstrating treatment of the disease. Alternatively, and in line with recent industry recommendations [
48], the therapeutic effects of
NTP42 should be investigated using a second distinct preclinical model, such as the Sugen 5416/hypoxia (SuHx)-induced PAH model. While both the MCT model and the SuHx model are widely reported to emulate PAH, they each display features both typical and atypical of the human disease and of each other [
47,
49‐
51]. For example, a key distinguishing feature between the SuHx and MCT-induced PAH models is the formation of occlusive neointimal or plexiform lesions, which are characteristic of the advanced lesions seen in PAH patients associated with the former, and widely associated with advanced cases of the human condition, but which is absent in the MCT-PAH model [
47,
49‐
51]. On the other hand, inflammation appears to be a key component of the disease that develops in the MCT-model, a feature that is less of a burden in the human PAH lung and that is virtually absent in the SuHx model [
47,
49‐
51].
Notwithstanding the significance of the findings of this study, it should also be acknowledged that the level of the MCT-induced disease, as demonstrated by the low but consistent increase in mPAP, RVSP and Fulton’s Index, was not as severe when compared with previous studies [
52,
53]. It is acknowledged that the level of PAH disease induced by MCT administration can vary significantly across species, strains and individual animals [
51,
54]. The more moderate PH induced in this model may account for the lack of significant benefit observed for
NTP42 in the Fulton’s Index parameter, and for the lack of benefit observed for Selexipag in the RVSP and Fulton’s Index parameter. Notably, the Selexipag animal group size was lower than the other treatment groups investigated, and this may also have had an effect on the statistical findings. Future preclinical investigations should be performed under more severe disease conditions, to further highlight the potential benefits of
NTP42 or indeed, of the comparator standard-of-care treatments.
A further acknowledged limitation of this study was that while
NTP42 treatment led to reductions in RVSP relative to the ‘MCT Only’ control, no significant changes were observed in RV hypertrophy, as assessed through measurement of the Fulton’s Index. In an additional follow-up study primarily focussing on the effect of
NTP42 on cardiac hypertrophy,
NTP42 treatment at 0.125 mg/kg QD significantly decreased the MCT-induced rise in the Fulton’s Index (
Supplemental Figure 5A). In additional assessments of cardiac hypertrophy within this experimental cohort, RV tissues were stained with
anti-CD31 antibody (
Supplemental Figure 5B) which enabled autologous quantitation of both CD31 positive vascular endothelial cells, a measure of vascularization/angiogenesis, and also of cardiomyocyte cross-sectional area, a measure of cardiac hypertrophy. Thus, measurement of these two parameters (i) cardiomyocyte cross-sectional area and (ii) vascularisation per unit area provides a direct assessment of ‘metabolic demand’ and ‘metabolic supply’, respectively, within the RV tissue and when expressed as a ratio, can be used as a measure, so-called Metabolic Index, of ‘Maladaptive hypertrophy’ (where ratio of (i)/(ii) is > 1) or ‘Adaptive hypertrophy’ (where ratio of (i)/(ii) is < 1). In accordance with the Fulton’s Index (
Supplemental Figure 5A) findings, treatment with
NTP42 at 0.125 mg/kg QD led to a significant decrease in cardiomyocyte cross-sectional area (
Supplemental Figure 5C). While treatment with
NTP42 at this dose had no significant effect on RV vascularization per se (
Supplemental Figure 5D), considering both RV hypertrophy and vascularisation together within a “Metabolic Index” parameter, treatment with
NTP42 led to a significant increase in this indicative parameter of cardiac adaptation (
Supplemental Figure 5E). Finally, the progress observed in the clinical management of PAH has been largely attributed to the development of combination therapy strategies, and on the staggered escalation of those dual- or even triple-combination therapies based on close monitoring and assessment of clinical outcomes in line with established treatment algorithms [
55‐
57]. Consistent with these clinical guidelines, it will be important to assess the efficacy of
NTP42 alongside other therapies such as PDE5 inhibitors (e.g. Sildenafil), prostacyclin analogues (e.g. Selexipag) or members of the ERA class of PAH treatments.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.