Use of multitarget tyrosine kinase inhibitors to attenuate platelet-derived growth factor signalling in lung disease

Abstract

Platelet-derived growth factors (PDGFs) and their receptors (PDGFRs) play a fundamental role in the embryonic development of the lung. Aberrant PDGF signalling has been documented convincingly in a large variety of pulmonary diseases, including idiopathic pulmonary arterial hypertension, lung cancer and lung fibrosis. Targeting PDGF signalling has been proven to be effective in these diseases. In clinical practice, the most effective way to block PDGF signalling is to inhibit the activity of the intracellular PDGFR kinases. Although the mechanism of action of such drugs is not specific for PDGF signalling, the medications have a broad therapeutic index that allows clinical use. The safety profile and therapeutic opportunities of these and future medications that target PDGFs and PDGFRs are reviewed.

Short abstract

An increasing role for PDGF signalling inhibitors in clinical trials for the treatment of various pulmonary diseases http://ow.ly/buaI30f9HcN

Introduction

Platelet-derived growth factors (PDGFs) and their receptors (PDGFRs) represent one of the most intensively studied families of signalling factors over the past four decades. PDGFs constitute a family of four gene products (PDGF-A–D) acting by means of two receptor tyrosine kinases, PDGFR-α and -β. The PDGF signalling pathway is found in most cell types.

In vivo studies have documented an important role of PDGF signalling in the normal development of several organs, such as the kidney, eye and lung. PDGF signalling plays an essential role in cell proliferation, differentiation, migration and survival [1–3]. In adults, PDGF signalling is involved in formation of de novo connective tissue during wound healing. Aberrant expression and signalling of PDGF ligands and receptors is associated with several connective tissue disorders, and lung diseases such as pulmonary arterial hypertension (PAH), lung cancer and idiopathic pulmonary fibrosis (IPF) [4]. This review first focuses on the PDGF signalling pathways that involve specific ligands and their receptors. This is followed by a description of the role of this important class of molecules in lung disease.

PDGF signalling pathways

The PDGFRs are transmembrane proteins belonging to the receptor tyrosine kinase (RTK) class. PDGF signalling is initiated by the binding of distinct dimeric PDGF ligands to the extracellular domain of two monomeric receptors at the same time, thereby inducing dimerisation of PDGFR and autophosphorylation of the tyrosine residues within its intracellular domain. Five different ligand isoforms (PDGF-AA, PDGF-BB, PDGF-CC, PDGF-DD and PDGF-AB) and two PDGFR isotypes (PDGFR-α and PDGFR-β) with three different PDGFR dimers (PDGFR-α/α, PDGFR-α/β and PDGFR-β/β) have been described (figure 1). The ligands bind to these receptor pairs with different affinities. PDGF-AA and PDGF-CC bind with high affinity to PDGFR-α/α in vitro and in vivo. PDGF-BB shows binding affinity to all three PDGFR dimers in vitro and signals through PDGFR-β/β in vivo. PDGF-AB binds to PDGFR-α/α and PDGFR-α/β in vitro. PDGF-DD shows affinity for PDGFR-β/β in vitro. PDGF-CC and -DD activate the heterodimer PDGFR-α/β in vitro [5–7]. In vivo binding affinity toward the different PDGFRs is largely unknown for PDGF-DD and -AB.

FIGURE 1.

Platelet-derived growth factor receptors (PDGFRs) and ligand patterns. PDGFRs are transmembrane proteins. The extracellular domain consists of five immunoglobulin-like domains; binding occurs at domains 2 and 3. The intracellular domain is a tyrosine kinase. There are three dimeric forms of PDGFRs (-αα, -ββ and -αβ). Five different ligand isoforms are known to bind to PDGFRs (AA, AB, BB, CC and DD).

Phosphorylation of the tyrosine residues within the intracellular domain results in transduction of signals via recruitment of surrounding proteins containing Src homology region 2 domains (figure 2). The two main intracellular signalling pathways activated by PDGF signalling are the phosphatidylinositol 3′-kinase/Akt (also known as protein kinase B)/mammalian target of rapamycin (PI3K/Akt/mTOR) pathway and the mitogen-activated protein kinase (MAPK) cascade pathway [8–10]. mTOR suppresses the synthesis of PDGFR. Although the mechanism has not been fully elucidated, data suggest that suppression occurs by reduced PDGFR transcription. Aberrant activation of this pathway is known to appear in several types of human cancer [11–13].

FIGURE 2.

Platelet-derived growth factor (PDGF) signalling pathway. Binding of distinct dimeric PDGF ligands to the respective receptor dimer leads to autophosphorylation of the tyrosine residues within their intracellular domains. Autophosphorylation then leads to transduction of signals via recruitment of surrounding proteins containing Src homology region 2. These include domains such as Grb2, Grb7, SHC, PI3K and GTPase-activating protein for Ras. Signal transduction occurs via two main pathways: 1) the PI3K pathway, which mediates Akt signalling for the promotion of cell survival and 2) the mitogen-activated protein kinase (MAPK) cascade. Hydrolytic conversion of RAS-guanosine diphosphate (GDP) to RAS-guanosine triphosphate (GTP) leads to activation of MAPK cascade members, such as extracellular signal-regulated kinase (ERK) or MAPK kinase (MEK) via RAF1, resulting in gene target transcription. This pathway is important for cell growth, proliferation, differentiation and migration. Son of sevenless (SOS) is a nucleotide exchange factor for Ras.

The MAPK cascade signalling pathway is important in stimulating cell migration, differentiation and proliferation. Conversion of RAS-guanosine diphosphate to RAS-guanosine triphosphate leads to activation of the MAPK cascade members via RAF 1. These proteins such as extracellular signal-regulated kinase or MAPK kinase result in gene target transcription.

Differences in signalling result from PDGFR-α and PDGFR-β activation. In vivo, embryos of transgenic mice lacking PDGFR-α showed deficiencies in a large number of mesenchymal cells, such as smooth muscle cell progenitors, whereas embryos lacking PDGFR-β are deficient in vascular smooth muscle cells and pericytes [14]. At a transcriptome level, PDGFR-α/α, PDGFR-α/β and PDGFR-ββ activated a set of 33, 25 and 15 genes, respectively [15].

The role of PDGF signalling in lung development

PDGF signalling plays a crucial role in the embryonic development of the lung. In vivo studies have documented an essential role of mesenchymal cells expressing PDGFR-α mRNA in postnatal alveolar septation. PDGF-A-null mice died during embryogenesis or shortly after birth, while surviving PDGF-A-deficient mice develop lung emphysema secondary to failure of alveolar septation caused by defective alveolar myofibroblast differentiation [3]. In addition, overexpression of PDGF-A in the lung epithelium resulted in perinatal death caused by fetal lung enlargement, failure of airspace development and mesenchymal cell hyperplasia [16].

PDGF signalling in lung disease

Aberrant PDGF signalling has been documented convincingly in a large variety of pulmonary diseases. In addition, the role of PDGF signalling inhibitors as a therapeutic option has been investigated in multiple clinical trials. PDGF inhibitors currently in use include DNA aptamers or soluble extracellular parts of the receptors that bind PDGF isoforms and thus prevent their binding to signalling receptors [17, 18], neutralising antibodies [18] or decoy receptors that sequester PDGFs and thus prevent their binding to and activation of PDGFRs [19]. Alternatively, the activity of PDGFRs can be inhibited by neutralising antibodies [20–22]. These types of antagonists have the advantage of being reasonably specific; however, they are expensive. One of the most effective ways to block PDGF signalling is to inhibit the activity of the intracellular PDGFR kinases. Several potent inhibitors of PDGFR kinases have been tested, including imatinib, linifanib, nintedanib and sorafenib [20, 23]. Table 1 provides a summary of completed clinical trials. Table 2 provides a summary of targets for each drug and the corresponding median inhibitory concentration (IC₅₀). The mechanism of action of these drugs is not specific for PDGF signalling.

TABLE 1.

Summary of completed clinical trials

Lung disease/documented PDGF-PDGFR involved in pathogenesis	Drug	Completed clinical study, study phase, study design and number of patients	Patient population	Intervention	Primary outcome	Results
IPAH  PDGF-BB  PDGFR-β	Imatinib mesylate	IMPRES [24] Phase 3, 24 weeks, multicentre, randomised, double blind, placebo controlled (n=202)	PVR ≥800 dyn·s·cm−5 symptomatic on ≥2 PAH therapies	Randomised to receive imatinib 200 mg once daily or placebo	Change in 6MWD	Mean treatment effect on 6MWD 32 m; p=0.002 PVR decreased by 379 dyn·s·cm−5 p<0.001 Serious AEs were more frequent with imatinib compared to placebo (44% versus 30%)
	Sorafenib	Gomberg-Maitland et al. [25] Phase 1b, 16 weeks, single centre, open label (n=22)	Advanced but stable PAH on parenteral prostanoids (with or without oral sildenafil)	Received sorafenib started at 200 mg daily then escalated to 400 mg twice daily	Safety and tolerability	Sorafenib was well tolerated at 200 mg twice daily AEs: moderate skin reactions on the hands and feet and alopecia
NSCLC  PDGF-A/B  PDGFR-α/β	Imatinib mesylate (in combination with paclitaxel)	Bauman et al. [26] Phase 2, multicentre, single stage, open label (n=34)	Age ≥70 years with untreated, stage IIIB/IV NSCLC and ECOG performance status 0–2	Received up to six 28-day cycles of imatinib and paclitaxel	RR	Met primary end-point; however, no improvement in overall survival and PFS. Regimen not recommended for further study
	Linifanib (in combination with carboplatin and paclitaxel)	Ramalingam et al. [27] Phase 2, multinational, open label (n=139)	Stage IIIB/IV nonsquamous NSCLC	Received carboplatin and paclitaxel plus placebo, linifanib 7.5 mg or linifanib 12.5 mg	PFS	Improved PFS. Higher incidence of AEs known to be associated with VEGF/PDGF inhibition
	Nintedanib (BIBF 1120) with pemetrexed	Ellis et al. [28] Phase 1, multicentre, open label (n=26)	Metastatic, unresectable, or locally advanced NSCLC and had to have relapsed during or following one prior platinum-based chemotherapy regimen	Received BIBF 1120 starting dose of 100 mg twice daily (days 2–21) with pemetrexed 500 mg·m−2 (day 1) over a 21-day cycle BIBF 1120 dose was escalated until the MTD was determined	Safety, tolerability and MTD	The MTD of BIBF 1120 in combination with standard-dose pemetrexed was 200 mg twice daily BIBF 1120 in this combination was tolerable, with promising signs of efficacy
	Nintedanib (in combination with docetaxel)	LUME-Lung 1 [29] Phase 3, multicentre, double blind, randomised, placebo controlled (n=1314)	Stage IIIb/IV recurrent NSCLC progressing after first-line chemotherapy	Randomised to docetaxel 75 mg·m−2 by i.v. infusion on day 1 plus nintedanib 200 mg twice daily orally or placebo on days 2–21, every 3 weeks Treatment was continued until unacceptable AEs or disease progression	PFS	PFS was significantly improved after a median follow-up of 7.1 months Improved overall survival in all patients with adenocarcinoma (12.6 versus 10.3 months); more pronounced in the subgroup of patients who progressed within 9 months after start of first-line therapy (10.9 versus 7.9 months) AEs: diarrhoea, reversible increase in ALT and AST, gastrointestinal side-effects
	Nintedanib (in combination with premetrexed)	LUME-Lung 2 [30] Phase 3, multicentre, double blind, randomised, placebo controlled (n=713)	Stage IIIb/IV or recurrent NSCLC previously treated with chemotherapy. ECOG performance score of 0–1	Randomised to pemetrexed 500 mg·m−2 i.v. on day 1, combined with either oral nintedanib 200 mg twice daily or placebo, given on days 2–21, every 3 weeks	PFS	Stopped early. ITT analysis of PFS favoured the treatment arm (median 4.4 months versus 3.6 months; HR 0.83, 95% CI 0.70–0.99; p=0.0435)
IPF  PDGF-AA, -BB and -CC	Imatinib mesylate (Gleevec)	Daniels et al. [31] Phase 2, 96 weeks, multicentre, randomised, double blind, placebo controlled (n=119)	Mild to moderate IPF diagnosed within 3–36 months of screening with clinical worsening in the past year (10% decline in FVC % pred or worsening chest radiograph or worsening dyspnoea)	Randomised to receive imatinib mesylate 600 mg (six tablets) orally once daily or placebo	Combined measure of disease progression (defined as 10% decline from baseline FVC) or death	No effect on survival or lung function
	Nintedanib (BIBF 1120)	TOMORROW [32] Phase 2, 12 months, multicentre, randomised, double blind, placebo controlled (n=432)	Age ≥40 years with <5 years diagnosis of IPF, FVC ≥50% pred, DLCO 30–79% pred, PaO2 ≥55 mmHg at <1500 m altitude or ≥50 mmHg at >1500 m altitude	Randomisation to receive one of four doses of BIBF 1120 (50 mg once a day, 50 mg twice a day, 100 mg twice a day or 150 mg twice a day) or placebo	Annual rate of decline in FVC	Compared to placebo, BIBF 1120 at a dose of 150 mg twice daily was associated with: a trend toward a reduction in the decline in lung function (0.06 versus 0.19 L·year-1); fewer acute exacerbations (2.4 versus 15.7 per 100 patient-years, p=0.02); preserved quality of life (−0.66 versus 5.46, p=0.007); gastrointestinal AEs, which led to more discontinuation in the 150 mg BIBF 1120 twice a day group; increase in liver aminotransferase levels
		INPULSIS-1 and 2 [33] Phase 3, 52 weeks, multicentre, randomised double blind, placebo controlled (n=1061)	Age ≥40 years with <5 years diagnosis of IPF, FVC ≥50% pred, DLCO 30–79% pred,	Randomised to receive 150 mg nintedanib twice daily or placebo	Annual rate of decline in FVC	INPULSIS-1 No difference in time to first acute exacerbation (HR 1.15, 95% CI 0.54–2.42) Comparable proportion of patients with at least one investigator-reported acute exacerbation (6.1% versus 5.4%, respectively, in the nintedanib and placebo groups) INPULSIS-2 Significant increase in time to first acute exacerbation (HR 0.38, 95% CI 0.19–0.77) Lower proportion of patients with at least one investigator-reported acute exacerbation in the nintedanib group (3.6%) compared with placebo (9.6%) Most common AEs: diarrhoea followed by nausea
	Pirfenidone	ASCEND [34] Phase 3, 52 weeks, multicentre, double blind, placebo-controlled (n=555)	Age 40–80 years, centrally confirmed diagnosis of IPF	Randomised to receive pirfenidone 2403 mg per day or placebo	Change in FVC or death at week 52	47.9% relative reduction in the proportion of patients who had an absolute decline of ≥10% of FVC % pred or who died No significant difference in dyspnoea scores (p=0.16) No significant difference in rates of death from any cause (p=0.10) or from IPF (p=0.23)
		CAPACITY 004 [35] Phase 3, 72 weeks, multicentre, double blind, placebo randomised controlled trial (n=435)	Age 40–80 years, diagnosis of IPF in the previous 48 months, FVC ≥50% and ≤90% pred, DLCO ≥35% and ≤90% pred 6MWT≥150 m	Randomised to receive pirfenidone 2403 mg·day−1, or pirfenidone 1197 mg·day−1 or placebo	Change in percentage of FVC % pred from baseline to week 72	Significant reduction in decline of FVC (p=0.001). Mean FVC change at week 72 was –8·0±16.5% in the pirfenidone 2403 mg·day−1 group and –12.4±18.5% in the placebo group (difference 4.4%, 95% CI 0.7–9.1). Outcomes in the pirfenidone 1197 mg·day−1 group were intermediate to the pirfenidone 2403 mg·day−1 and placebo groups AEs: patients in the pirfenidone 2403 mg·day−1 group had higher incidences of nausea, dyspepsia, vomiting, anorexia, photosensitivity and dizziness
		CAPACITY 006 [35] Phase 3, 72 weeks, multicentre, double blind, randomised, placebo controlled (n=344)	Age 40–80 years, diagnosis of IPF in the previous 48 months FVC ≥50% pred and ≤90% pred, DLCO ≥35% pred and ≤90% pred 6MWT ≥150 m	Randomised to receive pirfenidone 2403 mg·day−1 or placebo	Change in FVC % pred from baseline to week 72	Significant reduction in decline of FVC. Mean change at week 72 was –9.0±19.6% in patients in the pirfenidone 2403 mg·day−1 group and –9.6±19.1% in patients in the placebo group AEs: same as CAPACITY 004
HPS	Pirfenidone	Gahl et al. [36] 44 months, single centre, double blind, randomised controlled (n=21)	Confirmed diagnosis of HPS and FVC 40–75% pred All 21 subjects were Puerto Rican; 20 were homozygous for the 16-bp duplication in exon 15 of HPS1; one was homozygous for a 3904-bp deletion in HPS3	Randomised to receive pirfenidone 800 mg three times daily or placebo	Average rate of decline of FVC, FEV1, TLC and DLCO, measured as % pred, in the two treatment groups	5% difference in the yearly rate of FVC decline (p=0.001) In patients with an initial FVC >50% pred, pirfenidone slowed lost pulmonary function (FVC, FEV1, TLC and DLCO) at a rate of 8% per year, compared to placebo
		O'Brien et al. [37] Single centre, double blind, randomised, placebo controlled (n=35)	HPS-1 or -4 confirmed by molecular analysis and FVC 51–85% pred, regardless of radiographic evidence of fibrosis on HRCT	Randomised to receive pirfenidone or placebo. A dosage escalation schedule was employed to reach 801 mg pirfenidone three times daily	Rate of decline in FVC	Stopped early (after 12 months) due to futility No significant safety concerns
Diffuse cutaneous systemic sclerosis	Nintedanib	Spiera et al. [38] Phase 2a, 12 months, single centre, single arm, open label (n=30)	Fulfilled the American College of Rheumatology classification criteria for systemic sclerosis and had the diffuse subtype. Stable MRSS of ≥16 points in the month between screening and baseline visits, and had disease duration of <10 years DLCO ≥30% pred	Imatinib 400 mg daily	Change in the MRSS after 12 months of treatment	Improved MRSS 6.6 points or 22.4% (p=0.001) FVC improved by 6.4% pred (p=0.008)

PDGF: platelet-derived growth factor; PDGFR: PDGF receptor; IPAH: idiopathic pulmonary arterial hypertension; NSCLC: nonsmall cell lung cancer; IPF: idiopathic pulmonary fibrosis; HPS: Hermansky–Pudlak syndrome; PVR: pulmonary vascular resistance; PAH: pulmonary arterial hypertension; 6MWD: 6-min walking distance; AEs: adverse events; ECOG: Eastern Cooperative Oncology Group; RR: response rate; PFS: progression-free survival; VEGF: vascular endothelial growth factor; MTD: maximum tolerated dose; ALT: alanine aminotransferase; AST: aspartate aminotransferase; ITT: intention to treat; FVC: forced vital capacity; % pred: % predicted; D_LCO: diffusing capacity of the lung for carbon monoxide; P_aO2: arterial oxygen tension; 6MWT: 6-min walk test; FEV₁: forced expiratory volume in 1 s; TLC: total lung capacity; HRCT: high-resolution computed tomography; MRSS: modified Rodnan skin score.

TABLE 2.

Summary of drug targets and median inhibitory concentration (IC₅₀)

	Drug target	IC50 nM
Imatinib [39, 40]	PDGFR V-Abl C-Kit	100 600 100
Linifanib [41]	PDGFR-β KDR CSF-1R Flt-1/3	66 4 3 3/4
Nintedanib [42, 43]	PDGFR-α/β VEGFR1/2/3 FGFR1/2/3	59/65 34/13/13 69/37/108
PK10453 [44]	PDGFR-α/β	10.1/35
Sorafenib [43, 45]	PDGFR-β Raf-1 B-Raf VEGFR-2	57 6 22 90
Sunitinib [46]	PDGFR-β VEGFR2 (Flk-1)	2 80
Pirfenidone	TGF-β PDGF-A and -B	NA NA

PDGFR: platelet-derived growth factor receptor; VEGFR: vascular endothelial growth factor receptor; FGFR: fibroblast growth factor receptor; TGF: transforming growth factor; NA: not available.

Idiopathic PAH

Idiopathic PAH is a disease that involves vascular remodelling characterised by enhanced proliferation of pulmonary artery smooth muscle cells (PASMCs) and suppressed apoptosis. Previous studies have shown that PDGF and its receptors, particularly PDGF-BB and PDGFR-β are upregulated in lung tissues and PASMCs isolated from patients and animals with PAH [47–49]. PDGF-BB induces the proliferation and migration of PASMCs and has been proposed to be a key mediator in the progression of PAH [47]. Animal models have suggested that activated PDGFR-β is a key contributor to pulmonary vascular remodelling and idiopathic PAH [50]. Patients with PAH exhibit enhanced expression and phosphorylation of the PDGF-β receptor in remodelled pulmonary arterioles, particularly at the binding sites for phosphatidyl-inositol-3-kinase (PI3K) and phospholipase C (PLC)γ at tyrosine residues 751 and 1021, respectively. Selective disruption of PDGF-dependent PI3K and PLCγ activity is sufficient to abolish these pathogenic responses in vivo, identifying these signalling events as valuable targets for antiremodelling strategies in PAH [50].

Currently available PAH specific therapies, such as phosphodiesterase type 5 inhibitors, endothelin receptor antagonists or prostacyclin, inhibit vasoconstriction and improve endothelial dysfunction. However, their impact on vascular remodelling is not strong. Such compounds are not able to significantly inhibit the hyperproliferation of PASMCs or the fibrogenesis in the vascular wall [51].

The inhibition of PDGF signalling for the treatment of PAH has been investigated in several studies after the publication of a promising case report [52]. Imatinib is a selective inhibitor of the tyrosine kinase receptors, including PDGFR-β and c-Kit and their relevant ligands PDGF-BB and stem cell factor [53]. IMPRES (Imatinib in Pulmonary Arterial Hypertension, a Randomised, Efficacy Study) investigated the safety and efficacy of the tyrosine kinase inhibitor, imatinib mesylate, in advanced PAH over 24 weeks. Improvements of exercise capacity and haemodynamics were observed, but drug discontinuations and adverse effects including nausea, oedema, diarrhoea, and subdural haematoma in this anticoagulated population were common [24]. Whether this beneficial effect of imatinib on PAH is mediated through PDGF signalling is unclear. This therapy is unlikely to be advanced, because of the associated side-effects in this population.

Sorafenib is a multikinase/angiogenesis inhibitor that inhibits Raf1 kinase, a regulator of endothelial apoptosis, and inhibits the angiogenesis growth factor receptors vascular endothelial growth factor receptor (VEGFR)-2, VEGFR-3, PDGFR-β and c-Kit [54]. Gomberg-Maitland et al. [25] investigated the safety and tolerability of sorafenib in patients with advanced yet stable PAH on parenteral prostanoids (with or without oral sildenafil). The results at 200 mg twice daily demonstrated that the treatment was safe, with the most common adverse reactions being skin reactions occurring on the hands and feet, and alopecia. This small trial did not meet efficacy goals.

Medarametla et al. [44] tested the hypothesis that a novel, nonselective inhaled PDGF receptor inhibitor, PK10453, would decrease pulmonary hypertension both in the rat monocrotaline (MCT) model and the rat MCT plus pneumonectomy (MCT+PN) model of PAH. Results showed decreased progression of PAH in the rat MCT and MCT+PN models. The authors concluded that nonselective inhibition of both PDGFR-α and PDGFR-β may have a therapeutic advantage over inhaled imatinib in PAH. In this study imatinib was 8.5-fold more selective for PDGFR-α than for PDGFR-β (IC₅₀ 71 nM versus 607 nM, respectively). Most PAH-related cell-based studies interrogating the PDGFR pathway have used high doses of imatinib (5–10 μM), which would not allow a differentiation between inhibition of the PDGFR-α and PDGFR-β isoforms.

Overall, the PDGF signalling pathway seems to be a promising therapeutic target in PAH. Future studies will be required to balance efficacy and safety in this anticoagulated population.

Lung cancer

PDGF signalling is important for tumour growth, lymphangiogenesis and angiogenesis in vivo [53–55]. Aberrant PDGF signalling has been detected in several tumours, such as chronic myelomonocytic leukaemia; glial brain tumours; prostate, breast and lung adenocarcinoma; hepatocellular carcinoma; and nonsmall cell lung cancer (NSCLC) [20, 55–58].

Tumour expression of PDGFR-α/β and PDGF-A/B in NSCLC is associated with poor prognosis [59–61]. Inhibition of PDGFR-α and-β by therapeutic antibodies in NSCLC reduces tumour mass, highlighting the importance of the PDGF/PDGFR axis for tumour growth [62]. However, PDGFR expression patterns vary significantly in different NSCLC subtypes [61]. Preclinical models have demonstrated that inhibition of PDGF signalling with imatinib improved the drug delivery and efficacy of other chemotherapeutic agents in NSCLC, emphasising the potential role of imatinib as an adjunct to small-molecule or liposomal chemotherapy [63]. However, clinical trials with imatinib, showed little or no efficacy, and no improvement in progression-free survival (PFS) [26, 59, 64].

Linifanib (ABT-869) is an orally active, selective receptor tyrosine kinase inhibitor with IC₅₀ values in the low nanomolar range for VEGF (FLT1, KDR and FLT4) and PDGF (PDGFR-α and-β, CSF-1R, c-KIT and FLT3) receptors. A phase II study evaluated the safety and efficacy of linifanib in patients with stage IIIB/IV nonsquamous NSCLC. The study arms included carboplatin and paclitaxel with either linifanib 7.4 mg (arm A) or 12.5 mg (arm B) or placebo. Addition of linifanib to chemotherapy significantly improved PFS (arm A 5.4 months versus arm B 8.3 months; hazard ratio (HR) 0.51, p=0.02). However, there was increased toxicity for both doses reflective of known VEGF/PDGF inhibitory effects [27].

Inhibition of mTOR, a downstream molecule of the PI3K/Akt/mTOR pathway, with rapamycin, had been shown to induce growth arrest in the G1 phase of the cell cycle, and in some cases induced apoptosis in several tumour cell lines [65–68]. Jiang et al. [69] revealed the synergistic effect between mTOR complex 1/2 and glycolysis inhibitors, suggesting that the combined application of mTORC1/2 and glycolysis inhibitors may be a new promising approach to treat NSCLC. The safety and efficacy of newly developed inhibitors of mTOR, such as CCI-779, RAD001 and AP23573 in NSCLC are under clinical investigation.

Nintedanib, formerly called BIBF 1120, is a potent triple angiokinase inhibitor that targets VEGFR1/2/3, basic fibroblast growth factor receptor-1/2/3 and PDGFR-α/β signalling. Nintedanib has an interesting safety profile, as it does not lead to the typical side-effects of antiangiogenic drugs such as hypertension or hand–foot syndrome, probably because of its innovative triple-blocking mechanism of action [70]. In vitro and in vivo studies showed a marked inhibition of tumour growth in xenograft models of human NSCLC with nintedanib [28, 42]. The efficacy and safety of nintedanib in NSCLC has been evaluated in two phase III double blind, randomised, placebo-controlled clinical trials: LUME-Lung 1 and 2.

LUME-Lung 1 compared docetaxel plus placebo against docetaxel plus nintedanib. 1314 patients with stage IIIb/IV recurrent NSCLC progressing after first-line chemotherapy were included. After 7.1 months, the first preplanned primary analysis showed a higher PFS in the nintedanib arm (3.4 versus 2.7 months; HR 0.79, p=0.002). A subsequent analysis was performed at 31.7 months in order to evaluate the overall survival. These results were significant in all patients with adenocarcinoma (12.6 versus 10.3 months; HR 0.83, p=0.04). An even more significant result was observed in a subgroup of patients with adenocarcinoma who had progressed within 9 months after start of first-line therapy, achieving an improvement in overall survival of 4 months (10.9 versus 7.9 months; HR 0.75, p=0.007). The more common side-effects reported for the nintedanib plus docetaxel arm included diarrhoea (6.6% versus 2.6%), reversible increase in alanine aminotransferase (7.8% versus 0.9%), reversible increases in aspartate aminotransferase (3.4% versus 0.5%) and gastrointestinal side-effects, mostly attributed to nintedanib [29].

LUME-Lung 2 compared pemetrexed plus placebo against pemetrexed plus nintedanib in patients with stage IIIb/IV or recurrent NSCLC previously treated with chemotherapy. After randomising 713 of the 1300 planned patients, the trial was stopped due to the results of the preplanned data monitoring committee futility analysis, which suggested that the primary end-point would not be met for PFS. Ongoing patients were unblended, and follow-up continued per protocol. Subsequent intention-to-treat analysis of the primary end-point (PFS) favoured the treatment arm (median 4.4 versus 3.6 months; HR 0.83, 95% CI 0.7–0.99; p=0.04) [30].

In November 2014, the combination of nintedanib and docetaxel obtained European Medicines Agency approval as a second-line option for NSCLC patients with adenocarcinoma histology.

Lung fibrosis

Fibrotic diseases are characterised by active tissue remodelling involving proliferation of mesenchymal cell types such as myofibroblasts and accumulation of extracellular matrix components such as collagen. These lead to progressive scaring and loss of organ function. Upregulation of PDGF signalling has been linked with fibrotic diseases affecting the lung, kidney, liver, skin and heart [23, 71, 72]. Several environmental factors, such as asbestos and air pollutants, stimulate the expression of PDGFR-α in small-animal models [73–75]. An important role of PDGF signalling in the development of pulmonary fibrosis is suggested by studies of mouse models of silica, radiation and bleomycin-induced lung fibrosis [76–79]. PDGF-C is involved in the progression of pulmonary and cardiac fibrosis [79, 80].

IPF

IPF is a progressive fibrosing interstitial pneumonia of unknown cause, primarily occurring in older adults, and associated with the histopathological and/or radiological pattern of usual interstitial pneumonia. In IPF, transforming growth factor-β signal transduction promotes the expression of PDGF-B by regulatory T-cells (Tregs), which stimulate PDGF-B-mediated fibroblast proliferation [77]. Imatinib mesylate abolished fibroblast proliferation induced by Tregs in a mouse model of lung fibrosis induced by silica [77]. However, a multicentre randomised controlled trial showed that imatinib mesylate did not affect lung function or survival in patients with mild-to-moderate IPF [31]. Whether resistance to imatinib mesylate mediated by α1-acid glycoprotein, which is upregulated in patients with IPF, was responsible for this outcome is not known [81].

The most effective dose of nintedanib (BIBF 1120) in IPF was determined in a clinical phase 2 study, TOMORROW (To Improve Pulmonary Fibrosis with BIBF 1120). 432 patients underwent randomisation to receive one of four doses of BIBF 1120 (50 mg once a day, 50 mg twice a day, 100 mg twice a day or 150 mg twice a day) or placebo. The group taking the highest dose of BIBF 1120, 150 mg twice daily, showed a trend towards a slower decline in lung function and fewer exacerbations compared with placebo [32]. Based on these results, two double-blind, phase-3 clinical trials (INPULSIS-1 and INPULSIS-2) were conducted [33].

INPULSIS-1 and -2, which involved a total of 1061 patients with IPF randomly assigned to nintedanib 150 mg or placebo twice daily, met the primary end-point of reduction in the annual rate of decline in forced vital capacity (FVC) over 52 weeks. In these trials, nintedanib reduced the annual rate of FVC decline compared to placebo by 48% in INPULSIS-1 (−115 versus −240 mL·year⁻¹, respectively; 95% CI 78–173 mL·year⁻¹) and by 55% in INPULSIS-2 (−114 versus −207 mL·year⁻¹, respectively; 95% CI 45–143 mL·year⁻¹) [33].

In INPULSIS-1, there was no difference between the nintedanib and placebo groups in time to first acute exacerbation (HR 1.15, 95% CI 0.54–2.42), and the proportion of patients with at least one investigator-reported acute exacerbation was comparable between the nintedanib and placebo groups (61% versus 54%, respectively).

In INPULSIS-2, there was a significant increase in time to first acute exacerbation in the nintedanib group compared with placebo (HR 0.38, 95% CI 0.19–0.77). In addition, the proportion of patients with at least one investigator-reported acute exacerbation was lower in the nintedanib group (36%) compared with placebo (96%).

Diarrhoea was the most frequent adverse event in the nintedanib groups compared to the placebo groups, and was reported in 61.5% versus 18.6% of participants, respectively, in INPULSIS-1 and 63.2% versus 18.3% of participants, respectively, in INPULSIS-2. Nausea was the second most common adverse event among patients treated with nintedanib versus placebo, occurring in 22.7% versus 5.9% of participants, respectively, in INPULSIS-1 and 26.1% versus 7.3% of participants, respectively, in INPULSIS-2.

Pirfenidone is an antifibrotic agent that inhibits TGF-β-stimulated collagen synthesis, decreases extracellular matrix production and blocks fibroblast proliferation in vitro. It was demonstrated that pirfenidone inhibited synthesis of both PDGF-A and -B isoforms by lung macrophages [82], reduced inflammation and suppressed the bleomycin-induced increase in the levels of TGF-β [83]. How much of the antifibrotic effects are mediated through a PDGF pathway remains unclear.

Pirfenidone has been investigated for IPF since 1999 [84, 85]. ASCEND (Assessment of Pirfenidone to Confirm Efficacy and Safety in Idiopathic Pulmonary Fibrosis) was a phase-3 trial in 555 patients with IPF randomly assigned to receive oral pirfenidone (2403 mg per day) or placebo for 52 weeks [34]. Compared to the placebo group, pirfenidone resulted in a significant reduction in the 1-year rate of decline in FVC, but did not reduce dyspnoea.

Two concurrent, multicentre clinical trials (CAPACITY 004 and 006) studied pirfenidone in IPF; the primary end-point was change in FVC % predicted at week 72 [35]. 779 patients with mild-to-moderate IPF (i.e. FVC ≥50% pred and diffusing capacity of the lung for carbon monoxide (D_LCO) ≥35% pred) were randomly assigned in a 2:1:2 ratio to oral pirfenidone 2403 mg·day⁻¹, 1197 mg·day⁻¹ or placebo in the 004 trial and oral pirfenidone 2403 mg·day⁻¹ or placebo in the 006 trial. The higher dose of pirfenidone significantly decreased the percentage fall in FVC in the 004 trial (difference between groups 4.4%; p=0.001), but not in the 006 trial (difference between groups 0.6%; p=0.51). The higher dose of pirfenidone significantly reduced the decline in the 6-min walk test distance, a secondary end-point, in the 006 trial (absolute difference 32 m; p=0.0009), but not in the 004 trial.

In a prespecified analysis that pooled results of the ASCEND trial and CAPACITY 004 and 006 (n=1247 patients), pirfenidone decreased death from any cause relative to placebo (22 (3.5%) deaths in the pirfenidone group compared with 42 (6.7%) deaths in the placebo group; HR 0.52, 95% CI 0.31–0.87). As the ASCEND trial was 52 weeks in duration, the pooled survival analysis only considered data from the first 52 weeks of the CAPACITY trials (which were 72 weeks in duration). A separate pooled analysis considering all available data on all-cause mortality showed a trend favouring pirfenidone, but was not statistically significant (Kaplan–Meier estimate 0.75, 95% CI 0.51–1.11) [34].

The US Food and Drug Administration approved pirfenidone (Esbriet) and nintedanib (Ofev) for the treatment of IPF in October 2014.

Hermansky–Pudlak syndrome

Hermansky–Pudlak syndrome (HPS) is a rare autosomal recessive disorder characterised by tyrosinase-positive oculocutaneous albinism, a bleeding diathesis from platelet dysfunction and systemic complications associated with lysosomal dysfunction, including pulmonary fibrosis.

In a randomised placebo-controlled trial, treatment with pirfenidone (800 mg three times daily) for ≤44 months led to a 5% difference in the yearly rate of FVC decline in 11 pirfenidone-treated patients compared to 10 placebo-treated patients (p=0.001). Using data restricted to patients with an initial FVC of >50% pred, patients in the pirfenidone group lost pulmonary function (FVC, forced expiratory volume in 1 s, total lung capacity and D_LCO) at a rate of 8% per year slower than the placebo group. The authors concluded that pirfenidone slowed the progression of pulmonary fibrosis in patients with HPS who have significant residual lung function [36].

Following these initial promising results, the National Institutes of Health conducted a second randomised controlled study which enrolled 35 subjects with HPS-1 pulmonary fibrosis; 23 subjects received pirfenidone and 12 received placebo. The study was stopped due to futility. This was after an interim analysis, performed 12 months after 30 patients were enrolled showed no statistical difference between the placebo and prifenidone groups for the rate of decline in FVC (0.7% per year less for the pirfenidone group compared with the placebo group) [37].

Connective tissue disease

PDGF signalling plays an important role during embryonic development and contributes to the maintenance of many elements of the connective tissue matrix in adults. For example, elevated expression of PDGF and PDGFR has been found in scleroderma skin and lung tissues [86].

Divekar et al. [87] demonstrated that imatinib reduced interleukin (IL)-4-producing T-cells, but increased CD41 T-cells in the lungs of patients with systemic sclerosis (SSc). Therefore, targeting PDGF signalling to control immune cell populations in lung diseases might also be important.

In earlier experiences in clinical use of imatinib, several case reports showed the favourable activity of imatinib in patients with connective tissue diseases including SSc and mixed connective tissue disease (MCTD) [88–90]. Chung et al. [88] reported two cases of SSc treated with imatinib. Both patients showed improvement of skin tightening after the use of 200 mg·day⁻¹ of imatinib for 3 months. Sfikakis et al. [89] reported a severe case of SSc treated with 400 mg·day⁻¹ imatinib for 6 months. In this patient FVC improved from 68.0% pred to 88.3% pred. In addition, Distler et al. [90] reported the improvement of pulmonary function and high-resolution computed tomography findings in patients with MCTD with imatinib therapy. In a phase-IIa, single-arm, open-label clinical trial, 30 patients with diffuse cutaneous SSc were treated with 400 mg of imatinib daily. Improved modified Rodnan skin scores and FVC were seen [38]. Adverse events were common, but were mild to moderate. Some trials were terminated prematurely for safety reasons [91]. The most common adverse events reported included fluid retention, weakness and nausea or vomiting. Imatinib use in severe SSc requires further study.

Lymphangioleiomyomatosis

Lymphangioleiomyomatosis (LAM) is a rare and unusual cancer that involves the lungs either as part of tuberous sclerosis complex or in a sporadic form. Growth factors such as PDGF and epidermal growth factor have been shown to enhance LAM and renal angiomyolipoma (AML) cell proliferation in vitro [92, 93]. Shiomi et al. [94] demonstrated that LAM are cells of mesenchymal origin that stain positive for PDGFR-β. Imatinib mesylate, which targets PDGF, could completely block the growth of the LAM/AML cells resulting in cell death. Cells treated with rapamycin did not undergo cell death, although growth was inhibited. These findings suggest that imatinib could be a potential therapy in the treatment of LAM. Clinical trials have not been performed.

Acute respiratory distress syndrome-associated lung fibrosis

Acute respiratory distress syndrome (ARDS) is characterised by damage to the alveolar capillary membrane, oedema formation and repair of the alveolar–capillary membrane with a varying degree of fibrosis.

A recent study on human lung fibroblasts showed that PDGF and TGF-β1 regulate ARDS-associated lung fibrosis through distinct signalling pathway-mediated activation of fibrosis-related proteins [95]. The authors suggested that treatments with both PDGF and TGF-β1 antagonists may result in better antifibrotic outcomes for lung fibrosis induced by acute lung injury. Clinical trials have not been performed.

Summary

PDGF is a growth factor involved in many different lung diseases. Although the use of multitarget tyrosine kinase inhibitors to attenuate PDGF signalling is approved for lung cancer and pulmonary fibrosis, expanded uses of newer and more specific agents will enter clinical trials and find a place in the treatment of other pulmonary diseases.

Disclosures

C. Strange ERR-0061-2017_Strange ^{(1.2MB, pdf)}