Discussion
Our study showed that patients with end stage fibrotic ILD undergoing lung transplantation had significant and often severe pulmonary arterial vasculopathy in their explanted lungs. Unexpectedly, this arterial vasculopathy was present in lungs from patients with and without PH and had no correlation to the clinical severity of the underlying lung disease. To the best of our knowledge, this is the first study to describe a high frequency of severe vascular changes in advanced fibrotic ILD, with almost half of the patients showing vascular occlusion with intimal fibrosis and one-fifth showing plexiform lesions which are traditionally considered a hallmark only of WHO group 1 iPAH pathology.
Our patients with PH are classified as WHO Group 3, PH secondary to hypoxemic lung disease. One of the most important factors contributing to pulmonary arterial changes in patients with ILD is chronic hypoxia, which is a part of the terminology of WHO group 3 (PH due to lung diseases and/or hypoxia). Chronic hypoxia affects all three layers of the pulmonary vasculature: the intima, media and adventitia and their harbouring cells—the endothelial cells (EC), smooth muscle cells (SMC) and fibroblasts, respectively.19 However, hypoxia alone does not usually lead to high-grade vascular remodelling with intimal obliteration and plexiform lesions as seen in iPAH.20
Plexiform lesions are networks of vascular micro vessels that consist of two distinct populations of EC: normal phenotype that lines vessels and the proliferating, apoptosis-resistant phenotype at the core of the lesion.21 22 They are considered the hallmark of severe iPAH and were first described in 1927.12 23,12 In 1989, a landmark study by Pietra et al, of 48 patients with iPAH documented plexiform lesions in 52% of the total cohort.12 This study also demonstrated that presence of plexiform lesions portended a poorer prognosis. In 2012, Stacher et al evaluated explanted lungs from 62 patients with group 1 PAH and documented plexiform lesions in 90% overall and in all of 48 patients with idiopathic group 1 iPAH.11
Data obtained by computerised three-dimensional reconstructions of serial lung sections from patients with PH, suggested that a single plexiform lesion can occlude the entire length of an affected vessel, contributing to PH severity.21
Historically, pulmonary arterial vascular changes are less well documented and infrequently described in patients with fibrotic ILD. A Japanese lung necropsy study of 60 patients with emphysema, IPF or CPFE reported plexiform arterial lesions in 2 patients with CPFE and 1 patient with IPF and none in emphysema.16 In a similar study of explanted lungs of 70 patients transplanted for end stage COPD (chronic obstructive pulmonary disease), only 1 patient had plexiform lesions.17 Our data demonstrated much more common advanced vasculopathy including plexiform lesions in end stage fibrotic ILD lung explants obtained at time of transplantation.
Our hypothesis was that patients with severe PH would have more severe pulmonary arterial vascular changes but not plexiform lesions seen in group 1 PAH. We were expecting positive correlation between mPAP and pulmonary artery vasculopathy grade and only mild to moderate vascular changes. However, pretransplantation measured mPAP and DLCO failed to predict which patients had pulmonary arterial pathological changes. In addition, we found unexpectedly severe vascular change and a high prevalence of plexiform arterial lesions in patients with normal directly measured pulmonary pressures. We found that severe pulmonary arterial vasculopathy was common in patients with end stage fibrotic ILD and that these changes could not be predicted clinically before lung transplantation. The significance of these and our data is unclear. Is the presence of plexiform arterial lesions in fibrotic ILD an indicator a causal role for the pulmonary arterial vasculopathy in clinical ILD disease progression or if simply a morphological epimarker of the irreversible parenchymal distortion seen in end-stage fibrotic ILD?22 24–26 We speculate about possible explanations:
Geographic and temporal heterogeneity in pulmonary arterial vascular changes throughout the lung. Since the geographic distribution of parenchymal pathology is uneven within the lung, there might be significant heterogeneity in the distribution and severity of the vascular changes. In some patients, we may have found signs of severe local vasculopathy that had yet to cause a significant effect on the PVR as measured in RHC. This idea is supported by previous findings in iPAH histopathology regarding notable heterogeneity in the number and distribution patterns of plexiform lesions within a given PAH lung and between patients with PAH. This occurred without relation to other parameters of vascular remodelling or haemodynamics.11 25 Even if true, this does not account for advanced arterial vasculopathy and of plexiform lesions as a response to chronic hypoxia. Chronic hypoxic changes have been described to cause intimal fibrosis and medial and adventitial thickening, but not higher-grade vascular remodelling including plexiform lesions.27
Shared pathogenic mechanisms between fibrosis and pulmonary arterial vasculopathy. Many mediators including endothelin, VEGF (vascular endothelial growth factor), TGF-β (transfroming growth factor β), NO (nitric oxide), PGE2 (prostaglandin E2), BMPR2 (bone morphogenic protein receptor type 2) and IL-6 (interleukin 6) have been shown to play important roles in both the progression of fibrosis and the development of PH.28 The possibility of shared pathogenic mechanisms between IPF and PH was also demonstrated through microarray gene expression; patients with IPF patients, like patients with PH show an exaggerated expression of mediators of SMC and EC proliferation compared with controls, suggesting the existence of preclinical pulmonary vascular disease.29 The patients end stage fibrotic lung disease theoretically might have caused vascular changes unrelated to clinically measured PH by mPAP in RHC. In other words, the presence of the severe vascular changes might represent end stage lung disease rather than severe PH.
Interestingly, patients who were treated with antifibrotics before lung transplant had significantly lower mPAP and PVR. The role of pirfenidone and nintedanib in patients with IPF and PH is not clear since no data on these agents are currently available. Tyrosine kinase inhibitors such as Imatinib and nintedanib have shown to improve exercise capacity and haemodynamics in patients with advanced WHO group 1 PAH. Pirfenidone which combines antiproliferative and anti-inflammatory activity may have similar effects. However, no data currently support this hypothesis.30
There is no known effective therapy to prevent or treat PH in fibrotic ILD and treatment with pulmonary vasodilators is not recommended based on multiple clinical trials.31–44 Further molecular, biochemical and clinical characterisation subgroup of patients with ILD and high-grade vascular changes might have an important mechanistic, clinical and therapeutic significance. Clinical data including directly measured pulmonary haemodynamics failed to indicate patients at risk for pulmonary arterial vascular pathology. This highlights the challenges in identification of patients with fibrotic ILD who could potentially benefit from future therapeutic interventions.
Our study has several limitations. First, this is a retrospective study that has inherent limitations in patient selection. Second, despite our relatively large sample size as compared with other studies with explanted lungs, the size is still small with inherently reduced statistical power. Third, sampling error is possible and may underestimate the degree of significant vascular changes, especially with disease that is patchy in nature; however, we also view this as strength of our research as this may explain why plexiform pulmonary arterial vasculopathy has yet been better described in end stage ILD.
In summary, we found significant pulmonary arterial vasculopathy in patients with end stage ILD. The vasculopathy did not correlate with the clinical severity of the underlying lung disease and did not correlate with the presence and/or severity of PH as measured by RHC. We found a significant number of patients with end stage lung disease (21%) with plexiform lesions which are traditionally a hallmark of WHO group 1 PAH. The observed vasculopathy might be derived from severe and prolonged hypoxia in the complex biochemical environment of the diseased lung that involve mechanisms similar to those seen with WHO group 1 PAH. These findings indicate that advanced pulmonary arteriopathy is common and may develop in a heterogeneous regional pattern in advanced lung disease prior to the clinical detection of PH. Larger scales studies are needed to confirm our findings and conclusions.