PH has been a focus of new investigations due to its potential prognostic and therapeutic implications in IPF patients. Recent studies have demonstrated that severe PH in IPF can occur in the absence of advanced pulmonary dysfunction or hypoxemia [3
]. As more specific and effective treatment for PH has become available, early diagnosis and treatment of PH could improve the prognosis in the subset of IPF with PH. Right heart catheterization is the gold standard for the diagnosis of PH but is invasive and not widely used in IPF patients. We sought to identify the histologic features that could potentially be the predictors of PH or indicators for further clinical work-up in IPF patients, mainly using the sections of surgical lung biopsies that had been performed for the diagnosis.
It has been thought that PH in IPF develops as a consequence of progressive fibrosis and obliteration of alveoli with vascular changes occurring secondary to the reduction of the pulmonary vascular bed [3
]. However, this postulation has been questioned as many investigators explored the association of the remodeling of pulmonary vessel in PH associated with IPF. We also have noted that non-fibrotic areas in some lung biopsies of interstitial lung disease (ILD) cases show extraordinary capillary proliferations that were accompanied by clinical PH; they mimicked the changes of pulmonary capillary hemangiomatosis (PCH) with pulmonary veno-occlusive disease (PVOD)-like features in the background (unpublished observations). These cases also had hemosiderin depositions and muscularization of pulmonary arterioles in addition to PCH- or PVOD-like changes, which comprise the features of postcapillary PH. Since the distribution of fibrosis in usual interstitial pneumonia, histopathologic manifestation of IPF, coincides with the distribution of pulmonary veins and venules in the interlobular and interstitial septa of peripheral alveolar acini, these findings could be simply caused by fibrosis. However, one can also postulate that the fibrogenic stimuli involved in IPF might also affect remodeling of postcapillary pulmonary vessels, causing PVOD-like condition.
This led us to hypothesize that these histologic changes may be associated with PH in IPF, possibly independent of lung function or fibrosis. To test this hypothesis, we attempted to measure several histologic parameters in the surgical lung biopsies in the cohort of IPF patients diagnosed over a recent 10-year-period. It was very difficult to quantify these histologic features in an objective, systematic manner. Among these, we found that PCH-like changes by counting the number of endothelial cells on CD34 immunostaining (expressed as ASCD) and hemosiderin deposition assessed by iron stain were most feasible for quantitation in the exploration of this hypothesis. Although smoker's pigments in the alveolar macrophages can mimic hemosiderin, they are typically dustier and distinguishable from refractile, chunky iron pigments.
In our study, iron deposition in IPF was associated with higher RVSP independently of the degree of overall fibrosis or pulmonary function impairment as assessed by fibrosis score on the HRCT scan and FVC on PFT, respectively. Though not reaching the statistical significance, ASCD also showed a trend of association with RVSP on multivariate analysis. FVC% was associated with RVSP on univariate analysis but not on multivariate analysis, however. We analyzed the RVSP data as categorical variable by assigning the value of ≥ 40 mmHg as positive and <40 mmHg as negative for pulmonary hypertension. As was in the analysis using RVSP value as continuous variable, iron score 3 and ASCD score ≥ 1.5 showed the similar findings in both univariate and multivariate analyses (data not shown).
Our study demonstrated HRCT fibrosis score did not show any association with RVSP either on univariate or multivariate analysis, reiterating the conclusion of a previous study; a cross-sectional study of 65 patients with advanced IPF revealed that the extent of pulmonary fibrosis determined on chest HRCT did not predict PH in advanced IPF patients [15
]. A recent clinical study also has reported that there is a poor correlation between lung function measures and PH in IPF patients, suggesting that factors other than parenchymal fibrosis may play a role in the pathogenesis of PH [17
Histopathologic findings of pulmonary vessels in IPF were reported mostly based on vascular abnormalities seen in scarred areas and consist of intimal hyperplasia, fibrosis, and reduplication of the inner elastic lamina in the small muscular pulmonary arteries in fibrotic lobules [18
]. However, Colombat et al reported the vascular changes in the architecturally preserved lung tissues as intimal proliferation and fibrosis causing occlusion of venules and small pulmonary veins involving 65% of explant lung specimens from their 26 end-stage IPF cases [16
]. Other studies reported increased capillary density and angiogenesis in nonfibrotic lung tissues [19
]. A greater angiogenic response to IL-8 was detected in UIP lung tissue than in control tissue [22
]. A study shows that TGF-beta1-induced and hypoxia-induced VEGF expression by human lung epithelial cells may promote neovascularization, thereby contributing to the repair of injuries to the lung endothelium [23
The existence of neovascularization in IPF was originally described in 1963 by Turner-Warwick [24
]. Heterogeneous vascular remodeling in UIP has been reported by Ebina et al, who also identified increased alveolar septal vascular densities in less fibrotic areas [19
]. Several other authors have reported the abnormal vascular remodeling in ILDs although their findings in the literature seemed to be somewhat conflicting [20
]. Chronic inflammation is often associated with fibroproliferation and neovascularization histologically [27
]. Inflammation and angiogenesis are also closely related events and temporally coincide [27
]. It is possible that neovascularization in ILD is at least partly due to the overlapping properties of many inflammatory, fibrogenic and angiogenic cytokines involved in the pathogenesis of ILDs. A previous study has postulated that an imbalance between the levels of angiogenic and angiostatic chemokines may result in aberrant angiogenesis in both animal models and tissue specimens from IPF patients [28
A recent clinicopathologic study reported that occlusive venopathy in nonfibrotic lung tissue was observed in 65% of their 26 explanted lungs from end-stage IPF patients [16
]. As in our cases, this study also demonstrated that most of these patients had increased alveolar capillary densities with PCH-like changes, hemosiderin depositions and muscularized arterioles [16
]. Their statistical analysis showed that the mean pulmonary artery pressure (PAP) correlated with the macroscopic extent of fibrosis, but not with occlusive venopathy [16
]. However, this study only included the explanted lungs that would represent the end-stage fibrosis requiring lung transplantation. On the other hand, our cases were mainly composed of surgical biopsies for diagnostic purposes that would include both earlier as well as late stage of IPF cases. This could have been a factor for the differing results between the two studies.
Our retrospective study is mainly limited by the lack of confirmatory right heart catheterization. We used the hemodynamic data obtained by Doppler echocardiography that affords a reasonably good correlation with the right hear catheterization data [29
]. One may suspect the iron deposition or increased ASCD in IPF is non-specific, possibly secondary to left heart disease or age-related. In the present study, however; neither echocardial parameters for left heart function nor age of the patients correlated with iron deposition or ASCD, which would argue against such possibilities. We sampled all remaining gross specimens stored in the surgical pathology files for a complete microscopic examination of existing tissues for all surgical biopsies. We also thoroughly examined explanted lungs using multiple representative sections. Some cases still did not have sufficient microscopic fields to assess the changes in nonfibrotic lung tissues, however. We used both FVC% and HRCT fibrosis score to assess the severity of IPF, which should have given a more comprehensive evaluation for the severity than the assessment based on the given biopsy materials as used in a previous study [19
In summary, our study demonstrated that iron deposition and ASCD, histologic features associated with postcapillary remodeling, are also associated with RVSP in a non-selected cohort of IPF patients. Iron deposition and ASCD were associated or had a trend independent of the severity of disease assessed by FVC% and HRCT. These findings were not related to age or left heart function assessed by echocardiography.