Published ahead of print on January 25, 2007, doi:10.1164/rccm.200608-1153CC
© 2007 American Thoracic Society doi: 10.1164/rccm.200608-1153CC
Idiopathic Pulmonary Fibrosis and Pulmonary HypertensionConnecting the Dots1 Advanced Lung Disease and Transplant Program, Inova Fairfax Hospital, Falls Church, Virginia; 2 Division of Pulmonary, Allergy, and Critical Care Medicine, Duke University, Durham, North Carolina; and 3 Department of Pathology, Johns Hopkins University School of Medicine, Baltimore, Maryland Correspondence and requests for reprints should be addressed to Steven D. Nathan, M.D., Medical Director, Advanced Lung Disease and Transplant Program, Inova Heart & Vascular Institute, Inova Fairfax Hospital, 3300 Gallows Road, Falls Church, VA 22042. E-mail: steven.nathan{at}inova.com ABSTRACT Idiopathic pulmonary fibrosis (IPF) has a poor prognosis and a course that is unpredictable. Pulmonary hypertension may complicate the course of IPF and potentially impact prognosis. There are multiple factors that might influence the onset and severity of pulmonary hypertension in IPF. The relationship between the physiologic and pathobiologic manifestations of the progressive fibrotic process and interceding pulmonary hypertension has not been well defined. This article serves to explore these relationships and to hypothesize about the possible linkage between these entities. From a prognostic standpoint, recent evidence suggests this to be important to assess for pulmonary hypertension in patients with IPF. The appropriate triggers for evaluating for pulmonary hypertension and the best method of detection require further study. Despite the relative ease of noninvasive methods, such as echocardiography, right-heart catheterization remains the best diagnostic test. The appeal of pulmonary hypertension in IPF is that it may be an enticing therapeutic target in a disease that otherwise does not have any proven effective therapies. Which agent(s) might be useful and when they should be implemented mandate the appropriate studies being performed. Some of the data presented in this article have previously been reported in abstract form only.
Key Words: pulmonary fibrosis • hypertension, pulmonary • pulmonary function tests • cytokines Idiopathic pulmonary fibrosis (IPF) is a terminal condition in most patients and is characterized by progressive fibrosis and respiratory insufficiency with an associated median survival of 2.5 to 5 years (1, 2). The hypothesis of the underlying pathogenesis of IPF has evolved from that of an inflammatory condition, to one that is characterized by overexuberrant fibroproliferation (3, 4). Despite an increasing understanding of the cellular and molecular aspects integral to the evolution and propagation of this disease, the ability to predict the clinical course of individual patients remains imprecise (5, 6). Although serial restrictive physiology does portend a worse outcome, a sizable number of patients succumb without a prior documented decrement in their pulmonary function tests (PFTs) (5–8). This might be due to relatively infrequent monitoring of PFTs in the context of rapid unanticipated declines. Some of these declines and the associated mortality are attributable to acute exacerbations, which can complicate the course of the disease. However, this remains a poorly defined complication that is unlikely to explain all such acute or subacute progressive deteriorations (5). There is a growing appreciation for the role of interceding pulmonary hypertension (PH) in the course of many diffuse parenchymal lung disorders, including IPF (9, 10). This article attempts to explore the role of PH in IPF and hypothesize how PH might provide the link to explain some of the clinical manifestations and devastating course that afflict most of these patients. COURSE AND PROGNOSIS OF IPF
A prior commonly held perception regarding the prognosis of IPF has been that the extent of restrictive physiology will predict mortality. Studies that have assessed lung volumes at baseline have shown that this is a poor predictor of outcomes (6, 11–13). However, more recently, several studies have assessed serial change in forced vital capacity (FVC) and have shown this to be a good predictor of the subsequent disease course (7, 8). Although patients with a 10% decrement in their FVC tend to do worse than those with stable FVCs, the latter group is still at significant risk of succumbing to their disease. This concept was further emphasized from the first large, randomized, placebo-controlled multicenter study of interferon (IFN)-
One of the revelations from the natural history data, gleaned from the placebo arm of the IFN- Exercise capacity has been shown to be a better predictor of outcomes than PFTs in a number of independent studies (11, 12, 16–18). Lama and associates showed that desaturation to less than 89% on a six-minute-walk test (6MWT) had very good performance characteristics in discriminating between survivors and nonsurvivors (16). The 4-year survival in this study in the two groups was 69% (nondesaturators) versus 35% (desaturators). Distance as a predictor of outcomes did not perform as well as desaturation. This was subsequently verified in a follow-up study by the same group showing that any desaturation, even if the oxygen saturation remained greater than 88%, was associated with worse outcomes (18). Hallstrand and associates showed similar findings with their timed 6MWT. In this study of 28 patients, they also verified that PFTs correlated poorly with outcomes (17). More recently, Kawut and associates have shown that desaturation to less than 95% during steady-state exercise discriminated between survivors and nonsurvivors in 24 patients with IPF (11). They also showed that the distance walked had very good discriminatory power when a breakpoint of 350 m was used. In a retrospective review using the United Network for Organ Sharing (UNOS) database of patients with IPF listed for lung transplantation, a breakpoint of 207 m had the best discriminatory power in determining outcomes (19). More recently, a composite of 6MWT parameters was shown to have better performance characteristics in distinguishing survivors from nonsurvivors (12). The authors demonstrated that the distance walked performed better than the SpO2 nadir, but that a combination of distance and SpO2 nadir, in the form of the composite distance–saturation product, provided the greatest discriminatory power. These observations need confirmation with prospective trials. Nonetheless, the intriguing question raised by these studies relates to what important prognostic factor is captured by exercise studies that is not uncovered by standard PFTs. PREVALENCE OF PH IN IPF The prevalence of PH complicating the course of patients with IPF has been reported as occurring in 32 to 85% of patients (9, 10, 20, 21). Data from the UNOS showed that 45% of patients with IPF listed for transplant have right-heart catheterization evidence of PH, with 9% of patients having a mean pulmonary arterial pressure (mPAP) of greater than 40 mm Hg (22). This wide range in the prevalence of PH reported likely reflects the timing of the measurement during the course of the patient's disease, with patients who are later in their disease course manifesting more evidence of PH. In addition, most patients with IPF will manifest significant PH with exercise, which might be an earlier manifestation of those patients likely to develop PH (23). This concept of serial progression has been underscored in a retrospective review of 39 patients with IPF in whom serial right-heart catheterization measures of PAP were available (24). The first measure showed a prevalence of PH of 33%, whereas the second measurement in the same group performed immediately before transplant demonstrated a prevalence of 85%. A follow-up analysis of 63 patients with IPF verified this concept, with an initial prevalence of 41% increasing to more than 90% at follow-up (25). These findings should be validated in the controlled setting of a prospective study and be inclusive of patients with IPF of all ages and levels of severity. CORRELATES OF PH IN IPF
Intuitively, the severity of fibrosis and degree of restrictive physiology should correlate with the prevalence and degree of PH. However, it appears that PH may not correlate with lung volumes in patients with IPF. In a group of 51 patients with various forms of interstitial lung disease, including 24 with IPF, Kawut and associates have shown that whereas PAP correlates well with mortality, there is no difference in FVC between survivors and nonsurvivors. In a retrospective analysis of right-heart catheterization data from 79 patients, our group has shown that there is no difference in the FVC between those patients with and without PH (10). In a further analysis of 100 patients with IPF in whom right-heart catheterization data were available, patients were categorized in deciles by their closest available FVC. Interestingly, there did not appear to be a significant difference in the prevalence of PH across the five groups analyzed (FVC > 70%, 60–69%, 50–59%, 40–49%, and < 40%) (26). Paradoxically, mPAP and prevalence of PH in the group with the least restriction (FVC > 70%, n = 11; prevalence, 55%; mPAP = 30.5 mm Hg) appeared to be higher than that of the group with the worst restriction (FVC < 40%, n = 15; prevalence, 33%; mPAP = 21 mm Hg). This apparent paradox could be explained by survival bias because patients with severe restriction and significant PH are more likely to succumb to their disease. The numbers from this study are not only relatively small but also represent a single-center experience and therefore require further validation. Nonetheless, this finding is intriguing and supports the notion that factors aside from progressive fibrosis are responsible for PH in IPF. An alternate hypothesis is that the FVC does not accurately reflect the degree of fibrosis. Indeed, there is evidence to support this notion from the large prospective study of IFN- As opposed to lung volumes, exercise capacity does appear to have a significant association with PH in IPF. In the study by Kawut and associates, those patients with interstitial pneumonitis who died had poorer six-minute-walk distances than those who survived (11). In addition, the prevalence of PH was higher in the former group (57 vs. 19%). Although no direct comparison was done between those patients with and without PH, one can infer that those patients with PH have a lower exercise capacity. From among our 79 patients with right-heart catheterization data, there were 34 with evaluable, contemporaneous 6MWTs. Ten of these patients did not have PH, whereas 24 did. There was a significant difference between these two groups in the distance walked (366 vs. 144 m, p < 0.001) as well as the minimal SpO2 (88 vs. 80%, p < 0.001). Thus, the fundamental question that must be addressed is the relationship between the extent of interstitial remodeling and the development of PH. Conventional wisdom has suggested that PH in the context of IPF would develop as a consequence of the obliteration of alveoli with vascular changes occurring "secondary" to the alveolar remodeling. However, as investigators begin to explore the relationship between PH and IPF, this assumption must be questioned. Do patients with PH and IPF represent a unique spectrum of disease? Can important clues to pathogenesis be learned by a closer examination of the role of vascular remodeling in IPF? PH AS A PREDICTOR OF OUTCOMES One of the earlier studies suggesting that PH impacted the outcomes of patients with IPF was from King and associates, who demonstrated that the size of the PA segment, as measured on the plane chest radiograph, correlated with subsequent mortality (27). More recently, in a cohort of 88 patients with IPF, it has been shown that the systolic PAP (sPAP), as estimated by echocardiography, has a strong correlation with survival. Specifically, those patients with an estimated sPAP greater than 50 mm Hg had a median survival of only 0.7 year versus more than 4 years for those patients with an sPAP of less than 50 mm Hg (9). In our study of 79 patients with IPF in whom right-heart catheterization data were available, a significant difference in outcomes was demonstrated with a 1-year mortality rate of 28% in those patients with PH versus 5.5% in those without (p = 0.002) (10). PATHOGENESIS OF PH AND IMPLICATIONS FOR THERAPY IN IPF The complexity of pathologic and pathobiologic paradigms that characterize IPF might also apply to PH complicating the condition. Although most cases of PH in IPF are mild or moderate, a significant number of patients present with severe disease, with almost systemic levels of PAP (Figure 1). Questions related to the pathobiology of PH in IPF include whether there are differences in pulmonary vascular remodeling or different pathogenetic mechanisms associated with varying levels of PAP. Our current understanding of the mechanisms of PH revolve around pulmonary artery vasoconstriction or pulmonary artery remodeling. With regard to the former, this is mostly due to hypoxemia, both systemic and possibly local. Although this likely plays a role, it is unlikely to account for all the PH in IPF. Most of the studies that have documented the presence and severity of PH in IPF have done so under normoxic conditions. Given the present understanding of the effects of hypoxia and its impact on survival, measures to correct any significant desaturation in patients with IPF appear warranted. However, the role of oxygen free radicals in the genesis or perpetuation of the disease does raise theoretic concerns about potential deleterious effects of high supplemental oxygen concentrations (28).
It would appear, however, that given the extent of alveolar damage and abnormal incorporation of connective tissue and ongoing inflammation in IPF, pulmonary artery remodeling might play a more relevant role than vasoconstriction, yet the two pathogenetic processes might be intimately interrelated. Pulmonary vascular remodeling can be regarded on a global macroscopic or local microscopic basis. With regard to the former, the theme of vascular remodeling heterogeneity is fulfilled with both areas of vessel ablation and other areas of neovascularization (29–34). Although a proangiogenic environment exists in the IPF lung, the overall vessel density has been shown to be reduced with net vascular ablation (34). Vessel ablation has been shown to occur particularly within fibroblastic foci and in areas of honeycombing. Intuitively, one might expect that the extent of the latter should correlate with progressive restrictive physiology. Its apparent lack of correlation with PH suggests that other mechanisms are involved. Whether angiogenesis is a protective or harmful phenomenon in IPF remains controversial. One of the mechanisms whereby the balance between angiogenesis and angiostasis might influence outcomes is through the resultant effects on PH. This is an area that remains unexplored, but is likely more complicated than angiogenesis favoring pulmonary normotension and angiostasis shifting the balance to hypertensive physiology. Indeed, the morphology of new vessel formation has also been demonstrated to be different, with absence of an elastin layer, which conceivably may reduce vascular compliance and further contribute to the development of PH (34). Local vascular remodeling in PH can be broadly divided as based on the predominant vascular cells in the intimal, medial, and adventitial compartments in pulmonary arteries. At the present time, there are no quantitative or qualitative data that relate IPF and the severity of PH with adventitial changes or varying smooth muscle cell remodeling. Intimal lesions may impact the most on pulmonary vascular resistance. A recent preliminary gene microarray study revealed that a subset of patients with IPF and moderate/severe pulmonary hypertension (17 of 117 patients) showed down-regulation of a fraction of endothelial cell genes and up-regulation of the phospholipase A2 gene and other factors potentially involved in pulmonary vascular remodeling in IPF-associated PH (35). These data support that endothelial cell dysfunction may underlie the pathogenesis of IPF-associated PH and perhaps help to explain the apparent dissociation between the degree of fibrosis and PH (36).
Several mediators involved in experimental and/or idiopathic pulmonary arterial hypertension (IPAH) may potentially participate in the pathogenesis of PH associated with IPF. There is overproduction of profibrogenic leukotrienes, particularly by inflammatory cells, produced by the activation of 5-lipoxygenase (5-LO) (37). 5-LO expression is up-regulated in pulmonary arteries of patients with IPAH, particularly in alveolar macrophages and lesional endothelial cells (38). Of note, leukotrienes can up-regulate several potential mediators of both lung fibrosis and pulmonary vascular remodeling, such as tumor necrosis factor (TNF)-
Endothelin (ET)-1 is a strong candidate molecule in the pathogenesis of IPF-associated PH. ET-1 promotes pulmonary arterial vasoconstriction and induces pulmonary arterial smooth muscle cell growth (36). The rationale for ET-1 receptor blockade in IPAH relied on data showing enhanced ET-1 expression in plexiform lesions as compared with normal lungs (40). Moreover, IPF lungs also show increased expression of ET-1 and ET-converting enzyme as compared with normal lungs (40, 41). Furthermore, arterial ET-1 levels appear to correlate inversely with arterial oxygen and directly with PAP in patients with IPF (42). ET-1 has also been shown to have profibrotic properties and has been recently studied in this regard in patients with IPF without clinical PH (43). PDGF is another profibrotic cytokine that is currently being studied as a target of therapy in IPF (44). PDGF has also been noted to be up-regulated in PH. Recent insights into the potential role of PDGF in the pathogenesis of experimental PH demonstrating enhanced expression in lung tissue has led to preliminary investigations of the effect of its blockade in this condition (45, 46). TGF- These cytokines/growth factors not only might offer a link in the pathogenesis of IPF and associated PH but also potentially the opportunity for joint targeting of the fibrotic and pulmonary vascular components of the disease. Variable expression of these and other factors integral to vascular remodeling might explain the heterogeneity of PH in patients with IPF. IMPLICATIONS AND MONITORING OF PH IN IPF PH might provide some of the answers about the unpredictable course of patients with IPF. Specifically, what is the role of PH in those patients who die without a prior documented decline in their lung volumes? All acute decompensations are not necessarily acute exacerbations of IPF. Not all patients with acute decompensations truly manifest new alveolar infiltrates that define an acute exacerbation. For those who do not, what is the role of progressive PH in their decline and demise? Is it causative or consequential and therefore a surrogate of other events? Whatever its role and cause, PH appears to be an important determinant of disease outcomes and therefore a necessary measure to define disease severity. When and how it should be measured is open to question. Despite its appeal and the data from the Mayo group, echocardiography might not suffice (9). Arcasoy and associates have shown that echocardiography is inaccurate in estimating sPAP in patients with interstitial lung disease (48). In their study of 106 patients with interstitial lung disease, echocardiography provided an estimate of sPAP within 10 mm Hg in only 37% of the patients. It did tend to be more accurate if the sPAP was less than 45 mm Hg (72%) as opposed to greater than 45 mm Hg (31%). Nonetheless, the role of echocardiography as a screening tool for the presence of PH in IPF does warrant further study. Brain natriuretic peptide (BNP) has been evaluated in a small series in which it was demonstrated to have excellent performance characteristics (49). A more recent, larger study of patients with various forms of advanced lung disease provides additional support to the utility of BNP as a predictor of PH; however, this still requires further validation, ideally in the context of a large prospective study limited to patients with IPF (50). Exercise desaturation, a low diffusing capacity of carbon monoxide (DLCO) and need for supplemental oxygen might all be surrogate indicators of underlying PH. In our study, patients with the need for supplemental oxygen in conjunction with a DLCO of less than 40% predicted were 10 times more likely to have concomitant PH than those without either of these two features (10). However, verification of the appropriate BNP threshold, oxygen needs, and the DLCO, as well as other potential biomarkers that might indicate underlying PH, requires further study. Although right-heart catheterization does appear to provide important prognostic information, a stronger case for routinely advocating such invasive testing would be underscored if studies of PAH medications prove useful for PH in IPF. One will need to be cautious in the interpretation of such studies with regard to their endpoints. Although they might be effective in reducing the PAP, it is imperative that endpoints include improvements in exercise capacity, oxygen requirements, dypsnea, quality of life, and, most important, survival. Small pilot studies of sildenafil and inhaled iloprost have shown encouraging results; however, large phase 3 studies of these and other agents are required for subsequent validation (51, 52). The history of medicine is replete with physiologically enticing interventions that have ultimately been proven detrimental (53, 54). This underscores the need for the appropriate studies before the implementation of PH therapies for IPF can be broadly accepted. IPF has always been regarded as a disease based in the periphery. If such therapies do prove useful, then the pulmonary vasculature could well emerge at the center of the quest to prolong the lives of patients with this devastating condition. FOOTNOTES Conflict of Interest Statement: S.D.N. is a consultant and serves on an advisory board for Intermune, Inc., for which he received $5,000 in 2004 and $6,000 in 2005; he has received consultant and lecture fees from United Therapeutics ($1,500 in 2004), Actelion ($850 in 2004, $1,500 in 2005, $12,500 in 2006), and Cotherix ($1,000 in 2005 and $1,500 in 2006); he has received research grants from Intermune, United Therapeutics, Cotherix, Wyeth, and Novartis for studies in IPF and pulmonary hypertension. P.W.N. has served as a consultant for Intermune, for which he received $10,000 per year from 2004–2006; he has served as a consultant for Wyeth, Actelion, and Johnson & Johnson, for which he received $2,500/company in 2006; he has received lecture fees from Intermune totaling $6,000 per year in 2004–2006. R.M.T. received lecture fees related to a talk on COPD in a meeting organized by AstraZeneca aimed at fundamental cellular and molecular processes involved in COPD ($2,400 honorarium and travel expenses); he received funds from Quark Biotech related to the investigation of the role of a specific mediator of cellular stress, RTP801, discovered by the company, in COPD; this support consisted of a contract for the years 2005–2006 for $100,000 and a gift for a postdoctoral support during 2006–2007 for $100,000; neither support is related to the focus of the present paper. Received in original form August 15, 2006; accepted in final form January 25, 2007 REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1b, which included the serial prospective monitoring of 330 patients with IPF. These subjects had serial pulmonary function testing performed every 3 months for 1 year. Despite this close observation, 43% of the patients died without a prior documented 10% decrement in their FVC (
, platelet-derived growth factor (PDGF), and fibroblast growth factor (
, thus accounting for increased collagen deposition and enhanced pulmonary artery remodeling. Whether there is down-regulation of PGE2 or prostacyclin synthases in the interstitium or in pulmonary arteries of patients with IPF (as demonstrated in IPAH cases) remains undetermined (
