Clinical relevance of plasma prostaglandin F2α metabolite concentrations in patients with idiopathic pulmonary fibrosis.

BACKGROUND
Idiopathic pulmonary fibrosis (IPF) is a devastating lung disease of unknown etiology with few current treatment options. Recently, we determined an important role of prostaglandin F2α (PGF2α) in pulmonary fibrosis by using a bleomycin-induced pulmonary fibrosis model and found an abundance of PGF2α in bronchoalveolar lavage fluid of IPF patients. We investigated the role of PGF2α in human IPF by assessing plasma concentrations of 15-keto-dihydro PGF2α, a stable metabolite of PGF2α.


METHODS
We measured plasma concentrations of 15-keto-dihydro PGF2α in 91 IPF patients and compared these values with those of controls (n = 25). We further investigated the relationships of plasma 15-keto-dihydro PGF2α concentrations with disease severity and mortality.


RESULTS
Plasma concentrations of 15-keto-dihydro PGF2α were significantly higher in IPF patients than controls (p<0.001). Plasma concentrations of this metabolite were significantly correlated with forced expiratory volume in 1 second (Rs [correlation coefficient] = -0.34, p = 0.004), forced vital capacity (Rs = -0.33, p = 0.005), diffusing capacity for carbon monoxide (Rs = -0.36, p = 0.003), the composite physiologic index (Rs = 0.40, p = 0.001), 6-minute walk distance (Rs = -0.24, p = 0.04) and end-exercise oxygen saturation (Rs = -0.25, p = 0.04) when patients with emphysema were excluded. Multivariate analysis using stepwise Cox proportional hazards model showed that a higher composite physiologic index (relative risk = 1.049, p = 0.002) and plasma 15-keto-dihydro PGF2α concentrations (relative risk = 1.005, p = 0.002) were independently associated with an increased risk of mortality.


CONCLUSIONS
We demonstrated significant associations of plasma concentrations of PGF2α metabolites with disease severity and prognosis, which support a potential pathogenic role for PGF2α in human IPF.


Introduction
Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, irreversible and usually lethal lung disease of unknown etiology that has limited therapeutic options [1]. Although IPF carries an overall poor prognosis [2], the clinical course of individual patients varies from slow progression to acute decompensation and death [3,4]. A variety of pathways and mechanisms underlying pulmonary fibrosis have been identified [5], but have not been well evaluated in the clinical arena.
Prostaglandins (PGs) are oxygenated metabolites of arachidonic acid produced by sequential catalysis of cyclooxygenase (COX) and respective synthase, and they contribute to a variety of physiological responses and pathological processes [6]. PGE 2 and PGI 2 are considered to exert antifibrotic effects [7]. However, we recently reported that PGF 2a facilitates pulmonary fibrosis through PGF receptor (FP) in a murine bleomycin-induced pulmonary fibrosis model [8]. In addition, PGF 2a is abundant in bronchoalveolar lavage fluid of human subjects with IPF, and stimulates proliferation and collagen production of lung fibroblasts via FP [8]. This indicates that PGF 2a produced in the lung of IPF patients may contribute to disease progression.
Endogenous PGF 2a is swiftly degraded in various organs including the lung to 13,14-dihydro-15-keto PGF 2a  ), a stable metabolite of PGF 2a , which has a longer half-life in the circulation and has been used as a reliable indicator of in vivo PGF 2a biosynthesis [9]. In the current study, we measured plasma concentrations of this metabolite in a cohort of IPF patients to clarify their clinical relevance and prognostic value.

Ethics Statement
This study was approved by Kyoto University Graduate School and Faculty of Medicine Ethics Committee and the Ethics Committee of Tenri Hospital, and written informed consent was obtained from all patients.

Study Subjects
The study population consisted of 91 IPF patients who visited Kyoto University Hospital from February 2008 through August 2011, and those who visited Tenri Hospital from April 2006 through September 2008. At the time of study entry, all of study patients had not been receiving any specific treatment for IPF.
IPF was diagnosed on the basis of the current official joint statement on IPF [10]. In 28 patients, usual interstitial pneumonia was confirmed by surgical lung biopsy. Quantification of disease severity by pulmonary function tests [11], and suitability of prognostic indicators [12] may be confounded by coexistent emphysema in patients with IPF. Therefore we also performed a subgroup analysis of patients without emphysema. Emphysema on high-resolution computed tomography was defined as the presence of well-demarcated areas of decreased attenuation in comparison with contiguous normal lung, and delimited by a very thin (,1 mm) or no wall, with upper zone predominance [13]. Ageand body mass index-matched subjects without lung disease (n = 25) were recruited as controls.

Physiological Assessments
Pulmonary function tests were performed using CHESTAC system (Chest M.I. Inc., Tokyo, Japan). Diffusing capacity for carbon monoxide (DL CO ) was measured using the single-breath technique. Percent-predicted values were used for analyses. The composite physiologic index (CPI) was calculated as previously described [11]. Arterial blood gas analysis, including arterial partial pressure of oxygen (PaO 2 ) and arterial partial pressure of carbon dioxide (PaCO 2 ), was performed while patients were breathing room air at rest in the supine position. The alveolararterial oxygen pressure difference (A-aDO 2 ) was calculated according to a standard formula, using the respiratory exchange ratio of 0.8. Six-minute walk testing (6MWT) was performed as recommended by American Thoracic Society guidelines [14] and oxygen saturation was continuously monitored during 6MWT using a pulse oximeter (Pulsox-300i, Konica Minolta Inc., Osaka, Japan).

Blood Sample Collection and Laboratory Assessments
Samples of peripheral venous blood were collected in the morning before breakfast. Plasma concentrations of 15-ketodihydro PGF 2a were measured using an enzyme immunoassay kit (13,14-dihydro-15-keto Prostaglandin F 2a EIA kit; Cayman Chemical; Ann Arbor, MI, USA). Serum KL-6 levels were measured by a sandwich-type electrochemiluminescence immunoassay kit (Picolumi KL-6; Sanko Junyaku, Tokyo, Japan) and serum SP-D levels were measured by a sandwich-type enzyme immunoassay kit (SP-D kit Yamasa EIA II; Yamasa Shoyu, Chiba, Japan).

Doppler Echocardiography
Twenty-eight IPF patients recruited in Kyoto University Hospital underwent Doppler echocardiography at study entry. Doppler echocardiography was performed using conventional clinical echocardiographic equipment (Xario XG, Toshiba Medical Systems Co., Ltd, Tochigi, Japan). Systolic pulmonary arterial pressure (sPAP) at rest was calculated by the sum of estimated right atrial pressure and the transtricuspid gradient as previously reported [15]. Pulmonary hypertension (PH) was defined as sPAP of greater than or equal to 40 mmHg at rest [15].

Statistics
All statistical analyses were performed using JMP version 9 (SAS Institute, Cary, NC, USA). Continuous variables are expressed as mean6standard deviation. Comparisons of categorical data between two groups were performed by Fisher's exact probability tests. Continuous variables were compared with the unpaired t-test if normally distributed, and the Mann-Whitney U test when the distribution was not normal. Correlations between pairs of variables were analyzed by Spearman's rank correlation tests. For survival analysis, patients receiving lung transplantation during follow-up (n = 1) or subjects lost to follow-up (n = 10) were censored, and then the duration from entry to death, the transplant date, or the last visit was recorded. Mortality was first assessed for all risk factors using univariate Cox proportional hazard analysis, and then stepwise multivariate Cox proportional hazard analysis was performed to examine the prognostic predictive value of plasma 15-keto-dihydro PGF 2a concentrations while adjusting for other clinical predictors of cumulative mortality. Results of the regression analysis were presented in terms of relative risks (RRs) with corresponding 95% confidence intervals. Survival curves were obtained using the Kaplan-Meier method and the difference in survival rates between subgroups was calculated using a log-rank test. A p value less than 0.05 was considered to indicate statistical significance.

Plasma 15-keto-dihydro PGF 2a Concentrations Correlate with Disease Severity Indices
The clinical characteristics of patients and controls are summarised in Table 1 (see Table S1 for the patient characteristics in each hospital). The proportions of females (p = 0.03) and never smokers (p,0.001) were significantly lower in IPF patients than those in control subjects. Among all 91 patients, 16 had concurrent emphysema. Plasma concentrations of 15-keto-dihydro PGF 2a were significantly higher in IPF patients than those in control subjects, regardless of the presence of concurrent emphysema ( Figure 1, p,0.001).
To assess the clinical relevance of plasma 15-keto-dihydro PGF 2a concentrations, we investigated their relationships with disease duration, pulmonary function, arterial blood gas data and 6MWT data ( Table 2)

Plasma 15-keto-dihydro PGF 2a Concentrations Predict Mortality
During a mean follow-up period of 29.8 (range: 0.5-74.9) months, 22 of 91 patients died. The major cause of death was respiratory failure due to the progression of IPF (n = 15), 4 patients died of lung cancer and 3 patients died suddenly at home from an unknown cause. Although none of the patients received IPFspecific treatment at entry, some underwent treatment with pirfenidone (n = 13), prednisone (n = 25), cyclophosphamide (n = 3), azathioprine (n = 4) and cyclosporine (n = 6) during a follow-up period. Combination therapy with two or more medications was used in 14 patients.
We assessed the relationships between clinical measurements and mortality using Cox proportional hazard model (Table 3). Among overall patients (n = 91) as well as patients without emphysema (n = 75), low FEV 1 , low FVC, low DL CO , high CPI, short 6-minute walk distance and low end-exercise oxygen saturation were significantly associated with an increased risk of mortality. Plasma concentrations of 15-keto-dihydro PGF 2a were significantly associated with mortality (RR = 1.005, p,0.001), but sex, age and serum levels of KL-6 and SP-D were not. Multivariate analysis was performed using stepwise Cox proportional hazard model in which all significant predictors of mortality in univariate analyses were included as covariates. We found that high CPI (RR = 1.049, p = 0.002 in all patients; RR = 1.043, p = 0.016 in patients without emphysema) and increased plasma concentrations of 15-keto-dihydro PGF 2a (RR = 1.005, p = 0.002    in all patients; RR = 1.005, p = 0.003 in patients without emphysema) were the independent predictors of mortality. The independent association of plasma 15-keto-dihydro PGF 2a concentrations with mortality was reproducibly found in each cohort (see Table S2). Figure S1 shows the Kaplan-Meier survival curves when patients were divided into two groups, with plasma 15-ketodihydro PGF 2a concentrations above or below the median value (156 pg/mL), respectively. Patients with higher plasma 15-ketodihydro PGF 2a concentrations ($156 pg/mL) had a significantly worse prognosis than patients with lower concentrations (p = 0.04).

Plasma 15-keto-dihydro PGF 2a Concentrations are Associated with the Presence of Pulmonary Hypertension Assessed by Doppler Echocardiography
Finally, since PH is a significant comorbidity which can affect prognosis in patients with IPF [16,17], we compared plasma 15keto-dihydro PGF 2a concentrations according to the presence or absence of PH. Among patients with successful sPAP measurements, 8 of 28 had sPAP$40 mmHg, which fulfilled the definition of PH. Plasma concentrations of 15-keto-dihydro PGF 2a were significantly higher in patients with PH than in those without PH (3176291 pg/mL vs 172675 pg/mL, p = 0.04). However, there were no significant differences in serum levels of KL-6 and SP-D between patients with and without PH (15036444 U/mL vs 12526735 U/mL, p = 0.38; and 3676181 ng/mL vs 3476314 ng/mL, p = 0.87, respectively).

Discussion
We demonstrated raised concentrations of 15-keto-dihydro PGF 2a , a stable metabolite of PGF 2a , in the plasma of patients with IPF. Plasma concentrations of this metabolite were significantly correlated with indices of disease severity including FEV 1 , FVC, DL CO , CPI, 6-minute walk distance and end-exercise oxygen saturation, when patients with emphysema were excluded. Furthermore, in our cohort of patients, higher plasma 15-ketodihydro PGF 2a concentrations were significantly associated with an increased risk of mortality after adjusting for disease severity indices represented by CPI. Among patients with successful echocardiographic measurements, plasma 15-keto-dihydro PGF 2a concentrations were significantly higher in patients with PH than in those without PH.
Considering that IPF is a devastating lung disease with poor prognosis but highly variable clinical course, predicting which paths individual patients will take remains a central challenge for clinicians [18]. A primary finding of the present study was that plasma concentrations of PGF 2a metabolite were significantly correlated with the severity of IPF as well as mortality. However, the relationships of plasma 15-keto-dihydro PGF 2a concentrations with pulmonary function, arterial blood gas data and 6MWT data differed according to the presence of concomitant emphysema, suggesting that emphysema is a significant confounding comorbidity in IPF. FVC, DL CO , CPI and desaturation during 6MWT predicted mortality in our cohort of patients with IPF, which is consistent with previous studies [11,17,19,20]. However, even after adjusting for these physiologic predictors, plasma 15-ketodihydro PGF 2a concentrations were significantly associated with mortality. We have recently shown by both in vivo and in vitro approaches that PGF 2a -FP signaling mediates pulmonary fibrosis independently of transforming growth factor-b signaling [8], which is a well-known crucial pathway for fibrogenesis [21]. Raised concentrations of plasma PGF 2a metabolite may reflect an upregulation of PGF 2a -FP signaling pathway and subsequent fibrotic processes, which cannot be evaluated by a single measurement of pulmonary function or 6MWT. Indeed, the relationships between plasma concentration of PGF 2a metabolite and physiological parameters of IPF tended to be significant but relatively weak. Another possible explanation why plasma PGF 2a metabolite concentrations predicted mortality was the impact of comorbid PH. In subgroup analysis, plasma PGF 2a metabolite concentrations were significantly higher in patients with PH, whereas serum levels of KL-6 and SP-D did not differ between the subgroups. Since PH commonly complicates the course of IPF and potentially affects prognosis [16,17], the significant relationship of plasma PGF 2a metabolite concentrations with mortality may also reflect a higher frequency of concomitant PH in patients with higher plasma concentrations of PGF 2a metabolite.
Since PGF 2a is a locally bioactive hormone that is uncovered virtually in all tissues including lung [9], the source of PGF 2a or its metabolite in the plasma of pulmonary fibrosis patients is speculative. For example, macrophages activated by proinflammatory stimuli produce PGF 2a in large amounts [22]. PGF 2a is also produced by type II alveolar epithelial cells [23], and widely upregulated COX-2 expression was observed in metaplastic epithelium in pulmonary fibrous disorders [24]. In addition, PGF synthase exists in contractile interstitial cells of bovine lungs, which are considered to be a precursor of myofibroblasts, main contributing cell types to fibrogenesis [25]. Thus, increased epithelial permeability in the lung of IPF patients [26] may allow PGF 2a produced in the lungs to leak into the circulation, where it is instantly bioconverted through metabolism [9].
Another possible mechanism might be functional alteration in pulmonary arterial endothelium and/or smooth muscle cells, both of which play significant roles in the pathogenesis of pulmonary arterial hypertension [27]. Recent studies suggested that the pathogenesis of PH in IPF is a complicated interaction of epithelial cells, fibroblasts and vascular cells mediated by multiple factors including several growth factors [28], which can stimulate the expression of COX-2, thereby leading to the production of PGs [29]. The COX-2 pathway might be upregulated in pulmonary arterial endothelium and/or smooth muscle cells through hypoxic stimulus and/or the development of PH. This pathogenesis is worthy to be studied further because, in contrast, previously identified biomarkers of IPF, KL-6 and SP-D, were found to be mainly produced by alveolar epithelial cells [30,31].
Cytosolic phospholipase A 2 (cPLA 2 ), which cleaves phospholipids and yields arachidonic acids including PGs as well as lysophospholipids, might play a pivotal role in the pathogenesis of pulmonary fibrosis [32]. Thus, PGF 2a and lysophosphatidic acids, both derived from the breakdown of phospholipids by cPLA 2 and possible stimulators of fibrosis [8,33], might act complementarily to affect the progression of disease.
The present study has some limitations. Firstly, this is a relatively small cohort study with IPF patients alone. Since PGF 2a is produced by various cells and is implicated in the regulation of intricate pathophysiological processes [9], further studies on the utility of plasma 15-keto-dihydro PGF 2a in patients with systemic disease-associated interstitial lung diseases are warranted. At the time of blood sampling, all study participants were not receiving any specific treatment for IPF. Hence the possible effects of IPFspecific medication on the value and usefulness of plasma 15-ketodihydro PGF 2a should also be assessed for future clinical use. Secondly, since the presence and degree of PH were not evaluated by right heart catheterisation and echocardiographic data were not available in all patients, our current findings should be interpreted with caution and further studies are necessary.
In summary, we demonstrated significant associations of plasma concentrations of PGF 2a metabolites with disease severity and prognosis in IPF patients. Although further larger studies are needed to confirm the clinical utility of this metabolite, the current findings support a potential pathogenic role for PGF 2a in human IPF. Our current results also provide a more integrated understanding of this devastating and complex disease and may help in the development of new antifibrotic drugs.