Poor Outcome and Mortality in Patients with Lower Lung-Dominant Sarcoidosis

Background Pulmonary sarcoidosis predominantly affects the upper lung zones but sometimes affects the lower lung zones. We hypothesised that patients with lower lung zone-dominant sarcoidosis had lower baseline forced vital capacity, progressive restrictive lung function decline, and higher long-term mortality. Methods We retrospectively reviewed clinical data including the pulmonary function tests of 108 consecutive patients with pulmonary sarcoidosis pathologically confirmed by lung and/or mediastinal lymph node biopsy from 2004 to 2014 from our database. Results Eleven patients (10.2%) with lower lung zone-dominant sarcoidosis were compared with 97 patients with nonlower lung zone-dominant sarcoidosis. The median age of the patients with lower dominance was significantly older (71 vs. 56, p = 0.0005). The patient with lower dominance had a significantly lower baseline percent forced vital capacity (FVC) (96.0% vs. 103%, p = 0.022). The annual change in FVC was −112 mL in those with lower dominance vs. 0 mL in nonlower dominance (p = 0.0033). Fatal acute deterioration was observed in three patients (27%) in the lower dominant group. Overall survival in the lower dominant group was significantly worse. Conclusions Patients with lower lung zone-dominant sarcoidosis had an older age and lower baseline FVC with disease progression and acute deterioration associated with higher long-term mortality.


Introduction
Sarcoidosis is a systemic granulomatous disease of unknown cause that occurs in pulmonary sites in 90% of cases [1]. Pulmonary sarcoidosis usually afects the upper lung zones [2,3]; however, the lesions are sometimes distributed predominantly in the lower lung zones [2]. Te course of pulmonary sarcoidosis has been widely studied using clinical and pulmonary function follow-up [4,5], but, to our knowledge, no specifc study has been reported of sarcoidosis with lower lung zone dominance. In a previous report, we presented a patient of sarcoidosis with lower lung feld-dominant reticular shadows, who died from acute deterioration [6]. Although the autopsy fndings suggested that the sarcoidosis was complicated by usual interstitial pneumonia, we were not able to determine whether sarcoidosis was complicated by interstitial pneumonia while the patient was alive. Terefore, in a clinical setting, we have no choice but to treat lower lung zone-dominant sarcoidosis as a distinctive phenotype. We hypothesised that patients with lower lung zonedominant sarcoidosis led to decline in lung function with higher long-term mortality related to respiratory causes including acute deterioration episodes.

Subjects.
Te study analysed patients diagnosed with pulmonary sarcoidosis in the National Hospital Organization Kinki-Chuo Chest Medical Center between April 2004 and March 2014. Te diagnosis of sarcoidosis was based on a multidisciplinary diagnosis: consistent clinical features and bronchoalveolar lavage fuid analysis and confrmation of the presence of noncaseating epithelioid cell granulomas by lung and/or mediastinal lymph node biopsy, according to the WASOG (World Association of Sarcoidosis and Other Granulomatous Disorders) guidelines [7]. Tis study was approved by the Ethics Committee of the National Hospital Organization, Kinki-Chuo Chest Medical Center (approved on July 27, 2015; approval number 501).

Radiographic Assessments.
CT scanners (HiSpeed Advantage and Light-Speed 16; GE Healthcare, Milwaukee, WI, USA) were used for the present study. Tin-section CT examinations were performed at 120 kVp and 160 mAs, with 1.5 mm collimation at 15 mm intervals. Lung zone dominance was classifed into lower dominance and nonlower dominance by an expert chest physician (K. T.) and an expert chest radiologist (M. A.) without reference to the clinical or pulmonary function test results. Tey separately checked the images and discussed those cases where there was disagreement, approving the fnal radiographic assessment by consensus. Pulmonary sarcoidosis was classifed as lower dominance in chest radiography when the total extent of pulmonary lesions was predominant in the lower lung, with the lower lung being defned as below the level of the inferior pulmonary veins [8]. Te following CT patterns were recognised [9,10]: (1) nodules (defned as those with diameters <1 cm), which included centrilobular nodules, subpleural nodules, and nodules along bronchovascular bundle; (2) conglomeration, which was defned as a large opacity ≥3 cm in diameter that often surrounded and encompassed the bronchi and vessels; (3) thickening of bronchovascular bundle; (4) interlobular septal thickening; (5) air-space consolidation, which was defned as an area of increased attenuation with obscuring of the adjacent bronchial walls and vessels; (6) groundglass opacity, defned as an area of slightly increased attenuation in which the bronchial walls and vessels remained visible; (7) traction bronchiectasis, defned as irregular bronchial dilatation within areas of parenchymal abnormality; and (8) honeycombing, defned as an accumulation of cystic spaces with thickened walls. Pulmonary sarcoidosis with traction bronchiectasis, honeycombing, and cystic changes was defned as fbrotic sarcoidosis.

Assessment of Disease Progression.
Multiple pulmonary function measurements for each of the patients were available for analysis, and the information from each subject was summarised by using the estimated yearly rate of change (slope), which was calculated from a linear regression of the pulmonary function data [15,16].

Defnition of Acute Deterioration.
Te following features based on previously used criteria for acute exacerbation in idiopathic pulmonary fbrosis [17,18] were used to defne an acute deterioration event: (1) three conditions were met including (i) progressive worsening dyspnea within one month, (ii) new ground-glass opacities or consolidation evident on HRCT, and (iii) a reduction in arterial oxygen tension (PaO 2 ) at rest is of more than 10 Torr compared to previous measurements; (2) the exclusion of obvious causes of acutely impaired respiratory function, such as infection, pneumothorax, cancer, pulmonary embolism, or congestive cardiac failure.
2.6. Statistical Analysis. Diferences in clinical data between the nonlower dominance and the lower dominance were compared using the Mann-Whitney U test for continuous variables and Fischer's exact test for categorical parameters. Cox proportional hazards regression models were used to calculate hazard ratios for mortality. Survival curves were based on the Kaplan-Meier method and compared using the log-rank test. Statistical analysis was performed with JMP Statistics Program, version 11 (SAS Institute Inc., NC, USA). For all statistical analyses, p < 0.05 was considered signifcant.

Baseline Clinical Data of the Patients.
Among 139 patients identifed in the database, 31 were excluded (no granulomas were detected histologically in the lung: n � 14; incomplete medical record: n � 4; no bronchoscopic study: n � 1; CT not available: n � 1; cutaneous sarcoidosis; n � 1; and incomplete data in our hospital due to being diagnosed initially in other institutes: n � 10). Te remaining 108 patients were analysed, of whom 11 had lower dominance and 97 had nonlower dominance. At baseline, no signifcant diferences were observed in the male/female ratio or serum angiotensin-converting enzyme (ACE), whereas the lower dominant group had a signifcantly older age (71 vs. 56, p < 0.001), higher serum Krebs von den Lungen (KL-6) levels (1046 U/mL vs. 342 U/mL, p < 0.001), higher surfactant protein D (SP-D) levels (193 ng/mL vs. 53.4 ng/mL, p < 0.001), and lower baseline percent forced vital capacity (FVC) (96.0% vs. 103%, p � 0.022) ( Table 1). Table 2 shows the clinical characteristics of patients with lower lung zonedominant sarcoidosis. Representative chest X-ray and HRCT scans of a patient with lower lung zone-dominant sarcoidosis are shown in Figure 1. Existence of difuse pulmonary nodules and noncaseating granulomas raises the possibility that the patients with lower lung zone-dominant sarcoidosis might have chronic hypersensitivity pneumonitis. Te patients with lower lung zone-dominant sarcoidosis had extrapulmonary sarcoid lesions, high levels of serum ACE, noncaseating granulomas in mediastinal lymph nodes, or perilymphatic nodular pattern: nodules along the bronchovascular interstitium, interlobular septa, and subpleural lung regions (Table 2). Tese fndings therefore did not support the likelihood of chronic hypersensitivity pneumonitis. Six of 11 patients with the lower dominance had high serum B-type natriuretic peptide (BNP) or Nterminal-pro-BNP (NT-pro BNP) levels and had suspected pulmonary hypertension from their echocardiographic assessments. One of 97 patients with the nonlower dominance had high serum BNP levels and had suspected pulmonary hypertension from their echocardiographic assessment.

Radiographic and Histological Findings.
Interlobular septal thickening, ground glass opacities, and traction bronchiectasis were signifcantly more prevalent in the lower dominant group (64%, 82%, and 64%, respectively) than in the nonlower dominant group (6%, 22%, and 2%, respectively) (p < 0.001) ( Table 3). In the lower dominant group, noncaseating epithelioid cell granulomas were detected in transbronchial lung biopsy of the upper lobes in 10 of 11 patients and the lower lobes in 6 of 11 patients. Five patients had fbrosing mural alveolitis in the lower-lung zones ( Table 2). In the nonlower dominant group, noncaseating epithelioid cell granulomas were detected in transbronchial lung biopsies of the upper lobes in 41 of 97 patients, the lower lobes in 19, and lobes that were not otherwise specifed in 37; no patients had fbrosing mural alveolitis.

Outcome and Mortality.
Te median observation period of the lower dominant group, the nonlower dominant group, and both groups combined was 41 months (range 12-91 months), 58 months (range 12-140 months), and 58 months (range 12-140 months), respectively. Of the lower-dominant patients, fve (46%) were given corticosteroids and two (18%) were given immunosuppressants (one patient was given azathioprine, and the other was given cyclosporine) during the course of the disease. Te remaining six patients were given no treatment because three patients had pulmonary infections and the remaining three patients were stable. In contrast, in the nonlower dominant group, 4% of patients were given corticosteroids. Acute deterioration episodes were observed in three patients (27%) in the lower dominant group but in only one patient (1%) in the nonlower dominant group (Figure 3). Te lower dominant group showed signifcantly worse overall survival with median survival time of 60 months (p < 0.001) ( Figure 4). Mortality was caused by acute deterioration episodes (three patients) and chronic respiratory failure (one patient) in the lower dominant group and by lung cancer (one patient) in the nonlower dominant group. Annual FVC change was not signifcantly associated with overall survival (hazard ratio � 0.0301 and p � 0.0818).

Discussion
Te main results of the present study were as follows: (1) patients with lower lung zone-dominant sarcoidosis had lower baseline forced vital capacity with progressive decline in restrictive lung function decline; (2) patients with lower lung zone-dominant sarcoidosis had a signifcantly decreased survival as compared with the nonlower lung zonedominant group; and (3) mortality was mainly related to acute deterioration episodes.
Te lower lung dominance had signifcantly higher serum KL-6 and SP-D levels compared with the nonlower lung zone-dominant group. Serum KL-6 and/or SP-D can be elevated in patients with pulmonary granulomatous diseases including sarcoidosis and berylliosis [19][20][21]. Meanwhile, serum levels of KL-6 refect fbrotic processes and levels of SP-D are elevated in patients with idiopathic pulmonary fbrosis [22,23]. In the lower lung zone-dominant group, interlobular septal thickening, ground glass opacities, and traction bronchiectasis, which are putative features of fbrotic interstitial pneumonias [10], were more prevalent. Moreover, fve patients in the lower lung zone-dominant group had fbrosing alveolitis. Combined, the histopathology of lower lung dominance might include fbrosing interstitial pneumonia or lung injury.
We have presented patients with lower lung zonedominant sarcoidosis with fbrotic signs, including traction bronchiectasis in HRCT; most of them are old in age and male. In general, fbrotic interstitial lung diseases (ILDs) also involve lower lungs in elderly male patients and have poor prognosis and high mortality [24]. We reported a case of lower lung zone-dominant sarcoidosis, in which sarcoidosis was complicated by usual interstitial pneumonia [6]. Tis determination was based on the fact that the autopsy specimens showed difuse alveolar damage and no concentric fbrosis or no concentric lamellar calcifcations (Schumann bodies), which are characteristics of the fbrotic stage of sarcoidosis [25,26]. In contrast, Nobata et al. reported a case of lower lung zone-dominant sarcoidosis, in which they noted the possibility that sarcoidosis was at the fbrotic stage of sarcoidosis because the fbrosis was distributed along the bronchovascular bundles on surgical lung biopsy [3]. Only one autopsy specimen from the lower dominance was available in the present study. Further studies with larger numbers of the lower-dominant group diagnosed on surgical lung biopsy and/or autopsy are needed to determine whether sarcoidosis and another fbrotic ILD overlapped or whether the fbrosis was attributed to pulmonary sarcoidosis itself.
We showed that annual FVC change, a putative marker of disease progression, was not signifcantly associated with mortality, probably due to fatal acute deterioration in the lower dominant group. In progressive fbrosing interstitial lung disease, relative FVC decline ≧10% in the previous 24 months was shown to be associated with mortality [15].  FVC has been used to defne disease progression in recent trials on patients with progressive pulmonary fbrosis [27,28]. We also showed that the baseline FVC in the lower dominant group was signifcantly lower, and despite treatment with steroids and immunosuppressants, the yearly decline in FVC was more marked than that observed for the nonlower group. In addition, similar results were reported in sarcoid patients with pulmonary fbrosis, that is, stage IV sarcoidosis [29]. Moreover, we showed that the lower dominant group had higher mortality. We also showed that patients with fbrotic sarcoidosis in the lower dominant group had relatively higher mortality compared with those in the nonlower group, although the number of cases in both groups is limited and statistical analysis was not performed. Tus, when lower lung-dominance is seen in sarcoid patients, it necessitates careful follow-up with serial pulmonary function tests.
Recently completed randomised controlled trials have demonstrated the clinical efcacy of nintedanib, an antifbrotic agent, in patients with progressive fbrosing interstitial lung disease (PF-ILD) [28]. Moreover, in progressive pulmonary fbrosis, a conditional recommendation was made for nintedanib in the ATS/ERS/JRS/ALAT guideline [24]. Te use of antifbrotic agents might be warranted to slow down the decline in FVC and to reduce the mortality in the future.
Te present study has some limitations. First, this is a single-centre retrospective study. Despite this, we had access to all available consecutive data to maximised accurate diagnosis of pulmonary sarcoidosis with a distinct clinicopathological entity. In addition, the survival of our cohort was similar to that of other reports based on large national cohorts [5,29,30]. Furthermore, we were able to follow individual subjects longitudinally, afording us the  opportunity to detect acute deterioration of pulmonary sarcoidosis. Te second limitation is including a limited number of patients with the lower-dominance. Nonetheless, we showed signifcant statistical diferences in the annual rate of change in FVC as a disease progression measure and overall survival between the two groups. Further studies with larger numbers of the lower-dominant group are needed to confrm these results. Tird, not all patients in the lower-dominant group had surgical lung biopsy-confrmed fbrosing interstitial pneumonia in the lower-lung zones. However, transbronchial lung biopsy revealed that 5 out of 11 patients had fbrosing alveolitis in the lower-lung zones, raising the possibility that histologically, a signifcant number of patients in the lower-dominant group had fbrosing interstitial pneumonia. Finally, right heart catheterization for assessing pulmonary hypertension, which is sometimes observed in advanced pulmonary sarcoidosis or interstitial pneumonia [31][32][33][34], was not performed in our cohort. However, this would not afect the decline in FVC and the mortality because pulmonary hypertension could reduce only the DLco value and because no patients died of right-sided cardiac failure.

Conclusions
Patients with lower lung zone-dominant sarcoidosis had an older age and lower baseline FVC with disease progression and acute deterioration associated with higher long-term mortality related to respiratory causes, particularly acute deterioration. Further studies are warranted to determine whether sarcoidosis and another fbrotic ILD overlapped or whether the fbrosis was caused by pulmonary sarcoidosis itself.

Data Availability
Te data used to support the fndings of this study can be obtained from the corresponding author upon reasonable request.