Advertisement

Predictors of mortality in fibrosing pulmonary sarcoidosis

Open ArchivePublished:May 11, 2020DOI:https://doi.org/10.1016/j.rmed.2020.105997

      Highlights

      • Mortality in stage 4 sarcoidosis is associated with a CPI>40, fibrosis extent> 20% and pulmonary hypertension.
      • Patients native from Sub-Saharan Africa or French West Indies had a higher risk of mortality.
      • Replacing MPAD/AAD by MPAD/BSA in Walsh algorithm allows to accurately predict mortality.

      Abstract

      Introduction

      Pulmonary fibrosing sarcoidosis is associated with increased mortality. This study was aimed to explore the prognosis value of a panel of parameters for predicting mortality.

      Methods

      This retrospective study included 216 patients with confirmed stage 4 pulmonary sarcoidosis. Stage 4 diagnosis date served as baseline. The following information was systematically present at baseline: epidemiological characteristics; treatments; pulmonary function; composite physiologic index (CPI); systolic pulmonary artery pressure at echocardiography; pulmonary fibrosis extent, main pulmonary artery/ascending aorta diameters ratio (MPAD/AAD) and MPAD/body surface area (BSA) measured and calculated using computed tomography, Walsh's algorithm based on CPI, lung fibrosis extent and MPAD/AAD ratio, and modified Walsh's algorithm with MPAD/BSA replacing MPAD/AAD allowed to estimate good or bad prognosis profiles. The primary outcome of the study was all cause mortality and lung transplantation. The value of baseline parameters was tested as predictors of mortality using univariate and multivariate analyses.

      Results

      Median follow-up was 105 months. There were 41 deaths and 5 transplantations. At multivariate analysis, survival was independently predicted by several parameters including CPI, lung fibrosis extent, pulmonary hypertension at echography or MPAD/BSA ratio, Walsh's algorithm, and geographic origin. The modified Walsh's algorithm was most highly predictive.

      Conclusion

      Survival was best predicted by geographic origin, lung fibrosis extent, PH at echography or MPAD/BSA ratio, as well as by various scores among them the modified Walsh's algorithm had very high predictive value thanks to MPAD/BSA ratio which accurately predicted mortality.

      Keywords

      1. Introduction

      A higher risk of mortality has been evidenced in sarcoidosis in several studies [
      • Rossides M.
      • Kullberg S.
      • Askling J.
      • Eklund A.
      • Grunewald J.
      • Arkema E.V.
      Sarcoidosis mortality in Sweden: a population-based cohort study.
      ,
      • Jamilloux Y.
      • Maucort-Boulch D.
      • Kerever S.
      • Gerfaud-Valentin M.
      • Broussolle C.
      • Eb M.
      • et al.
      Sarcoidosis-related mortality in France: a multiple-cause-of-death analysis.
      ,
      • Swigris J.J.
      • Olson A.L.
      • Huie T.J.
      • Fernandez-Perez E.R.
      • Solomon J.
      • Sprunger D.
      • et al.
      Sarcoidosis-related mortality in the United States from 1988 to 2007.
      ,
      • Reich J.M.
      Mortality of intrathoracic sarcoidosis in referral vs population-based settings: influence of stage, ethnicity, and corticosteroid therapy.
      ,
      • Park J.E.
      • Kim Y.S.
      • Kang M.J.
      • Kim C.J.
      • Han C.H.
      • Lee S.M.
      • et al.
      Prevalence, incidence, and mortality of sarcoidosis in Korea, 2003-2015: a nationwide population-based study.
      ,
      • Gribbin J.
      • Hubbard R.B.
      • Le Jeune I.
      • Smith C.J.P.
      • West J.
      • Tata L.J.
      Incidence and mortality of idiopathic pulmonary fibrosis and sarcoidosis in the UK.
      ]. An excessive mortality rate does not apply to every sarcoidosis patient but concerns a subgroup of patients. This point has been very clearly shown in Rossides’ study where mortality was increased in patients requiring sarcoidosis-specific treatment in the first 3 months while other patients had a normal survival [
      • Rossides M.
      • Kullberg S.
      • Askling J.
      • Eklund A.
      • Grunewald J.
      • Arkema E.V.
      Sarcoidosis mortality in Sweden: a population-based cohort study.
      ]. In Western studies, death was often found associated to a significantly deteriorated pulmonary function [
      • Baughman R.P.
      • Winget D.B.
      • Bowen E.H.
      • Lower E.E.
      Predicting respiratory failure in sarcoidosis patients.
      ,
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ], pulmonary fibrosis [
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ], and pre-capillary pulmonary hypertension [
      • Baughman R.P.
      • Engel P.J.
      • Taylor L.
      • Lower E.E.
      Survival in sarcoidosis-associated pulmonary hypertension: the importance of hemodynamic evaluation.
      ,
      • Boucly A.
      • Cottin V.
      • Nunes H.
      • Jaïs X.
      • Tazi A.
      • Prévôt G.
      • et al.
      Management and long-term outcomes of sarcoidosis-associated pulmonary hypertension.
      ,
      • Mirsaeidi M.
      • Machado R.F.
      • Schraufnagel D.
      • Sweiss N.J.
      • Baughman R.P.
      Racial difference in sarcoidosis mortality in the United States.
      ] while cardiac sarcoidosis, particularly in Japan [
      • Iwai K.
      • Sekiguti M.
      • Hosoda Y.
      • DeRemee R.A.
      • Tazelaar H.D.
      • Sharma O.P.
      • et al.
      Racial difference in cardiac sarcoidosis incidence observed at autopsy.
      ,
      • Yazaki Y.
      • Isobe M.
      • Hiroe M.
      • Morimoto S.
      • Hiramitsu S.
      • Nakano T.
      • et al.
      Prognostic determinants of long-term survival in Japanese patients with cardiac sarcoidosis treated with prednisone.
      ] plays also a role [
      • Zhou Y.
      • Lower E.E.
      • Li H.-P.
      • Costea A.
      • Attari M.
      • Baughman R.P.
      Cardiac sarcoidosis: the impact of age and implanted devices on survival.
      ]. The long-term use of treatments like corticosteroids, as shown in other chronic diseases like in rheumatoid arthritis [
      • Chester Wasko M.
      • Dasgupta A.
      • Ilse Sears G.
      • Fries J.F.
      • Ward M.M.
      Prednisone use and risk of mortality in patients with rheumatoid arthritis: moderation by use of disease-modifying antirheumatic drugs.
      ,
      • del Rincón I.
      • Battafarano D.F.
      • Restrepo J.F.
      • Erikson J.M.
      • Escalante A.
      Glucocorticoid dose thresholds associated with all-cause and cardiovascular mortality in rheumatoid arthritis.
      ,
      • Movahedi M.
      • Costello R.
      • Lunt M.
      • Pye S.R.
      • Sergeant J.C.
      • Dixon W.G.
      Oral glucocorticoid therapy and all-cause and cause-specific mortality in patients with rheumatoid arthritis: a retrospective cohort study.
      ] or chronic obstructive pulmonary disease (COPD) might also increase mortality [
      • Schols A.M.
      • Wesseling G.
      • Kester A.D.
      • de Vries G.
      • Mostert R.
      • Slangen J.
      • et al.
      Dose dependent increased mortality risk in COPD patients treated with oral glucocorticoids.
      ]. Comorbidities could also have an impact on mortality since the burden of comorbidities might be increased in sarcoidosis patients, particularly coronary artery disease, congestive heart disease, stroke, depression, diabetes and osteoporotic fractures, all conditions which per se have been associated to mortality risk [
      • Ungprasert P.
      • Matteson E.L.
      • Crowson C.S.
      Increased risk of multimorbidity in patients with sarcoidosis: a population-based cohort study 1976 to 2013.
      ,
      • Nowiński A.
      • Puścińska E.
      • Goljan A.
      • Peradzynska J.
      • Bednarek M.
      • Korzybski D.
      • et al.
      The influence of comorbidities on mortality in sarcoidosis: a observational prospective cohort study.
      ].
      Several factors may be simultaneously present in a same sarcoidosis patient (e.g., marked pulmonary function impairment, pulmonary fibrosis, pulmonary hypertension, cardiac sarcoidosis or long-term corticosteroid treatment). Thus, there is a need for determining exact elementary factors independently predictive of mortality. This is mandatory to identify as early as possible patients with high risk pulmonary sarcoidosis. Recently, predictors of mortality in pulmonary sarcoidosis were investigated in two large cohorts from referral centre populations [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. Both studies showed that an extent of pulmonary fibrosis higher than 20% on thoracic high resolution computed tomography (HRCT) and pulmonary hypertension (PH) were independent predictive factors of mortality. PH was assessed using different means -one measuring the main pulmonary artery diameter to ascending aorta diameter ratio (MPAD/AAD) [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ], and the other right heart catheterisation- [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. Contradictory results were found for composite physiologic index (CPI) a weighted index of pulmonary function variables that correlates with the extent of interstitial disease on HRCT [
      • Wells A.U.
      • Desai S.R.
      • Rubens M.B.
      • Goh N.S.L.
      • Cramer D.
      • Nicholson A.G.
      • et al.
      Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed by computed tomography.
      ]. An algorithm combining CPI and the extent of pulmonary fibrosis on HRCT and MPAD/AAD ratio (“Walsh's algorithm”) [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ] or right catheterisation [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ] was predictive of mortality in both studies, but only in univariate analysis in Kirkil's study [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. Moreover, age was predictive in multivariate analysis and being black or having stage 4 chest radiographs in univariate analysis only in the Kirkil's study [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ].
      The aim of our study was to assess, in an extra cohort made of stage 4 sarcoidosis in a tertiary centre for sarcoidosis in Greater Paris (France), the value of predicting mortality factors including those tested in Walsh's and Kirkil's studies [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ], particularly the algorithm defined by Walsh but also some other factors like PH estimated through echocardiography, frequently used in practice, and MPAD/body surface area ratio (BSA), which has been shown a better surrogate of PH measured by right catherization in sarcoidosis [
      • Huitema M.P.
      • Spee M.
      • Vorselaars V.M.M.
      • Boerman S.
      • Snijder R.J.
      • van Es H.W.
      • et al.
      Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
      ] Moreover, comorbidities and treatment were also investigated. As many sarcoidosis patients with increased mortality risk have pulmonary fibrosis [
      • Baughman R.P.
      • Winget D.B.
      • Bowen E.H.
      • Lower E.E.
      Predicting respiratory failure in sarcoidosis patients.
      ,
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ], the study was focused on a group of stage 4 sarcoidosis patients.

      2. Methods

      2.1 Patients

      This retrospective study was conducted in Pulmonary Department of Avicenne universitary hospital. The study received institutional review board approval (CLEA-2016-036). The requirement for signed informed consent was waived according to French legislation. The database of patients admitted to the pulmonary department with confirmed stage 4 sarcoidosis was searched for the 1988–2017 period. A minor part of the patients studied had been recruited for previous studies with a different design [
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ,
      • Uzunhan Y.
      • Nunes H.
      • Jeny F.
      • Lacroix M.
      • Brun S.
      • Brillet P.-Y.
      • et al.
      Chronic pulmonary aspergillosis complicating sarcoidosis.
      ]. Diagnosis of sarcoidosis was confirmed according to the 1999 statement on sarcoidosis [
      Statement on sarcoidosis. Joint statement of the American thoracic society (ATS), the European respiratory society (ERS) and the world association of sarcoidosis and other granulomatous disorders (WASOG) adopted by the ATS board of directors and by the ERS executive committee.
      ] and evidence of noncaseating granulomas was required. All patients included in the study had a pulmonary function test, chest HRCT and an echocardiography with an estimation of systolic pulmonary arterial pressure available at the time of stage 4 diagnosis.

      2.2 Epidemiological data and investigations at baseline

      Stage 4 diagnosis date was chosen as baseline. The following information was collected at baseline: age, sex, geographical origin, body mass index (BMI), body surface area (BSA) calculated from the simplified formula of Mosteller [
      • Mosteller R.D.
      Simplified calculation of body-surface area.
      ] BSA(m2)=Height(cm)xWeight(kg)3600, the presence or not of diabetes mellitus, cardiac sarcoidosis, treatment at baseline consisting of corticosteroids (<or >10 mg/day prednisone equivalent) or its combination with an immunosuppressant therapy. Smoking status was registered as “current smoker”, “ever smoker” and “never smoker”. “Ever smoker” applied when patients smoked more than one cigarette a day for more than 1 year.
      The following data were also collected at baseline: forced vital capacity (FVC), forced expiratory volume in 1s (FEV1) and diffusing capacity of the lung for carbon monoxide (DLCO) were expressed as % of predicted values. We calculated the CPI with the following formula: CPI=91(0,65×percentpredictedDLCO)(0,53×percentpredictedFVC)+(0,34×percentpredictedFEV1) for every patient [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ].
      PH was defined by estimated systolic pulmonary artery pressure >35 mmHg on doppler echocardiography.
      Chest radiography and HRCT were reviewed in concert by two radiologists (PYB and ML with a 15 and 3 years’ experience on interstitial pneumonia imaging respectively). CT lung fibrosis extent was determined at 3 levels (upper, middle, lower part of the lungs) on both lungs according to Sumikawa [
      • Sumikawa H.
      • Johkoh T.
      • Colby T.V.
      • Ichikado K.
      • Suga M.
      • Taniguchi H.
      • et al.
      Computed tomography findings in pathological usual interstitial pneumonia: relationship to survival.
      ]. The extent of fibrosis was rounded to the nearest 5%. Main pulmonary artery diameter (MPAD) and ascending aorta diameter (AAD) were assessed on HRCT. We also calculated MPAD/BSA ratio known to be also predictive of PH in sarcoidosis [
      • Huitema M.P.
      • Spee M.
      • Vorselaars V.M.M.
      • Boerman S.
      • Snijder R.J.
      • van Es H.W.
      • et al.
      Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
      ]. MPAD/AAD >1 and MPAD/BSA>16 were used as cut-off according to previous study [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ,
      • Huitema M.P.
      • Spee M.
      • Vorselaars V.M.M.
      • Boerman S.
      • Snijder R.J.
      • van Es H.W.
      • et al.
      Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
      ].
      The Walsh's algorithm [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ] was applied (Fig. 1). CPI higher than 40 allowed to directly identify patients with a poor prognosis. Patients with CPI ≤40 were classified with a poor prognosis if they had on HRCT either MPAD/AAD>1, and/or the presence of pulmonary fibrosis >20%. All other patients were classified with a good prognosis. A “modified Walsh's algorithm” derived from Walsh's algorithm used MPAD/BSA>16 ratio instead of MPAD/AAD>1 ratio (Fig. 2).
      Fig. 1
      Fig. 1Flow chart of patient selection and their classification according to Walsh's algorithm
      Abbreviations: PFT pulmonary function test; HRCT: High resolution of Computed tomography, CPI: composite physiologic index; MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter.
      Fig. 2
      Fig. 2Flow chart of patient selection and their classification according to modified Walsh's algorithm
      Abbreviations: PFT pulmonary function test; HRCT: High resolution of Computed tomography, CPI: composite physiologic index; MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter.

      2.3 Outcomes

      The primary outcome was all cause mortality and lung transplantation. We calculated the survival period from baseline to the date of death or lung transplantation, or in the case of survivors to the last known contact. Vital status was known for all patients at the end of the study period.

      2.4 Statistical analysis

      Data were given as means (SD), medians (Q1; Q3) or as proportions expressed in percentages where appropriate. Mortality was assessed using a multivariate Cox proportional hazards model, with adjustments for age and gender (forced-in covariates). The proportional hazards assumption was checked using statistical tests based on the scaled Schoenfeld residuals. To avoid collinearity, different models were elaborated: Walsh's algorithms (original and modified) were analysed separately and apart CPI, extent of fibrosis or pulmonary hypertension assessment. PH at echography was analysed separately from MPAD/BSA, also because of collinearity. Candidate covariates were included for multivariate analyses if p value was <0.2. Covariates were selected with backward elimination. Survival curves were based on the Kaplan–Meier method. Analysis was performed using STATA (version 13, StataCorp, College Station, Texas). P-values <0,05 were considered statistically significant.

      3. Results

      3.1 Flow chart and population characteristics

      Fig. 1 shows the flow chart with a classification of patients at baseline according to the Walsh's algorithm while Fig. 2 shows the classification of patients at baseline according to the modified Walsh's algorithm (Fig. 1, Fig. 2). We identified 298 stage 4 sarcoidosis patients referred to Pulmonology Department in Avicenne Hospital investigated from 1988 to 2016. Eighty-two patients were not included: 47 had no CT, 9 had no echocardiography, 12 had no CT and no echocardiography, 7 had no pulmonary function test and stage 4 was not confirmed in 7 patients by radiologists involved in the study. Thus, 216 patients made up the final study population. Thirty-five patients had never been treated, 38 were treated by prednisone with a daily dose lower than 10 mg, 79 were treated with a daily dose higher than 10 mg, 60 took immunosuppressive drugs while treatment was unknown for 4 patients.
      Table 1 presents patients’ characteristics collected at baseline. Forty-one patients died and 5 underwent lung transplantation. Survival at 1, 5, 10 and 15 years from baseline were respectively at 98.1%, 94.4%, 83.5% and 75.4%.
      Table 1Patients’ characteristics at stage 4 sarcoidosis diagnosis (n = 216).
      Age46 (37–54)
      Sex (male)118 (54.6%)
      Geographical origin
      Europe113 (52.3%)
      Sub-Saharan Africa or French West Indies57 (26.4%)
      North Africa39 (18%)
      Asia5 (2.3%)
      Not stated2 (0.9%)
      Number of deaths/lung transplantation41 (19%)/5 (2.3%)
      Treatment
      No treatment35 (16.2%)
      Prednisone ≤ 10 mg/day38 (17.6%)
      Prednisone > 10 mg/day79 (36.6%)
      Immunosuppressive drugs60 (27.8%)
      Not stated4 (1.9%)
      Smoking status
      Never133 (61.6%)
      Ever58 (26.8%)
      Current19 (8.8%)
      Not stated6 (2.8%)
      Diabetes35 (16.2%)
      BMI23.9 (21.2; 27.2)
      Follow up (time in months)105 (64.5; 187)
      Cardiac sarcoidosis30 (13.9%)
      Pulmonary function
      FVC % predicted69.2 (19.9)
      FEV1% predicted61.3 (18.8)
      DLCO % predicted52.8 (18.9)
      CPI41 (29; 54)
      O2 supplementation23 (10.6%)
      PH at echography38 (17.6%)
      HRCT
      Extent of fibrosis (% of total lung)15 (10; 25)
      MPAD/AAD1 (0.9; 1.1)
      MPAD/BSA16,5 (14.7; 18.9)
      Data are presented as n (%) or mean (sd) or median (Q1; Q3). FEV1: forced expiratory volume in 1 s; FVC: forced vital capacity; DLCO: diffusing capacity of the lung for carbon monoxide; O2: oxygen; CPI: composite physiologic index (91.0−(0.65 × DLCO)−(0.53 × FVC)+(0.34 × FEV1); MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter; BSA: body surface area; HRCT: High resolution computed tomography; BMI Body mass index.
      Median follow up was around 8 years. Of note, patients native from Sub-Saharan Africa or French West Indies had a higher mortality risk than other patients of the study despite a trend to be younger (see electronic supplement Table 1). They also had more often diabetes mellitus. Moreover, they had a higher CPI score, a higher pulmonary fibrosis extent, while PH and O2 supplementation were no more frequent than in other patients. A higher proportion of them also had a bad prognosis according to Walsh's algorithm (see electronic supplement Table 1).

      3.2 Univariate analysis

      Univariate cox regression survival analysis for “all cause” mortality was performed for different explanatory variables and are exhibited in Table 2.
      Table 2Univariate cox-regression survival analysis for all-cause mortality.
      VariableHR95% CIP value
      Age (years)1.021.00; 1.040.016
      Sex (male)1.410.78; 2.550.249
      Geographical origin
      Europe0.720.40; 1.290.27
      Sub-Saharan Africa or FWI2.121.18; 3.800.012
      North Africa0.510.2; 1.290.16
      Treatment0.820.6; 1.110.20
      Smoker (ever and current)0.820.42; 1.430.41
      Diabetes1.830.87; 3.830.11
      BMI0.950.88; 1.020.14
      Cardiac sarcoidosis1.200.47; 3.000.71
      O2 supplementation5.072.64; 9.720.001
      Fibrosis > 20%2.801.49; 5.280.001
      PH on echocardiography3.421.89; 6.200.001
      MPAD/AAD > 11.490.83; 2.670.18
      MPAD/BSA (mm.m-2) > 162.611.26; 5.420.01
      CPI > 403.781.87; 7.640.001
      Walsh poor prognosis5.541.71; 17.940.004
      Modified Walsh poor prognosis*11.001.51; 80.060.02
      Abbreviations: HR: Hazard ratio; CI confidence interval; O2: oxygen; CPI: composite physiologic index; MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter; BSA: body surface area. FWI: French West Indies, BMI Body mass index, PH: pulmonary hypertension; * modified Walsh's algorithm” by integrating MPAD/BSA>16 instead of MPAD/AAD>1.
      Being native from Sub-Saharan Africa or French West Indies, oxygen supplementation, extent of pulmonary fibrosis >20%, PH at echocardiography, and the different prognostic scores (CPI, Walsh's algorithm and modified Walsh's algorithm) were all predictive of mortality. However, other tested parameters, including sex, tobacco smoking, diabetes mellitus, sarcoidosis treatment at baseline or associated cardiac sarcoidosis were not found predictive of mortality. The proportional hazards assumption was not satisfied for age, making this factor uninterpretable in cox analysis.
      The MPAD/BSA ratio was a mortality prognostic factor as well as PH at echocardiography while MPAD/AAD ratio was not. Median MPAD/BSA ratio was significantly higher in patients with echocardiography PH versus patients without (19.36 versus 16.26 p < 0,0001) (Mann Whitney test); median MPAD/AAD ratio was also significantly higher for patients with PH versus patients without (1,06 versus 1) (p < 0,001) (Mann Whitney test)). The previous cut-offs described by Walsh et al. [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ] for CPI> 40; MPAD/AAD> 1 and extent of fibrosis >20% were applied in our population in a log rank analysis (Fig. 3, Fig. 4) as other binary variables. Patients with a bad prognosis according to Walsh's algorithm or to a “modified Walsh's algorithm” by integrating MPAD/BSA>16 instead of MPAD/AAD>1 had a lower survival than those with a good prognosis (Fig. 4).
      Fig. 3
      Fig. 3Comparison of survival for various factors represented by Kaplan Meier curve (P represented the result of log rank test)
      MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter; BSA: body surface area. PH: Pulmonary hypertension on echocardiography defined by pulmonary artery pressure estimation>35 mmHg.
      Fig. 4
      Fig. 4Comparison of survival for scoring represented by Kaplan Meier curve
      P represented the result of log rank test. CPI Composite physiologic index. Walsh's algorithm according to Ref. [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ]. Modified Walsh's algorithm, used MPAD/BSA ratio instead of MPAD/AAD ratio.

      3.3 Multivariate analysis

      The different models of multivariate analysis are shown in Table 3. To avoid collinearity CPI, Walsh's algorithm and modified Walsh's algorithm results were analysed separately.
      Table 3Multivariate cox regression survival analysis for all-cause mortality in different models.
      VariableHR%95 CIP value
      Model 1Walsh poor prognosis13.63.09; 59.25<0.001
      Age1.051.02; 1.07<0.001
      Sex (Male)1.410.77; 2.560.26
      Sub-Saharan Africa or FWI2.291.25; 4.180.007
      Model 2Modified Walsh poor prognosis*514.42; 561.30.001
      Age1.061.03; 1.09<0.001
      Sex (Male)1.590.87; 2.900.15
      Sub-Saharan Africa or FWI2.521.35; 4.710.04
      Model 3Age1.041.02; 1.060.001
      Sex (Male)1.710.92; 3.140.09
      CPI>403.311.51; 7.260.003
      PH at echocardiography2.221.15; 4.280.02
      Fibrosis >20%1.981.02; 3.840.04
      Abbreviations: HR: Hazard ratio; CI confidence interval, CPI: composite physiologic index; * modified Walsh's algorithm” by integrating MPAD/BSA>16 instead of MPAD/AAD>1; MPAD: mean pulmonary artery diameter; AAD: ascending aorta diameter; BSA: body surface area FWI: French West Indies; PH: Pulmonary hypertension at echography defined by pulmonary artery pressure estimation>35 mmHg
      The independent factors predictive of mortality in model 1 and 2 were Sub-Saharan Africa or French West Indies origin, Walsh poor prognosis and modified Walsh poor prognosis algorithm. Modified Walsh algorithm was highly predictive with a hazard ratio of 51 (IC 95% (4.42; 561.3) p = 0.001).
      In model 3, PH at echocardiography (or according to MPAD/BSA>16 see electronic supplement Table n°2), CPI > 40, and fibrosis>20% were independent risk factors of mortality.
      All these factors were predictive of mortality independently from age and sex, nevertheless the proportional hazards assumption was not satisfied for age.

      4. Discussion

      Attention is paid to advanced pulmonary sarcoidosis with important consequences on treatment. Following two important studies evidencing predictive factors of mortality in pulmonary sarcoidosis [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ], we found interesting to look for predictive factors including most of those explored in these studies but also other ones including: (i) evaluation of pulmonary hypertension by echocardiography, largely used in clinical practice (ii) MPAD/BSA ratio that has recently been shown very promising for predicting PH at right catheterisation [
      • Huitema M.P.
      • Spee M.
      • Vorselaars V.M.M.
      • Boerman S.
      • Snijder R.J.
      • van Es H.W.
      • et al.
      Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
      ], and (iii) factors possibly involved in mortality risk like sarcoidosis treatments, associated cardiac sarcoidosis or diabetes mellitus [
      • Baughman R.P.
      • Winget D.B.
      • Bowen E.H.
      • Lower E.E.
      Predicting respiratory failure in sarcoidosis patients.
      ,
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ].
      Our study showed that: (i) CPI>40, pulmonary fibrosis extent> 20% at HRCT and PH at echocardiography or using MPAD/BSA ratio were independent predictors of mortality; (ii) Walsh's algorithm was confirmed predictive of mortality; (iii) strikingly, modified Walsh's algorithm taking into account MPAD/BSA instead of MPAD/AAD was highly predictive of mortality, with a higher hazard ratio and this was linked to the fact that the MPAD/BSA ratio was a good predictor of mortality while the MPAD/AAD ratio was not; (iv) patients native from Sub-Saharan Africa or French West Indies had a higher risk of mortality; (v) all other parameters, including sarcoidosis treatment, associated cardiac sarcoidosis and diabetes mellitus were not predictive.
      Different approaches are used to explore mortality in sarcoidosis: autopsy studies [
      • Iwai K.
      • Sekiguti M.
      • Hosoda Y.
      • DeRemee R.A.
      • Tazelaar H.D.
      • Sharma O.P.
      • et al.
      Racial difference in cardiac sarcoidosis incidence observed at autopsy.
      ], centre cohort studies [
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ,
      • Ungprasert P.
      • Matteson E.L.
      • Crowson C.S.
      Increased risk of multimorbidity in patients with sarcoidosis: a population-based cohort study 1976 to 2013.
      ,
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ], nation-wide register-based studies [
      • Rossides M.
      • Kullberg S.
      • Askling J.
      • Eklund A.
      • Grunewald J.
      • Arkema E.V.
      Sarcoidosis mortality in Sweden: a population-based cohort study.
      ] and death-certificate based studies [
      • Jamilloux Y.
      • Maucort-Boulch D.
      • Kerever S.
      • Gerfaud-Valentin M.
      • Broussolle C.
      • Eb M.
      • et al.
      Sarcoidosis-related mortality in France: a multiple-cause-of-death analysis.
      ,
      • Swigris J.J.
      • Olson A.L.
      • Huie T.J.
      • Fernandez-Perez E.R.
      • Solomon J.
      • Sprunger D.
      • et al.
      Sarcoidosis-related mortality in the United States from 1988 to 2007.
      ,
      • Mirsaeidi M.
      • Machado R.F.
      • Schraufnagel D.
      • Sweiss N.J.
      • Baughman R.P.
      Racial difference in sarcoidosis mortality in the United States.
      ]. The epidemiology of sarcoidosis is very variable according to geography, gender and race [
      • Dumas O.
      • Abramovitz L.
      • Wiley A.S.
      • Cozier Y.C.
      • Camargo C.A.
      Epidemiology of sarcoidosis in a prospective cohort study of U.S. Women.
      ,
      • Duchemann B.
      • Annesi-Maesano I.
      • Jacobe de Naurois C.
      • Sanyal S.
      • Brillet P.-Y.
      • Brauner M.
      • et al.
      Prevalence and incidence of interstitial lung diseases in a multi-ethnic county of Greater Paris.
      ,
      • Arkema E.V.
      • Grunewald J.
      • Kullberg S.
      • Eklund A.
      • Askling J.
      Sarcoidosis incidence and prevalence: a nationwide register-based assessment in Sweden.
      ], so are mortality rate and causes [
      • Park J.E.
      • Kim Y.S.
      • Kang M.J.
      • Kim C.J.
      • Han C.H.
      • Lee S.M.
      • et al.
      Prevalence, incidence, and mortality of sarcoidosis in Korea, 2003-2015: a nationwide population-based study.
      ,
      • Mirsaeidi M.
      • Machado R.F.
      • Schraufnagel D.
      • Sweiss N.J.
      • Baughman R.P.
      Racial difference in sarcoidosis mortality in the United States.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ,
      • Dumas O.
      • Abramovitz L.
      • Wiley A.S.
      • Cozier Y.C.
      • Camargo C.A.
      Epidemiology of sarcoidosis in a prospective cohort study of U.S. Women.
      ]. In our study, natives from Sub-Saharan Africa or French West Indies had an increased sarcoidosis-related risk of mortality compared to grouped other patients, most of them born in Europe. This result is in line with Kirkil's and with Mirsaeidi's studies evidencing a higher mortality risk in African Americans than in Caucasians, even though contrary to Mirsaeidi patients native from Sub-Saharan Africa or French West Indies had no trend to have more PH than others in our series [
      • Mirsaeidi M.
      • Machado R.F.
      • Schraufnagel D.
      • Sweiss N.J.
      • Baughman R.P.
      Racial difference in sarcoidosis mortality in the United States.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. In our study, 35% of Sub-Saharan Africans or French West Indian died or were transplanted vs 16.3% of other patients. At baseline, they tended to be younger, less frequently smokers and had more severe pulmonary fibrosis with both higher fibrosis extent percentage and higher CPI but the proportion of patients with PH, according either to echocardiography or to MPAD/BSA ratio was not significantly different. A worse prognosis predictor using Walsh's algorithm was also evidenced in this group of patients.
      To be noted, we did not find a predictive value for mortality when considering baseline treatment or cardiac sarcoidosis. The rarity of cardiac sarcoidosis might explain this negative result. The adverse impact of treatments, particularly corticosteroids might have been underestimated. However, death in patients with advanced pulmonary sarcoidosis is usually linked to underlying sarcoidosis [
      • Nardi A.
      • Brillet P.-Y.
      • Letoumelin P.
      • Girard F.
      • Brauner M.
      • Uzunhan Y.
      • et al.
      Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
      ,
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. The evidence of an increased mortality risk due to long-term corticosteroid treatment observed in rheumatoid arthritis and COPD is not necessarily transposable to sarcoidosis but deserves further investigation on larger series with a longer follow up.
      Our study patients were recruited in a sarcoidosis tertiary referral centre in Greater Paris, France. Compared to those in Walsh's and Kirkil's studies, our patients had more advanced pulmonary disease: mean CPI was 41 instead of 34.3; Oxygen supplementation concerned 10.6% instead of 8%; a 18.3% fibrosis extent was observed instead of 15%, and 21.3% died instead of 20% in Walsh'study. In Kirkil's study, 8.4% of patients died. Interestingly, in Kirkil's study, despite a higher mortality in radiological stage 4, some patients at other radiological stages, particularly stage 3, died. However, these two studies used as baseline the diagnosis of sarcoidosis to calculate survival, and not the stage 4 diagnosis as in our study, making a methodological difference for comparing survival.
      The impact of PH on mortality was confirmed in our study, interestingly while we used another approach: in Walsh's study PH diagnosis was based on MPAD/AAD ratio [
      • Walsh S.L.
      • Wells A.U.
      • Sverzellati N.
      • Keir G.J.
      • Calandriello L.
      • Antoniou K.M.
      • et al.
      An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
      ], while it relied on right catheterisation in Kirkil's study [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. In our study, it was evidenced through echocardiography and MPAD/BSA ratio. Interestingly, in our cohort MPAD/BSA ratio, well correlated to echocardiography, was predictive of mortality while MPAD/AAD ratio despite also correlated to PH at echography was not. It is in line with the study by Huitema et al. [
      • Huitema M.P.
      • Spee M.
      • Vorselaars V.M.M.
      • Boerman S.
      • Snijder R.J.
      • van Es H.W.
      • et al.
      Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
      ] indicating that MPAD/BSA ratio might be the most accurate predictor of PH on CT in patients with pulmonary sarcoidosis.
      When using models for multivariate survival analysis, the modified Walsh's algorithm Hazard ratio was found very impressive. It must be emphasized that it is an easy-practice tool. Thus, the measure of MPAD/BSA might be recommended in clinical practice allowing to improve Walsh's algorithm. It might help the detection of patients more at risk and help consider patients for transplantation. Several patients who died from pulmonary sarcoidosis might have benefited from this algorithm leading to a lung transplantation all the more as they were often in the age range making it possible [
      • Taimeh Z.
      • Hertz M.I.
      • Shumway S.
      • Pritzker M.
      Lung transplantation for pulmonary sarcoidosis. Twenty-five years of experience in the USA.
      ]. Moreover, the opportunity to select patients with sarcoidosis pulmonary fibrosis and a bad prognosis on modified Walsh's algorithm might facilitate the design of new treatment trials, for example with antifibrotic drugs.
      There were several limitations in our study. It was retrospective and monocentric.
      Only patients with stage 4 patients were studied. This condition is associated to increased mortality but it has been shown that an increased risk of mortality may also be encountered in some stage 3 patients particularly those with associated PH [
      • Kirkil G.
      • Lower E.E.
      • Baughman R.P.
      Predictors of mortality in pulmonary sarcoidosis.
      ]. While pulmonary hypertension is most often associated with pulmonary fibrosis in sarcoidosis, a significant proportion of cases with pulmonary hypertension may have less advanced radiologic changes. In the multi-national registry for sarcoidosis associated pulmonary hypertension nearly 25% of patients had a radiological stage 2 or 3 [
      • Shlobin O.A.
      • Kouranos V.
      • Barnett S.D.
      • Alhamad E.H.
      • Culver D.A.
      • Barney J.
      • et al.
      Physiological predictors of survival in patients with sarcoidosis associated pulmonary hypertension: results from an international registry.
      ].
      The diagnosis of pulmonary hypertension in our study was not determined by right heart catherization, an invasive procedure, but only estimated with a systolic pulmonary artery pressure>35 mmHg at echocardiography. Measurement of systolic pulmonary artery pressure was not possible in some cases, and the diagnosis of pulmonary hypertension on echography may have been underestimated. Similarly, we did not take into account the presence of indirect signs of pulmonary hypertension such as right ventricle dilatation.
      As our assessment of treatments at baseline was retrospective, some information like cumulative doses of prednisone were not available, preventing us from estimating with accuracy the impact of treatments on mortality. Finally, some points were not investigated like personal income or occupational and environmental risks which might have given interesting information and might have impacted the mortality risk [
      • Harper L.J.
      • Gerke A.K.
      • Wang X.-F.
      • Ribeiro Neto M.L.
      • Baughman R.P.
      • Beyer K.
      • et al.
      Income and other contributors to poor outcomes in US sarcoidosis patients.
      ].

      5. Conclusion

      This study on radiological stage 4 sarcoidosis clearly showed that mortality could be predicted in patients with CPI>40, lung fibrosis extent >20% and PH. High MPAD/BSA ratio at HRCT appeared a good predictor of PH and of mortality. Modified Walsh's algorithm using MPAD/BSA ratio was shown a very accurate predictive index of mortality.

      CRediT authorship contribution statement

      Florence Jeny: Writing - original draft, Conceptualization, Formal analysis, Investigation, Visualization. Yurdagül Uzunhan: Writing - original draft, Conceptualization, Formal analysis, Investigation, Visualization. Maxime Lacroix: Writing - review & editing, Investigation. Thomas Gille: Writing - review & editing, Investigation. Pierre-Yves Brillet: Writing - review & editing, Investigation, Resources. Annelyse Nardi: Writing - review & editing, Investigation, Resources. Diane Bouvry: Writing - review & editing, Resources. Carole Planès: Writing - review & editing, Resources. Hilario Nunes: Writing - review & editing, Conceptualization, Resources, Supervision. Dominique Valeyre: Writing - original draft, Writing - review & editing, Conceptualization, Resources, Supervision.

      Declaration of competing interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Rossides M.
        • Kullberg S.
        • Askling J.
        • Eklund A.
        • Grunewald J.
        • Arkema E.V.
        Sarcoidosis mortality in Sweden: a population-based cohort study.
        Eur. Respir. J. 2018 Feb 1; 51: 1701815
        • Jamilloux Y.
        • Maucort-Boulch D.
        • Kerever S.
        • Gerfaud-Valentin M.
        • Broussolle C.
        • Eb M.
        • et al.
        Sarcoidosis-related mortality in France: a multiple-cause-of-death analysis.
        Eur. Respir. J. 2016; 48: 1700-1709
        • Swigris J.J.
        • Olson A.L.
        • Huie T.J.
        • Fernandez-Perez E.R.
        • Solomon J.
        • Sprunger D.
        • et al.
        Sarcoidosis-related mortality in the United States from 1988 to 2007.
        Am. J. Respir. Crit. Care Med. 2011 Jun 1; 183: 1524-1530
        • Reich J.M.
        Mortality of intrathoracic sarcoidosis in referral vs population-based settings: influence of stage, ethnicity, and corticosteroid therapy.
        Chest. 2002 Jan; 121: 32-39
        • Park J.E.
        • Kim Y.S.
        • Kang M.J.
        • Kim C.J.
        • Han C.H.
        • Lee S.M.
        • et al.
        Prevalence, incidence, and mortality of sarcoidosis in Korea, 2003-2015: a nationwide population-based study.
        Respir. Med. 2018 Nov; 144S: S28-S34
        • Gribbin J.
        • Hubbard R.B.
        • Le Jeune I.
        • Smith C.J.P.
        • West J.
        • Tata L.J.
        Incidence and mortality of idiopathic pulmonary fibrosis and sarcoidosis in the UK.
        Thorax. 2006 Nov; 61: 980-985
        • Baughman R.P.
        • Winget D.B.
        • Bowen E.H.
        • Lower E.E.
        Predicting respiratory failure in sarcoidosis patients.
        Sarcoidosis Vasc. Diffuse Lung Dis. Off. J. WASOG. 1997 Sep; 14: 154-158
        • Nardi A.
        • Brillet P.-Y.
        • Letoumelin P.
        • Girard F.
        • Brauner M.
        • Uzunhan Y.
        • et al.
        Stage IV sarcoidosis: comparison of survival with the general population and causes of death.
        Eur. Respir. J. 2011 Dec; 38: 1368-1373
        • Baughman R.P.
        • Engel P.J.
        • Taylor L.
        • Lower E.E.
        Survival in sarcoidosis-associated pulmonary hypertension: the importance of hemodynamic evaluation.
        Chest. 2010 Nov 1; 138: 1078-1085
        • Boucly A.
        • Cottin V.
        • Nunes H.
        • Jaïs X.
        • Tazi A.
        • Prévôt G.
        • et al.
        Management and long-term outcomes of sarcoidosis-associated pulmonary hypertension.
        Eur. Respir. J. 2017; 50
        • Mirsaeidi M.
        • Machado R.F.
        • Schraufnagel D.
        • Sweiss N.J.
        • Baughman R.P.
        Racial difference in sarcoidosis mortality in the United States.
        Chest. 2015 Feb; 147: 438-449
        • Iwai K.
        • Sekiguti M.
        • Hosoda Y.
        • DeRemee R.A.
        • Tazelaar H.D.
        • Sharma O.P.
        • et al.
        Racial difference in cardiac sarcoidosis incidence observed at autopsy.
        Sarcoidosis. 1994 Mar; 11: 26-31
        • Yazaki Y.
        • Isobe M.
        • Hiroe M.
        • Morimoto S.
        • Hiramitsu S.
        • Nakano T.
        • et al.
        Prognostic determinants of long-term survival in Japanese patients with cardiac sarcoidosis treated with prednisone.
        Am. J. Cardiol. 2001 Nov 1; 88: 1006-1010
        • Zhou Y.
        • Lower E.E.
        • Li H.-P.
        • Costea A.
        • Attari M.
        • Baughman R.P.
        Cardiac sarcoidosis: the impact of age and implanted devices on survival.
        Chest. 2017 Jan; 151: 139-148
        • Chester Wasko M.
        • Dasgupta A.
        • Ilse Sears G.
        • Fries J.F.
        • Ward M.M.
        Prednisone use and risk of mortality in patients with rheumatoid arthritis: moderation by use of disease-modifying antirheumatic drugs.
        Arthritis Care Res. 2016; 68: 706-710
        • del Rincón I.
        • Battafarano D.F.
        • Restrepo J.F.
        • Erikson J.M.
        • Escalante A.
        Glucocorticoid dose thresholds associated with all-cause and cardiovascular mortality in rheumatoid arthritis.
        Arthritis Rheumatol. Hoboken NJ. 2014 Feb; 66: 264-272
        • Movahedi M.
        • Costello R.
        • Lunt M.
        • Pye S.R.
        • Sergeant J.C.
        • Dixon W.G.
        Oral glucocorticoid therapy and all-cause and cause-specific mortality in patients with rheumatoid arthritis: a retrospective cohort study.
        Eur. J. Epidemiol. 2016; 31: 1045-1055
        • Schols A.M.
        • Wesseling G.
        • Kester A.D.
        • de Vries G.
        • Mostert R.
        • Slangen J.
        • et al.
        Dose dependent increased mortality risk in COPD patients treated with oral glucocorticoids.
        Eur. Respir. J. 2001 Mar; 17: 337-342
        • Ungprasert P.
        • Matteson E.L.
        • Crowson C.S.
        Increased risk of multimorbidity in patients with sarcoidosis: a population-based cohort study 1976 to 2013.
        Mayo Clin. Proc. 2017 Dec; 92: 1791-1799
        • Nowiński A.
        • Puścińska E.
        • Goljan A.
        • Peradzynska J.
        • Bednarek M.
        • Korzybski D.
        • et al.
        The influence of comorbidities on mortality in sarcoidosis: a observational prospective cohort study.
        Clin. Res. J. 2017 Sep; 11: 648-656
        • Walsh S.L.
        • Wells A.U.
        • Sverzellati N.
        • Keir G.J.
        • Calandriello L.
        • Antoniou K.M.
        • et al.
        An integrated clinicoradiological staging system for pulmonary sarcoidosis: a case-cohort study.
        Lancet.Respir. Med. 2014 Feb; 2: 123-130
        • Kirkil G.
        • Lower E.E.
        • Baughman R.P.
        Predictors of mortality in pulmonary sarcoidosis.
        Chest. 2018 Jan; 153: 105-113
        • Wells A.U.
        • Desai S.R.
        • Rubens M.B.
        • Goh N.S.L.
        • Cramer D.
        • Nicholson A.G.
        • et al.
        Idiopathic pulmonary fibrosis: a composite physiologic index derived from disease extent observed by computed tomography.
        Am. J. Respir. Crit. Care Med. 2003 Apr 1; 167: 962-969
        • Huitema M.P.
        • Spee M.
        • Vorselaars V.M.M.
        • Boerman S.
        • Snijder R.J.
        • van Es H.W.
        • et al.
        Pulmonary artery diameter to predict pulmonary hypertension in pulmonary sarcoidosis.
        Eur. Respir. J. 2016 Feb; 47: 673-676
        • Uzunhan Y.
        • Nunes H.
        • Jeny F.
        • Lacroix M.
        • Brun S.
        • Brillet P.-Y.
        • et al.
        Chronic pulmonary aspergillosis complicating sarcoidosis.
        Eur. Respir. J. 2017; 49
      1. Statement on sarcoidosis. Joint statement of the American thoracic society (ATS), the European respiratory society (ERS) and the world association of sarcoidosis and other granulomatous disorders (WASOG) adopted by the ATS board of directors and by the ERS executive committee.
        Am. J. Respir. Crit. Care Med. 1999 Aug; 160 (February 1999): 736-755
        • Mosteller R.D.
        Simplified calculation of body-surface area.
        N. Engl. J. Med. 1987 Oct 22; 317: 1098
        • Sumikawa H.
        • Johkoh T.
        • Colby T.V.
        • Ichikado K.
        • Suga M.
        • Taniguchi H.
        • et al.
        Computed tomography findings in pathological usual interstitial pneumonia: relationship to survival.
        Am. J. Respir. Crit. Care Med. 2008 Feb 15; 177: 433-439
        • Dumas O.
        • Abramovitz L.
        • Wiley A.S.
        • Cozier Y.C.
        • Camargo C.A.
        Epidemiology of sarcoidosis in a prospective cohort study of U.S. Women.
        Ann. Am. Thorac. Soc. 2016 Jan; 13: 67-71
        • Duchemann B.
        • Annesi-Maesano I.
        • Jacobe de Naurois C.
        • Sanyal S.
        • Brillet P.-Y.
        • Brauner M.
        • et al.
        Prevalence and incidence of interstitial lung diseases in a multi-ethnic county of Greater Paris.
        Eur. Respir. J. 2017; 50
        • Arkema E.V.
        • Grunewald J.
        • Kullberg S.
        • Eklund A.
        • Askling J.
        Sarcoidosis incidence and prevalence: a nationwide register-based assessment in Sweden.
        Eur. Respir. J. 2016; 48: 1690-1699
        • Taimeh Z.
        • Hertz M.I.
        • Shumway S.
        • Pritzker M.
        Lung transplantation for pulmonary sarcoidosis. Twenty-five years of experience in the USA.
        Thorax. 2016 Apr; 71: 378-379
        • Shlobin O.A.
        • Kouranos V.
        • Barnett S.D.
        • Alhamad E.H.
        • Culver D.A.
        • Barney J.
        • et al.
        Physiological predictors of survival in patients with sarcoidosis associated pulmonary hypertension: results from an international registry.
        Eur. Respir. J. 2020 Mar 26; 1901747
        • Harper L.J.
        • Gerke A.K.
        • Wang X.-F.
        • Ribeiro Neto M.L.
        • Baughman R.P.
        • Beyer K.
        • et al.
        Income and other contributors to poor outcomes in US sarcoidosis patients.
        Am. J. Respir. Crit. Care Med. 2019 Apr 15; 201: 955-964