Lung involvement in primary Sjögren's syndrome: Correlation between high-resolution computed tomography score and mortality : Journal of the Chinese Medical Association

Secondary Logo

Journal Logo

Original Article

Lung involvement in primary Sjögren's syndrome: Correlation between high-resolution computed tomography score and mortality

Chen, Ming-Hana,b,c; Chou, Hsiao-Pingd; Lai, Chien-Chihb; Chen, Yu-Dongd; Chen, Ming-Huangc; Lin, Hsiao-Yib,c; Huang, De-Fengb,c,*

Author Information
Journal of the Chinese Medical Association: February 2014 - Volume 77 - Issue 2 - p 75-82
doi: 10.1016/j.jcma.2013.11.001


    1. Introduction

    Primary Sjögren's syndrome (pSS) is a systemic autoimmune disease that predominantly impairs the functioning of the exocrine glands due to focal lymphocytic infiltration and usually presents clinically as persistent dryness of the mouth and eyes.1 Systemic involvement of the musculoskeletal, cutaneous, pulmonary, gastrointestinal, renal, and/or nervous organs/systems is present in 50–80% of pSS patients. Interstitial lung disease (ILD) and small airway abnormalities are the main pathologic findings in pSS with pulmonary involvement.2–5 Lung involvement in particular is a very important determinant of the patient's clinical status, quality of life, and outcome.

    The frequency of lung involvement in patients with pSS varies, ranging from 9% to 90% in Caucasians.6,7 The reasons for this variation remain unclear but may include underdiagnosis due to insignificant symptoms, ethnic, environmental, or genetic factors, or statistical error resulting from limited sample sizes. Chest radiographs, pulmonary function tests (PFTs), and high-resolution computed tomography (HRCT) of the lung are used to investigate the presence and severity of involvement; HRCT is a relatively noninvasive method and is currently the most important way to detect early lung parenchymal abnormalities and decreased lung function. HRCT has been proven to be very sensitive for lung involvement in patients with pSS, even in patients without pulmonary symptoms.3,8 Common HRCT findings of pSS in the lungs include ground-glass attenuation, bronchiectasis, a reticular pattern, or a honeycomb appearance.9,10 The HRCT scoring method has been used to evaluate lung diseases and is a good predictor of patient outcome. HRCT score has only been described in 14 pSS patients with pulmonary symptoms, but its correlation with outcome has not been previously investigated.10

    The prognosis of patients with pSS is good unless organ involvement manifests. The major causes of mortality of pSS patients are cardiovascular diseases (32.4–52.9%) and hematologic malignancy (17.6–23.5%), which can be either related or unrelated to pSS.11,12 However, the influence of lung involvement on the outcomes of patients with pSS have yet to be well elucidated. Ito et al9 analyzed PFT and HRCT in 33 pSS patients, and found that low partial oxygen pressures and the presence of honeycombing were related to higher mortality. The aim of this study was to determine the clinical characteristics of lung involvement in Taiwanese patients with pSS. Their PFT results, HRCT findings, HRCT scores, and causes of death were recorded. Risk factors for mortality in pSS patients were also investigated.

    2. Methods

    2.1. Patients

    The Taipei Veterans' General Hospital (Taipei, Taiwan) is a 3000-bed acute care hospital that serves as a tertiary referral center. Patients treated at this hospital with a diagnostic code for pSS (ICD-9 710.2) between January 1996 and December 2009 were recruited for this study. All patients satisfied the 2002 American–European Revised Classification Criteria for pSS.13 To arrive at a definitive diagnosis, it is essential to meet: (1) at least four of six criteria, of which criteria IV and VI are positive; or (2) at least three of the III–VI criteria. The criteria are: (I) ocular symptoms of decreased tear production; (II) oral symptoms of decreased saliva production; (III) objective evidence of ocular involvement, Schirmer's test ≤5 mm/5 minutes, or rose bengal score ≥4; (IV) salivary gland histopathology with lymphocytes focus score ≥1; (V) objective evidence of salivary gland involvement (either unstimulated whole salivary flow <1.5 mL in 15 minutes, diffuse sialectasis by parotid sialography, or salivary scintigraphy showing delayed uptake); and (VI) presence of autoantibodies (anti-SSA/Ro, anti-SSB/La, or both). In addition, these illnesses mimicking SS, such as past history of head and neck irradiation treatment or pre-existing lymphoma, hepatitis B or C infection, acquired immunodeficiency disease, sarcoidosis, or graft versus host disease, should be excluded. Patients with other connective tissue diseases that could have induced secondary SS, such as systemic lupus erythematosus, rheumatoid arthritis, polymyositis, dermatomyositis, scleroderma, primary biliary cirrhosis, and mixed connective tissue disease, were excluded.14

    The clinical and immunological profiles, PFT results, and HRCT findings of all patients were analyzed. Forty-four pSS patients were diagnosed as having coexisting lung involvement, defined as the persistence of pulmonary symptoms, such as dyspnea or cough, with coexisting abnormal PFT and/or HRCT findings for longer than 2 weeks.4,15 Other medical illnesses that could cause pulmonary manifestations, such as chronic heart failure, infection, hyperthyroidism, arrhythmia, and bronchial asthma, were excluded by clinical assessment and other laboratory investigations, including echocardiographic studies, electrocardiography, and thyroid function tests.

    This study was approved by the institutional ethics committee of Taipei Veterans' General Hospital, Taiwan (2013-03-028BC).

    2.2. PFT and radiologic analysis

    Routinely evaluated PFT included forced expiratory volume in the first second (FEV1), forced vital capacity (FVC), carbon monoxide-diffusing capacity (DLCO), residual volume (RV), and peak expiratory flow (PEF). All HRCT examinations were performed using contiguous axial 5 mm section thickness through the lungs on a Toshiba Aquilion 64 CT scanner (Toshiba, Tokyo, Japan). HRCT findings were read by experienced chest radiologists and categorized according to the classification of CT patterns described by the American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias.16 HRCT scoring was performed according to the definitions reported previously by Schurawitzki et al.17 Briefly, each of the six lobes of the lung with the lingula counted as a separate lobe was assessed based on the percentage of affected lung parenchyma. Grades were defined as follows: Grade 0, no involvement; Grade 1, involvement up to 25% of a lobe; Grade 2, involvement between 26% and 50% of a lobe; and Grade 3, involvement >50% of a lobe. The HRCT score was obtained by combining grades over all six “lobes”, with a possible range of 0–18. The severity of lung involvement was classified according to the HRCT, where 0–6 was defined as mild, 7–12 was defined as moderate, and 13–18 was defined as severe.10

    2.3. Statistical analysis

    The goodness-of-fit test, Chi-square test, and/or Fisher's exact test were used to compare categorical variables. Spearman rank correlations were calculated between HRCT scores and PFT values. Overall survival rate was estimated using the Kaplan–Meier method, and survival curves were compared using the log-rank test. A p value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA).

    3. Results

    3.1. Clinical features and PFT results of patients with pSS with lung involvement

    Among our 44 pSS patients with lung involvement (17 male and 27 female), the mean age at diagnosis was 66.5 years (Table 1). Twelve patients, all of whom were male, had a history of smoking. An antinuclear antibody (ANA) titer of >1:160 was noted in 63.6% of all patients and rheumatoid factor (RF) was positive in 11 of 40 (27.5%) patients. As shown in Tables 1 and 2, PFT was performed and HRCT scans were obtained from all 44 patients with lung involvement. All patients (44 of 44, 100%) showed restrictive-type results with reduced DLCO (<75% of the predicted value). In addition, 15 (34.1%) patients had PEF results <60% of the predicted value, 13 (29.5%) patients had a reduced FVC, 12 (27.3%) individuals had a reduced FEV1, four (9.3%) individuals had a reduced RV, and only one (2.3%) patient had a reduced FEV1/FVC (<60% of the predicted value).

    Table 1:
    Demographic, immunologic profiles, and PFTs in 44 pSS patients with lung involvement.
    Table 2:
    Results of HRCT scores and PFTs in 44 pSS patients with lung involvement.

    3.2. HRCT findings

    The HRCT features are summarized in Table 3, and some patterns are shown in Fig. 1. All scans showed a mixed pattern. The most frequent HRCT finding was interstitial linear opacities (88.6%), followed by interlobular septal thickening (81.8%), and ground-glass opacities (77.3%). The distribution of parenchymal lung involvement was uneven, and was found to be most pronounced in the lower lobes (upper; middle; and lower lobes = 28.6%; 33.3%; and 38.1%, respectively). Analysis of the HRCT scores showed that 20 (45.5%) patients had mild, 17 (38.6%) patients had moderate, and 7 (15.9%) patients had severe lung involvement.

    Table 3:
    HRCT features in 44 pSS patients with lung involvement.
    Fig. 1:
    High-resolution computed tomography (HRCT) scan showing interstitial linear opacities (arrow) (A), interlobular septal thickening (arrow) (B), reticular pattern (arrow) and honeycombing appearance (arrowhead) (C), and airspace consolidation (arrow) (D).

    3.3. Correlation between HRCT score and PFT results

    As shown in Fig. 2, investigation of the association between total HRCT scores and PFT results revealed a negative correlation between HRCT scores and DLCO (r = −0.376, p = 0.012), but no correlation between HRCT scores and FEV1 (r = 0.031, p = 0.842), FVC (r = −0.142, p = 0.357), FEV1/FVC (r = 0.031, p = 0.159), RV (r = −0.261, p = 0.091), or PEF (r = 0.138, p = 0.373).

    Fig. 2:
    Spearman's correlation between high-resolution computed tomography (HRCT) scores and pulmonary function tests (PFTs) of % of predicted value for carbon monoxide-diffusing capacity (DLCO) in primary Sjögren's syndrome (pSS) patients with lung involvement. *Significant p value.

    3.4. Outcome analysis

    Twelve (27.3%) of the 44 patients died during the follow-up period (mean follow-up time, 3.73 years; range, from 2 months to 16 years). The causes of death among the patients with pSS in this study included respiratory failure (n = 8, 66.7%), pneumonia with respiratory failure (n = 3, 25.0%), and sepsis (n = 1, 8.3%). PFT and HRCT scores were compared between those of the 44 patients with lung involvement who died during follow-up and those who survived (Table 4). Patients who died during follow-up had lower predicted values for FEV1 (63.1 ± 19.4% vs. 79.0 ± 22.7%, p = 0.017), FVC (58.7 ± 20.4% vs. 77.1 ± 17.5%, p = 0.005), and PEF (54.3 ± 20.5% vs. 72.0 ± 24.8%, p = 0.035) as well as higher HRCT scores (9.2 ± 5.7 vs. 5.2 ± 3.5, p = 0.033) than those who survived. There was no significant difference between deceased and surviving patients for other PFT results and clinical profiles, including predicted values for DLCO, FEV1/FVC, RV, age at diagnosis, gender, smoking history, positive ANA (>1:160), and positive RF (all p > 0.05).

    Table 4:
    Comparison of clinical profile, PFT results, and HRCT scores between expired and survival groups in 44 pSS patients with lung involvement.

    Kaplan–Meier survival analysis showed that median overall survival was shorter among patients whose FEV1 value was <60% of the predicted value than those whose FEV1 value was >60% of the predicted value (p = 0.005; Fig. 3A). Patients with reduced FVC values (<60% of the predicted value) had shorter median overall survival than those with FVC values >60% of the predicted value (p < 0.001; Fig. 3B). Similarly, patients with reduced PEF values (<60% of the predicted value) had shorter median overall survival than those with PEF values >60% of the predicted value (p = 0.021; Fig. 3C). Furthermore, patients with severe lung involvement (high HRCT scores of 13–18) had shorter median overall survival than those with mild or moderate lung involvement (HRCT scores of 0–12; p < 0.001; Fig. 3D). Multivariate analysis adjusted for FEV1, FVC, and PEF, showed that severe lung involvement (HRCT score 13–18) was an independent risk factor for mortality in patients with lung involvement (p = 0.007, odds ratio = 40.154, 95% CI = 2.747–586.991). In addition, being aged > 65 years, gender, and smoking history were not major predictors of mortality in pSS patients with lung involvement.

    Fig. 3:
    Kaplan–Meier survival curves of primary Sjögren's syndrome (pSS) patients with lung involvement. Cumulative overall survival between (A) forced expiratory volume in the first second (FEV1) ≥ and <60% of the predicted value; (B) forced vital capacity (FVC) ≥ and <60% of the predicted value; (C) peak expiratory flow (PEF) ≥ and <60% of the predicted value. (D) Mild or moderate lung involvement and severe lung involvement were compared. A log-rank test was used in this analysis.

    4. Discussion

    The clinical course of patients with pSS is usually confined to the sicca symptoms unless the extraglandular organs are involved. This study analyzed the clinical characteristics, HRCT findings, and PFT results in pSS patients with lung involvement in Taiwan. Several studies have reported PFT and HRCT findings in pSS patients as summarized in Table 5.4,8,9,13,18,19 Such HRCT findings in pSS included ground-glass opacities, interstitial linear opacities, interlobular septal thickening, nodules, and cysts.3,4,10,18,19 The reported incidence of these HRCT findings are varied; for example, ground-glass opacities range from 10.8% to 91.7%, interlobular septal thickening range from 24.3% to 55.0%, nodules range from 24.3% to 78.3%, and cysts range from 9.7% to 39.3%. However, the ground-glass opacities pattern seems to be seen less frequently in patients without pulmonary symptom, suggesting that less severe lung involvement goes hand-in-hand with low prevalence of ground-glass opacities. All of our patients presented with lung involvement; the most frequent HRCT findings were interstitial linear opacities (88.6%), interlobular septal thickening (81.8%), ground-glass opacities (77.3%), and cysts (72.7%). The discrepancy of the above results could be partially explained by a difference in race, gender distribution, duration of disease, age at examination, and severity of lung injury.

    Table 5:
    Summary of HRCT studies in patients with pSS.a
    Table 5:

    PFT in pSS can be restrictive or obstructive, even in patients without pulmonary symptoms.4,8 All of our pSS patients, having been presented with lung involvement, had low DLCO (<75% of predicted vital capacity value), whereas <50% had decreased FEV1 and FVC (<60% of predicted vital capacity value). As shown in Table 5, Shi and colleagues4 reported that PFT results of pSS patients with ILD had a restrictive pattern. Another study determining the PFT results of 33 pSS patients found that over half had restrictive change (<80% of predicted vital capacity value).9 An obstructive defect of PFT was mainly observed in pSS patients with small airway disease.2,4 Compared to these studies, restrictive change was noted in only 8.1% of pSS patients without pulmonary symptoms in the Uffmann et al8 study, indicating that more serious restrictive-type impairment was noted in pulmonary function of patients with lung involvement than in patients without lung involvement. A limited number of studies focusing on the relationship between HRCT score and PFT have been conducted. Our current study revealed that impairment of DLCO was associated with high HRCT scores. This finding was similar to the report by Yazisiz et al,10 in which negative correlations between HRCT scores and PFT results were found in 14 Turkish pSS patients with lung involvement. Taking these findings together, we postulated that higher HRCT score is related to more severe lung impairment.

    Previous studies have reported that the standardized mortality ratio of patients with pSS is slightly higher than that of age- and gender-matched populations, ranging from 1.17 to 2.07.11,12,20,21 A prospective cohort study comprising 484 Swedish patients with pSS with a median follow-up time of 7 years showed that the all-cause mortality was no higher in patients with pSS than in the general population.12 The exact relationship between pulmonary involvement and mortality rate among pSS patients remains to be explored, and more studies should be conducted to further elucidate the connection. Some reports have suggested that pulmonary manifestations of pSS do not worsen the prognosis. A 10-year follow-up of 30 British patients with pSS reported by Davidson et al22 found that most patents had stable pulmonary function, and mortality was low (15.4%). The mortality rate of our Taiwanese pSS patients with lung involvement was 27.3%, similar to the observation of Ito et al9; in their analysis of 33 Japanese and Korean pSS patients with lung involvement, 10 (30.3%) expired. In the study by Davidson et al,22 four (15.4%) of 26 British patients with pSS died during the 10-year follow-up period: one patient died of respiratory disease (fibrosing alveolitis complicated by bronchopneumonia) and three patients died of cerebrovascular events and cancer.22 In our study, the most common cause of death was respiratory failure (11 of 12 patients, 91.7%) with or without coexisting pneumonia (n = 3). Taken together, the outcome of pSS with lung involvement in an Asian population seems to be unfavorable.

    In addition to exploring the relationship between HRCT score and PFT, our current study demonstrated that the mortality rate was higher in patients with high HRCT scores than those with low HRCT scores in pSS patients with lung involvement. This is the first study to show HRCT to be a good predictor of survival in patients with pSS with lung involvement. A previous study suggested that the presence of microscopic honeycombing was independently associated with mortality.9 The percentage of patients in our study whose HRCT scans showed a honeycomb pattern was 65.9%, higher than reported in other studies (range, 23.5–42.8%).3,10 This may explain why the mortality rate of our patients was higher than other previous reports. Even patients with low FEV1, FVC, or PEF presented shorter median overall survival; multivariate analysis adjusted for these PFT results showed that high HRCT score was an independent risk factor for mortality, suggesting that HRCT findings can better reflect the severity of lung involvement in patients with pSS.

    There are some limitations in this study. First, patients without pulmonary symptoms were not included. Second, this is not a prospective study. Furthermore, another limitation was the small sample size.

    In conclusion, the presence of lung involvement with a high HRCT score or poor PFT was determined to be a risk factor for mortality in patients with pSS in Taiwan.


    1. Fox RI. Sjögren's syndrome. Lancet. 2005;366:321-331.
    2. Papiris SA, Maniati M, Constantopoulos SH, Roussos C, Moutsopoulos HM, Skopouli FN. Lung involvement in primary Sjögren's syndrome is mainly related to the small airway disease. Ann Rheum Dis. 1999;58:61-64.
    3. Franquet T, Gimenez A, Monill JM, Diaz C, Geli C. Primary Sjögren's syndrome and associated lung disease: CT findings in 50 patients. AJR Am J Roentgenol. 1997;169:655-658.
    4. Shi JH, Liu HR, Xu WB, Feng RE, Zhang ZH, Tian XL, et al. Pulmonary manifestations of Sjögren's syndrome. Respiration. 2009;78:377-386.
    5. Chen CY, Chen YM, Yen SH, Perng RP. Lung cancer associated with rheumatoid arthritis does not shorten life expectancy. J Chin Med Assoc. 2005;68:216-220.
    6. Deheinzelin D, Capelozzi VL, Kairalla RA, Barbas Filho JV, Saldiva PH, de Carvalho CR. Interstitial lung disease in primary Sjögren's syndrome. Clinical-pathological evaluation and response to treatment. Am J Respir Crit Care Med. 1996;154:794-799.
    7. Strimlan CV, Rosenow EC 3rd, Divertie MB, Harrison EG Jr. Pulmonary manifestations of Sjögren's syndrome. Chest. 1976;70:354-361.
    8. Uffmann M, Kiener HP, Bankier AA, Baldt MM, Zontsich T, Herold CJ. Lung manifestation in asymptomatic patients with primary Sjögren syndrome: assessment with high resolution CT and pulmonary function tests. J Thorac Imaging. 2001;16:282-289.
    9. Ito I, Nagai S, Kitaichi M, Nicholson AG, Johkoh T, Noma S, et al. Pulmonary manifestations of primary Sjögren's syndrome: a clinical, radiologic, and pathologic study. Am J Respir Crit Care Med. 2005;171:632-638.
    10. Yazisiz V, Arslan G, Ozbudak IH, Turker S, Erbasan F, Avci AB, et al. Lung involvement in patients with primary Sjögren's syndrome: what are the predictors? Rheumatol Int. 2010;30:1317-1324.
    11. Pertovaara M, Pukkala E, Laippala P, Miettinen A, Pasternack A. A longitudinal cohort study of Finnish patients with primary Sjögren's syndrome: clinical, immunological, and epidemiological aspects. Ann Rheum Dis. 2001;60:467-472.
    12. Theander E, Manthorpe R, Jacobsson LT. Mortality and causes of death in primary Sjögren's syndrome: a prospective cohort study. Arthritis Rheum. 2004;50:1262-1269.
    13. Vitali C, Bombardieri S, Jonsson R, Moutsopoulos HM, Alexander EL, Carsons SE, et al. Classification criteria for Sjögren's syndrome: a revised version of the European criteria proposed by the American-European Consensus Group. Ann Rheum Dis. 2002;61:554-558.
    14. Oxholm P. Primary Sjögren's syndrome–clinical and laboratory markers of disease activity. Semin Arthritis Rheum. 1992;22:114-126.
    15. Constantopoulos SH, Papadimitriou CS, Moutsopoulos HM. Respiratory manifestations in primary Sjögren's syndrome. A clinical, functional, and histologic study. Chest. 1985;88:226-229.
    16. ATS/ERS. American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias. This joint statement of the American Thoracic Society (ATS) and the European Respiratory Society (ERS) was adopted by the ATS board of directors, June 2001 and by the ERS Executive Committee, June 2001. Am J Respir Crit Care Med. 2002;165:277-304.
    17. Scatshurawitzki H, Stiglbauer R, Graninger W, Herold C, Polzleitner D, Burghuber OC, et al. Interstitial lung disease in progressive systemic sclerosis: high-resolution CT versus radiography. Radiology. 1990;176:755-759.
    18. Koyama M, Johkoh T, Honda O, Mihara N, Kozuka T, Tomiyama N, et al. Pulmonary involvement in primary Sjögren's syndrome: spectrum of pulmonary abnormalities and computed tomography findings in 60 patients. J Thorac Imaging. 2001;16:290-296.
    19. Watanabe M, Naniwa T, Hara M, Arakawa T, Maeda T. Pulmonary manifestations in Sjögren's syndrome: correlation analysis between chest computed tomographic findings and clinical subsets with poor prognosis in 80 patients. J Rheumatol. 2010;37:365-373.
    20. Martens PB, Pillemer SR, Jacobsson LT, O'Fallon WM, Matteson EL. Survivorship in a population based cohort of patients with Sjögren's syndrome, 1976–1992. J Rheumatol. 1999;26:1296-1300.
    21. Skopouli FN, Dafni U, Ioannidis JP, Moutsopoulos HM. Clinical evolution, and morbidity and mortality of primary Sjögren's syndrome. Semin Arthritis Rheum. 2000;29:296-304.
    22. Davidson BK, Kelly CA, Griffiths ID. Ten year follow up of pulmonary function in patients with primary Sjögren's syndrome. Ann Rheum Dis. 2000;59:709-712.

    HRCT score; lung involvement; mortality; primary Sjögren's syndrome; pulmonary function test

    © 2014 by Lippincott Williams & Wilkins, Inc.