This retrospective study started with identification of pSS patients by searching the electronic medical records. We enrolled patients with pSS who were admitted to the Department of Rheumatology at Yueyang Hospital of Integrated Traditional Chinese and Western Medicine, Shanghai University of Traditional Chinese Medicine between October 2017 and January 2022.

Inclusion criteria for participants were as follows: 1) age over 18 years old; 2) meeting the 2012 American College of Rheumatology (ACR) recommended SS classification criteria (Shiboski et al., 2012); 3) presence of respiratory symptoms or lung abnormalities on HRCT; 4) at least one HRCT examination within 1 year.

The exclusion criteria were as follows: 1) secondary dry syndrome or combined with other rheumatic diseases; 2) combined pulmonary infections, chronic obstructive pulmonary disease, pneumoconiosis and lung diseases such as tumors, tuberculosis, and pulmonary nodules; 3) patients with severe primary diseases of the heart, lungs, liver, kidneys and other organs; 4) patients with incomplete clinical information. Then all patients were divided into pSS-ILD group and pSS-non-ILD group based on whether their HRCT features showed ILD or not. pSS patients whose imaging features supporting ILD in HRCT (e.g., ground-glass opacity, reticulation, consolidation, nodules, traction bronchiectasis, honeycombing) were diagnosed as pSS-ILD (Lohrmann et al., 2004). Altogether 256 pSS patients were screened in this study and divided into the pSS-ILD (102 cases) and pSS-non-ILD groups (154 cases). This study was approved by the Ethics Committee of Shanghai University of Traditional Chinese Medicine (No. 2018-014) and was registered at Chinese Clinical Trial Registry (No. ChiCTR2000037057).

Demographic and clinical data were collected from patients’ medical records, including age, sex, disease duration. Clinical features associated with pSS and pSS-ILD were fully recorded as well, such as xerostomia, xerophthalmia, cough, shortness of breath. Moreover, the blood biochemical data were also collected, including complete blood cell count, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), blood lipids, rheumatoid factor (RF), immunoglobulin, complement 3 (C3), complement 4 (C4), antinuclear antibody (ANA) spectrum, cytokine.

Blood biochemical data testing methods: Complete blood cell count: The BC-6800 automatic hematology analyzer (mindray, China) was used for testing. The mindray series analyzer’s auto hematology analyzer reagent (mindray, China) was used.

Erythrocyte sedimentation rate (ESR): The TEST 1 Automatic Rapid Blood Sedimentation Analyzer (Alifax, Italy) was used for determination. EDTA anticoagulant was used for blood collection.

C-reactive protein (CRP): The high-sensitivity human C-reactive protein detection kit (mindray, China) was used for the determination of C-reactive protein in human whole blood samples by latex-enhanced immunoscattering turbidimetric assay.

Blood lipids: Enzyme method was used for testing, including total cholesterol (TC), triglycerides (TG), high-density lipoprotein (HDL), and low-density lipoprotein (LDL) testing. Reagent brands: glucose oxidase-cholesterol dehydrogenase method TG kit, glucose oxidase-peroxidase method TC kit, Direct HDL-cholesterol (HDL-c) assays Kit and Direct LDL-c kit (Beckman, United States).

Rheumatoid factor (RF): The N Latex RF Kit (Siemens healthineers, Germany) was used for testing.

Immunoglobulin (Ig), complement 3 (C3), complement 4 (C4): The Scattering turbidimetric method was used for testing. The reagent kit brand was N Antiserum to Immunoglobulin A (IgA), immunoglobulin M (IgM), immunoglobulin G (IgG), complement C3 and complement C4 reagent kit (Siemens healthineers, Germany).

Antinuclear antibody (ANA) spectrum: The EUROLINE ANA profile 3 (Euroimmun Medizinische Labordiagnostika, Germany) was used for testing.

Cytokines: Luminex^® xMAP (Luminex Corporation, United States) technology was used for quantitative detection of tumor necrosis factor-α (TNF-α), Interleukin 2 receptor (IL-2R), IL-6 and IL-8. The reagent kit brand was 12 Human Cytokine assay kit (RAISECARE, China).

In addition, all patients underwent HRCT examination. Approach to Radiological Analysis: The Philips iCT 256 spiral computed tomography (CT) scanner was used for image analysis. The tracheal level was selected as the starting position, and the scan was conducted downwards until the bottom of the lungs (excluding the lung apex). The scan parameters were as follows: pitch 1, reconstruction slice thickness of 1.0 mm, interlayer spacing of 1.0 mm, tube current of 40–120 mA, voltage of 120 KV, and reconstruction with a wide window level setting of 2000 HU/-600 HU. All imaging data were reviewed by two experienced radiologists in batches. For high-resolution CT images, a qualitative analysis method and reproducible calculated measurement parameters, including the morphology, extent, and severity of interstitial lung disease (such as linear, nodular, reticular, honeycombing, traction bronchiectasis lesions) were used for classification, counting, and scoring. The maximum diameter and area of all lesions were measured, and the CT density value (CT value) was calculated to evaluate the degree and type of lesion. During the analysis process, independent assessments were performed first, followed by a consensus reached through discussion.

For HRCT abnormalities, different HRCT features of pSS-ILD patients were image-typed according to the American Thoracic Society/European Respiratory Society (2013) International Multidisciplinary Classification of idiopathic interstitial pneumonia (Travis et al., 2013), including Non-specific interstitial pneumonitis (NSIP), Unusual interstitial pneumonia (UIP), Lymphocytic interstitial pneumonitis (LIP), Organizing pneumonia (OP). Additionally, we refer to the literature to stage the severity of fibrosis by chest HRCT in pSS-ILD patients into stages 1, 2, and 3 (Qiu et al., 2008). Stage 1 is defined as early or active lesion with ground glass shadow, increased interlobular stroma and thickened interlobular septa; stage 2 is defined as progressive or chronic prolonged lesion with stage 1 manifestations and interface signs, subpleural arcuate shadow, intrapulmonary lamellar solidity and small nodular shadow; stage 3 is defined as advanced lesion with stage 1, stage 2 and reticular shadow, fibrous streak shadow, honeycomb shadow, multiple large alveoli, distended bronchial dilatation and diffuse fibrosis. Determination of HRCT results were determined by both radiologists and rheumatologist.

All data were statistically analyzed using SPSS 24.0 statistical software. Continuous variables were presented as mean ± standard deviation (SD) for normally distributed data; if the variance had chi-square, two independent samples t-test was used, and if the variance was not chi-square, approximate t-test was used. Median (interquartile spacing) was used for non-normally distributed data and non-parametric test was used for statistical analysis. Categorical variables were presented as the percentages and were compared using chi-square test or Fisher’s exact test. Logistic regression analysis was used to identify the factors associated with the development of ILD in patients with pSS. All reported p-values were two sided and p < 0.05 was considered to be statistically significant.

This study included a total of 256 patients with pSS. Of these, there were 102 patients in the pSS- ILD group, with a mean age of 63.36 ± 9.22 years, of whom 66 (64.71%) were over 60 years old. The pSS-non-ILD group consisted of 154 patients, with a mean age of 57.16 ± 12.72 years, of whom 73 (47.40%) were over 60 years old. Generally, there were significant differences between the pSS-ILD and total pSS-non-ILD groups related to age (p < 0.01) and disease duration (p < 0.001) characteristics studied (Table 1).

Next, clinical symptoms of the two groups were analyzed. Among 256 patients with pSS, the most common clinical manifestations were xerostomia, xerophthalmia, and arthralgias. Fever, fatigue, Raynaud’s phenomenon, cough and sputum, chest tightness and shortness of breath, skin rash, muscle aches and pains have been observed in a few patients. Patients with pSS-ILD have a significantly higher proportion of clinical manifestations of xerophthalmia (85.2%), cough (25.4%), and shortness of breath (22.5%) compared with patients without ILD (p < 0.05) (Table 1).

As shown in Table 1, ESR, CRP, white blood cell count (WBC), neutrophil/lymphocyte ratio (NLR), TC, and TG were significantly higher in the pSS-ILD group than in the pSS-non-ILD group (p < 0.05). As for the immunologic index, the positivity for different autoantibodies in pSS-ILD groups was as follows: 87 cases (85.29%) for ANA, 65 cases (63.73%) for Ro52, 65 cases (63.73%) for Ro60, 39 cases (38.24%) for SSB, and 9 cases (8.82%) for anti-centromere antibodies. By chi-square test, patients in the pSS-ILD group had a higher rate of anti-Ro52⁺ and a lower rate of anti-centromere⁺ than those in the pSS-non-ILD group, and the differences between the two groups were statistically significant (p < 0.05) (Table 1).

In addition, among all pSS patients, cytokines were measured and analyzed in 222 patients showed that IL-6 was significantly higher in the pSS-ILD group than in the pSS-non-ILD group (p < 0.05), while no statistically significant differences between the remaining cytokines in the two groups (p > 0.05) (Table 1).

HRCT provides superior evaluation of the lung compared with conventional radiographs. If HRCT findings are characteristic, lung biopsy is not necessary to confirm ILD in humans. In our study, all 256 patients underwent HRCT and 102 of them showed interstitial lung lesions on HRCT. For HRCT abnormalities, the percentages of different patterns were showed in Table 2. Of note, ground glass-like shadow described as a hazy increased opacity of the lungs was the most common, followed by patchy solid shadow. The percentages of different stages were as follows: 38 patients (34.3%) were in stage 1 (early stage), 35 patients (34.3%) were in stage 2 (chronic prolongation), and 29 patients (28.4%) were in stage 3 (late stage). And then HRCT imaging classification of pSS-ILD patients were showed as follows: 74 cases (72.5%) with NSIP, 10 cases (9.8%) with UIP, 6 cases (5.9%) with LIP, and 2 cases (2.0%) with OP. However, due to atypical HRCT imaging manifestations that could not be clearly defined to which imaging type they belonged, or the presence of two or more types of imaging manifestations at the same time, they were categorized as other in our study results (10 cases, 9.8%) (Table 2).

Additionally, we investigated the features for ILD detected by HRCT in pSS patients with or without respiratory symptoms. It is noteworthy that 50 percentage of patients without obvious pulmonary symptoms exhibited abnormalities, consistent with ILD symptoms, in HRCT-scan. We further observed the characteristics of HRCT in these patients without respiratory symptoms (imaging performance, staging and classification). Obviously, HRCT features in pSS-ILD patients without obvious pulmonary symptoms are predominantly early compared to patients with obvious pulmonary symptoms (Figure 1). This suggests that regular HRCT should also be performed in patients with pSS without respiratory symptoms.

Furthermore, we also performed logistic regression analysis to investigate the independent risk factors of interstitial lung lesions in patients with pSS. In the univariate logistic regression analysis comparing the pSS-ILD cohort with the control cohort, several variables showed statistically significant odds ratios. The pSS-ILD cohort had an odds ratio (OR) of 2.034 (95% CI 1.216–3.414, p = 0.007) for age>60 years, 1.006 (95% CI 1.003–1.010, p < 0.001) for disease duration, 0.103 (95% CI 0.024–0.446, p = 0.002) for shortness of breath. Among the laboratory metrics, the following ORs were observed: ESR (OR = 1.010, 95% CI 1.000–1.020, p = 0.046); CRP (OR = 1.951, 95% CI 1.008–3.776, p = 0.047); WBC (OR = 1.150, 95% CI 1.032–1.281, p = 0.011); NLR (OR = 1.179, 95% CI 1.043–1.333, p = 0.008); TG (OR = 1.793, 95% CI 1.191–2.700, p = 0.005); TC (OR = 1.329, 95% CI 1.038–1.702, p = 0.024); Anti-Ro52⁺ (OR = 1.949, 95% CI 1.167–3.255, p = 0.011); Anti-centromere⁺ (OR = 0.369, 95% CI 0.168–0.811, p = 0.013).

Based on the results and clinical relevance, 11 variables were a step forward included in the multivariate logistic regression analysis. The finally results showed that older than 60 years (OR = 2.381, 95% IC 1.302–4.355, p < 0.05), longer duration of disease (OR = 1.007, 95% IC 1.003–1.012, p < 0.001), higher TG (OR = 1.652, 95% CI 1.014–2.691, p < 0.05), higher TC (OR = 1.424, 95% CI 1.059–1.915, p < 0.05) and anti-Ro52⁺ (OR = 1.949, 95% CI 1.167–3.255, p = 0.011) were risk factors for ILD in pSS patients. Interestingly, we also found a protective factor for ILD in pSS patients was anti-centromere⁺ (OR = 0.402, 95% CI 0.17–0.935, p < 0.05) (Table 3).