Blood KL-6 predicts prognosis in primary Sjögren’s syndrome-associated interstitial lung disease

Interstitial lung disease associated with primary Sjögren’s syndrome (SJS-ILD) has a variable clinical course. We aimed to investigate the role of blood biomarkers in predicting prognosis for SJS-ILD. Clinical data of 46 SJS-ILD patients were retrospectively reviewed. Plasma biomarker levels, including Krebs von den Lungen-6 (KL-6), CC chemokine ligand 18 (CCL18), chitinase-3-like-1 (YKL-40), interleukin-4 receptor alpha (IL-4Ra), and matrix metalloproteinase-7 (MMP-7) were measured using the multiplex Luminex assays (R&D Systems, Minneapolis, USA). The median follow-up period was 69.0 months. The mean age of the patients was 59.4 years; 17.4% were men. The KL-6 level was significantly higher in non-survivors (n = 12; 119.6 vs. 59.5 pg/mL, P = 0.037) than survivors (n = 34), while the levels of the other biomarkers did not differ. Receiver operating characteristic analysis indicated that KL-6 shows the best performance for predicting survival (area under the curve = 0.705, P = 0.037; best cut-off value = 53.5 pg/mL). Multivariable Cox analysis that was adjusted by age and diffusing capacity for carbon monoxide suggested a high KL-6 level (> 53.5 pg/mL) as an independent prognostic factor for survival (hazard ratio = 5.939, 95% confidence interval 1.312–26.881, P = 0.021). Our results suggest that blood KL-6 might be a useful in predicting the prognosis for patients with SJS-ILD.

Clinical data. The clinical and survival data of all patients were retrospectively collected from medical records, and/or the records of the National Health Insurance Service of Korea. To measure disease activity in patients with primary SJS, EULAR Sjögren's Syndrome Disease Activity Index (ESSDAI) scores at the time of ILD diagnosis was calculated 33 . Spirometry, total lung capacity (TLC) and the diffusing capacity of the lung for carbon monoxide (DL CO ) were measured according to the recommendations from the American Thoracic Society (ATS)/European Respiratory Society (ERS). The results were presented as percentages of the normal predicted values [34][35][36] . A six-minute walk test (6MWT) was performed according to ERS/ATS recommendations 37 , and bronchoalveolar lavage (BAL) was performed according to the ATS guildelines 38 . Data from follow-up assessments at 3-6-month intervals or from hospitalization events were reviewed to determine the development of acute exacerbation (AE). AE was defined according to the criteria suggested by Collard et al. 39  www.nature.com/scientificreports/ is worsening dyspnea within 30 days with new bilateral lung infiltration with no evidence of infection or other alternative causes for the dyspnea (e.g., pulmonary embolism or heart failure).
Measurement of blood biomarkers. Blood samples were obtained at the time of diagnosis by venipuncture and immediately centrifuged. The separated plasma samples were then stored at − 80 °C until biomarker measurement. The plasma levels of KL-6, CCL18, YKL-40, IL-4Ra, and MMP-7 were measured using the multiplex Luminex assays (R&D Systems, Minneapolis, USA) in accordance with the manufacturer's instructions.
HRCT evaluation. HRCT scans were obtained in accordance with standard protocols and reviewed by a radiologist (J.C.) who was blinded to the clinical information. The HRCT patterns were categorized into usual interstitial pneumonia (UIP), probable UIP, indeterminate for UIP, or an alternative diagnosis based on the 2018 Fleischner Society guidelines 3 . A UIP pattern was defined by a subpleural and basal predominance of reticular abnormalities, honeycombing with or without traction bronchiectasis, and the absence of inconsistent findings with a UIP pattern such as extensive ground-glass opacities (GGO), micro-nodules, discrete cysts, or segmental/ lobar consolidations 40 Table S2). The non-survivors had lower TLC and lymphocyte levels in their BAL fluid, and showed positive anti SS-A/ Ro and a UIP pattern on the HRCT more frequently than the survivors ( Table 1). Total of 39 patients (84.8%) received steroid and/or immunosuppressants (median treatment duration: 15 months [interquartile range: 6-33 months]. However, there was no difference between non-survivors and survivors in terms of the number of patients treated, the initial steroid dose, and the duration of the treatment given. Comparison of blood biomarkers. The level of KL-6 was significantly elevated in the non-survivors (119.6 vs 59.5 pg/mL, P = 0.037) as compared to the survivors (Table 2). However, no significant differences were observed between the two groups for the other biomarker levels. KL-6 was the most significant predictor of mortality (area under a curve [AUC] = 0.705, 95% confidence interval [CI] 0.509-0.901, P = 0.037) using ROC analysis for 10-year survival, and the optimal cut-off value was 53.5 pg/mL (sensitivity = 66.7%, specificity = 79.4%) ( The unadjusted Cox proportional hazards model showed that age, smoking status, higher C-reactive protein levels, lower DL CO and TLC, a shorter six-minute walk test distance (6MWD), a UIP pattern on the HRCT, and a higher level of KL-6 (> 53.5 pg/mL) were significantly associated with 10-year mortality ( Table 3). In the multivariable analysis adjusted by age and DL CO , a high KL-6 level (> 53.5 pg/mL) was independently associated with a poor prognosis (hazard ratio [HR] = 5.939, 95% CI 1.312-26.881, P = 0.021) ( Table 4). However, no association was observed between any of the other biomarkers and mortality in patients with SJS-ILD.
Correlation between KL-6 and physiological parameters. Significant (Fig. 3). However, no significant correlation was observed between the other biomarkers and lung function or exercise capacity.

Discussion
In this study, the prediction of survival using various biomarkers was evaluated for patients with primary SJS-ILD. We focused on evaluating whether biomarkers that were previously studied in other CTD-ILDs are useful in patients with SJS-ILD in terms of the severity and prognosis for ILD. KL-6 was found to be superior to CCL18, YKL-40, IL-4Ra, and MMP-7 in predicting the prognosis of patients with SJS-ILD. High KL-6 levels were independently associated with an increased risk of 10-year mortality in patients with SJS-ILD when adjusted by age and DL CO . Patients with high KL-6 levels showed poorer lung function, with more frequent AE and death than those with low KL-6 levels. Moreover, KL-6 levels were found to correlate with the severity of the disease in patients with SJS-ILD. KL-6 showed the best performance in predicting 10-year mortality as compared to the other biomarkers. A high KL-6 level (> 53.5 pg/mL) was found to be an independent risk factor for mortality after adjustment for age and DL CO when using multivariable Cox analysis. These findings are consistent with those in previous reports 12, 19 . Kamiya et al. studied 99 patients with SJS-ILD (with a median follow-up period of 5.97 years) and reported that Table 1. Comparison between the baseline characteristics of non-survivors and survivors in patients with SJS-ILD. Data are presented as mean ± standard deviation or number (%), unless otherwise indicated. 6MWD six-minute walk test distance, 6MWT the lowest SpO2 lowest oxygen saturation during the six-minute walking test, ANA anti-nuclear antibody, BAL bronchoalveolar lavage, DL CO diffusing capacity of the lung for carbon monoxide, ESSDAI European Alliance of Associations for Rheumatology Sjögren's Syndrome Disease Activity Index, FVC forced vital capacity, HRCT high resolution computed tomography, ILD interstitial lung disease, IM immunosuppressants, SJS Sjögren syndrome, TLC total lung capacity, UIP usual interstitial pneumonia. a The immunosuppressants included azathioprine (n = 20), cyclosporine (n = 16), mycophenolate mofetil (n = 21), and cyclophosphamide (n = 6). b Prednisolone equivalent dose. c Median (interquartile range).    In this study, patients with high KL-6 levels experienced AE more frequently, and KL-6 levels were negatively correlated with lung function and exercise capacity, suggesting the value of KL-6 for predicting AE and disease severity. Previous reports support our findings 17,43 . Ohshimo et al. studied 77 patients with IPF and reported that a high KL-6 level (≥ 1300 U/mL) was an independent risk factor for the development of AE (HR = 11.8, 95% CI 1.43-97.8, P = 0.022) when adjusted for age, sex, smoking history, and treatment 43 . They also showed that baseline KL-6 levels were significantly higher in patients who experienced AE than those who did not (2528 ± 1645 U/mL vs. 1584 ± 1000 U/mL, P < 0.0001) 43  , and also showed that the semiquantitative grades of ILD on the HRCT (grade 1, 0-25%; grade 2, 26-50%; grade 3, 51-75%; grade 4, 76-100%) were significantly proportional to serum KL-6 levels, from which grades could be successfully differentiated (grades 1 vs. 2, P = 0.022; grades 2 vs. 3, P < 0.001; grades 3 vs. 4, P = 0.002) 17 . They also showed that serum KL-6 level had a moderate negative correlation with both FVC% (r = − 0.399, P < 0.001) and DL CO % (r = − 0.578, P < 0.001) 17 . Table 4. Risk factors for mortality in patients with SJS-ILD assessed using a multivariable Cox proportional hazards model. CCL18 CC chemokine ligand 18, CI confidence interval, HR hazard ratio, IL-4Ra interleukin-4 receptor alpha, ILD interstitial lung disease, KL-6 Krebs von den Lungen-6, MMP-7 matrix metalloproteinase-7, SJS Sjögren's syndrome, YKL-40 chitinase-3-like-1. a Adjusted by age and DL CO .  Table 5. Comparison between the clinical characteristics for high and low KL-6 groups in patients with SJS-ILD. Data are presented as mean ± standard deviation or number (%), unless otherwise indicated. a The immunosuppressants included azathioprine (n = 20), cyclosporine (n = 16), mycophenolate mofetil (n = 21), and cyclophosphamide (n = 6). b Prednisolone equivalent dose. c Median (interquartile range). 6MWD six-minute walk test distance, 6MWT the lowest SpO2 lowest oxygen saturation during the six-minute walking test, ANA anti-nuclear antibody, BAL bronchoalveolar lavage, DL CO diffusing capacity of the lung for carbon monoxide, ESSDAI European Alliance of Associations for Rheumatology Sjögren's Syndrome Disease Activity Index, FVC forced vital capacity, HRCT high resolution computed tomography, ILD interstitial lung disease, KL-6 Krebs von den Lungen-6, SJS Sjögren syndrome, TLC total lung capacity, UIP usual interstitial pneumonia.
In this study, positive anti-SSA was more frequent in survivors but was not associated with prognosis in the unadjusted Cox analysis. Moreover, there was no difference in the frequency of positive autoantibodies such as ANA and anti-SSB between the non-survivors and survivors. The previous study support our findings; Gao et al.,  , P = 0.812) was not different between the non-survivors and survivors, suggesting that autoantibodies were not associated with prognosis of ILD 44 . However, there is also a contradictory report; Boitiaux et al., in 45 newly diagnosed patients with idiopathic interstitial pneumonia, showed that the anti-SSA (+) group (n = 15) had lower vital capacity (63 ± 22 vs. 87 ± 23% predicted, P = 0.006) and more frequent GGO (87 vs 67%, P = 0.001) and reticulation (33 vs 21%, P = 0.030) on HRCT than the anti-SSA(−) group (n = 30) 45 . Due to these inconsistent findings, it is still insufficient to draw a conclusion about the association between autoantibodies and prognosis of SJS-ILD. Our study has some limitations. First, this was a retrospective study conducted at a single center, which might lead to selection biases or a lack in generalizability. However, the baseline characteristics of our patients were comparable to those in other studies 12,17 . Second, we did not include a validation cohort to confirm our findings. Our sample size was relatively small; however, the rarity of SJS-ILD means that the number of patients studied is not considered low, and the long-term observation in our study can provide important insights into this rare condition. Finally, detailed treatment information such as type, dose, timing, and the duration over which medication was given were not considered in the analysis of the prognostic factors. However, the treatment given to survivors and non-survivors did not differ.
In conclusion, our results suggest that blood KL-6 might be a useful predictor of the prognosis for patients with SJS-ILD.

Data availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.