Diagnostic Performance of CLEIA Versus FEIA for KL-6 Peripheral and Alveolar Concentrations in Fibrotic Interstitial Lung Diseases: A Multicentre Study

Abstract

Background: Interstitial lung diseases (ILD) is a group of lung disorders characterized by interstitial lung thickening due to inflammatory and fibrotic processes. Krebs von den Lungen-6 (KL-6) is a molecule secreted by damaged type II alveolar pneumocytes in the alveolar space. The goal of the present study was to compare two detection methods of KL-6 in both bronchoalveolar lavage (BAL) and serum from ILD patients at the moment of diagnosis.

Methods: Patients with suspicious of ILD and followed at two Italian referral centres for rare lung diseases were included in the study. BAL fluid and serum were collected and analysed by chemiluminescent enzyme immunoassay (CLEIA) and fluorescent enzyme immunoassay (FEIA) methods provided by Tosoh Biosciences.

Results: A total of 158 (mean age ± standard deviation, 61.5 ± 13.7, 65 females) patients were enrolled. A total of, 36 had diagnosis of idiopathic pulmonary fibrosis (IPF), 74 sarcoidosis, 15 connective tissue disease-ILD (CTD-ILD) and 33 other ILD. Diagnostic agreement between two methods was demonstrated for both BAL (r = 0.707, p < 0.0001) and serum (r = 0.816, p < 0.0001). BAL KL-6 values were lower than serum (p < 0.0001). IPF patients had higher serum KL-6 concentration than other ILDs (p = 0.0294), while BAL KL-6 values were lower in IPF than in non-IPF (p = 0.0023).

Conclusion: This study explored KL-6 concentrations through the CLEIA method in serum and BAL of patients with various ILDs, showing significant differences of biomarkers concentrations between IPF and other non-IPF ILDs. Our findings are promising as they provided further knowledge concerning KL-6 expression across different ILDs and may suggest its utility in differential diagnosis.

Keywords: Krebs von den Lungen‐6; bronchoalveolar lavage; diagnosis; interstitial lung diseases; pulmonary fibrosis.

FIGURE 1

Flowchart of the study design from the enrolment of ILD patients in the two ILD referral centers at Siena and Foggia University for collection and processing of serum and BAL samples to the KL‐6 analysis through the two methods FEIA and CLEIA. BALF, bronchoalveolar lavage fluid; CLEIA, AIA‐CL300 based on chemiluminescent enzyme immunoassay; FEIA, AIA‐360 based on fluorescent enzyme immunoassay; ILD, interstitial lung diseases; KL‐6, Krebs von den Lungen‐6.

FIGURE 2

Comparative analysis and receiver operating characteristic (ROC) curve analysis using BAL and serum KL‐6 concentrations, detected by CLEIA method, in ILD patients stratified according to diagnosis. (a) Lower BAL‐KL‐6 concentrations in IPF patients than sarcoidosis (p = 0.0165) and other ILD (p = 0.0338) groups. (b) AUROC of 72.6% (p = 0.0008) identified BAL‐KL‐6 cut‐off value of 248.3 U/mL to distinguish IPF and sarcoidosis patients, while cut‐off value of 462.4 U/mL distinguished IPF and other ILDs with AUROC of 75.4% (p = 0.0009). (c) Lower serum KL‐6 concentrations in sarcoidosis than IPF (p < 0.0001), CTD‐ILD (p = 0.0132) and other ILD (p = 0.0009). (d) AUROC of 87.4% (p < 0.0001) identifies cut‐off of 1070 U/mL (SE 67.65% and SP 89.29%) to distinguish IPF and sarcoidosis, whereas 596.8 U/mL and 748.4 U/mL were the best to distinguish patients with sarcoidosis from those with other ILDs (AUROC 74.5%, p = 0.0002) and CTD‐ILD (AUROC 77.7%, p = 0.0014), respectively.

FIGURE 3

Spearman correlation tests between clinical features and serum and BAL CLEIA KL‐6 concentrations. Serum concentrations were inversely correlated with (a) DLCO% (r = −0.3011, p = 0.0318) and (b) FVC% (r = −0.3744, p = 0.0041), while they were directly correlated with (c) age (r = 0.4267, p < 0.0001). BAL KL‐6 concentrations were directly correlated with (d) alveolar lymphocyte percentages (r = 0.2912, p = 0.0049) and inversely correlated with (e) alveolar neutrophil percentages (r = −0.2630, p = 0.0113).

FIGURE 4

Comparative analysis and receiver operating characteristic (ROC) curve analysis stratifying patients according to IPF and non‐IPF using BAL and serum KL‐6 concentrations. BAL KL‐6 values (a) were lower in IPF than non‐IPF patients (p = 0.0023), while serum KL‐6 concentrations (b) were higher in IPF than non‐IPF patients (p = 0.0294). AUROC of 70.2% (p = 0.0026) identify the best BAL KL‐6 values (c) cut‐off value of 248.3 U/mL to distinguish IPF and non‐IPF patients. Serum KL‐6 cut‐off value (d) of 1038 U/mL serum allow to distinguish IPF and non‐IPF (AUROC 64.35%, p = 0.0298).