Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Diagnostic potential of genomic blood biomarkers of pulmonary fibrosis in a prospective cohort

  • Daniel He,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliations Department of Medicine, Division of Respiratory Medicine, University of British Columbia, Vancouver, BC, Canada, Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, Prevention of Organ Failure (PROOF) Centre of Excellence, Vancouver, BC, Canada

    ⨯
  • Casey P. Shannon,

    Roles Formal analysis, Supervision, Writing – review & editing

    Affiliations Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, Prevention of Organ Failure (PROOF) Centre of Excellence, Vancouver, BC, Canada, Providence Health Care Research Institute, Providence Research, Vancouver, BC, Canada

    ⨯
  • Jeremy A. Hirota,

    Roles Conceptualization, Investigation, Methodology, Resources, Supervision, Writing – review & editing

    Affiliations Department of Medicine, Division of Respiratory Medicine, University of British Columbia, Vancouver, BC, Canada, Department of Medicine, McMaster University, Hamilton, ON, Canada, Department of Medicine, Firestone Institute for Respiratory Health–Division of Respirology, McMaster University, Hamilton, ON, Canada, Department of Biology, University of Waterloo, Waterloo, Ontario, Canada

    ⨯
  • Kjetil Ask,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliations Department of Medicine, McMaster University, Hamilton, ON, Canada, Department of Medicine, Firestone Institute for Respiratory Health–Division of Respirology, McMaster University, Hamilton, ON, Canada

    ⨯
  • Christopher J. Ryerson,

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing

    Affiliations Department of Medicine, Division of Respiratory Medicine, University of British Columbia, Vancouver, BC, Canada, Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, Providence Health Care Research Institute, Providence Research, Vancouver, BC, Canada

    ⨯
  • Scott J. Tebbutt

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing

    scott.tebbutt@hli.ubc.ca

    Affiliations Department of Medicine, Division of Respiratory Medicine, University of British Columbia, Vancouver, BC, Canada, Centre for Heart Lung Innovation, St Paul’s Hospital, Vancouver, BC, Canada, Prevention of Organ Failure (PROOF) Centre of Excellence, Vancouver, BC, Canada, Providence Health Care Research Institute, Providence Research, Vancouver, BC, Canada

    ⨯

Abstract

Fibrotic interstitial lung diseases (ILDs) result from excessive deposition of extracellular matrix (ECM) proteins in the lung, causing irreversible damage to the lung architecture. Clinical management of ILDs differs depending on the diagnosis, but differentiation between subtypes can be difficult and better clinical biomarkers are needed. In this study, we use a 166-gene NanoString assay to investigate whether there are ILD subtype-specific transcripts in whole blood. We identified one transcript, killer cell lectin like receptor 1 (KLRF1), as differentially expressed between idiopathic pulmonary fibrosis (IPF) and systemic sclerosis-associated ILD (SSc-ILD), and identified two transcripts (VCAN, LTK) associated with IPF expression against other ILD subtypes. These findings were validated by examining their expression in ILD lung, with KLRF1 expression significantly higher in SSc-ILD compared to IPF and hypersensitivity pneumonitis (HP) samples. Taken together, this pilot study provides support for the use of the peripheral transcriptome in identifying diagnostic biomarkers of ILD with biological relevance.

Citation: He D, Shannon CP, Hirota JA, Ask K, Ryerson CJ, Tebbutt SJ (2024) Diagnostic potential of genomic blood biomarkers of pulmonary fibrosis in a prospective cohort. PLoS ONE 19(12): e0314876. https://doi.org/10.1371/journal.pone.0314876

Editor: Yoshiaki Zaizen, Kurume University School of Medicine: Kurume Daigaku Igakubu Daigakuin Igaku Kenkyuka, JAPAN

Received: June 6, 2024; Accepted: November 18, 2024; Published: December 3, 2024

Copyright: © 2024 He et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All files have been uploaded to GEO and are available as GSE267652.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Interstitial lung diseases (ILDs) are a heterogenous group of diseases affecting the lung parenchyma through fibrotic and/or inflammatory processes. In fibrotic ILDs, excess deposition of extracellular matrix (ECM) proteins in the pulmonary interstitium result in distortion of the lung architecture, irreversibly impairing gas exchange and ultimately causing early mortality [1]. Early diagnosis is a major clinical challenge in the management of ILD, and patients with ILD are often at risk of misdiagnosis, inappropriate or delayed treatment, and worse outcomes as a result [2].

Given the scarcity of disease biomarkers in ILD, investigations into molecular biomarkers of ILD (particularly circulating biomarkers) are an ongoing area of interest [3–5]. A 52-gene signature derived from peripheral blood mononuclear cell (PBMC) transcriptomic profiling has been shown to be predictive of transplant-free survival in patients with IPF [6,7], and transcriptomic profiling has also identified genes associated with decline in lung function in the ILD subtypes idiopathic pulmonary fibrosis (IPF) and hypersensitivity pneumonitis (HP) [8–10].

While it is evident that the peripheral transcriptome is associated with prognosis in ILD, to our knowledge, no investigations have been performed regarding its diagnostic potential. The objective of this study was to evaluate the ability of the whole blood transcriptome in discriminating between ILD subtypes. In this pilot study, we profiled whole blood RNA obtained from patients with ILD (n = 59) using a previously published NanoString assay designed to discriminate between the early- and late-phase asthmatic response [11]. Given that fibrosis is a TH2-driven process like asthma [12], we reasoned that the transcripts profiled in the assay might also differentiate between other TH2-like disease subtypes.

Materials and methods

Study population and sample processing

Patients were diagnosed with fibrotic ILD (idiopathic pulmonary fibrosis, IPF, n = 22; fibrotic hypersensitivity pneumonitis, HP, n = 14; systemic sclerosis-associated ILD, SSc-ILD, n = 20; interstitial pneumonia with autoimmune features, IPAF, n = 3) in accordance with diagnostic guidelines and prospectively recruited with written informed consent (ethics protocol H09-00748) between 2012 and 2018 at St. Paul’s Hospital (Vancouver, BC, Canada). At time of sampling, patients were either undergoing anti-fibrotic (pirfenidone, nintedanib), anti-inflammatory (prednisone, mycophenolate, N-acetylcysteine), or no pharmacotherapy. Whole blood samples were collected in PAXgene Blood RNA Tubes (BD Biosciences, Mississauga, ON, Canada) and extracted for RNA using the PAXgene Blood miRNA Kit (PreAnalytiX, Hombrechtikon, Switzerland). RNA quality was confirmed by the RNA 6000 Nano Kit (Agilent, Santa Clara, California, USA) via RNA integrity numbers obtained from Bioanalyzer analysis. Blood transcript expression was profiled using 100 ng of purified RNA with a custom NanoString nCounter Elements assay (NanoString Technologies, Seattle, WA, USA) that measured 166 transcripts [11]. Samples were randomized across six NanoString cartridges (12 samples per cartridge), with remaining assay slots used for sample replicates to assess data reproducibility. Data are available via the Gene Expression Omnibus (GEO) accession number GSE267652.

RNAscope

Custom RNAscope™ probes for KLRF1 (276–375 of NM_016523.1), LTK (2419–2518 of NM_001135685.1), and VCAN (transcript variant 3; 1356–1455 of uc003kij.3) were developed by Advanced Cell Diagnostics (ACD, Newark, California). Tissue microarray (TMA) sections (4μM-thick) were obtained from a formalin-fixed and paraffin-embedded TMA block containing cores from archived donor lung tissue samples of patients with IPF, HP, and SSc-ILD collected at McMaster University. In situ hybridization (ISH) was performed using a Leica BOND RX autostainer with associated Leica reagent kits (Concord, Ontario, Canada) on TMA sections. Slides were scanned at 40X magnification with an Olympus VS120-L100 Virtual Slide System for image capture and a Leica Aperio ScanScope AT2 for quantification. Image analysis was performed using QuPath (v0.4.3), and probe quantification was determined using H-scores as indicated by manufacturer protocols.

Data analysis

NanoString data underwent quality assessment by examining field of view (FOV) ratio (number of FOV images successfully captured), binding density (image saturation), linearity of positive controls, and limit of detection via negative controls. Data were filtered for lowly abundant features, then normalized using normalization factors derived from the geometric mean of 10 housekeeping transcripts. Differential expression analysis of transcripts was performed using the linear models for microarray (limma) R package (v3.54.2), while biomarker panels were developed using sparse partial least squares discriminant analysis (sPLSDA), elastic net, and random forest algorithms from the ‘mixOmics’ (v6.22.0) and ‘tidymodels’ (v1.0.0) R packages. Image analysis was performed using QuPath (v0.4.3) by automatically identifying individual cells and setting probe-specific thresholds based on signal strength. Cells without a visible nucleus were excluded from analysis. H-scores were determined by grouping cells into 5 bins based on the numbers of dots per cell (0, 1–3, 4–9, 10–15, >15) and multiplying the total percentage of cells in each bin by 0 to 4. To test for statistical significance of H-scores, a linear mixed-effects model (‘lme4’ v1.1–32) was fitted with the formula H-Score ~ Diagnosis + (1|Patient) and pairwise comparisons between diagnoses to determine statistical significance with Tukey’s post-hoc adjustment were performed using the ‘emmeans’ (v1.8.5) R package. Fisher’s exact test was used to test for differences in demographic count variables between diagnoses, while the F-test was used to test for differences in demographic continuous variables between diagnoses. All analysis was performed in the R (version 4.2.3) statistical computing environment.

Results

We profiled the expression of 166 RNA transcripts in a cohort of patients with IPF, systemic sclerosis-associated ILD (SSc-ILD), HP, or interstitial pneumonia with autoimmune features (IPAF) (Table 1). No significant differences were identified between subtypes for race, percent predicted forced vital capacity (FVC%) or diffusing capacity of the lung for carbon monoxide (DLCO%), smoking status, or treatment status, but patients with SSc-ILD and IPAF were younger at time of sampling and patients with IPF were predominantly male which is in line with previously reported cohorts [13].

thumbnail
Download:
Table 1. Summary table of clinical demographics of patients probed for whole blood RNA expression.

https://doi.org/10.1371/journal.pone.0314876.t001

To identify variables that might affect RNA expression, we performed a principal component (PC) analysis and correlated the PCs with demographic variables (S1 Fig). After adjusting for race and treatment status, we identified 1 differentially expressed transcript at an FDR of 0.05 when comparing IPF to SSc-ILD (KLRF1, FDR = 0.03) (S1 Table and Fig 1). While no transcripts were significantly associated with FVC%, we identified 13 transcripts that were significantly associated with DLCO% (S1 Table). Next, we examined the utility of RNA profiling in developing biomarker panels to differentiate between ILD subtypes through the use of three learning algorithms (sparse partial least squares discriminant analysis (sPLSDA), elastic net, and random forest). The best-performing models using 5-fold cross-validation repeated ten times were IPF vs SSc-ILD (sPLSDA) and IPF vs SSc-ILD (elastic net), which yielded respective accuracies of 0.74 ± 0.01 and 0.80 ± 0.02 (S2 Table).

thumbnail
Download:
Fig 1. Expression of KLRF1, LTK, VCAN in whole blood (log2 RNA expression) or lung tissue microarray sections (H-Score) containing samples from patients with idiopathic pulmonary fibrosis (IPF), hypersensitivity pneumonitis (HP), and systemic sclerosis-associated interstitial lung disease (SSc-ILD).

Statistical significance of blood RNA expression was determined using differential expression analysis via ‘limma’ and corrected using false discovery rate (FDR), with statistical significance indicated by FDR<0.05 (***). To determine statistical significance of H-scores, a linear mixed-effects model was fitted and pairwise comparisons between diagnoses were performed and corrected for using Tukey’s post-hoc adjustment (* denotes p<0.05).

https://doi.org/10.1371/journal.pone.0314876.g001

To investigate the lung expression of the three most highly weighted genes (KLRF1, LTK, VCAN) identified by our biomarker panels when comparing IPF vs non-IPF ILD (S3 Table), we used ISH on TMA sections consisting of twelve cores from five patients with IPF, ten cores from three patients with HP, and four cores from two patients with SSc-ILD (Fig 2), all of which showed no significant between-group differences in demographic or clinical variables (Table 2). Transcript abundance was determined semi-quantitatively by calculating H-scores for each marker per core, which is a weighted total of the percentage of cells expressing the probe at varying levels. Using a linear mixed-effects analysis, the H-score of KLRF1 was higher in SSc-ILD compared to IPF (p = 0.02) and HP (p = 0.01), while LTK was lower in IPF compared to SSc-ILD (p = 0.03) (Fig 1).

thumbnail
Download:
Fig 2. Histological staining of tissue microarray cores taken from donor lungs of patients with idiopathic pulmonary fibrosis (IPF), hypersensitivity pneumonitis (HP), and systemic sclerosis-associated interstitial lung disease (SSc-ILD).

Arrows indicate areas of positive staining.

https://doi.org/10.1371/journal.pone.0314876.g002

thumbnail
Download:
Table 2. Summary table of clinical and demographic characteristics of patients with ILD profiled for lung ISH. Data for race, DLCO%, and treatment status was not available.

https://doi.org/10.1371/journal.pone.0314876.t002

Discussion

The 2018 ATS/ERS/JRS/ALAT diagnostic guidelines for IPF made a recommendation to investigate novel molecular biomarkers that could be integrated into clinical diagnosis [14]; in the 2022 update, a recommendation was not made for or against a genomic classifier identifying a usual interstitial pneumonia pattern diagnostic of IPF in transbronchial biopsies, thus highlighting the continued need for research into molecular biomarkers [15]. Although previous studies have investigated the peripheral transcriptome in relation to prognosis [6,8,10], no study has looked at its diagnostic potential. In this pilot study, we demonstrate that the peripheral blood RNA can differentiate between ILD subtypes, in particular IPF and SSc-ILD, through differential expression and learning algorithms utilizing gene expression data obtained from a custom NanoString assay. The blood gene expression data show association with DLCO% predicted but not FVC% predicted, which was previously noted when examining differential expression of the peripheral transcriptome in mild and severe IPF [8]. Furthermore, we identify KLRF1 expression as a significant in the differentiation of IPF and SSc-ILD in both whole blood and lung, and down-regulation of LTK as being associated with IPF.

Killer cell lectin like receptor F1 (KLRF1, also known as NKp80) is expressed in T cells, monocytes, and CD56bright natural killer (NK cells), and is a marker of NK cell maturity [16–18], inducing cytotoxic activity and calcium mobilization [19]. The dysfunction of NK cells in systemic sclerosis and autoimmunity has been well-documented [20–22], and differences in their abundance have been observed between ILD subtypes [23,24]. We found that KLRF1 expression was associated with lower DLCO%, and its expression was lower in the peripheral blood of patients with SSc-ILD compared to IPF and HP yet the opposite effect was observed when probing lung expression. Previous reports have identified a reduction in abundance and function of NK cells in IPF lungs concurrent with an increase in abundance in blood compared to healthy controls [25]. Single-cell RNA-sequencing data available via the IPF Cell Atlas [26] suggests higher KLRF1 expression in SSc-ILD T/NK subsets compared to other ILDs. Given the disparity between single-cell and bulk RNA data, further investigations are required to delineate the mechanisms surrounding NK cell recruitment to the ILD lung and its role in ILD pathogenesis.

LTK and VCAN were differentially expressed at a nominal p-value of 0.05 between IPF and non-IPF ILDs (but not at FDR<0.05), and were two highly weighted coefficients in our elastic net and sPLSDA biomarker panels. LTK is highly expressed in plasmacytoid dendritic cells (pDCs) (Human Protein Atlas, IPF Cell Atlas), and decreased circulating pDCs have been reported in IPF patients [27], which is consistent with the decreased LTK expression observed in this study in blood and lung. VCAN encodes for versican, an ECM proteoglycan that is localized to fibroblastic foci in the lung, myofibroblasts, and monocytes [26,28]. We were unable to identify a statistically significant increase in VCAN expression in IPF compared to HP or SSc-ILD lung and blood samples; however, given its detectable expression in both lung and peripheral blood, and prior studies showing circulating levels of versican degradation products are associated with increased mortality in idiopathic ILDs [29], further studies into its gene expression as a prognostic biomarker may be warranted.

In this pilot study, we report the first between-subtype comparison of the peripheral ILD transcriptome using a discrete set of genes, which validates performing more extensive genome-wide analysis of peripheral ILD subtype-specific profiles and correlations of identified biomarkers in lung tissue. In particular, molecules detected in the blood that are associated with immunity and fibrosis are detectable in the lung, thereby suggesting a test based on peripheral transcriptomics is feasible for diagnostic biomarkers with biological relevance, similar to prognostic markers in Herazo-Maya et al. [7]. Given the heterogeneity within ILD subtypes, future analysis of the peripheral transcriptome should also consider radiological pattern, progressive phenotype, and other related clinical factors [30].

Supporting information

S1 Table. List of transcripts analyzed for differential expression between ILD subtypes and with lung function.

Sheets with “_cov” denote comparisons made after adjusting for race and treatment status.

https://doi.org/10.1371/journal.pone.0314876.s001

(XLSX)

S2 Table. Classification model performance on cross-validation for selected learning algorithms.

Data are reported as accuracy ± standard error.

https://doi.org/10.1371/journal.pone.0314876.s002

(CSV)

S3 Table. Coefficient weights of biomarker panels developed for each between subtype comparison.

https://doi.org/10.1371/journal.pone.0314876.s003

(XLSX)

S1 Fig. Correlation of principal component analysis (PCA) components (PCs) with demographic and clinical metadata.

Values shown are Pearson correlation coefficients.

https://doi.org/10.1371/journal.pone.0314876.s004

(TIF)

References

  1. 1. Renzoni EA, Poletti V, Mackintosh JA. Disease pathology in fibrotic interstitial lung disease: is it all about usual interstitial pneumonia? The Lancet. 2021 Oct 16;398(10309):1437–49. pmid:34499865
  2. 2. Spagnolo P, Ryerson CJ, Putman R, Oldham J, Salisbury M, Sverzellati N, et al. Early diagnosis of fibrotic interstitial lung disease: challenges and opportunities. The Lancet Respiratory Medicine. 2021 Sep 1;9(9):1065–76. pmid:34331867
  3. 3. Clynick B, Corte TJ, Jo HE, Stewart I, Glaspole IN, Grainge C, et al. Biomarker signatures for progressive idiopathic pulmonary fibrosis. Eur Respir J. 2022 Mar;59(3):2101181. pmid:34675050
  4. 4. Axelsson GT, Gudmundsson G, Pratte KA, Aspelund T, Putman RK, Sanders JL, et al. The Proteomic Profile of Interstitial Lung Abnormalities. Am J Respir Crit Care Med. 2022 Aug 1;206(3):337–46. pmid:35438610
  5. 5. Bowman WS, Newton CA, Linderholm AL, Neely ML, Pugashetti JV, Kaul B, et al. Proteomic biomarkers of progressive fibrosing interstitial lung disease: a multicentre cohort analysis. Lancet Respir Med. 2022 Jun;10(6):593–602. pmid:35063079
  6. 6. Herazo-Maya JD, Noth I, Duncan SR, Kim S, Ma SF, Tseng GC, et al. Peripheral Blood Mononuclear Cell Gene Expression Profiles Predict Poor Outcome in Idiopathic Pulmonary Fibrosis. Sci Transl Med. 2013 Oct 2;5(205):205ra136. pmid:24089408
  7. 7. Herazo-Maya JD, Sun J, Molyneaux PL, Li Q, Villalba JA, Tzouvelekis A, et al. Validating a 52-gene risk profile for outcome prediction in Idiopathic Pulmonary Fibrosis: an international multicentre cohort study. Lancet Respir Med. 2017 Nov;5(11):857–68.
  8. 8. Yang IV, Luna LG, Cotter J, Talbert J, Leach SM, Kidd R, et al. The peripheral blood transcriptome identifies the presence and extent of disease in idiopathic pulmonary fibrosis. PLoS One. 2012;7(6):e37708. pmid:22761659
  9. 9. Huang Y, Oldham JM, Ma SF, Unterman A, Liao SY, Barros AJ, et al. Blood Transcriptomics Predicts Progression of Pulmonary Fibrosis and Associated Natural Killer Cells. Am J Respir Crit Care Med. 2021 Jul 15;204(2):197–208. pmid:33689671
  10. 10. Fernández Pérez ER, Harmacek LD, O’Connor BP, Danhorn T, Vestal B, Maier LA, et al. Prognostic accuracy of a peripheral blood transcriptome signature in chronic hypersensitivity pneumonitis. Thorax. 2022 Jan;77(1):86–90. pmid:34183448
  11. 11. Singh A, Shannon CP, Kim YW, Yang CX, Balshaw R, Cohen Freue GV, et al. Novel Blood-based Transcriptional Biomarker Panels Predict the Late-Phase Asthmatic Response. Am J Respir Crit Care Med. 2018 Feb 15;197(4):450–62. pmid:29087730
  12. 12. Spagnolo P, Tonelli R, Samarelli AV, Castelli G, Cocconcelli E, Petrarulo S, et al. The role of immune response in the pathogenesis of idiopathic pulmonary fibrosis: far beyond the Th1/Th2 imbalance. Expert Opin Ther Targets. 2022 Jul;26(7):617–31. pmid:35983984
  13. 13. Zaman T, Moua T, Vittinghoff E, Ryu JH, Collard HR, Lee JS. Differences in Clinical Characteristics and Outcomes Between Men and Women With Idiopathic Pulmonary Fibrosis: A Multicenter Retrospective Cohort Study. Chest. 2020 Jul;158(1):245–51. pmid:32084394
  14. 14. Raghu G, Remy-Jardin M, Myers JL, Richeldi L, Ryerson CJ, Lederer DJ, et al. Diagnosis of Idiopathic Pulmonary Fibrosis. An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline. Am J Respir Crit Care Med. 2018 Sep 1;198(5):e44–68. pmid:30168753
  15. 15. Raghu G, Remy-Jardin M, Richeldi L, Thomson CC, Inoue Y, Johkoh T, et al. Idiopathic Pulmonary Fibrosis (an Update) and Progressive Pulmonary Fibrosis in Adults: An Official ATS/ERS/JRS/ALAT Clinical Practice Guideline. Am J Respir Crit Care Med. 2022 May;205(9):e18–47. pmid:35486072
  16. 16. Roda-Navarro P, Arce I, Renedo M, Montgomery K, Kucherlapati R, Fernández-Ruiz E. Human KLRF1, a novel member of the killer cell lectin-like receptor gene family: molecular characterization, genomic structure, physical mapping to the NK gene complex and expression analysis. Eur J Immunol. 2000 Feb;30(2):568–76. pmid:10671213
  17. 17. Freud AG, Keller KA, Scoville SD, Mundy-Bosse BL, Cheng S, Youssef Y, et al. NKp80 Defines a Critical Step during Human Natural Killer Cell Development. Cell Rep. 2016 Jul 12;16(2):379–91. pmid:27373165
  18. 18. Mukherjee N, Ji N, Tan X, Chen CL, Noel ODV, Rodriguez-Padron M, et al. KLRF1, a novel marker of CD56bright NK cells, predicts improved survival for patients with locally advanced bladder cancer. Cancer Medicine. 2023;12(7):8970–80. pmid:36583228
  19. 19. Vitale M, Falco M, Castriconi R, Parolini S, Zambello R, Semenzato G, et al. Identification of NKp80, a novel triggering molecule expressed by human NK cells. Eur J Immunol. 2001 Jan;31(1):233–42. pmid:11265639
  20. 20. Horikawa M, Hasegawa M, Komura K, Hayakawa I, Yanaba K, Matsushita T, et al. Abnormal natural killer cell function in systemic sclerosis: altered cytokine production and defective killing activity. J Invest Dermatol. 2005 Oct;125(4):731–7. pmid:16185273
  21. 21. Benyamine A, Magalon J, Sabatier F, Lyonnet L, Robert S, Dumoulin C, et al. Natural Killer Cells Exhibit a Peculiar Phenotypic Profile in Systemic Sclerosis and Are Potent Inducers of Endothelial Microparticles Release. Front Immunol. 2018 Jul 18;9:1665. pmid:30072999
  22. 22. Liu M, Liang S, Zhang C. NK Cells in Autoimmune Diseases: Protective or Pathogenic? Frontiers in Immunology [Internet]. 2021 [cited 2024 Jan 18];12. Available from: https://www.frontiersin.org/articles/10.3389/fimmu.2021.624687
  23. 23. Esposito I, Perna F, Ponticiello A, Perrella M, Gilli M, Sanduzzi A. Natural killer cells in Bal and peripheral blood of patients with idiopathic pulmonary fibrosis (IPF). Int J Immunopathol Pharmacol. 2005;18(3):541–5. pmid:16164835
  24. 24. Bergantini L, d’Alessandro M, Cameli P, Otranto A, Finco T, Curatola G, et al. Prognostic role of NK cell percentages in bronchoalveolar lavage from patients with different fibrotic interstitial lung diseases. Clinical Immunology. 2021 Sep 1;230:108827. pmid:34428741
  25. 25. Cruz T, Jia M, Sembrat J, Tabib T, Agostino N, Bruno TC, et al. Reduced Proportion and Activity of Natural Killer Cells in the Lung of Patients with Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med. 2021 Sep;204(5):608–10. pmid:34077698
  26. 26. Neumark N, Cosme C, Rose KA, Kaminski N. The Idiopathic Pulmonary Fibrosis Cell Atlas. American Journal of Physiology-Lung Cellular and Molecular Physiology. 2020 Dec;319(6):L887–92. pmid:32996785
  27. 27. Galati D, Zanotta S, Polistina GE, Coppola A, Capitelli L, Bocchino M. Circulating dendritic cells are severely decreased in idiopathic pulmonary fibrosis with a potential value for prognosis prediction. Clin Immunol. 2020 Jun;215:108454. pmid:32376503
  28. 28. Bensadoun ES, Burke AK, Hogg JC, Roberts CR. Proteoglycan deposition in pulmonary fibrosis. Am J Respir Crit Care Med. 1996 Dec;154(6 Pt 1):1819–28. pmid:8970376
  29. 29. Sand JMB, Tanino Y, Karsdal MA, Nikaido T, Misa K, Sato Y, et al. A Serological Biomarker of Versican Degradation is Associated with Mortality Following Acute Exacerbations of Idiopathic Interstitial Pneumonia. Respiratory Research. 2018 May 4;19(1):82. pmid:29728109
  30. 30. Ryerson CJ. Lumpers versus splitters: What to do with suspected idiopathic pulmonary fibrosis? Respirology. 2019;24(4):300–1. pmid:30440108