Pulmonary Research · en · 10 min

By Theresa M. Whitford · April 14, 2026

Biomarkers in interstitial lung disease (ILD) offer a lens into the biology of disparate diseases that share clinical features, from idiopathic pulmonary f…

Biomarkers in interstitial lung disease (ILD) offer a lens into the biology of disparate diseases that share clinical features, from idiopathic pulmonary fibrosis (IPF) to hypersensitivity pneumonitis and connective tissue disease–associated ILD. This piece surveys promising biomarkers and their roles in diagnosis and prognosis, focusing on data as of late 2025. The emphasis is on interpretive value rather than therapeutic guidance, highlighting how biomarkers refine disease characterization in a field that blends radiology, histology, and clinical trajectory.

Biomarker categories and the diagnostic landscape

Biomarkers in ILD span genetic signals, circulating proteins, cellular phenotypes, and imaging-derived features. A structured view helps clinicians parse heterogeneity in ILD presentation and progression. In IPF, for example, circulating proteins linked to epithelial injury and fibrogenesis—such as matrix metalloproteinases and cytokines—correlate with radiographic extent and physiological decline. Data points that have gained traction include:

• KL-6 (MUC1) levels showing elevations in patients with ILD, with mean serum values in IPF cohorts around 900–1200 U/mL versus 300–600 U/mL in controls across multiple studies (n ≈ 20–40 cohorts, sample sizes ranging 120–900 participants per study).
• Surfactant protein A and D (SP-A, SP-D) elevations associated with disease activity; SP-D in IPF patients has reported mean levels exceeding 100 ng/mL in several cohorts (n ≈ 6–12, N ≈ 250–650 total) and correlates with decline in diffusing capacity.

Beyond single markers, multi-analyte panels and composite scores are increasingly used to augment diagnostic confidence, especially in indeterminate cases where HRCT patterns overlap across ILD subtypes. Evidence supports that integrating serologic panels with clinical context improves diagnostic agreement among multidisciplinary teams by roughly 15–25 percentage points in cohort studies (n ≈ 5–10 centers, total patients ≈ 300–700). At the same time, the field remains cautious about overinterpreting any single biomarker, given disease phase and comorbidities can modulate baseline levels.

Genetic and epigenetic signals: predisposition and prognosis

Genetic markers offer a window into disease susceptibility and trajectory. Telomere biology and mucin gene variants have emerged as robust signals in IPF and related ILDs. As of late 2025, several findings stand out:

• TERT and TERC mutations persist as strong associations with familial IPF and sporadic disease risk, with penetrance estimates in independent cohorts around 10–15% for familial cases and a relative risk increase of 2.0–3.0 for carriers in population studies (n ≈ 1,000–2,500 tested across multiple registries).
• TERC variant rs2853669 and other promoter region polymorphisms show modest associations with disease onset and progression, contributing to risk models that explain up to 8–12% of observed phenotypic variance in IPF cohorts (n ≈ 500–1,200).

Epigenetic patterns—DNA methylation signatures in alveolar epithelial cells and peripheral blood mononuclear cells—have demonstrated that certain methylation clocks correlate with decline in forced vital capacity (FVC) over 12–24 months. In a meta-analysis pooling 3,000+ ILD samples across 7 studies, methylation age acceleration aligned with worse prognosis, with hazard ratios for mortality around 1.8–2.2 in higher clock-advanced groups (p < 0.01).

Genetic and epigenetic markers do not function as stand-alone diagnostics but contribute to risk stratification, enabling earlier specialist referral and more tailored follow-up plans in complex ILD presentations. In registries that integrate genetic data with imaging and physiology, combined models yield improved discrimination between fibrotic ILD subtypes, with area under the receiver operating characteristic (AUC) values rising from 0.70–0.75 (clinical variables alone) to 0.78–0.82 (multimodal biomarker panels) in IPF-dominant cohorts (n ≈ 400–1,200).

Protein-based circulating biomarkers: epithelial injury, fibrogenesis, and immune signals

The circulating proteome offers a dynamic readout of ongoing injury, repair, and remodeling in the lung. Several proteins have accumulated robust data across ILD subtypes, though performance often depends on disease stage and comorbid lung conditions. Notable signals include:

• KL-6 and soluble mucin-like glycoproteins: meta-analytic summaries place mean serum KL-6 elevations in IPF around 1,000–1,400 U/mL (vs. 350–700 U/mL in non-ILD controls), with higher levels predicting more rapid FVC decline in a subset of patients (n ≈ 300–1,200 pooled across studies).
• SP-D and SP-A: IPF and other fibrotic ILDs show SP-D elevations often exceeding 150–250 ng/mL, and SP-A levels above 60–100 ng/mL have been associated with progression risk in several prospective cohorts (n ≈ 150–500 participants per study).
• Cytokines and chemokines such as CCL2, CXCL8, TNF-α, and TGF-β1: elevations correlate with radiographic fibrosis score and qMRI-derived microstructural changes, with TNF-α and TGF-β1 showing dose-response relationships to progression rates in IPF-like phenotypes (n ≈ 200–600; p < 0.05 in most cohorts).

Recent efforts emphasize panels rather than single markers. For instance, a 2023–2024 multicenter study integrating KL-6, MMP-7, SP-D, and \u03b3-glutamyl transpeptidase did better at distinguishing IPF from nonspecific interstitial pneumonia (NSIP) than any single protein alone (AUC increase from 0.72 to 0.84 in cross-validation, n ≈ 350). In prognostic contexts, a two-marker approach combining MMP-7 and SP-D predicted two-year mortality with a hazard ratio around 2.4 (p < 0.01) in IPF cohorts (n ≈ 180).

Imaging biomarkers: radiomics and biological signatures in HRCT

Imaging remains central to ILD diagnosis, and advances in imaging biomarkers align radiology with biology. Radiomics—quantitative feature extraction from HRCT—converts patterns into numerical predictors that researchers correlate with pathology and outcomes. As of late 2025, imaging biomarkers show several concrete advantages:

• I-ROI fibrosis score using machine-learning–derived texture features correlates with longitudinal decline in FVC and DLCO. In IPF cohorts (n ≈ 400–800), radiomic signatures predicted 12-month FVC decline with R^2 values around 0.32–0.45, outperforming standard visual fibrosis extent in some models.
• Imaging-paired biomarkers combine serum signals with radiomic phenotypes to improve diagnostic alignment among ILD subtypes. In a 2022–2024 international dataset (n ≈ 600), a multimodal score integrating SP-D level and texture heterogeneity achieved a 0.88 AUC for IPF diagnosis versus non-IPF fibrotic ILD, compared with 0.74 for serology alone and 0.80 for radiology alone.

Additionally, quantitative CT metrics such as baseline high-attenuation area, honeycombing extent, and lung segmental volumes have shown prognostic value. A 2024 meta-analysis encompassing IPF and fibrotic ILDs found that honeycombing proportion exceeding 40% of total lung involvement predicted mortality within 3 years with a pooled hazard ratio of 2.1 (95% CI: 1.6–2.8, n ≈ 15 cohorts; total IPF patients ≈ 2,200). When combined with serum markers like KL-6, the predictive performance exceeded either domain alone in several datasets (AUC improvement of 0.05–0.10 in cross-validation).

Cellular biomarkers: alveolar and immune cell phenotypes in blood and BAL

Cell-based biomarkers capture shifts in the immune milieu and epithelial injury responses. Blood-based cellular panels and bronchoalveolar lavage (BAL) cellularity provide complementary perspectives to protein and genetic markers.

• and macrophage polarization patterns show correlations with progression in fibrosing ILD subtypes. In a multicenter BAL study (n ≈ 400), higher neutrophil percentages (>3–5%) related to faster FVC decline over 12–24 months (p < 0.01).
• T cell phenotypes including increased circulating Th17/Treg ratios have been observed in progressive ILD phenotypes, with ratios predicting worsened outcomes in IPF and chronic hypersensitivity pneumonitis in cohorts totaling about 250–500 patients (p < 0.05).

Single-cell approaches and multiplexed flow cytometry have begun to disentangle heterogeneity within alveolar macrophages and fibrocyte populations. Early signals indicate that certain macrophage transcriptional programs align with rapid radiographic progression, but results require replication in larger cohorts (n > 1,000) and standardization of BAL sampling across centers.

Prognostic value: how biomarkers shape risk stratification and follow-up

Prognostic biomarkers aim to translate biology into expected clinical trajectories, guiding monitoring intensity and timing of specialist referrals. Several entries stand out for their reproducibility and range of applicability:

• Composite prognostic models incorporating clinical variables (age, sex, baseline FVC), imaging measures (extent of fibrosis, honeycombing), and serum markers (MMP-7, SP-D, KL-6) show improved discrimination for two-year mortality in IPF, with AUCs rising from 0.65–0.70 for clinical models to 0.78–0.85 for combined models (n ≈ 600–1,000 across multiple registries).
• Rate of decline markers such as 3–6 month FVC decline coupled with baseline KL-6 levels predict two-year survival better than baseline FVC alone in several IPF cohorts (hazard ratios 1.9–2.6 for rapid decliners, n ≈ 300–800).

In non-IPF fibrotic ILD, biomarkers that mirror immune activation and epithelial damage tend to exhibit more modest prognostic power when evaluated in isolation but gain value within multimodal models. This is particularly relevant for mixed phenotypes like progressive fibrosing ILD, where integrating serologic panels with HRCT-derived texture features yields higher prognostic accuracy than either domain individually (AUC improvements of 0.04–0.12 in validation cohorts, n ≈ 350–900).

methodological considerations: standardization, reproducibility, and interpretation

As biomarkers migrate from research to practice, standardization becomes the gatekeeper of clinical utility. Several issues shape interpretation and eventual adoption:

• Analytical variability across laboratories for KL-6, SP-D, and MMP-7 can exceed 20–30% in some settings, complicating threshold-based decisions. Harmonization efforts and reference ranges derived from multicenter validation cohorts help mitigate drift (n ≈ 15–20 labs participating in cross-lab comparisons since 2023).
• Biologic variability influenced by age, smoking status, chronic liver disease, infection, and concomitant autoimmune conditions can confound biomarker signals. For instance, KL-6 can be modestly elevated in hepatobiliary disease, necessitating clinical context for interpretation (observed in 8–12% of ILD patients in several cohorts).
• Temporal dynamics matter: serial measurements better reflect trajectory than single time-point values. In IPF, longitudinal trends in SP-D and MMP-7 over 12 months correlated more strongly with FVC decline than baseline values alone in meta-analytic data (n ≈ 1,200 across studies).

The path to clinical deployment involves interlaboratory standardization, defined cutoffs validated in independent cohorts, and clear integration guidelines within multidisciplinary diagnostic pathways. In practice, biomarkers function best as components of a broader decision framework rather than stand-alone criteria.

Clinical applicability and research directions: what clinicians should watch

Although this overview emphasizes diagnostic and prognostic relevance rather than treatment, biomarkers influence clinical decision-making by clarifying disease behavior and enabling stratified follow-up. Key takeaways for 2025 and beyond include:

• Adoption of panel-based approaches appears more robust than single-marker testing for differentiating IPF from other fibrotic ILDs and for prognostication in heterogeneous cohorts. Expect greater uptake of multi-analyte assays in high-volume ILD centers with established biobanking and data infrastructure (n ≈ 20–30 centers in international networks).
• Integration with imaging and physiology enhances predictive accuracy; combined models incorporating radiomics with serology outperform either modality alone in several validation datasets (AUC improvement of 0.05–0.15 noted in IPF cohorts, n ≈ 400–900).
• Personalized surveillance will increasingly align biomarker trajectories with follow-up frequency, potentially lowering unnecessary imaging in slow progressors and prompting closer monitoring in rapid decliners, with data supporting 6–12 month intervals as a practical cadence in many IPF-dominant populations (n ≈ 1,000–2,000 across longitudinal cohorts).

Future research priorities include expanding diverse population representation, refining thresholds for biomarker panels across etiologies, and validating novel signals from single-cell and spatial transcriptomics in large, real-world ILD cohorts (targeting n > 2,000 participants across 5–7 registries by 2026–2027).

Implications for policy, registries, and data sharing

As biomarker science matures, data-sharing frameworks and regulatory considerations influence translational speed. Several policy-relevant trends are evident as of late 2025:

• Data harmonization mandates promote cross-study comparability of biomarker assays, with consensus on pre-analytical conditions, unit reporting, and reference ranges. Initiatives reporting harmonized protocols across 10–15 centers have reduced inter-study variability for KL-6 and SP-D by approximately 12–18% in external validation exercises (n ≈ 600–1,000 samples).
• Registries with integrated biomarker data enable mechanistic correlative studies and real-world prognostication. Large ILD registries now routinely collect serology, genetics, imaging-derived metrics, and BAL-derived data, with multi-center analyses achieving statistical power to detect modest effect sizes (hazard ratios 1.3–1.7 for specific markers in defined subgroups; n ≈ 2,000+ participants).
• Ethical and equity considerations surface around access to advanced biomarker testing and potential disparities in specialist availability. Policymakers and professional societies emphasize equitable access and the avoidance of overmedicalization in populations with limited resources.

The momentum toward multi-modal biomarker integration represents a maturing diagnostic paradigm rather than a shift away from clinical acumen. The strongest evidence supports using biomarkers to sharpen diagnostic confidence and to inform prognosis within a multidisciplinary ILD framework, while always weighing individual patient context and radiographic patterns.

As the ILD field evolves, the role of biomarkers will likely broaden from static tests to dynamic, integrated prognostic tools that blend molecular signals with imaging phenotypes and clinical trajectories. This evolution promises a more nuanced map of interstitial disease biology, where a patient’s biomarker signature helps illuminate the pace and pattern of decline, guiding careful observation, timely referrals, and refined stratification of care pathways within Pneuma Health Journal’s pulmonary research discourse.