A novel, non-spirometric “BMP index” for predicting severe adverse outcomes in fibrosing interstitial lung diseases

TL;DR

A new non-invasive 'BMP index' combining BMI, mucin-1, and PTX3 levels helps predict severe outcomes in fibrosing interstitial lung diseases.

Contribution

The BMP index is a novel composite biomarker for predicting acute exacerbation and early death in fibrosing interstitial lung diseases.

Findings

The BMP index effectively stratified risk across derivation, validation, and combined cohorts.

Higher BMP scores correlated with increased risks of acute exacerbation and early death.

The BMP index outperformed existing predictors in robustness and predictive accuracy.

Abstract

Fibrosing interstitial lung diseases (fILDs) exhibit diverse clinical courses and prognoses, with no established biomarkers for predicting acute exacerbation (AE) and early death. This study introduces a novel composite “BMP index”, integrating body mass index (BMI), mucin-1, and pentraxin 3 (PTX3) to assess AE and early death risk in fILDs. Between January, 2017, and May, 2025, we enrolled 366 patients with fILDs diagnosed, managed and followed at three hospitals in southern Taiwan. Using stratified randomization with a ratio 3:2, participants were assigned to derivation (218 patients) and validation (148 patients) cohorts. Baseline circulating mucin-1 and PTX3 levels were measured via enzyme-linked immunosorbent assay. One point was assigned for each of the three components of the BMP index: BMI < 24 kg/m2, mucin-1 ≥ 2.5 ng/mL, and PTX3 ≥ 2.2 ng/mL, yielding a total score from 0 to 3. The BMP index consistently stratified risk across the derivation, validation, and combined whole cohorts, with higher scores indicating increased AE and early death risks. Sensitivity analyses and comparative performance assessments confirming its robustness and superior predictive performance compared to multiple recently reported predictors. Our findings support the BMP index as a non-spirometric, noninvasive, radiation-free tool for early clinical risk stratification in fILDs.

The online version contains supplementary material available at 10.1038/s41598-025-32619-1.

Figures