Human epididymis protein 4: a potential prognostic biomarker for rheumatoid arthritis-associated interstitial lung disease
摘要
This study aimed to assess the impact of human epididymis protein 4 (HE4) on the progression and mortality of interstitial lung disease (ILD) in rheumatoid arthritis (RA) patients.
MethodsA retrospective 5-year follow-up study was conducted on RA-ILD patients. Serum HE4 levels were measured using electrochemiluminescence immunoassay to investigate their associations with clinical parameters and prognosis.
ResultsA total of 136 consecutive RA-ILD patients were enrolled. The 5-year survival rate was 91.91%. Patients with ILD progression exhibited higher baseline serum HE4 levels than those without progression (152.44 pmol/L vs. 88.95 pmol/L, P < 0.001). During follow-up, serum HE4 levels progressively increased with disease progression in patients with RA-ILD. Patients who died had higher baseline serum HE4 levels than survivors (229.00 pmol/L vs. 111.80 pmol/L, P < 0.001). Baseline serum HE4 levels were significantly negatively correlated with pulmonary function parameters and positively correlated with inflammatory indicators. ROC analysis showed models incorporating HE4 presented better predictive efficacy for progression and mortality than the conventional model. Multivariate regression further validated that elevated serum HE4 levels (≥ 110.30 pmol/L) served as an independent risk factor for ILD progression (odds ratio [OR] = 10.083, 95% confidence interval [CI]: 2.660–38.224, P = 0.001). Kaplan–Meier analysis, as an exploratory analysis, showed that high serum HE4 levels (≥ 131.95 pmol/L) were correlated with elevated mortality risk (hazard ratio (HR) = 7.822, 95% CI: 2.347–26.070, P < 0.001). Notably, multivariable Cox regression failed to confirm its independent prognostic value for mortality.
ConclusionsSerum HE4 shows promise as a potential prognostic biomarker for RA-ILD, particularly for evaluating the risk of ILD progression. Given the limitations of this retrospective single-center study and lack of external validation, its value for predicting mortality remains to be further verified in larger cohorts.