Clinical significance of cytokeratin 19 fragment in COVID-19 patients: a retrospective study

Simei Shen, Dandan Wu, Haoran Xie, Haiyan He, Yihua Wang, Xuedong LvFrontiers in Public Health 2025 · DOI

Abstract

CYFRA 21-1 may be a potential diagnostic and prognostic indicator of COVID-19 pneumonia. Prospective multicenter studies are needed to confirm its clinical value.

Introduction

Accurately distinguishing COVID-19 from community-acquired pneumonia (CAP) and identifying patients at high risk for severe respiratory failure remain significant clinical hurdles. While traditional inflammatory markers like C-reactive protein (CRP) provide general insights into systemic inflammation, they often lack the specificity to reflect direct lung tissue injury. This study explores the clinical utility of CYFRA 21-1—a fragment of cytokeratin 19 released during alveolar epithelial cell death—as a biomarker for COVID-19 diagnosis and risk stratification.

Methodology

This retrospective study analyzed data from 252 adult patients at Nantong First People’s Hospital (2022–2023). The cohort included three distinct groups: mild non-COVID-19 CAP (n=86), mild COVID-19 (n=100), and severe COVID-19 (n=66). Researchers employed multivariate logistic regression to identify independent risk factors and Receiver Operating Characteristic (ROC) curves to assess predictive accuracy. Patients with known lung tumors or structural lung diseases were excluded to isolate the impact of acute infection on biomarker levels.

Results

Diagnostic Differentiation

CYFRA 21-1 levels demonstrated a clear progressive increase corresponding to disease severity. Levels were significantly higher in COVID-19 patients compared to those with standard CAP, suggesting that SARS-CoV-2 causes more profound epithelial disruption. Multivariate analysis confirmed CYFRA 21-1 as an independent diagnostic factor for COVID-19 (Odds Ratio = 2.369).

CYFRA 21-1 levels across CAP, Mild COVID-19, and Severe COVID-19 groups.

Severity and Mortality Prediction

The biomarker showed exceptional performance in predicting disease progression. For identifying patients likely to develop severe COVID-19, CYFRA 21-1 achieved an Area Under the Curve (AUC) of 0.913, indicating high sensitivity and specificity.

ROC curve for CYFRA 21-1 in predicting COVID-19 severity (AUC = 0.913).

Furthermore, in the severe patient group, CYFRA 21-1 was significantly elevated in non-survivors compared to survivors, serving as an independent predictor of in-hospital mortality (OR = 1.109). A strong negative correlation was also observed between CYFRA 21-1 levels and the Oxygenation Index (OI), providing a biological link between the marker and clinical respiratory impairment.

Discussion and Conclusion

By repurposing CYFRA 21-1—traditionally an oncology marker—clinicians may gain a more precise tool for monitoring alveolar damage in COVID-19. Its correlation with oxygenation levels and its high predictive value for severity make it a promising candidate for early risk stratification.

Despite these results, the study's single-center, retrospective nature and the exclusion of high-risk patients with pre-existing lung conditions are notable limitations. Future prospective research is necessary to validate these findings across broader populations and to determine if CYFRA 21-1 can assist in tracking long-term pulmonary complications such as fibrosis.

Summary updated

Abstract

CYFRA 21-1 may be a potential diagnostic and prognostic indicator of COVID-19 pneumonia. Prospective multicenter studies are needed to confirm its clinical value.
BibTeX
@inproceedings{shen2025clinical,
  author = {Simei Shen and Dandan Wu and Haoran Xie and Haiyan He and Yihua Wang and Xuedong Lv},
  title = {{Clinical significance of cytokeratin 19 fragment in COVID-19 patients: a retrospective study}},
  booktitle = {Frontiers in Public Health},
  year = {2025},
  doi = {10.3389/fpubh.2025.1738947},
}
Figure