# Prognostic Value of Routine Hematological Markers and ECOG Performance in Predicting Overall Survival in Lung Cancer—A Retrospective Cohort Study and Literature Review

## TL;DR

This study shows that blood markers like anemia and high white blood cell count can help predict survival in lung cancer patients, especially when combined with tumor stage.

## Contribution

The study introduces a cluster analysis approach to refine risk stratification using hematological markers and tumor stage in lung cancer.

## Key findings

- Anemia and leukocytosis were independent predictors of shorter survival in lung cancer patients.
- Cluster analysis identified two subgroups with significantly different mortality risks based on tumor stage and blood markers.
- Early-stage lung cancer patients had nearly double the median overall survival compared to advanced-stage patients.

## Abstract

Background: Accurate prognostic assessment in lung cancer is challenging, especially in settings with limited access to molecular testing. Routine hematological markers may complement TNM staging in identifying high-risk patients. Methods: We retrospectively analyzed 304 patients with histologically confirmed lung cancer to evaluate the prognostic value of hematological and inflammatory markers in relation to tumor stage and ECOG performance status. Survival was estimated using Kaplan–Meier analysis and independent predictors were identified through Cox regression. Cluster analysis integrated hematological markers with tumor characteristics. Results: Most patients presented with advanced disease (61.2% T4; 57.1% metastatic). Early stages (≤IIC) were associated with nearly double the median overall survival compared with stages ≥IIIA (p = 0.001). Nodal involvement and metastases further stratified outcomes. Among hematological markers, anemia (Hb < 11.85 g/dL) and leukocytosis (WBC > 11.71 × 109/L) predicted shorter survival and remained independent predictors in the multivariable Cox model (HR 1.70, p < 0.001; HR 1.54, p = 0.004), along with T4 stage (HR 1.47, p = 0.015). PLT count and fibrinogen were significant in univariate analysis but not after adjustment, while CRP and D-dimer showed no association. Cluster analysis identified two subgroups, with patients in Cluster 2 (T4 stage, anemia, leukocytosis) experiencing more than double the mortality risk compared with Cluster 1 (HR 2.33, p < 0.001). Conclusions: TNM stage remains the dominant prognostic factor in lung cancer. However, Hb and WBC count provide additional prognostic information, and cluster analysis may refine risk stratification by capturing clinically relevant heterogeneity.