Dynamic liver dysfunction predicts poor survival in patients with EGFR-mutant non-small cell lung cancer and liver metastases treated with EGFR tyrosine kinase inhibitors
By: Zhang, Wen, Wu, Xuemei, Sun, Xiaorong, Wen, Jian, He, Xiaoli, Zhang, Mingzhou, Wang, Guansong, Xu, Zhi

BioMed Central
2026-01-23; doi: 10.1186/s12885-026-15616-z

Abstract

Background

Liver metastasis is an adverse prognostic factor in patients with epidermal growth factor receptor (EGFR)-mutant non-small cell lung cancer (NSCLC). While baseline liver burden is a known risk factor, the prognostic significance of dynamic liver function changes during targeted therapy remains understudied. This study aimed to evaluate the prognostic value of Consecutive Liver Function Abnormalities (CLFA) in this population.

Methods

We retrospectively analyzed 82 patients with EGFR-mutant NSCLC and liver metastases receiving first-line EGFR-TKIs. CLFA was defined as the presence of any abnormal liver function parameter (ALT, AST, ALP, GGT, TBIL, or ALB) at three consecutive time points (baseline, 6 weeks, and 12 weeks). Metabolic confounders and hepatic tumor burden were evaluated. To address immortal time bias and distinguish dynamic deterioration from baseline impairment, time-dependent Cox regression, landmark analysis (12-week), and sensitivity analysis (excluding patients with baseline abnormalities) were performed.

Results

CLFA was identified in 31 patients (37.8%). Baseline metabolic characteristics were balanced between groups. Hepatotoxicity was predominantly mild (CTCAE Grade 1–2), with no dose reductions mandated by liver injury. Patients with CLFA had significantly shorter overall survival (OS) (median 20.4 vs. 34.3 months, Log-rank P < 0.001) and liver-specific progression-free survival (Log-rank P = 0.038). In multivariate analysis adjusting for potential confounders, including liver metastatic burden, CLFA remained an independent prognostic marker for poor OS (HR 2.56, 95% CI 1.25–5.24, P = 0.010). This association was robustly confirmed by time-dependent Cox regression (HR 2.61, P = 0.010) and landmark analysis (P = 0.0017). Notably, in a sensitivity analysis restricted to patients with normal baseline liver function, the subsequent development of CLFA was still associated with a two-fold increase in mortality risk (HR 2.03), indicating a deleterious effect of acquired liver injury independent of baseline status.

Conclusions

CLFA serves as a robust, dynamic prognostic marker independent of baseline tumor burden or initial liver function. The acquisition of persistent low-grade liver dysfunction signifies an aggressive clinical trajectory associated with significantly inferior survival, distinct from acute hepatotoxicity. These findings underscore the value of CLFA for risk stratification beyond standard safety monitoring.







Copyright 2026 InterMDnet | Privacy Policy | Disclaimer | System Requirements