Utility of serum CYFRA 21-1 as a prognostic biomarker in ALK-positive non-small cell lung cancer treated with ALK-TKIs: a retrospective cohort study.

Background: Anaplastic lymphoma kinase (ALK) fusion gene-positive non-small cell lung cancer (NSCLC) represents 3-7% of all cases and causes oncogene addiction. Although ALK tyrosine kinase inhibitors (ALK-TKIs) are effective for treating ALK-positive NSCLC, some patients still show suboptimal responses and poor outcomes. Clinically simple and detectable biomarkers for this group are limited. Carcinoembryonic antigen (CEA) and cytokeratin 19 fragment (CYFRA 21-1 and CYFRA) are widely used tumor markers in NSCLC. Elevated CEA levels are linked to tumor progression and resistance to cell death, while CYFRA is widely expressed in poorly differentiated squamous cell carcinomas. CYFRA has been identified as a prognostic factor in epidermal growth factor receptor (EGFR)-positive NSCLC, but its role in ALK-positive NSCLC remains unclear. Therefore, we retrospectively assessed the value of CEA and CYFRA as predictive biomarkers in patients with ALK-positive NSCLC treated with ALK-TKIs.

Methods: This retrospective study analyzed 197 patients with advanced or recurrent ALK-positive NSCLC, who were diagnosed across 13 institutions in Japan and received their first ALK-TKI between July 1, 2014 and December 31, 2022. Eligible patients had measurable lesions based on the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Patients were stratified based on normal (≤5.0 and ≤3.5 ng/mL) or high (>5.0 and >3.5 ng/mL) baseline serum CEA and CYFRA levels. Serum CYFRA and CEA levels, which were measured using commercially available immunoassays per standard institutional protocols. The primary endpoint was progression-free survival (PFS) with initial ALK-TKI therapy, and secondary endpoints included overall survival (OS) and objective response rate (ORR). Multivariate Cox regression models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).

Results: Among 152 patients with available CYFRA data, 91 (59.9%) had normal CYFRA levels and 61 (40.1%) had elevated levels. In this analysis, patients in the high CYFRA group had significantly shorter median PFS and OS (9.27 and 28.8 months, respectively) than those in the normal CYFRA group (42.0 and 143.3 months, respectively). Multivariate analysis confirmed that high CYFRA levels were independent predictors of poor PFS (HR: 2.35, 95% CI: 1.50-3.68, P<0.001) and OS (HR: 3.28, 95% CI: 1.89-5.70, P<0.001). Furthermore, the high CYFRA group had lower ORR and complete response (CR) rates, compared with the normal CYFRA group. In contrast, no significant differences in PFS or OS were observed between patients with normal and elevated CEA levels.

Conclusions: Elevated CYFRA levels correlate with reduced PFS and OS in ALK-positive NSCLC, indicating potential as a prognostic biomarker. Given CYFRA's association with tumor heterogeneity, which reduces ALK-TKI efficacy, its assessment could aid in risk stratification and treatment planning. Additional research is needed to verify the predictive value of CYFRA for ALK-TKI efficacy.