Identifying immunohistochemical biomarkers panel for non-small cell lung cancer in optimizing treatment and forecasting efficacy

Abstract

Abstract Background Chemotherapy and immunotherapy for non-small-cell lung cancer (NSCLC) are gaining momentum. However, its long-term efficacy remains limited to only a small fraction of patients. Hence, it is crucial to identify reliable immunohistochemical biomarkers to facilitate the formulation of optimal treatment strategies and to predict therapeutic outcomes. Methods We retrospectively analyzed a cohort of 140 patients diagnosed with NSCLC who received chemotherapy or immunotherapy. Using bioinformatics analysis and machine learning techniques, we assessed the role of immunohistochemical biomarkers and clinical characteristics in developing a predictive model for treatment options and outcomes in this population. Results Our research has found that immunohistochemical biomarkers can accurately predict treatment regimens and progression-free survival in NSCLC patients with an accuracy rate of 82.1%. We identified an exclusive detection panel for the six vital biomarkers. Of particular note is the role of programmed cell death protein 1 ligand 1 (PD-L1) expression in guiding treatment selection, with high expression predicting better outcomes in the immunotherapy group at a cut-off value of 50%. Non-squamous patients who tested positive for thyroid transcription factor 1 had a longer median progression-free survival, while squamous patients who tested positive for p63 protein or cytokeratin 5/6 expression had a longer median progression-free survival. Conclusions The results of our study are highly encouraging, as they revealed a significant correlation between immunohistochemical biomarkers, therapeutic regimens, and prognosis. These findings indicate that our immunohistochemical detection panel has great potential for facilitating customization of therapeutic regimens to improve patient care. The insights gained from this study could help clinicians optimize treatment protocols and ultimately enhance clinical outcomes.

Keywords

• NSCLC
• Immunohistochemical biomarkers
• Therapeutic decision-making
• Prognostic prediction
• Machine learning