Prognostic and predictive biomarkers associated with radiotherapy and chemotherapy response in non-small cell lung cancer patients

Lung cancer is the largest single cause of cancer-related mortality worldwide. Over 85% of these lesions correspond to non-small cell lung cancer (NSCLC). Lung cancer patients receive different treatments depending on their detailed clinical-pathological context. Radiotherapy and adjuvant chemotherapy scheme is mainly tailored to the clinical stage (TNM). Therefore, our thesis aim was to identify biomarkers to predict overall survival (OS), and treatment efficacy in radiotherapy/chemotherapy-treated lung cancer patients.

We designed retrospective studies with two different patient cohorts. The first one included 126 NSCLC cases with a history of radiotherapy treatment with curative intent. The second one included 1025 surgical specimens from primary pulmonary adenocarcinoma surgeries performed with curative intent that were available and complete in terms of tissue sample. Our endpoints were OS and proportion tumour response at 100 days (PTR-100), biomarkers analysed were LZIC, Ki67, p53, growth-pattern gradings including predominant pattern, worst patter, IASLC grading, and composite risk category (CRC).

We were able to classify tumours as radioresistant and radiosensitive ones according to PTR-100. LZIC, Ki67, and tumour nuclear size which showed to be independent biomarkers to predict radiotherapy-treatment tumour response. Manual and automated Ki67 scores predicted poor prognosis. In addition, high-risk manual Ki67 and high-risk stroma automated Ki67 predicted chemotherapy effects. CRC showed to be an independent prognostic variable in non-mucinous and mucinous adenocarcinomas. Furthermore, high-risk CRC-3X was predictive marker to chemotherapy effects in non-mucinous adenocarcinomas.

Our thesis results suggest that LZIC, Ki67 and tumour nuclear size could be predictive markers to stratify better patients with NSCLC to undergo to radiotherapy. We proposed CRC as a novel histological growth-pattern grading to predict poor prognosis in lung adenocarcinoma, and chemotherapy efficacy in non-mucinous adenocarcinomas. Finally, high-risk manual Ki67 and high-risk stroma Ki67 would be useful biomarkers to predict beneficial chemotherapy-treatment in lung adenocarcinomas.