Evolution of acquired resistance in a ROS1(+) KRAS G12C(+) NSCLC through the MAPK pathway

Cross-attention enables deep learning on limited omics-imaging-clinical data of 130 lung cancer patients

Deep-learning tools that extract prognostic factors derived from multi-omics data have recently contributed to individualized predictions of survival outcomes. However, the limited size of integrated omics-imaging-clinical datasets poses challenges. Here, we propose two biologically interpretable and robust deep-learning architectures for survival prediction of non-small cell lung cancer (NSCLC) patients, learning simultaneously from computed tomography (CT) scan images, gene expression data, and clinical information. The proposed models integrate patient-specific clinical, transcriptomic, and imaging data and incorporate Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome pathway information, adding biological knowledge within the learning process to extract prognostic gene biomarkers and molecular pathways. While both models accurately stratify patients in high- and low-risk groups when trained on a dataset of only 130 patients, introducing a cross-attention mechanism in a sparse autoencoder significantly improves the performance, highlighting tumor regions and NSCLC-related genes as potential biomarkers and thus offering a significant methodological advancement when learning from small imaging-omics-clinical samples.

Genomic Landscape of NSCLC in the Republic of Ireland

Introduction

The identification of genomic "targets" through next-generation sequencing (NGS) of patient's NSCLC tumors has resulted in a rapid expansion of targeted treatment options for selected patients. This retrospective study aims to identify the proportion of patients with advanced NSCLC in the Republic of Ireland whose tumors harbor actionable genomic alterations through broad NGS panel testing.

Methods

Institutional review board approval was obtained before study initiation. Patients with NSCLC whose tumors underwent genomic testing through the largest available NGS panel at a nationally funded Cancer Molecular Diagnostics laboratory (St. James's Hospital) between June 2017 and June 2022 were identified. Patient demographics and tumor-related data were collected by retrospective review from all cancer centers in Ireland, referring to the Cancer Molecular Diagnostics laboratory. A total of 203 (9%) tumor samples were excluded due to insufficient neoplastic cell content. Genomic data were collected through retrospective search of Ion Reporter software. The spectrum and proportion of patients with oncogenic driver mutations were evaluated using descriptive statistics (SPSS version 29.0).

Results

In total, 2052 patients were identified. Patients were referred from 23 different hospital sites and all four geographic regions (Leinster = 1091, 53%; Munster = 763, 37.2%; Connacht = 191, 9.3%; Ulster = 7, 0.3%). Median age was 69 (range: 26-94) years; 53% were male. The most common tumor histologic subtype was adenocarcinoma (77%, n = 1577). An actionable genomic alteration was identified in 1099 cases (53%), the most common of which was KRAS (n = 657, 32%). Less frequently, NSCLC tumors harbored the following: MET exon 14 skipping (n = 53, 2.6%), MET amplification (n = 26, 1.3%), EGFR (n = 181, 8.8%), HER2 (n = 35, 1.7%), and BRAF (n = 72, 3.5%) mutations. Fusions were detected in 76 patients (3.7%) including ALK (n = 44, 58%), RET (n = 11, 14.5%), ROS1 (n = 16, 21%), and FGFR3 (n = 5, 6.6%), whereas no NTRK fusion was identified. Co-alterations were detected in 114 patients (5.6%), the most common of which was KRAS/PIK3CA (n = 19, 17%), EGFR/PIK3CA (n = 10, 8.5%), and KRAS/IDH1 (n = 9, 8%). Other co-alterations of interest identified included KRAS G12A/ROS1 fusion (n = 1) and KRAS G12C/BRAF G469A (n = 2).

Conclusions

This is the first retrospective study to comprehensively characterize the genomic landscape of NSCLC in Ireland, using the broadest available NGS panel. Actionable alterations were identified in 53.4% of the patients, and KRAS was the most common oncogenic driver alteration. Our study revealed a lower prevalence of patients whose tumor harbors ALK, ROS1, and RET fusions, compared with similar data sets.

Clinical significance of TP53 alterations in advanced NSCLC patients treated with EGFR, ALK and ROS1 tyrosine kinase inhibitors: An update

 The development of targeted therapies for non-small cell lung cancer (NSCLC), such as the epidermal growth factor receptor (EGFR), anaplastic lymphoma receptor tyrosine kinase (ALK), and ROS proto-oncogene 1 (ROS1), has improved patients' prognosis and significantly extended progression-free survival. However, it remains unclear why some patients do not benefit from the treatment as much or have a rapid disease progression. It is considered that, apart from the oncogenic driver gene, molecular alterations in a number of caretaker and gatekeeper genes significantly impact the efficacy of targeted therapies. The tumor protein 53 (TP53) gene is one of the most frequently mutated genes in NSCLC. To date, numerous studies have investigated the influence of various TP53 alterations on patient prognosis and responsiveness to therapies targeting EGFR, ALK, or ROS1. This review focuses on the latest data concerning the role of TP53 alterations as prognostic and/or predictive biomarkers for EGFR, ALK, and ROS1 tyrosine kinase inhibitors (TKIs) in advanced NSCLC patients. Since the presence of TP53 mutations in NSCLC has been linked to its decreased responsiveness to EGFR, ALK, and ROS1 targeted therapy in most of the referenced studies, the review also discusses the impact of TP53 mutations on treatment resistance. It seems plausible that assessing the TP53 mutation status could aid in patient stratification for optimal clinical decision-making. However, drawing meaningful conclusions about the clinical value of the TP53 co-mutations in EGFR-, ALK- or ROS1-positive NSCLC is hampered mainly by an insufficient knowledge regarding the functional consequences of the TP53 alterations. The integration of next-generation sequencing into the routine molecular diagnostics of cancer patients will facilitate the detection and identification of targetable genetic alterations along with co-occurring TP53 variants. This advancement holds the potential to accelerate understanding of the biological and clinical role of p53 in targeted therapies for NSCLC.

Targeted therapy for non-small-cell lung cancer: New insights into regulated cell death combined with immunotherapy

Non-small-cell lung cancer (NSCLC), which has a high rate of metastatic spread and drug resistance, is the most common subtype of lung cancer. Therefore, NSCLC patients have a very poor prognosis and a very low chance of survival. Human cancers are closely linked to regulated cell death (RCD), such as apoptosis, autophagy, ferroptosis, pyroptosis, and necroptosis. Currently, small-molecule compounds targeting various types of RCD have shown potential as anticancer treatments. Moreover, RCD appears to be a specific part of the antitumor immune response; hence, the combination of RCD and immunotherapy might increase the inhibitory effect of therapy on tumor growth. In this review, we summarize small-molecule compounds used for the treatment of NSCLC by focusing on RCD and pharmacological systems. In addition, we describe the current research status of an immunotherapy combined with an RCD-based regimen for NSCLC, providing new ideas for targeting RCD pathways in combination with immunotherapy for patients with NSCLC in the future.

Screening biomarkers for predicting the efficacy of immunotherapy in patients with PD-L1 overexpression

PURPOSE

Immunotherapy plays an important role in non-small cell lung cancer (NSCLC); in particular, immune checkpoint inhibitors (ICIs) therapy has good therapeutic effects in PD-L1-positive patients. This study aims to screen NSCLC patients with PD-L1-positive expression and select effective biomarkers for ICI immunotherapy.

METHODS

Collected tumor samples from the Affiliated Cancer Hospital of Xinjiang Medical University and 117 patients with stage III-IV NSCLC were included in the study. All patients were on first- or second-line therapy and not on targeted therapy. Based on the molecular profiles and clinical features, we screened biomarkers for predicting the efficacy of immunotherapy in patients with PD-L1 overexpression.

RESULTS

117 NSCLC patients receiving ICIs immunotherapy were enrolled. First, we found that immunotherapy was more effective in patients with positive PD-L1 expression. Second, we found that ROS1 gene mutations, KRAS gene mutations, tumor stage, and the endocrine system diseases history are independent prognostic factors for PD-L1 positive patients. Then we combined independent risk factors and constructed a new Nomogram to predict the therapeutic efficacy of ICIs immunotherapy in PD-L1 positive patients. The Nomogram integrates these factors into a prediction model, and the predicted C-statistic of 3 months, 6 months and 12 months are 0.85, 0.84 and 0.85, which represents the high predictive accuracy of the model.

CONCLUSIONS

We have established a model that can predict the efficacy of ICIs immunotherapy in PD-L1 positive patients. The model consists of ROS1 gene mutations, KRAS gene mutations, tumor staging, and endocrine system disease history, and has good predictive ability.