Exploring Molecular Genetic Alterations and RAF Fusions in Melanoma: A Belvarafenib Expanded Access Program in Patients with RAS/RAF-Mutant Melanoma.

BACKGROUND: Melanoma incidence is on the rise in East Asia, yet studies of the molecular landscape are lacking in this population. We examined patients with melanoma who underwent next-generation sequencing (NGS) at a single tertiary center in South Korea, focusing on patients harboring NRAS or RAF alterations who received belvarafenib, a pan-RAF dimer inhibitor, through the Expanded Access Program (EAP). PATIENTS AND METHODS: Data were collected from 192 patients with melanoma who underwent NGS between November 2017 and May 2023. Variant call format data were obtained and annotated. Patients in the EAP received 450 mg twice daily doses of belvarafenib. RESULTS: Alterations in the RAS/RTK pathway were the most prevalent, with BRAF and NRAS alteration rates of 22.4% and 17.7%, respectively. NGS enabled additional detection of fusion mutations, including 6 BRAF and 1 RAF1 fusion. Sixteen patients with NRAS or RAF alterations received belvarafenib through the EAP, and disease control was observed in 50%, with 2 patients demonstrating remarkable responses. CONCLUSIONS: Our study highlights the value of NGS in detecting BRAF, NRAS mutations and RAF fusions, expanding possibilities for targeted therapies in malignant melanoma. Belvarafenib showed clinical benefit in patients harboring these alterations. Ongoing trials will provide further insights into the safety and efficacy of belvarafenib.

Synthetic Data Improve Survival Status Prediction Models in Early-Onset Colorectal Cancer.

PURPOSE: In artificial intelligence-based modeling, working with a limited number of patient groups is challenging. This retrospective study aimed to evaluate whether applying synthetic data generation methods to the clinical data of small patient groups can enhance the performance of prediction models. MATERIALS AND METHODS: A data set collected by the Cancer Registry Library Project from the Yonsei Cancer Center (YCC), Severance Hospital, between January 2008 and October 2020 was reviewed. Patients with colorectal cancer younger than 50 years who started their initial treatment at YCC were included. A Bayesian network-based synthesizing model was used to generate a synthetic data set, combined with the differential privacy (DP) method. RESULTS: A synthetic population of 5,005 was generated from a data set of 1,253 patients with 93 clinical features. The Hellinger distance and correlation difference metric were below 0.3 and 0.5, respectively, indicating no statistical difference. The overall survival by disease stage did not differ between the synthetic and original populations. Training with the synthetic data and validating with the original data showed the highest performances of 0.850, 0.836, and 0.790 for the Decision Tree, Random Forest, and XGBoost models, respectively. Comparison of synthetic data sets with different epsilon parameters from the original data sets showed improved performance >0.1%. For extremely small data sets, models using synthetic data outperformed those using only original data sets. The reidentification risk measures demonstrated that the epsilons between 0.1 and 100 fell below the baseline, indicating a preserved privacy state. CONCLUSION: The synthetic data generation approach enhances predictive modeling performance by maintaining statistical and clinical integrity, and simultaneously reduces privacy risks through the application of DP techniques.

Association of neutrophil-to-lymphocyte ratio, radiotherapy fractionation/technique, and risk of development of distant metastasis among patients with locally advanced rectal cancer.

BACKGROUND: We investigated the prognostic impact of the neutrophil-to-lymphocyte ratio (NLR) in patients with locally advanced rectal cancer (LARC) and whether modifiable factors in radiotherapy (RT) influenced the NLR. METHODS: Data of 1386 patients who were treated with neoadjuvant RT and concurrent or sequential chemotherapy for LARC between 2006 and 2019 were evaluated. Most patients (97.8%) were treated with long-course RT (LCRT; 50-50.4 Gy in 25-28 fractions) using three-dimensional conformal radiotherapy (3D-CRT) (n = 851) or helical tomotherapy (n = 504), and 30 patients underwent short-course RT (SCRT; 25 Gy in 5 fractions, followed by XELOX administration for 6 weeks). Absolute neutrophil and lymphocyte counts were obtained at initial diagnosis, before and during the preoperative RT course, and after preoperative concurrent chemoradiotherapy. The primary endpoint was distant metastasis-free survival (DMFS). RESULTS: The median follow-up time was 61.3 (4.1-173.7) months; the 5-year DMFS was 80.1% and was significantly associated with the NLR after RT but not before. A post-RT NLR ≥ 4 independently correlated with worse DMFS (hazard ratio, 1.42; 95% confidence interval, 1.12-1.80), along with higher ypT and ypN stages. Post-RT NLR (≥ 4) more frequently increased following LCRT (vs. SCRT, odds ratio [OR] 2.77, p = 0.012) or helical tomotherapy (vs. 3D-CRT, OR 1.29, p < 0.001). CONCLUSIONS: Increased NLR after neoadjuvant RT is associated with increased distant metastasis risk and poor survival outcome in patients with LARC. Moreover, high NLR following RT is directly related to RT fractionation, delivery modality, and tumor characteristics. These results are hypothesis-generating only, and confirmatory studies are required.

Deep-Learning-Based Natural Language Processing of Serial Free-Text Radiological Reports for Predicting Rectal Cancer Patient Survival.

Most electronic medical records, such as free-text radiological reports, are unstructured; however, the methodological approaches to analyzing these accumulating unstructured records are limited. This article proposes a deep-transfer-learning-based natural language processing model that analyzes serial magnetic resonance imaging reports of rectal cancer patients and predicts their overall survival. To evaluate the model, a retrospective cohort study of 4,338 rectal cancer patients was conducted. The experimental results revealed that the proposed model utilizing pre-trained clinical linguistic knowledge could predict the overall survival of patients without any structured information and was superior to the carcinoembryonic antigen in predicting survival. The deep-transfer-learning model using free-text radiological reports can predict the survival of patients with rectal cancer, thereby increasing the utility of unstructured medical big data.