publications

2026

1. To be published: Subtyping Limited Stage Small Cell Lung Cancer (LS-SCLC) Reveals Inflammatory Subtypes with Improved Chemoradiotherapy Outcomes

   Rafael Oliveira, Tarje Halvorsen, Kristin Killingberg, and 2 more authors

   2026

   Abs

   Background: Concurrent chemoradiotherapy (CRT) remains the primary treatment for limited stage (LS) small cell lung cancer (SCLC). Although most patients respond, the majority relapse and die from the disease, and treatment-related toxicity is common. Molecular subtyping has been extensively investigated in SCLC, but its clinical relevance remains uncertain, and most studies have focused on systemic treatment of extensive-stage (ES) SCLC. We analyzed baseline samples from a subset of participants (n=77) in a randomized trial (n=170) comparing high-dose (60 Gy in 40 fractions) with standard-dose (45 Gy in 30 fractions) twice-daily (BID) thoracic radiotherapy (TRT) given concurrently with platinum-etoposide chemotherapy for LS SCLC. This study aims to characterize molecular subtypes of LS-SCLC, particularly SCLC-A (ASCL1-driven), SCLC-N (NEUROD1-driven), SCLC-I-NE (inflamed neuroendocrine) and SCLC-I-nNE (inflamed non-neuroendocrine), and evaluate their association with time to progression (TTP) and overall survival (OS). Methods: Molecular subtypes were defined in 81 samples using non-negative matrix factorization (NMF) on gene expression data, supported by differential expression and known SCLC markers. Results: Inflamed subtypes were associated with better outcomes from CRT. Specifically, SCLC-I-NE was associated with longer TTP (HR 0.10, 95% CI 0.01–0.77, p=0.027), while SCLC-I-nNE was associated with improved OS (HR 0.16, 95% CI 0.03–0.77, p=0.023). In contrast, SCLC-N was consistently associated with worse outcomes, showing shorter TTP (HR 6.77, 95% CI 1.46–31.4, p=0.015) and poorer OS (HR 4.87, 95% CI 1.24–19.1, p=0.023). Multivariable analyses adjusting for clinical covariates demonstrated similar trends. Conclusions: Among patients receiving chemoradiotherapy for LS SCLC, there were indications of improved outcomes among those with tumors of inflamed subtypes, whereas SCLC-N was linked to unfavorable TTP and OS.

2. To be published: Genomic Radiosensitivity Reveal Subtype Differences and Predict Outcomes in SCLC

   Rafael Oliveira, Tarje Halvorsen, Kristin Killingberg, and 2 more authors

   2026

   Abs

2024

1. Radiomics and Radiogenomics Platforms Integrating Machine Learning Techniques: A Review

   Rafael Oliveira and Beatriz Martinho

   Information Systems and Technologies, Feb 2024

   Abs DOI Bib

   Radiomics and radiogenomics are still new fields to be explored in oncology, although there are several platforms and tools already developed, or in development. This review aimed to identify the current state of the art of radiomics and radiogenomics in terms of platforms and tools integrating machine learning techniques to support the analysis of medical imaging studies of oncological patients. A systematic literature search was performed, and 19 studies were included in the review. These studies were published between 2016 and 2022 and considered several platforms and tools integrating machine learning techniques for the analysis of medical imaging studies of oncological patients together with molecular and clinical endpoints. However, the included articles present scarce evidence related to the translation of the reported platforms and tools to the clinical practice.

   @article{ro,
     title = {Radiomics and Radiogenomics Platforms Integrating Machine Learning Techniques: A Review},
     author = {Oliveira, Rafael and Martinho, Beatriz},
     year = {2024},
     month = feb,
     journal = {Information Systems and Technologies},
     doi = {10.1007/978-3-031-45648-0_42},
     url = {https://link.springer.com/chapter/10.1007/978-3-031-45648-0_42},
     publisher = {Information Systems and Technologies,}
   }