ESTRO 2024 - Abstract Book

S5348

Radiobiology - Tumour biology

ESTRO 2024

798

Proffered Paper

Landscape of targeted drug/radiation responses across genomically diverse 3D NSCLC and HNSCC models

Henning Willers 1 , Xiao Pan 1 , Eva Volckova 1 , Nathalie Borgeaud 2 , Hanne Galleiske 2 , Daniel Zips 2,3 , Anjali Shinde 1 , Ibrahim Chamseddine 1 , Clemens Grassberger 1 , Beow Y Yeap 4 , Jong Kung 1 , Malcolm Mattes 5 , Michael Baumann 6 , Aaron Hata 7 , Christopher Ott 7 , Nils Cordes 2 , Cyril Benes 7 , Mechthild Krause 2 1 Massachusetts General Hospital, Radiation Oncology, Boston, USA. 2 University Hospital Carl Gustav Carus, Technische Universität Dresden, Helmholtz-Zentrum Dresden-Rossendorf, OncoRay – National Center for Radiation Research in Oncology, Dresden, Germany. 3 Charite, Radiation Oncology and Radiotherapy, Berlin, Germany. 4 Massachusetts General Hospital, Cancer Center, Boston, USA. 5 Rutgers Cancer Institute, Radiation Oncology, New Brunswick, USA. 6 German Cancer Research Center, Management Board, Heidelberg, Germany. 7 Massachusetts General Hospital, Center for Cancer Research, Charlestown, USA

Purpose/Objective:

Clinical translation of molecular targeted radiosensitizers remains challenging. To improve the robustness of preclinical data, we sought to examine the radiosensitizing properties of an unbiased panel of agents across genomically diverse non-small cell lung carcinoma (NSCLC) and head and neck squamous cell carcinoma (HNSCC) models.

Material/Methods:

We adapted a unique robotic high-throughput platform for testing ionizing radiation (IR) plus drug combinations in 3D cultures derived from NSCLC, HNSCC, and other cancer types. All models were annotated, mycoplasma free, and authenticated. Targeted agents covered a broad range of cellular pathways and had a clinical track record. Signal windows were established for all 3D growth formats. Drugs were added at four clinically relevant concentrations spanning 1.5 logs 24 hours prior to IR 2 Gy exposure of 384-well plates. Five days after IR or sham treatment, cell viability was assessed, an endpoint previously benchmarked using colony formation. Each data point was based on at least 8-12 technical replicates from two biological repeats. Radiosensitization factors were derived for each drug concentration using established protocol. A multistep statistical method combining Wilcoxon signed-rank and Spearman rank-order correlation tests together with adjusted Benjamini-Hochberg corrections for false discovery was employed to identify radiosensitization hits. Select secondary screens as well as in vitro and in vivo validation experiments of nominated combinations were carried out.

Results:

We screened 53 3D tumor models with IR combined with up to 42 drugs per cell line, for a total of 8,400 different combinations of drug concentration/cell line. IR/drug sensitivity patterns were comparable between NSCLC and HNSCC. For both cancer types, the majority of agents radiosensitized only a minority of models, with a total of 30% statistically significant hits. For example, in 32 NSCLC 3D sphere models, the data confirmed a previously identified