Abstract Background:In breast cancer (BC), radiotherapy (RT) is used adjuvantly to prevent recurrence and also in the palliative setting. Clinical signs of RT response are often not apparent for several weeks post-treatment and we currently lack tools to predict or monitor tumor response to RT early during treatment. The aim was to identify tumor-secreted biomarkers whose release reflects response to RT, which could be monitored during treatment in the blood or intratumorally by an implantable biosensor, currently under development within the Implantable Microsystems for Personalised Anti-Cancer Therapy (IMPACT) program. Methods: A series of experiments assessed the effect of different radiation doses (2-10Gy) on 3 human BC cell lines – MDA-MB-231 (ER-), MCF-7 (ER+) and HBL-100 (ER-) –, 1 canine breast cancer and 2 sheep lung cancer lines. Culture media was collected from each dose experiment at a range of post-radiation time-points (1-24 hours). Proteins were isolated from collected media for secretome mass spectrometry (MS) analysis. A subset of treatment/time conditions were repeated in the same BC cell lines and radioresistant (RR) derivatives from which RNA was extracted and analysed using Lexogen QuantSeq for whole-genome transcriptomics.In-lab candidate biomarker validation was carried out using immuhistochemistry (IHC), immunofluorescence (IF) and western blotting (WB) using validated antibodies. Levels of candidate biomarkers were also assessed in normal and untreated BC tissues using IHC. ELISA-based methods are currently under investigation for detection of the lead candidate biomarkers in the blood of large animal cancer models treated with RT. Results: Biomarker discovery using the MS data revealed 4 promising candidates: EIF3G, SEC24C, YBX3 and TK1. These are released from BC and animal cancer cells sensitive to radiation in a dose-dependent manner 24 hours after treatment. Analysis of the transcriptomic data showed an 8-fold higher expression of the genes encoding the 4 candidates in the radio-sensitive parental cell lines compared to the RR cell lines. IF and WB confirmed lower intracellular expression of the 4 proteins in RR cells compared to the parental lines. WB of collected culture media confirmed release of each of the 4 candidates 24 hours after a 2Gy dose of radiation in only the parental lines. GAPDH was not found in these media samples, demonstrating that protein release was not due to cell lysis. Conclusions: · We have identified 4 promising biomarkers which are released from cancer cells sensitive to RT and not released from RR derivatives. · All 4 candidates are released 24 hours after a 2Gy radiation dose, which fits with the current clinical dosing schedule where radiation is administered at 24 hour intervals. Ongoing work will elucidate if these biomarkers can be reliably detected in blood or intratumorally using implantable biosensors. · There are currently no validated predictive tools to monitor RT response during treatment. If successfully validated, these biomarkers could have a clinical role in personalising RT dosing schedules and durations for solid tumors in the neoadjuvant and palliative setting, thus optimising treatment and preventing the administration of ineffective RT and its associated side effects. Citation Format: Meehan J, Gray M, Turnbull AK, Martinez-Perez C, Bonello M, Ward C, Langdon SP, McLaughlin S, MacLennan M, Dixon JM, Wills J, Quinn N, Finich AJ, von Kriegsheim A, Cameron D, Kunkler IH, Murray A, Argyle D. Tumor-secreted predictive biomarkers of response to radiotherapy in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-12-24.
Read full abstract