3042 - Genomic Biomarkers of Radiation Sensitivity

Purpose/Objective(s): Radiation dosing has historically taken a one size fits all approach and is typically based on clinical factors such as tumor histology or tumor size. Some efforts have been made to better personalize radiation dose, either through escalation or de-escalation, with mixed results. Recently, there have been translational efforts to understand the relationship between tumor genomics and radiation response in the hopes of better personalizing care, either by augmenting dose in situations where resistance is predicted to increase local control, or de-escalating dose in situations of sensitivity to allow for decreased toxicity. Here, we investigate the role of tumor mutational profile and risk of local recurrence following radiation therapy.

Materials/Methods: Patients treated with a course of radiation therapy who underwent next generation sequencing of tumor specimens were included. Follow up imaging was assessed for local recurrence, which was defined as growth of the radiated area on two successive scans. ClinVar database was used to assess for pathogenic mutations. Cumulative incidence curves were calculated and stratified by mutational status to assess impact on rates of local control and adjusted for radiation dose.

Results: A total of 398 patients were included in the analysis. Tumor histologies included lung (38.7%), pancreas (29.9%), colorectal (17.1%), and prostate (11.3%). Cumulative incidence of a local failure at 24 months was 39.4%. TP53 mutations were associated with higher cumulative incidence of local failure at 24 months post radiation (46.2% vs 27.7%, p = 0.01). Similarly, alterations within SMAD4 (50% vs 38%, p = 0.03) were also associated with higher rates of local recurrence post radiation. BRCA1/2 mutations were associated with lower incidence of local recurrence at 24 months post radiation (13.5% vs 40.9%, p = 0.03). Wnt pathway (37.4% vs 40%, p = 0.40), KRAS (43.9% vs 35.6%, p = 0.26) and PIK3 pathway (38.7% vs 40%, p = 0.99) mutations were not associated with differences in rates of local recurrence. TP53 (HR 1.45, p = 0.05), SMAD4 (HR 1.54, p = 0.08), and BRCA1/2 (HR 0.32, p = 0.05) alterations remained associated with risk of local recurrence after adjusting for biological equivalent dose (BED).

Conclusion: Incidence of local recurrence post radiation varied based on tumor mutational profile. TP53 and SMAD4 mutations were associated with radioresistance while BRCA1/2 mutations were associated with radiosensitivity. These findings might allow for future radiation dose personalization and need to be validated in a prospective manner.