Click here to


Are you sure ?

Yes, do it No, cancel

Predicting Tumor Control Using Geometric Features of Hypoxia Measured with EPRI

H Smith1*, I Gertsenshteyn2,3, B Epel2,3, E Barth2,3, M Maggio3, S Sundramoorthy2,3, H Halpern2,3, (1) Department of Radiology, University of Chicago (2) Department of Radiation and Cellular Oncology, University of Chicago (3) National Institutes of Health Center for EPR Imaging In Vivo Physiology, University of Chicago, Chicago,IL


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: determine correlations between geometric features of hypoxia in FSa fibrosarcomas and tumor control probability in mice after radiation therapy treatment.

Methods: Oxygen images of 250 – 450 mm³ FSa tumors in mice (n=55) were obtained with electron paramagnetic resonance imaging (EPRI). Hypoxic regions were defined by voxels with pO2 = 10 mmHg. Tumors were treated with either a boost dose of radiation to hypoxic regions or a boost dose to an equal volume of well-oxygenated tumor. Local tumor recurrences were tracked over 90 days. Surface-to-volume ratio (SVR) and irregularity of the largest hypoxic volume were calculated in MATLAB and divided into “high” and “low” groups using the median values as the threshold. Kaplan-Meier curves were generated for each feature to identify whether SVR or irregularity were predictors for tumor control for both radiation treatment methods.

Results: FSa tumors with low SVR hypoxic regions had significantly higher tumor control probability when including both treatment types (p=0.02) and for hypoxia boost treatments (p=0.002); this correlation was not significant for well-oxygenated tumor boost cases. Irregularity of the hypoxic region did not significantly correlate with tumor control probability.

Conclusion: Of the two geometric features, SVR was a better predictor of tumor recurrence, while irregularity of the hypoxic region was not predictive of tumor control for the thresholds used in this study. Specifically, low SVR hypoxia was correlated with greater tumor control. More data are needed to verify any significance of irregularity correlation with tumor control, and further study of other geometric features of hypoxia are currently underway.

Download ePoster [PDF]

Funding Support, Disclosures, and Conflict of Interest: Funding: P41EB002034, R01CA098575, R01CA236385, P30CA014599, and the AAPM/RSNA Graduate Fellowship. US patents 8,664,955 and 9,392,957 were awarded to Howard Halpern for aspects of the pO2 imaging technology. Drs. Halpern and Epel are members of O2M to market the pO2 imaging technology.


Hypoxia, Quantitative Imaging, Image Analysis


IM/TH- Image Analysis (Single Modality or Multi-Modality): Imaging biomarkers and radiomics

Contact Email