Room: AAPM ePoster Library
Purpose: This research was conducted to determine if hyperpolarized magnetic resonance imaging (MRI) could effectively inform on tumor viability throughout tumor regression following radiotherapy to the point of relapse and whether it was more sensitive than conventional anatomical MRI techniques.
Methods: Athymic mice orthotopically implanted with patient-derived glioblastoma (GSC 8-11) were used as the model in this study and possessed a median survival time of 34 days. Each week, tumor volume was measured using T2-weighted MRI, and cellular energy metabolism was measured in vivo using hyperpolarized MRI. Specifically, hyperpolarized [1-¹³C]pyruvate was injected, and its uptake and conversion to lactate were dynamically measured to calculate nLac which is used as a biomarker in these imaging studies. On Days 25 and 27, tumors were irradiated to 5 Gy. T2-weighted and hyperpolarized MRI were performed on the treated mice throughout tumor regression up to early signs of relapse on Day 68. One-way ANOVA and linear mixed models determined statistically significant differences in these assays between time-points.
Results: Following radiotherapy, tumor growth was shunted and even appeared to slightly decrease by Day 62 before starting to regrow. However, no significant differences in volume compared to pre-treatment values were observed. From the hyperpolarized MRI experiments, nLac in treated mice was significantly reduced on Days 48 and 55 compared to Days 28 and 34. The tumors then began to recur, and nLac was significantly increased by Day 68 compared to Days 48 and 55.
Conclusion: This data demonstrates that hyperpolarized MRI can detect significant changes in tumor metabolism following radiotherapy as the tumor responds throughout regression as well as in the early stages of relapse as the tumor begins to recur. Furthermore, this technique is more sensitive than conventional T2-weighted MRI as changes in volume over this time-course occurred later and were not statistically significant.
Funding Support, Disclosures, and Conflict of Interest: TS was supported by CPRIT Research Training Grant Award (RP170067)