American Association for Cancer Research, Clinical Cancer Research, 6(27), p. 1641-1649, 2020
DOI: 10.1158/1078-0432.ccr-20-3580
Full text: Download
Abstract Purpose: Insufficient control of infiltrative glioblastoma (GBM) cells is a major cause of treatment failure and tumor recurrence. Hence, detailed insights into pathophysiologic changes that precede GBM recurrence are needed to develop more precise neuroimaging modalities for tailored diagnostic monitoring and therapeutic approaches. Experimental Design: Overall, 168 physiologic MRI follow-up examinations of 56 patients with GBM who developed recurrence after standard therapy were retrospectively evaluated, that is, two post–standard-therapeutic follow-ups before and one at radiological recurrence. MRI biomarkers for microvascular architecture and perfusion, neovascularization activity, oxygen metabolism, and hypoxia were determined for brain areas that developed in the further course into recurrence and for the recurrent GBM itself. The temporal pattern of biomarker changes was fitted with locally estimated scatterplot smoothing functions and analyzed for pathophysiologic changes preceding radiological GBM recurrence. Results: Our MRI approach demonstrated early pathophysiologic changes prior to radiological GBM recurrence in all patients. Analysis of the time courses revealed a model for the pathophysiology of GBM recurrence: 190 days prior to radiological recurrence, vascular cooption by GBM cells induced vessel regression, detected as decreasing vessel density/perfusion and increasing hypoxia. Seventy days later, neovascularization activity was upregulated, which reincreased vessel density and perfusion. Hypoxia, however, continued to intensify for 30 days and peaked 90 days before radiological recurrence. Conclusions: Hypoxia may represent an early sign for GBM recurrence. This might become useful in the development of new combined diagnostic–therapeutic approaches for tailored clinical management of recurrent GBM. Further preclinical and in-human studies are required for validation and evaluation.