Evaluation of tumour vascular distribution and function using immunohistochemistry and BOLD fMRI with carbogen inhalation. Issue 12 (December 2016)

• Record Type: Journal Article
• Authors: Xu, J.; Chen, A.; Xiao, J.; Jiang, Z.; Tian, Y.; Tang, Q.; Cao, P.; Dai, Y.; Krainik, A.; Shen, J.
• Is Part Of: Clinical radiology. Volume 71:Issue 12(2016)
• Page Start: 1255
• Page End: 1262
• Publication Date: 2016-12
• Subjects: Medical radiology -- Periodicals; Radiotherapy -- Periodicals; Radiotherapy -- Periodicals; Radiology -- Periodicals; Societies, Medical -- Periodicals; Medical radiology; Radiotherapy; Electronic journals; Periodicals; 616.0757
• DOI: 10.1016/j.crad.2016.04.012
• Languages: English
• ISSNs: 0009-9260

Abstract

Abstract : Aim: To evaluate oxygenation changes in rat subcutaneous C6 gliomas using blood-oxygen-level dependent (BOLD) functional magnetic resonance imaging (fMRI) combined with non-haemodynamic response function (non-HRF) analysis. Materials And Methods: BOLD fMRI were performed during carbogen inhalation in 20 Wistar rats bearing gliomas. Statistical maps of spatial oxygenation changes were computed by a dedicated non-HRF analysis algorithm. Three types of regions of interest (ROIs) were defined: (1) maximum re-oxygenation zone (ROImax ), (2) re-oxygenation zones that were less than the maximum re-oxygenation (ROInon-max ), and (3) zones without significant re-oxygenation (ROInone ). The values of percent BOLD signal change (PSC), percent enhancement (ΔSI), and significant re-oxygenation (T) were extracted from each ROI. Tumours were sectioned for histology using the fMRI scan orientation and were stained with haematoxylin and eosin and CD105. The number of microvessels (MVN) in each ROI was counted. Differences and correlations among the values for T, PSC, ΔSI, and MVN were determined. Results: After carbogen inhalation, the PSC significantly increased in the ROImax areas ( p <0.01) located in the tumour parenchyma. No changes occurred in any of the ROInone areas (20/20). Some changes occurred in a minority of the ROInon-max areas (3/60) corresponding to tumour necrosis. MVN and PSC ( R =0.59, p =0.01) were significantly correlated in the ROImax areas. In the ROInon-max areas, MVN was significantly correlated with PSC ( R =0.55, p =0.00) and ΔSI ( R =0.37, p =0.00). Conclusions: Statistical maps obtained via BOLD fMRI with non-HRF analysis can assess the re-oxygenation of gliomas. Highlights: BOLD fMRI non-invasively detects oxygenation in the tumor to evaluate hypoxia associated with poor prognoses of brain tumor. Statistical maps from a non-hemodynamic-response-function analysis contain spatial oxygenation and its change extent in tumor. Monitoring tumor oxygenation contributes to implementation of oxygen sensitization treatment. Hypoxia in tumors can be used to project the recurrence of cancer.