Altered Regional Brain Glucose Metabolism in Diffuse Large B-Cell Lymphoma Patients Treated With Cyclophosphamide, Epirubicin, Vincristine, and Prednisone: An Fluorodeoxyglucose Positron Emission Tomography Study of 205 Cases

Abstract

Background: A growing number of neuroimaging studies reported that chemotherapy might impair brain functions, leading to persistent cognitive alterations in a subset of cancer patients. The present study aimed to investigate the regional brain glucose metabolism differences between diffuse large B cell lymphoma (DLBCL) patients treated with cyclophosphamide, epirubicin, vincristine, and prednisone and controls using positron emission tomography with ¹⁸F-labeled fluoro-2-deoxyglucose integrated with computed tomography (¹⁸F-FDG PET/CT) scanning.

Methods: We analyzed ¹⁸F-FDG PET data from 205 right-handed subjects (for avoiding the influence of handedness factors on brain function), including 105 post-chemotherapy DLBCL patients and 100 controls. The two groups had similar average age, gender ratio, and years of education. First, we compared the regional brain glucose metabolism using a voxel-based two-sample t-test. Second, we compared the interregional correlation. Finally, we investigated the correlations between the regional brain glucose metabolism and the number of chemotherapy cycles.

Results: Compared with the controls, the post-chemotherapy group showed higher metabolism in the right hippocampus and parahippocampal gyrus (region of interest (ROI) 1) and the left hippocampus (ROI 2), and lower metabolism in the left medial orbitofrontal gyrus (ROI 3), the left medial superior frontal gyrus (ROI 4), and the left superior frontal gyrus (ROI 5). The two groups had different interregional correlations between ROI 3 and ROI 5. In some brain regions-mainly located in the bilateral frontal gyrus-the number of chemotherapy cycles was positively correlated with the regional brain glucose metabolism. Meanwhile, in some bilateral hippocampus regions, these two parameters were negatively correlated.

Conclusion: The present study provides solid data on the regional brain glucose metabolism differences between post-chemotherapy DLBCL patients and controls. These results should improve our understanding of human brain functions alterations in post-chemotherapy DLBCL patients and suggest that ¹⁸F-FDG PET/CT scanning is a valuable neuroimaging technology for studying chemotherapy-induced brain function changes.

Keywords: 18F-FDG; PET; brain glucose metabolism; chemotherapy; diffuse large B cell lymphoma.

FIGURE 1

Difference maps of regional brain glucose metabolism between the post-chemotherapy group and the control group. Red to yellow areas represent the regions where the brain glucose metabolism values were higher in the post-chemotherapy group, while blue to cyan areas represent the opposite (p < 0.05, FDR correction). (A) Axial image showing the ROI 1 (white arrow) in the right hippocampus and parahippocampal gyrus, and ROI 2 (black arrow) in the left hippocampus. (B) Coronal image showing the ROI 1 and the ROI 2. (C) Sagittal image showing the ROI 1. (D) Axial image showing the ROI 3 in the left medial orbitofrontal gyrus. (E) Axial image showing the ROI 4 in the left medial superior frontal gyrus. (F) Axial image showing the ROI 5 in the left superior frontal gyrus. (G) Sagittal image showing ROI 3, 4, and 5 from bottom to top.

FIGURE 2

Differences in interregional correction between the post-chemotherapy group and the control group. Regional brain glucose metabolism values in the ROI 3 correlated with those in the ROI 5 for the control group (r = 0.44, p < 0.0001). No significant correlation emerged in the post-chemotherapy group (r = 0.19, p = 0.059). The interaction linear-model analysis showed that the post-chemotherapy group and the control group had different interregional correlation of regional brain glucose metabolism values (uncorrected p < 0.05). LAC: lymphoma patients after chemotherapy.

FIGURE 3

Correlation between the regional brain glucose metabolism values and the number of chemotherapy cycles. (A–C) Positive correlation between the number of chemotherapy cycles and the regional brain glucose metabolism in the bilateral medial superior frontal gyrus, the bilateral superior frontal gyrus, the right middle frontal gyrus, the bilateral anterior cingulate cortex, and the bilateral supplementary motor area. (D–G) Negative correlation between the number of chemotherapy cycles and the regional brain glucose metabolism in the bilateral hippocampus, the bilateral orbitofrontal cortex, and the bilateral cerebellum.