Purpose

To investigate the topological organization alteration of whole-brain functional network in patients with postherpetic neuralgia (PHN) after repetitive transcranial magnetic stimulation (rTMS) treatment and assess whether the functional alteration could be used as a neural biomarker for the post-treatment evaluation.

Materials and methods

Totally 10 patients (male/female:5/5,Age:68.4±4.72) with PHN who received rTMS treatments were recruited in this study. High-resolution T1-weighted and functional MRI images were acquired before and 2 weeks after the rTMS. All the images were acquired based on a 3T scanner (Ingenia, Philips, The Netherlands) and a 15-channel head coil for MR signal reception. For 3D T1w imaging, the protocols were as follows: 3D FFE acquisition, Flip Angle=15, TR/TE=4.5ms/2.1ms, FOV= 240mm*240mm, Image resolution=1mm*1mm, Slice thickness=1mm with no slice gap, Matrix size=240*240, Slice number=160, Sense factor=1.3, phase encoding direction=AP, NSA=1, scan time=4min23s. For fMRI imaging, the protocols were as follows: Multi-slice FE-EPI acquisition with SPIR fat suppression, Flip Angle=90, TR/TE=2000ms/35ms, FOV=230mm*230mm, Image resolution=3mm*3mm , Slice thickness=3.5mm with no slice gap, Matrix size=76*76, Slice number=33, Sense factor=2, Phase encoding direction=AP, NSA=1, scan time=240s with 120 dynamic volumes. The subjects were asked to keep eyes closed and think nothing when scanned. Then, whole-brain functional connectivity networks were constructed by using bi-variates Pearson’s correlation. Network parameters including small-worldness, characteristic path length, clustering coefficient, global efficiency, node degree, node betweenness, node efficiency which can reflect the topology of brain network were calculated to investigate the characteristics of whole-brain functional communication. Non-parametric paired signed rank tests were performed for the above network parameters with sex and age as co-variates. A p<0.05 (with FDR correction for multi-comparison analysis) indicated a significant statistical difference. Correlation analysis were performed between the network parameters and clinical variables.

Results

After the rTMS treatment, there showed some significant differences for characteristic path length, clustering coefficient, global efficiency derived from the networks at some specific network sparsity but showed no significant difference for small-worldness. For node degree, there showed significantly increasing degrees in the brain regions of DCG.L and CRBL10.R and decreasing degrees in the brain regions of PreCG.L, PreCG.R, ACG.R, SMA.L, PCL.L, Vermis3 and Vermis45. For node bewteenness, there showed significantly increasing values in the brain regions of SMG.R, CAU.R, ITG.R, OLF.R, TPOmid.R, TPOsup.R and decreasing values in the brain regions of PCL.L, THA.L, LING.L, SFGmed.L, AMYG.R, PHG.L and Vermis3. For node efficiency, there showed significantly decreasing values in the brain regions of PreCG.L, PreCG.R, ACG.L,ACG.R, SMA.L,SMA.R, PCL.L,Vermis3 and Vermis45 and showed no regions with increasing values. For pre- and post-treatment small-worldness, characteristic path length, clustering coefficient, global and local efficiency, there showed no significant correlation with pre-treatment VAS, post-treatment VAS, ΔVAS (pre-treatment VAS minus post-treatment VAS), or pain duration. For node degree, node betweenness and node efficiency of pre-treatment, post-treatment and the difference value of pre- and post-treatment, there showed some significant correlation (P<0.05) with pre-treatment VAS, ΔVAS (pre-treatment VAS minus post-treatment VAS), or pain duration. The correlation of node-level parameters with VAS, ΔVAS and pain duration were list in Table 2-4 respectively.

Discussion

In general, brain functional network has the small-world architecture with high local clustering coefficient and small characteristic path length which supports a well-balance of brain functional segregation and integration. The results showed no any changes of small-worldnees after rTMS which indicated that the brain network might keep a stable functional communication condition even some treatment was given. This finding might be supported by the previous works which found no small-worldnees changes in the PHN patients compared to normal controls. In this work, node degree was defined as the total weighted value of the direct connection strength between the present node with all other nodes. Node betweeness was defied as the ratio of shortest information transferring path number between any other two nodes that pass through the present node and the total shortest information transferring path number between the two nodes. Node efficiency was defined as the averaged value of the reciprocal of the distance between the present node and the other node in a network. All these node level changes were showed in a decreasing trend which indicated that the rTMS treatment could modulate and reorganize the sensory-motor network, emotion and affective/memory circuits through the simulation on motor cortex.

Conclusion

The state of whole-brain functional network was altered in patients with postherpetic neuralgia after rTMS treatment. And these alterations might be used a neural biomarker for the post-treatment evaluation.

Acknowledgements

No acknowledgement found.