Background & Aims

Neurofibromatosis type 1 (NF1) is a genetic disorder characterized by a high tumor burden in the nervous system and a variety of symptoms, including neuropathic pain. At least one third of all NF1 patients suffer from typical painful NF-associated neuropathy. Pain research must focus, among other things, on the structure of the dorsal root ganglia (DRG), which is essential for the transmission and processing of pain from the periphery to the central nervous system. Understanding changes in the DRG in NF1 patients with pain is key to understanding the onset and resolution of pain in NF1. Previous studies have suggested DRG involvement in neurofibromatosis, but a detailed evaluation specific to NF1 is needed. We used DRG MRI, specifically T2 mapping, to investigate lumbosacral DRG changes and their association with pain in NF1 patients in vivo. Our goal was to identify potential imaging biomarkers for neuropathic pain in NF1 at the DRG level.

Methods

Twenty NF1 patients were included in the study and divided into groups according to whether they had pain (NF1p: n=9) or not (NF1n: n=11). Twenty healthy subjects without pain syndrome or any form of neurological or metabolic disease served as the control group (hc). All participants underwent imaging of the lumbosacral DRGs L5 and S1 using a 3 Tesla scanner. The DRG MRI protocol included T2 mapping using a multi-echo spin-echo sequence (?TE range 15-150 ms). Image post-processing was performed using MATLAB. Maps for T2 relaxation time (T2) and initial magnitude of magnetization as an indicator of proton spin density (PD) were fitted. 3D DRG ROIs were selected voxel-wise by two independent and blinded raters. An average value for the parameters DRG volume, DRG T2 and DRG PD was calculated for each subject. Statistical analyses included group comparisons using Kruskal-Wallis test and post-hoc pairwise Wilcoxon test with Benjamini-Hochberg p-value adjustment and ROC analysis.

Results

The DRG of NF1p, NF1n, and hc showed significant differences in volume (p<.001), T2 (p=.01), and PD (p=.01). The NF1p group had significantly higher values for volume (median 2819.8 vs. 647.4 mm3, p<.001), T2 (124.2 vs. 96.4 ms, p=.01), and PD (98.4 vs. 85.6 a.u., p=.02) compared to hc, and also a significantly increased PD (98.4 vs. 84.7 a.u., p=.02) compared to NF1n. DRG volume was significantly higher for the NF1n group compared to hc (1716.1 vs. 647.4 mm3, p<.001). T2 tended to be higher for the NF1n group compared to hc (104.6 vs. 96.4 ms, p=n.s.). PD was similar for these both groups (84.7 vs. 85.6 a.u., p=n.s.). For discriminating between pain and no pain in NF1 patients, the PD parameter (AUC=.83) was superior to T2 (AUC=.71) and volume (AUC=.68). A PD threshold of 83.8 a.u. was identified to offer a balanced trade-off between sensitivity (80.0%) and specificity (88.9%) for discriminating NF1p from NF1n.

Conclusions

Our exploratory study identified significant differences in DRG characteristics between NF1 patients with or without pain and healthy individuals. Changes in DRG morphology or volume may be due to tumor infiltration, as seen in previous histologic studies. Notably, 30-50% of NF1 patients have plexiform neurofibromas and these individuals are more likely to suffer from polyneuropathy. T2 mapping revealed changes in the microstructure of the DRG. In the NF1p group, both T2 and PD values were increased. Increased T2 is primarily associated with an increase in free water protons or endoneurial edema, whereas changes in PD suggest microstructural changes in the extracellular nerve matrix caused by increased plasma protein leakage and an inflammatory environment. These mechanisms are likely to contribute to the development of neuropathic pain in NF1 patients.

References

– Godel, T. et al. Dorsal root ganglia volume differentiates schwannomatosis and neurofibromatosis 2. Ann. Neurol. 83, 854–857 (2018).
– Hallisey, M. et al. Plexiform neurofibroma involving a dorsal root ganglion. Skeletal Radiol. 18, 314–317 (1989).
– Matthies, C. et al. Neuropathy in neurofibromatosis types 1, 2 and 3: a study protocol to address a neglected feature. Brain Spine 1, 100763 (2021).
– Neumann, D. et al. A. Simple recipe for accurate T(2) quantification with multi spin-echo acquisitions. MAGMA 27, 567–577 (2014).
– Sollmann, N. et al. High Isotropic Resolution T2 Mapping of the Lumbosacral Plexus with T2-Prepared 3D Turbo Spin Echo. Clin. Neuroradiol. 29, 223–230 (2019).

Presenting Author

Simon Weiner

Poster Authors

Simon Weiner

MD

University Hospital Würzburg

Lead Author

Eva Rampeltshammer

Dr. med.

University Hospital Wuerzburg

Lead Author

Thomas Kampf

University Hospital Wuerzburg, Neuroradiology

Lead Author

Cordula Matthies

Prof. Dr. med.

University Hospital Wuerzburg

Lead Author

Mirko Pham

Prof. Dr. MD

Institute for neuroradiology, university hospital Würzburg, Germany

Lead Author

Magnus Schindehütte

Neuroradiologie Universitätsklinikum Würzburg

Lead Author

Topics

  • Pain Imaging