Painful-diabetic sensorimotor peripheral neuropathy (Painful-DPSN) is a common chronic complication of diabetes mellitus. Unfortunately, the condition is poorly understood and is inadequately treated. The studies included within this thesis aimed to improve our understanding of the mechanisms underlying the condition and its treatment, with a particular focus upon the central nervous system.
Cross-sectional multi-modal imaging studies were performed to determine the cerebral structural and functional alterations in the brain in DSPN. In the largest cerebral imaging in DSPN, volumetric analysis showed group differences in cortical thickness at the primary somatosensory, primary motor and the insular cortices. There was greater reduction in cortical thickness in painless- compared with painful-DSPN at the primary somatosensory and insular cortices; however, at the primary motor cortex the reduction in cortical thickness was similar for both DSPN groups compared to the non-neuropathy groups. The cortical thickness correlated with age, as well as measures of neuropathy severity. Moreover, in sub-group analysis of painful-DSPN participants there was a significant reduction in anterior cingulate cortical thickness in the irritable nociceptor compared with the non-irritable nociceptor phenotype.
Proton and 31-phosphorus magnetic resonance spectroscopy (1H- and 31P-MRS, respectively) neurometabolites were then investigated to determine cerebral neuronal function in the primary somatosensory cortex and thalamus in participants with type 2 diabetes mellitus (T2DM) and DSPN. This study demonstrated reduced NAA:Cho, indicating neuronal dysfunction, in the dominant hemisphere thalamus in painless-DSPN compared to painful-DSPN and HV. The NAA:Cho in painful-DSPN was comparable with HV at this region. Additionally, the NAA:Cho at the dominant thalamus correlated with body mass index, blood glucose taken just before the magnetic resonance imaging (MRI) and peroneal motor nerve conduction velocity. Also, were no group differences in any neurometabolites at the somatosensory cortex. These results suggest that there is preservation of neuronal function in the thalamus in painful-DSPN, perhaps as a result of persisting painful neuronal signalling, essential for the perception of pain.
Using 31P-MRS at the somatosensory cortex, the phosphocreatine to ATP ratio (PCr:ATP) a marker of energy usage, was numerically the lowest in painful-DSPN, reaching significance compared with HV and painless-DSPN. The PCr:ATP ratio correlated with a number of measures of neuropathic pain and was lower in participants with a higher Numeric Rating Scale (NRS) pain score during MRI, indicating that there is higher energy usage with higher levels of pain. The pattern of thalamic 31P-MRS metabolite ratios was different in the thalamus, with markers of mitochondrial function, the inorganic phosphate (Pi) to PCr ratio (Pi:PCr) and Pi:ATP, being numerically the highest in painless-DSPN and reaching significance versus HV and painful-DSPN. These ratios correlated more with metabolic measures, rather than neuropathy/neuropathic pain measures. There were also correlations seen with 31P and 1H neurometabolites, indicating a correlation with neuronal function and cerebral energetics. In this first ever study using 31P-MRS in clinical DSPN, these results suggest differing bioenergetic processes occurring in the thalamus and the somatosensory cortex. Mitochondrial dysfunction predominated at the thalamus, with preservation of function in painful-DSPN probably due to persisting neuronal impulses. At the somatosensory cortex, however, there was a diabetes effect upon this region with evidence of reduced energy usage in painless-DSPN.
Two pre-post neuroimaging studies were undertaken in participants undergoing the OPTION-DM trial, a neuropathic pain treatment randomised controlled trial to determine the most efficacious pathway for the management of painful-DSPN. Cerebral images were taken at the end of a treatment pathway within the study when participants were optimally treated for their neuropathic pain, and again one-week later after medication had been withdrawn. There were no significant differences in neurotransmitter levels; although, methodological limitations may have contributed to the negative findings in the study. However, resting state function MRI analysis showed a significant increase in functional connectivity between the left thalamus and somatosensory cortex and left thalamus and insular cortex after withdrawal of neuropathic pain medications. The change in thalamic to somatosensory cortical functional connectivity also correlated with severity of baseline pain numeric rating and baseline total neuropathic pain symptom inventory scores. This study therefore further highlights the importance of the function of the thalamus as a pivotal pain processing centre in painful-DSPN. Moreover, functional connectivity of the thalamus to other regions of the brain may act as biomarkers of pain in clinical trials. However, future validation is required in prospective studies.
The final study in this thesis performed thigh skin biopsies in four well characterized groups with the aim to determine whether neuronal or vascular markers in the skin differentiate painful- and painless-DSPN. The study demonstrated there was a significant increase in the vascular marker von Willebrand Factor (vWF) in painful-DSPN compared with all other groups. This confirms the findings of a previous study within the research group when biopsies were taken at the ankle. Further mechanistic studies are necessary to determine the cause of the elevated vWF in the skin of patients with painful-DSPN; however, vWF may act as a peripheral marker of painful-DSPN in the future.
Overall, these studies highlight key mechanisms of cerebral involvement in painless- and painful-DSPN. Further mechanistic studies, particularly prospective with longitudinal imaging are necessary to determine the underlying mechanisms of neuroplasticity in DSPN
