Mechanisms of Axonal Pathophysiology in Diabetes and Chronic Kidney Disease
Tushar Issar
A thesis submitted in fulfilment of the requirement for the
degree of Doctor of Philosophy
Prince of Wales Clinical School
University of New South Wales
July 2020 Thesis/Dissertation Sheet Australia's Global University
Surname/Family Name lssar Given Name/s Tushar Abbreviation for degree as give in the University calendar PhD Faculty Medicine School Prince of Wales Clinical School Thesis Title Mechanisms of axonal pathophysiology in diabetes and chronic kidney disease
Abstract 350 words maximum: (PLEASE TYPE) Peripheral neuropathy is a common and debilitating complication of diabetes and chronic kidney disease (CKD). The pathophysiological mechanisms contributing to peripheral neuropathy in these conditions remain unclear. In this thesis, axonal excitability studies and nerve ultrasonography were utilised to assess peripheral nerve structure and function in human subjects with either diabetes, CKD, or both to investigate the mechanisms underlying peripheral nerve dysfunction in each condition.
To commence, it was essential that an instrument to assess the severity of neuropathy in patients CKD with and without type 2 diabetes (T2DM) was formally validated. Having validated The Total Neuropathy Score in Chapter 1, axon al excitability studies were then utilised to determine the relative contributions of T2DM and CKD underlying nerve dysfunction in diabetic kidney disease in Chapter 2. It was established that CKD, and not diabetes, underlies axonal pathophysiology in patients with diabetic kidney disease.
Studies were then conducted in autoimmune diabetes. In type 1 diabetes (TlDM), despite good glycaemic control as measured by HbA1c, development of peripheral neuropathy frequently occurs. In search of an explanation, the association between acute glucose control and nerve structure and function was explored in Chapter 3. Greater acute glucose variability and longer time spent in hyperglycaemia were associated with worse nerve excitability measures, altered corneal nerve morphology, and a higher number of corneal micro-neuromas. In Chapter 4, the mechanisms underlying axonal dysfunction in a recently recognised form of autoimmune diabetes known as latent autoimmune diabetes in adults (LADA) were then investigated. The basis of nerve dysfunction in LADA was different to TlDM and T2DM, and LADA patients exhibited more severe changes in nerve excitability and ultrasound measures.
Investigations were then undertaken in T2DM. There is a strong association between the metabolic syndrome (Mets) and the development of peripheral neuropathy. To explain this relationship, the effect of the metabolic syndrome (MetS) in T2DM was examined in Chapter 5. A reduction in the function of the Na+/K+ pump was found to explain the more severe changes in nerve structure and function in T2DM patients with Mets compared to patients with T2DM alone. Finally, in exploration of potential neuroprotective options for peripheral neuropathy, the effect of anti-diabetic medication on nerve function was then investigated in Chapter 6. Exenatide treatment was associated with better nerve function in cross-sectional and prospective studies. Prominent abnormalities remained in patients receiving SGLT2 inhibitor or DPP-4 inhibitor therapy.
Declaration relating to disposition of project thesis/dissertation
I hereby grant to lhe University of New South Wales or its agents a non-exclusive licence to archive and lo make available (including to members of the public) my thesis or disserlation in whole or in part in lhe University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation. such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).
.. , , , ...... ,44, .. (•., ..... ,,,,,, ... ,.... , .... ,,,, ...... §/-?:(:.::��( ...... , .. . Si nature Date The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years can be made when submitting the final copies of your thesis to the UNSW Library. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research. ORIGINALITY STATEMENT
‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’
Signed ……………………………………………......
Date ……………………………………………...... COPYRIGHT STATEMENT
‘I hereby grant the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).’
‘For any substantial portions of copyright material used in this thesis, written permission for use has been obtained, or the copyright material is removed from the final public version of the thesis.’
Signed ……………………………………………......
Date ……………………………………………......
AUTHENTICITY STATEMENT ‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis.’
Signed ……………………………………………......
Date ……………………………………………...... ( ...... � � INCLUSION OF PUBLICATIONS STATEMENT ; Austral1a·s Global Ut:��yv University
UNSW is supportive of candidates publishing their research results during their candidature as detailed in the UNSW Thesis Examination Procedure.
Publications can be used in their thesis in lieu of a Chapter if: • The candidate contributed greater than 50% of the content in the publication and is the "primary author", ie. the candidate was responsible primarily for the planning, execution and preparation of the work for publication • The candidate has approval to include the publication in their thesis in lieu of a Chapter from their supervisor and Postgraduate Coordinator. • The publication is not subject to any obligations or contractual agreements with a third party that would constrain its inclusion in the thesis
Please indicate whether this thesis contains published material or not:
This thesis contains no publications, either published or submitted for publication □ (if this box is checked, you may delete all the material on page 2)
Some of the work described in this thesis has been published and it has been documented in the relevant Chapters with acknowledgement □ (if this box is checked, you may delete all the material on page 2)
This thesis has publications (either published or submitted for publication) incorporated into it in lieu of a chapter and the details are presented below
CANDIDATE'S DECLARATION I declare that: • I have complied with the UNSW Thesis Examination Procedure • where I have used a publication in lieu of a Chapter, the listed publication(s) below meet(s) the requirements to be included in the thesis. Candidate's Name Date (dd/mm/yy) Tushar lssar S/t/::2.i?zr POSTGRADUATE COORDINATOR'S DECLARATION To only be filled in where publications are used in lieu of Chapters I declare that:
G the information below is accurate G where listed publication(s) have been used in lieu of Chapter(s), their use complies with the UNSW Thesis Examination Procedure G the minimum requirements for the format of the thesis have been met. PGC's Name PGC's Signature Date (dd/mm/yy)
For each publication incorporated into the thesis in lieu of a Chapter, provide all of the requested details and signatures required Details of publication #1: Full title: The utility of the Total Neuropathy Score as an instrument to assess neuropathy severity in chronic kidney disease: A validation study Authors: lssar T, Arnold R, Kwai, NCG, Pussell, BA, Endre, ZH, Poynten, AM, Kiernan, MC, Krishnan, AV Journal or book name: Clinical Neurophysiology Volume/page numbers: Volume 129, Issue 5, Pages 889-894 Date accepted/ published: Accepted 4th February 2018 Status Published X Accepted and In ress The Candidate's Contribution to the Work Tushar lssar was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 1. Validates the Total Neuropathy Score as a means to assess peripheral neuropathy in chronic kidney disease, which was required for subsequent studies. PRIMARY SUPERVISOR'S DECLARATION I declare that: • the information above is accurate • this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter • All of the co-authors of the publication have reviewed the above information and have a reed to its veracit b si a 'Co-Author Authorisation' form. Primary Supervisor's name Primary Su isor's signature Date (dd/mm/yy) Arun Krishnan c7)/ OrX Io{J Details of publication #2: Full title: Relative contributions of diabetes and chronic kidney disease to neuropathy development in diabetic nephropathy patients. Authors: lssar T, Arnold R, Kwai NCG, Walker S, Yan A, Borire AA, Poynten AM, Pussell BA, Endre ZH, Kiernan MC, Krishnan AV Journalor book name: Clinical Neurophysiology Volume/page numbers: Volume 130, Issue 11, Pages 2088-2095 Date accepted/ published: Accepted 12th August 2019 Status Published X Accepted and In In progress ress submitted The Candidate's Contribution to the Work Tushar lssar was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 2. Investigates the effects of diabetes and chronic kidney disease on peripheral nerve dysfunction in diabetic kidney disease.
PRIMARYI declare that: SUPERVISOR'S DECLARATION • the information above is accurate . this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter . All of the co-authors of the publication have reviewed the above information and have aqreed to its veracity by siqninq a 'Co-Author Authorisation' form.
PrimaryArun Krishnan Supervisor's name Primary Date ( dd/mm/yy) Supervisor's signature 1 1510� / � I
FullDetails title: of Associations publication between #3: acute glucose control and peripheral nerve structure and function in type 1 diabetes Authors: lssar T, Tummanapalli SS, Kwai NCG, Chiang JCB, Arnold R, Poynten AM, Markoulli M, Krishnan AV Journal or book name: Diabetic Medicine Volume/page numbers: Online (ahead of print) Date accepted/ published: Accepted 9th April 2020 Accepted and In In progress Published X \ Status I I press I (submitted) I TusharThe Candidate's lssar was responsibleContribution for to the the study Work design, recruitment, data collection (with exception of in-vivo corneal confocal microscopy), data interpretation, and the manuscript composition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 3. Explores the association between acute glucose variation and peripheral nerve structure and function in type 1 diabetes.
PRIMARYI declare that: SUPERVISOR'S DECLARATION . the information above is accurate . this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter . All of the co-authors of the publication have reviewed the above information and have aqreed to its veracity by siqninq a 'Co-Author Authorisation' form.
PrimaryArun Krishnan Supervisor's name Primary Su,ltervisor's signature Date (dd/mm/yy) tJ
FullDetails title: of Altered publication peripheral #4: nerve structure and function in latent autoimmune diabetes in adults Authors: lssar T, Yan A, Kwai NCG, Poynten AM, Borire AB, Arnold R, Krishnan AV Journalor book name: Diabetes/Metabolism Research and Reviews Volume/page numbers: Volume 36, Issue 3, e3260 Date accepted/ published: Accepted 27th November 2019 \ Published X \ Accepted and In \ 'n progress Status press I (submitted) I The Candidate's Contribution to the Work Tushar lssar was responsible for the study design, recruitment, data collection, data inter retation, and the manuscri t com osition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 4. Investigates the pathophysiological changes in peripheral nerve structure and function in Latent Autoimmune Diabetes in Adults in comparison with type 1 and 2 diabetes PRIMARY SUPERVISOR'S DECLARATION I declare that: • the information above is accurate • this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter • All of the co-authors of the publication have reviewed the above information and have a reed to its veracit b si a 'Co-Author Authorisation' form. Primary Supervisor's name Primary Sup rvisor's signature Date (dd/mm/yy) Arun Krishnan (Jf7 0 o2 Id I' Details of publication #5: Full title: Impact of the metabolic syndrome on peripheral nerve structure and function in type 2 diabetes Authors: lssar T, Tummanapalli SS, Borire AB, Kwai NCG, Poynten AM, Arnold R, Markoulli M, Krishnan AV Journal or book name: Volume/page numbers: Date accepted/ published: Status Published Accepted and In In progress X ress submitted The Candidate's Contribution to the Work Tushar lssar was responsible for the study design, recruitment, data collection (with exception of in-vivo corneal confocal microscopy), data interpretation, and the manuscript com osition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 5. Investigates the pathophysiological changes in peripheral nerve structure and function in type 2 diabetes with the metabolic syndrome in comparison to type 2 diabetes alone. PRIMARY SUPERVISOR'S DECLARATION I declare that: • the information above is accurate • this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter • All of the co-authors of the publication have reviewed the above information and have a reed to its veracit b si a 'Co-Author Authorisation' form. Primary Supervisor's name Primary Su Date (dd/mm/yy) Arun Krishnan ()� 7rz; I¢2 I - Details of publication #6: Full title: Effect of exenatide on peripheral nerve function in type 2 diabetes Authors: lssar T, Kwai NCG, Poynten AM, Arnold R, Milner KL, Krishnan, AV Journal or book name: Volume/page numbers: Date accepted/ published: Status Published Accepted and In In progress X ress submitted The Candidate's Contribution to the Work Tushar lssar was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition. Location of the work in the thesis and/or how the work is incorporated in the thesis: Chapter 6. Explores the effect of exenatide, SGL T2 inhibition, or DPP-4 inhibition on peripheral nerve function in type 2 diabetes. PRIMARY SUPERVISOR'S DECLARATION I declare that: • the information above is accurate . this has been discussed with the PGC and it is agreed that this publication can be included in this thesis in lieu of a Chapter . All of the co-authors of the publication have reviewed the above information and have aqreed to its veracity by siqning a 'Co-Author Authorisation' form. Primary Supervisor's name Primary Supervisor's signature Date (dd/mm/yy) Arun Krishnan o� lo;;_/ �I Acknowledgements
This thesis would not have been achievable without the guidance of my supervisors. I am forever grateful to Professor Arun Krishnan for his wisdom, enthusiasm, inspiration, and endless support. I will always remember his tremendous efforts throughout my career, and I owe my achievements to him. I am eternally thankful to have had the support of Dr Ann Poynten. Her invaluable insight and admirable expertise made the studies of this thesis possible. I am especially appreciative for her kindness, patience, and generosity in providing me the facilities to conduct research. I would also like to express my gratitude to Dr Natalie Kwai and Dr Ria Arnold for their valued advice throughout my studies and instrumental efforts in imparting their technical knowledge and skills.
I would like to thank all of my research colleagues at the Institute of
Neurological Sciences for making my time there so enjoyable. Shyam and
Aimy, thank you for your valuable assistance and support. It was a such a pleasure to work with you both and I wish you every success in the future. I would also like to thank Terry, Jeremy, Kimberly, and Vincent for providing endless laughter and positivity throughout the difficult periods that comes with research.
This thesis would not have been possible without the help of several other important people. I would like to express my appreciation for having the privilege to learn from Professor Hugh Bostock. His generosity in taking the
ii time educate me in neurophysiology is an experience I will carry throughout my career and his contributions to Science will forever be an inspiration. I would also like to thank Dr Adeniyi Borire and Dr Maria Markoulli for their expertise and contributions to my manuscripts. I am grateful for the technical support of Mr Tony Yakoubi, who was always willing to lend a helping hand. I would like to thank Mrs Analiza Santiago, Mrs Isabell Ghikas, and Mrs
Collette Brodie for their assistance with recruiting study participants. I would like to express my sincere gratitude for the patients that generously donated hours of their time for these studies.
Finally, this body of work would not have been possible without the support of my loving family. My parents, Arvind and Meena, have made tremendous sacrifices throughout their lives to support me in a countless number of ways.
My brother Aarush and my sister Misha, thank you for your endless support and encouragement. You all mean everything to me.
iii Abstract
Peripheral neuropathy is a common and debilitating complication of diabetes and chronic kidney disease (CKD). The pathophysiological mechanisms contributing to peripheral neuropathy in these conditions remain unclear. In this thesis, axonal excitability studies and nerve ultrasonography were utilised to assess peripheral nerve structure and function in human subjects with either diabetes, CKD, or both to investigate the mechanisms underlying peripheral nerve dysfunction in each condition.
To commence, it was essential that an instrument to assess the severity of neuropathy in patients CKD with and without type 2 diabetes (T2DM) was formally validated. Having validated The Total Neuropathy Score in Chapter
1, axonal excitability studies were then utilised to determine the relative contributions of T2DM and CKD underlying nerve dysfunction in diabetic kidney disease in Chapter 2. It was established that CKD, and not diabetes, underlies axonal pathophysiology in patients with diabetic kidney disease.
Studies were then conducted in autoimmune diabetes. In type 1 diabetes
(T1DM), despite good glycaemic control as measured by HbA1c, development of peripheral neuropathy frequently occurs. In search of an explanation, the association between acute glucose control and nerve structure and function was explored in Chapter 3. Greater acute glucose variability and longer time spent in hyperglycaemia were associated with worse nerve excitability measures, altered corneal nerve morphology, and a higher number of corneal micro-neuromas. In Chapter 4, the mechanisms underlying axonal dysfunction
iv in a recently recognised form of autoimmune diabetes known as latent autoimmune diabetes in adults (LADA) were then investigated. The basis of nerve dysfunction in LADA was different to T1DM and T2DM, and LADA patients exhibited more severe changes in nerve excitability and ultrasound measures.
Investigations were then undertaken in T2DM. There is a strong association between the metabolic syndrome (MetS) and the development of peripheral neuropathy. To explain this relationship, the effect of the metabolic syndrome
(MetS) in T2DM was examined in Chapter 5. A reduction in the function of the Na+/K+ pump was found to explain the more severe changes in nerve structure and function in T2DM patients with MetS compared to patients with
T2DM alone. Finally, in exploration of potential neuroprotective options for peripheral neuropathy, the effect of anti-diabetic medication on nerve function was then investigated in Chapter 6. Exenatide treatment was associated with better nerve function in cross-sectional and prospective studies. Prominent abnormalities remained in patients receiving SGLT2 inhibitor or DPP-4 inhibitor therapy.
v Publications
Publications resulting from studies undertaken in this doctorate
Chapter 1
Issar, T., Arnold, R., Kwai, N. C. G., Pussell, B. A., Endre, Z. H., Poynten, A. M., . . .
Krishnan, A. V. (2018). The utility of the Total Neuropathy Score as an instrument to
assess neuropathy severity in chronic kidney disease: A validation study. Clinical
Neurophysiology, 129(5), 889-894. doi: 10.1016/j.clinph.2018.02.120
Chapter 2
Issar, T., Arnold, R., Kwai, N. C. G., Walker, S., Yan, A., Borire, A. A., . . . Krishnan, A. V.
(2019). Relative contributions of diabetes and chronic kidney disease to neuropathy
development in diabetic nephropathy patients. Clinical Neurophysiology, 130(11),
2088-2095. doi: 10.1016/j.clinph.2019.08.005
Chapter 3
Issar, T., Tummanapalli, S. S., Kwai, N. C. G., Chiang, J. C. B., Arnold, R., Poynten, A. M.,
. . . Krishnan, A. V. (2020). Associations between acute glucose control and
peripheral nerve structure and function in type 1 diabetes. Diabetic Medicine. doi:
10.1111/dme.14306
Chapter 4
Issar, T., Yan, A., Kwai, N. C. G., Poynten, A. M., Borire, A. A., Arnold, R., & Krishnan, A.
V. (2020). Altered peripheral nerve structure and function in latent autoimmune
diabetes in adults. Diabetes/Metabolism Research and Reviews, 36(3), e3260. doi:
10.1002/dmrr.3260
vi Associated publications achieved during candidature
Habeych, M., Trinh, T., Issar, T., Kwai, N. C. G., & Krishnan, A.V. (2020). Motor
unit number estimation of facial muscles using the M Scan-Fit method.
Muscle & Nerve, In press doi: 10.1002/mus.27010
Tummanapalli, S. S., Issar, T., Kwai, N. C. G., Pisarcikova, J., Poynten, A. M., Krishnan, A.
V., . . . Markoulli, M. (2019). A Comparative Study on the Diagnostic Utility of
Corneal Confocal Microscopy and Tear Neuromediator Levels in Diabetic Peripheral
Neuropathy. Current Eye Research, 45(8), 921-903.
doi: 10.1080/02713683.2019.1705984
Tummanapalli, S. S., Issar, T., Yan, A., Kwai, N. C. G., Poynten, A. M., Krishnan, A. V., . . .
Markoulli, M. (2019). Corneal nerve fiber loss in diabetes with chronic kidney
disease. The Ocular Surface, 18(1), 178-185. doi: 10.1016/j.jtos.2019.11.010
Tummanapalli, S. S., Issar, T., Kwai, N. C. G., Poynten, A., Krishnan, A. V., Willcox, M., &
Markoulli, M. (2019). Association of corneal nerve loss with markers of axonal ion
channel dysfunction in type 1 diabetes. Clinical Neurophysiology, 131(1), 145-154.
doi: 10.1016/j.clinph.2019.09.029
Tummanapalli, S. S., Willcox, M. D. P., Issar, T., Yan, A., Pisarcikova, J., Kwai, N. C. G., . .
. Markoulli, M. (2019). Tear film substance P: A potential biomarker for diabetic
peripheral neuropathy. The Ocular Surface, 17(4), 690-698. doi:
10.1016/j.jtos.2019.08.010
vii Tummanapalli, S. S., Willcox, M. D. P., Issar, T., Kwai, N. C. G., Poynten, A. M., Krishnan,
A. V., . . . Markoulli, M. (2019). The Effect of Age, Gender and Body Mass Index on
Tear Film Neuromediators and Corneal Nerves. Current Eye Research, 45(4), 411-
418. doi: 10.1080/02713683.2019.1666998
Borire, A. A., Issar, T., Kwai, N. C. G., Visser, L. H., Simon, N. G. C., Poynten, A. M., . . .
Krishnan, A. V. (2019). Sonographic assessment of nerve blood flow in diabetic
neuropathy. Diabetic Medicine, 37(2), 343-349. doi: 10.1111/dme.14085
Yan, A., Issar, T., Tummanapalli, S. S., Markoulli, M., Kwai, N. C. G., Poynten, A. M., &
Krishnan, A. V. (2019). Relationship between corneal confocal microscopy and
markers of peripheral nerve structure and function in Type 2 diabetes. Diabetic
Medicine, 37(2), 326-334. doi: 10.1111/dme.13952
Borire, A. A., Issar, T., Kwai, N. C. G., Visser, L. H., Simon, N. G., Poynten, A. M., . . .
Krishnan, A. V. (2018). Correlation between markers of peripheral nerve function
and structure in type 1 diabetes. Diabetes/ Metabolism Research Reviews, 34(7),
e3028. doi: 10.1002/dmrr.3028
Arnold, R., Issar, T., Krishnan, A. V., & Pussell, B. A. (2016). Neurological complications in
chronic kidney disease. JRSM Cardiovascular Disease, 5, 2048004016677687. doi:
10.1177/2048004016677687
viii Awards and Presentations
Awards
1. William McIlrath Travel Scholarship 2018
2. Prince of Wales Hospital Equipment Grant 2017, 2018
3. JRSM’s Third Most Downloaded Publication 2017
4. Prince of Wales Hospital Travel Grant 2016, 2017
5. Australian Postgraduate Award Scholarship 2016–2018
Presentations
International Diabetes Federation Congress 2019
Poster presentation: “The effect of hyperglycaemia and glucose variability on peripheral nerve structure and function”
Tow Coast Association Awards 2018
Oral presentation: “Mechanisms of axonal dysfunction in diabetes and chronic kidney disease”
Australian Diabetes Society Annual Scientific Meeting 2016, 2017 o Oral presentation: “GLP-1 agonism alters peripheral nerve function in patients with type 2 diabetes” o Poster presentation: “Effects of GLP-1 agonism on peripheral nerve function in type 2 diabetes”
ix Table of Contents
Acknowledgements...... ii Abstract...... iv Publications...... vi Awards...... ix Presentations...... ix Table of Contents...... x Index of Tables...... xii Index of Figures...... xiv Abbreviations...... xvii Literature Review...... 1 Diabetes Mellitus...... 2 Type 1 Diabetes Mellitus...... 3 Latent Autoimmune Diabetes of Adults...... 4 Type 2 Diabetes Mellitus and the Metabolic Syndrome...... 6 Chronic Kidney Disease...... 8 Diabetic Kidney Disease...... 9 The Human Nervous System...... 11 The Peripheral Axon...... 15 Structure of Peripheral Axons...... 15 Function of Peripheral Axons...... 19 Ion Channels, Pumps, and Exchangers...... 23 Clinical Assessment of Peripheral Nerve Structure and Function...... 38 Axonal Excitability...... 41 Nerve Ultrasonography...... 58 In-vivo Corneal Confocal Microscopy...... 61 Diabetic Neuropathy...... 65 Uraemic Neuropathy...... 75 Methodology...... 80 Recruitment of Patients and Control Subjects...... 81 Equipment and Materials...... 82
x Axonal Excitability Assessment and Mathematical Modelling...... 84 Nerve Ultrasonography...... 92 In-vivo Corneal Confocal Microscopy...... 93 Nerve Conduction Studies...... 95 Clinical Assessment of Neuropathy...... 95 Glucose Variability...... 99 Metabolic Syndrome Components...... 101 Statistical Analyses...... 102 Chapter 1 – Validation of the Total Neuropathy Score as a means to assess peripheral neuropathy in chronic kidney disease...... 103 Chapter 2 – Investigation of the effect of diabetes and chronic kidney disease on axonal pathophysiology...... 120 Chapter 3 – Association between acute glucose control and axonal function and structure in type 1 diabetes...... 143 Chapter 4 – Axonal pathophysiology in Latent Autoimmune Diabetes of Adulthood...... 165 Chapter 5 – Pathophysiological mechanisms underlying altered axonal function and structure in type 2 diabetes with metabolic syndrome...... 187 Chapter 6 – Effect of exenatide on axonal function in type 2 diabetes...... 210 Summary and Future Directions...... 230 References...... 238
xi Index of Tables
Literature Review and Methodology
Table 1. Classification of chronic kidney disease...... 9
Table 2. Nerve excitability testing paradigms...... 53
Chapter 1
Table 1.1. Components of the total neuropathy score...... 107
Table 1.2. Subject demographics...... 112
Table 1.3. Internal consistency of the TNS...... 114
Table 1.4. Comparison of TNS item scores between groups...... 115
Table 1.5. Comparison of TNS item scores between subgroups and controls...... 116
Chapter 2
Table 2.1. Subject demographics...... 131
Table 2.2. Total neuropathy score and subscore comparison...... 132
Table 2.3. Nerve excitability findings...... 134
Table 2.4. Modelled parameters between DKD, CKD, T2DM, and control subjects...... 138
xii Chapter 3
Table 3.1. Participant demographics...... 154
Table 3.2. Partial correlations between acute glycaemic metrics and measures of peripheral nerve structure and function...... 156
Chapter 4
Table 4.1. Participant characteristics...... 176
Table 4.2. Modelled parameters for LADA, type 1 diabetes, type 2 diabetes and control cohorts...... 182
Chapter 5
Table 5.1. Participant demographics...... 199
Table 5.2. Corneal confocal microscopy cohorts...... 202
Table 5.3. Corneal confocal microscopy correlations...... 204
Chapter 6
Table 6.1. Subject demographics and clinical measures...... 220
Table 6.2. Prospective exenatide cohort clinical measures at baseline and at 3-month follow-up...... 224
xiii Index of Figures
Literature Review and Methodology
Figure 1. Schematic of a motor neuron and domains of a myelinated axon...... 14
Figure 2. The action potential...... 21
Figure 3. Distribution of ion channels, pumps, and exchangers in a myelinated axon...... 25
Figure 4. Stimulus-response curve...... 43
Figure 5. Strength-duration relationship...... 45
Figure 6. Threshold electrotonus...... 47
Figure 7. Current-threshold relationship...... 50
Figure 8. Recovery cycle...... 51
Figure 9. Sonograph of healthy and neuropathic median nerve...... 61
Figure 10. Confocal micrograph of healthy and neuropathic cornea...... 64
Figure 11. Pathogenesis of distal symmetric polyneuropathy...... 68
Figure 12. Median motor nerve excitability set-up...... 86
Figure 13. Mathematical model parameters...... 91
Figure 14. Median nerve ultrasonography...... 93
Figure 15. Corneal confocal microscopy...... 94
Figure 16. Total Neuropathy Score...... 96
Figure 17. Analysis of continuous glucose monitoring recordings...... 100
xiv Chapter 1
Figure 1.1. Histogram of TNS results for all patients with CKD compared with controls....115
Chapter 2
Figure 2.1. Schematic of the peripheral nerve membrane highlighting the two compartments of the axon and the parameters that were modelled...... 129
Figure 2.2. Mean excitability data for patients with T2DM, CKD or DKD and healthy controls...... 133
Figure 2.3. Summary of mathematical modelling of nerve excitability data for DKD,
CKD, and T2DM...... 137
Chapter 3
Figure 3.1. Receiver operating characteristic analysis of composite nerve excitability score obtained from participants with diabetes and people without diabetes...... 157
Figure 3.2. Corneal confocal image highlighting a microneuroma of a corneal nerve fibre and representative image of the inferior whorl from a participant with type 1 diabetes...... 159
Chapter 4
Figure 4.1. Median nerve cross-sectional area measurements...... 177
Figure 4.2. Group comparison of nerve excitability parameters...... 179
Figure 4.3. Group comparison of nerve excitability recordings...... 180
xv Chapter 5
Figure 5.1. Median nerve ultrasonography highlighting differences in median nerve cross-sectional area between participants with type 2 diabetes and metabolic syndrome, type 2 diabetes alone, and healthy controls...... 201
Figure 5.2. Representative corneal confocal micrographs comparing differences in nerve structure in the central and corresponding inferior whorl region of the cornea for participants with type 2 diabetes and metabolic syndrome, type 2 diabetes alone, and healthy controls...... 203
Chapter 6
Figure 6.1. Group comparison of nerve excitability measures...... 222
Figure 6.2. Nerve excitability recordings showing threshold electrotonus and the recovery cycle of a 72-year-old male patient with mild neuropathy...... 225
xvi Abbreviations
CCM corneal confocal microscopy
CKD chronic kidney disease
CNS central nervous system
CONGA continuous overall net glycaemic action
CSII continuous subcutaneous insulin infusion
DKD diabetic kidney disease
DPP-4 dipeptidyl peptidase-4
DSPN distal symmetric polyneuropathy
GLP-1 glucagon-like peptide-1
HCN hyperpolarisation-activated cyclic nucleotide-gated
Ih mixed cation conductance through HCN
I/V current-threshold relationship
KIR inwardly rectifying potassium channel
KV voltage-gated potassium channel
LADA latent autoimmune diabetes of adulthood
MDII multiple daily insulin infusion
Nap persistent voltage-gated sodium channel
Nat transient voltage-gated sodium channel
NaV voltage-gated sodium channel
NCS nerve conduction studies
PKC protein kinase C
PNS peripheral nervous system
T1DM type 1 diabetes
T2DM type 2 diabetes
xvii TEd depolarising (conditioning current) threshold electrotonus
TEh hyperpolarising (conditioning current) threshold electrotonus
TNS total neuropathy score
SDTC strength-duration time constant (τSD)
SGLT2 sodium-glucose co-transporter-2
xviii Literature Review
Literature Review
1 Literature Review
Diabetes Mellitus
Diabetes Mellitus (diabetes) defines a state of hyperglycaemia due to reduced, altered or absent insulin secretion and or action. The prevalence of diabetes in adults was estimated to be 463 million in 2019 and is projected to increase to 700 million by
2045 (Saeedi et al., 2019). Diabetes can be classified into the following general categories: type 1 diabetes (due to cellular-mediated autoimmune β-cell destruction, usually leading to absolute insulin deficiency), type 2 diabetes (due to progressive loss β-cell insulin secretion frequently on a background of insulin resistance), gestational diabetes (diabetes diagnosed in second or third trimester of pregnancy that was not overt diabetes prior to gestation) and specific types of diabetes due other causes such as monogenic diabetes syndromes (e.g. maturity-onset diabetes of the young), diseases of the exocrine pancreas (e.g. cystic fibrosis and pancreatitis), and drug- or chemical-induced diabetes (e.g. glucocorticoid use and after organ transplantation) (American Diabetes Association, 2020). This thesis will primarily focus on type 1 and type 2 diabetes. It is important to note that type 1 diabetes and type 2 diabetes are heterogeneous diseases in which clinical presentation and disease progression may vary considerably.
The diagnosis of diabetes may be based on plasma glucose criteria, either fasting plasma glucose (≥7 mmol/L; 126 mg/dL) or 2 hour plasma glucose during a 75g oral glucose tolerance test (≥11.1 mmol/L; 200 mg/dL), or glycated haemoglobin criteria
(≥6.5%; 48 mmol/mol) (American Diabetes Association, 2020). In a patient with classical symptoms of hyperglycaemia or hyperglycaemic crisis, a random plasma glucose ≥ 11.1 mmol/L (200 mg/dL) can be used (American Diabetes Association,
2020). In cases where type 1 diabetes requires confirmation, islet cell autoantibodies
2 Literature Review and autoantibodies to glutamate decarboxylase (GAD65), insulin, the tyrosine phosphatases (IA-2 and IA-2β), and zinc transporter 8 (ZnT8) may be used
(American Diabetes Association, 2020).
Once hyperglycaemia occurs, patients with all forms of diabetes are at risk for developing the same chronic complications, including nephropathy, retinopathy, and neuropathy. Rates of complication progression may differ between diabetes type
(American Diabetes Association, 2020). Diabetes is a leading cause of disability and death globally (International Diabetes Federation, 2015, Vos et al., 2016, Zheng et al.,
2018). In Australia, diabetes and resulting comorbidities account for an annual cost of approximately A$15 billion (Lee et al., 2013). This estimated value is inclusive of the cost for medication and adjuvant therapy directly targeting diabetes, government subsidies, and cost to employers and it is projected to increase (Schofield et al., 2017).
Treatment of diabetes is focused not only on glycaemic control but management of co-morbidities and etiological factors. Clinically, this may be achieved through a combination of drug therapy, insulin supplementation and rigorous lifestyle management involving a number of health care professionals.
Type 1 Diabetes Mellitus
The typical clinical course of type 1 diabetes includes a preclinical phase, presentation of diabetes (at which time patients are usually symptomatic of hyperglycaemia), a partial remission or honeymoon phase, and a continuing requirement for insulin therapy (Haller et al., 2005). Type 1 diabetes is characterised by a specific autoimmune destruction of the Islets of Langerhans in the pancreas, which was first described in the early 1900s (Opie, 1901). Autoimmune destruction of β-cells, the
3 Literature Review insulin secreting cells of the pancreas, provokes the total loss of insulin and co- released C-peptide initiating the requirement for exogenous insulin. In the absence of insulin therapy, patients with type 1 diabetes will eventually progress to metabolic decompensation and life‐threatening diabetic ketoacidosis (Craig et al., 2011). Type 1 diabetes typically arises in childhood or adolescence, but onset can occur in adulthood
(Pozzilli and Pieralice, 2018). Patients with an adult-onset of autoimmune diabetes that do not require insulin-therapy for at least 6 months after diagnosis are distinguished as having latent autoimmune diabetes of adulthood (LADA) (Pozzilli and Pieralice, 2018).
Treatment of Type 1 Diabetes
Type 1 diabetes management involves exogenous synthetic insulin therapy with the goal of mimicking normal physiological insulin secretions patterns (Chiang et al.,
2018). Insulin may be delivered through multiple daily insulin injections (MDII) or continuous subcutaneous insulin infusion (CSII) via pump therapy (Chiang et al.,
2018). MDII involves injecting a slow acting insulin analogue that delivers basal levels of insulin and a rapid acting analogue before meals for carbohydrate clearance with food intake (Chiang et al., 2018). CSII utilises fast acting insulin analogues that automatically perfuse insulin at a basal rate through a cannula placed subcutaneously
(Pozzilli et al., 2016).
Latent Autoimmune Diabetes of Adulthood
Adult-onset autoimmune diabetes encompasses a wide spectrum of heterogeneous genotypes and phenotypes, ranging from classic adult-onset type 1 diabetes to LADA
(Buzzetti et al., 2017). LADA was a term introduced in the 1990s to define a
4 Literature Review subgroup of patients who had non-insulin requiring diabetes that was initially thought to be type 2 diabetes but had autoimmune markers of type 1 diabetes (Tuomi et al.,
1993, Zimmet et al., 1994). In 2005, the Immunology of Diabetes Society proposed three main criteria for the diagnosis of LADA: adult age of onset (>30 years), presence of any islet cell autoantibody, and absence of insulin requirement for at least
6 months after diagnosis (Fourlanos et al., 2005). The major criticism of this criteria is the subjectivity of the clinician’s decision to commence insulin treatment. LADA is more heterogenous than youth-onset autoimmune diabetes and shares genetic, clinical, and metabolic features with both type 1 and type 2 diabetes, which suggests that LADA is an admixture of the two major types of diabetes (Cervin et al., 2008,
Liu et al., 2015, Pozzilli and Pieralice, 2018). Patients with LADA have highly variable rates of β-cell destruction as well as different degrees of insulin resistance and autoimmunity (Pozzilli and Pieralice, 2018). Whether LADA is a distinct disease syndrome or part of an autoimmune continuum is yet to be confirmed (Laugesen et al., 2015).
Treatment of Latent Autoimmune Diabetes of Adulthood
To date, no specific guidelines for the treatment of patients with LADA have been published. Treatment of adults with LADA is currently guided by the clinical intuition and expertise of the physician (Buzzetti et al., 2017). Most patients with LADA are initially treated with therapies intended for type 2 diabetes (Pozzilli and Pieralice,
2018). This approach might result in rapid progression to an insulin-dependent state, especially in patients who have high GAD autoantibody titres (Zampetti et al., 2014).
5 Literature Review
Type 2 Diabetes Mellitus and the Metabolic Syndrome
Type 2 diabetes accounts for 90-95% of all diabetes and encompasses individuals with relative (rather than absolute) insulin deficiency and peripheral insulin resistance
(American Diabetes Association, 2020). Although the specific aetiologies are not known, autoimmune destruction of β-cells does not occur and these individuals may not need insulin treatment to survive (American Diabetes Association, 2020).
However, β-cells dysfunction does occur and results in reduced insulin release, which is insufficient for maintaining normal glucose levels (Zheng et al., 2018). The main drivers of the type 2 diabetes globally are the rise in obesity, a sedentary lifestyle, energy-dense diets and population ageing (Zheng et al., 2018). Type 2 diabetes is often associated with a strong genetic predisposition or family history (Vujkovic et al., 2020). The risk factors for developing type 2 diabetes increases with age, obesity
(especially in abdominal region), lack of physical activity, hypertension and dyslipidaemia (American Diabetes Association, 2020). Elevated fasting glucose, abdominal obesity, hypertension, and dyslipidaemia comprise the ‘metabolic syndrome’ and increases an individual’s risk of diabetes, heart disease, stroke, as well as peripheral neuropathy (Cortez et al., 2014, Nilsson et al., 2019).
Treatment of Type 2 Diabetes
Type 2 diabetes management targets not only hyperglycaemia but also the other factors of the metabolic syndrome. Lifestyle modification, including dietary advice and diabetes education with an emphasis on physical activity, are recommended as first-line therapies from diagnosis or in conjunction for patients requiring glucose- lowering medication or metabolic surgery (Davies et al., 2018). Glucose-lowering medication includes metformin, sodium-glucose co-transporter-2 (SGLT2) inhibitors,
6 Literature Review glucagon-like peptide-1 (GLP-1) receptor agonists, dipeptidyl peptidase-4 (DPP-4) inhibitors, thiazolidinediones, sulfonylureas, and insulin (Davies et al., 2018).
Metformin is a biguanide and remains the first-line medication for type 2 diabetes
(Davies et al., 2018). Metformin is an oral medication that primarily inhibits hepatic glucose production and has other mechanisms of action to reduce plasma glucose but may result in lower serum vitamin B12 concentration (Thulé, 2012). SGLT2 inhibitors are oral medications that reduce plasma glucose by enhancing urinary excretion of glucose and their glucose-lowering efficacy is dependent on renal function (Davies et al., 2018). GLP-1 receptor agonists are incretin mimetics and are delivered via subcutaneous injection. They stimulate insulin secretion, reduce glucagon secretion, improve satiety, and promote weight loss (Thulé, 2012). GLP-1 is degraded by DPP-4 and the action of DPP-4 limits the serum half-life of endogenous GLP-1 to under 2 minutes (Holst, 2007). DPP-4 inhibitors are administered orally and like GLP-1 receptors agonists, increase insulin secretion and reduce glucagon secretion but have inferior glucose-lowering efficacy (Davies et al., 2018). Thiazolidinediones are oral medications that increase insulin sensitivity and high-density lipoprotein (Davies et al., 2018). Sulfonylureas are insulin secretagogues that are administered orally and stimulate insulin secretion from pancreatic β-cells (Thulé, 2012). Finally, insulin therapy may be advised in type 2 diabetes in the event of β-cell loss and resultant demise of appropriate insulin secretion. Typically, patients utilise longer or intermediate acting insulin analogues, however rapid-acting insulin formulations are used in patients not meeting glycaemic targets (Davies et al., 2018).
7 Literature Review
Chronic Kidney Disease
Chronic kidney disease (CKD) is a significant global health concern and disease burden, prevalence, and mortality rates are rising (Bikbov et al., 2020, Coresh, 2017).
The estimated global prevalence of CKD is between 10–15% (Bikbov et al., 2020,
Coresh, 2017, Hill et al., 2016). As well as being an important risk factor cardiovascular disease, CKD has a major effect on global health as a direct cause of morbidity and mortality (Bikbov et al., 2020). In 2017, CKD resulted in 35.8 million disability adjusted life-years.
CKD encompasses a continuum of disease from mild kidney damage to end-stage kidney disease, which requires renal replacement therapy in the form of dialysis or renal transplantation. Disease severity is classified using a five-stage system based on the estimated glomerular filtration rate (eGFR), which is an indirect measure of how well the kidneys filter wastes from the blood (Table 1) (Bowling et al., 2011, Chadban and Ierino, 2005, National Kidney Foundation, 2002). This estimation is calculated from age, sex, and creatinine clearance using the Modification of Diet in Renal
Disease formula (National Kidney Foundation, 2002). The majority of CKD patients worldwide are in stage 3 (Hill et al., 2016, Vos et al., 2016). CKD is diagnosed by the presence of kidney damage, manifested by abnormal albumin excretion or decreased eGFR, that persists for more than three months (Levey et al., 2005). The aetiology of
CKD may be due to a primary renal disorder or as a complication of a multisystem disorder. Diabetes is the leading cause of CKD as well as end-stage kidney disease and accounts for more than half of the global prevalence (Bikbov et al., 2020,
National Kidney Foundation, 2012, Stanton, 2014, Vos et al., 2016). Complications of
8 Literature Review
CKD such as hypertension, anaemia, and metabolic bone disease become apparent with declining eGFR (Holley, 2011, Inker et al., 2011, Thomas et al., 2008).
Table 1. Classification of chronic kidney disease Stage 1 Evidence of kidney damage with normal eGFR >90 mL/min/1.732 Stage 2 Evidence of kidney damage with mild reduction of eGFR 60–89 mL/min/1.732 Stage 3 Moderately reduced eGFR 30–59 mL/min/1.732 Stage 4 Severely reduced eGFR 15–29 mL/min/1.732 Stage 5 Renal failure or dialysis eGFR < 15 mL/min/1.732
Classification as defined by the National Kidney Disease Outcomes and Quality Initiative clinical practice guidelines (National Kidney Foundation, 2002).
Treatment of Chronic Kidney Disease
Only a small proportion of patients with mild or moderate CKD will progress to end- stage kidney disease (Hallan et al., 2006). Management of CKD primarily involves prevention of complications and identifying patients at risk of disease progression
(Fraser and Blakeman, 2016). At end-stage kidney disease, renal replacement therapy in the form of haemodialysis, peritoneal dialysis, or renal transplantation is required to sustain life.
Diabetic Kidney Disease
CKD caused by diabetes is known as diabetic kidney disease (DKD) or diabetic nephropathy. It is estimated that 30% of patients with type 1 diabetes or 40% of patients with type 2 diabetes will develop DKD during their lifetimes (Alicic et al.,
9 Literature Review
2017). For the majority of these patients, DKD will develop within 10 years of diagnosis (Couser et al., 2011). Cardiac autonomic neuropathy and certain plasma biomarkers have been identified as risk factors associated with rapid decline in eGFR in type 2 diabetes (Peters et al., 2017, Tahrani et al., 2014). Genetic risk factors for the development of DKD and rapid decline of eGFR have emerged but these require further validation (Davoudi and Sobrin, 2015, Jiang et al., 2019). Other clinical risk factors include increased albuminuria, hyperglycaemia, hypertension, dyslipidaemia, obesity, and smoking (Hussain et al., 2020). In the management of diabetic kidney disease, SGLT2 inhibitors and GLP-1 receptor antagonists have been shown to slow the progression of kidney disease (Ninčević et al., 2019). Oxidative stress in the kidney from hyperglycaemia and dyslipidaemia is postulated to be the key component of in the development diabetic nephropathy (Eid et al., 2019, Forbes et al., 2008,
Savelieff et al., 2020). Insulin resistance has also been identified as a major determinant of DKD via a number of pathogenic pathways (Karalliedde and Gnudi,
2016).
10 Literature Review
The Human Nervous System
The human nervous system is comprised of billions of neurons responsible for the transmission and processing of electrical signals around the body. These signals may arise internally, to coordinate actions around the body, or externally, to perceive the surrounding environment. Although structurally and functionally linked, the human nervous system can be divided into two anatomically distinct components, the central nervous system (CNS) and the peripheral nervous system (PNS). The CNS is comprised of the brain and spinal cord and functions to integrate information and coordinate activity. The PNS serves to connect the CNS to the body and encompasses the remaining elements of the nervous system, which includes the cranial nerves (III–
XII), spinal nerves, peripheral nerves, sensory receptors, and neuromuscular junctions
(Catala and Kubis, 2013).
At the cellular level, the nervous system is composed of neurons and neuroglia, which differ in structure and function. Neurons are electrically excitable cells which are capable of receiving and sending information. Neurons serve as the primary signalling unit of the nervous system and enable the communication between the CNS and the rest of the body via the PNS. Different types of neurons exist throughout the nervous system and they are typically classified into three types depending on their function.
Sensory neurons respond to stimuli in the periphery and send signals to the spinal cord and brain, where the information is perceived by other sensory neurons. Motor neurons in the brain send commands to other motor neurons in the spinal cord where they exert control over muscles and glands. Interneurons connect neurons in different regions of the brain or spinal cord. Structurally, the typical neuron consists of a cell body (soma) which gives rise to several short processes known as dendrites and a
11 Literature Review relatively longer, single tubular process termed the axon (Figure 1). In the PNS, the soma of motor neurons are located in the ventral horn of the spinal cord, under the protection of the blood-brain barrier, whereas the soma of sensory neurons are situated in the dorsal root ganglion and afforded no protection from the blood-brain barrier. In the context of peripheral nerve injury, this is especially important because it renders sensory neurons especially susceptible to damage. The soma houses the nucleus. The rough endoplasmic reticulum, Golgi bodies, and other organelles necessary for the maintenance of cellular function are present in the soma, dendrites, and axon (Berthold et al., 2005, Catala and Kubis, 2013). The dendrites are the principle structures that receive input from other neurons or sensory receptors and conduct this signal towards the soma. The axon is a highly specialised region that serves as a conduit for impulse conduction away from the soma and towards the dendrites of other neurons, muscles or glands. Depending on the type of neuron, axons may range in length from less than a millimetre to over a metre long and account for at least 95% of the cell mass in long neurons (Bear et al., 2016, Miller,
2014). The neuronal cell plasma membrane, which is approximately 5 nm thick, contains proteins essential for neuronal function and protein composition of the membrane varies between the soma, dendrites, and axon (Bear et al., 2016). In contrast, neuroglia do not directly participate in electrical signalling and information processing but they provide crucial structural and metabolic support for neurons. For example, Schwann cells in the PNS form the myelin seen on axons greater than 1–2
µm in diameter and provides an insulating sheath that permits saltatory conduction and the rapid and reliable propagation of impulses (Catala and Kubis, 2013, Fitzhugh,
1962, Huxley and Stämpeli, 1949, Kuriscák et al., 2002). Schwann cells are also involved in the metabolic support of axons and protecting them against toxic
12 Literature Review molecules and reactive oxygen species (Beirowski et al., 2014, Vincent et al., 2011,
Wang et al., 2012).
13 Literature Review
Paranode Juxtaparanode Internode
Node
Soma Axon hillock
Axon Axoglial Periaxonal Dendrite junction space Axon terminals Initial segment Myelin Node of Action sheath Ranvier potential
Figure 1. Schematic of a motor neuron and domains of a myelinated axon. The axon is
ensheathed in myelin segments which allows action potential propagation from node to node,
known as saltatory conduction.
14 Literature Review
The Peripheral Axon
The primary role of the PNS is to facilitate the transmission and receipt of electrical signals between CNS and the body. This communication is facilitated by electrical impulses, known as action potentials, which travel along the axon to permit the communication between the neurons and muscles, glands, or other neurons. Another important function of the axon is that it provides a conduit for synthesised proteins that have important structural and functional roles to be transported (Wang and He,
2013). The rapid and reliable transmission of action potentials is heavily dependent on the strategic distribution of myelin and proteins such as voltage-gated ions channels, ion pumps, and ion exchangers along specific domains of the axon (Barnett and
Larkman, 2007, Kursula, 2014). These structural and molecular factors are important determinants in the generation and conductance of action potentials, a property known as axonal excitability. Assessment of axonal excitability has provided important insight into nerve function in healthy and diseased states (Kiernan et al., 2020).
Structure of Peripheral Axons
The axon becomes physiologically distinct from the soma at the axon hillock
(Figure 1) (Leterrier, 2018). The axon hillock marks the beginning of the axon and delineates the boundary in the cytoplasm where the rough endoplasmic reticulum and
Golgi apparatus of the soma is not continuous with the axon (Wang and He, 2013).
The axon hillock is followed by the initial segment and is the region where an impulse is first generated and travels toward the axon terminals (Leterrier, 2018, McCormick,
2013). While ribosomes are found in the axon, the absence of the other cellular organelles important for protein synthesis renders the axon greatly dependent on the soma for trophic support, which is important in the context of sensory nerves in which
15 Literature Review the soma is outside the protection of the blood-brain barrier (Miller, 2014). This dependence leaves distal regions of the longest axons, such as the axons in the sciatic nerve, susceptible to toxic and metabolic injury (Berthold et al., 2005).
Myelinated Fibres
In the PNS, axons greater than 1–2 µm in diameter are insulated with consecutive myelin segments over the plasma membrane of the axon (Catala and Kubis, 2013).
The axonal membrane, known as the axolemma, is therefore encased by a myelin sheath, which isolates the fibre from neighbouring axons. These successive myelin segments are Schwann cells wrapped in a multilamellar, spiral fashion and spaced approximately 1 µm apart from adjacent Schwann cells, leaving regions of the axolemma without myelin (Freeman et al., 2016). These myelin-free, segments of axolemma are known as the nodes of Ranvier and are exposed to extracellular environment (Berthold et al., 2005). Despite these regularly spaced intervals in the myelin sheath, approximately 99.9% of the axonal membrane is covered in myelin and is termed the internode (Salzer et al., 2008). Myelin greatly increases the resistance between cytoplasm of the axon and extracellular space and also reduces current loss along the axolemma of the internode (Ramahi and Ruff, 2014). This effect of myelin enables almost all the current produced by one node of Ranvier to excite the adjacent node and therefore increase conduction velocity by what is termed saltatory conduction (Ramahi and Ruff, 2014).
Domains of Myelin
Myelin segments range from 300-2000 µm in length and can be subdivided into distinct domains, each with a specific protein composition and function (Figure 1)
16 Literature Review
(Berthold et al., 2005, McCormick, 2013). The node of Ranvier is of particular importance in myelinated fibres. In addition to being exposed to the extracellular environment, this region is distinguished from other domains as it contains an immense density (approximately 1000/µm2) of voltage-gated sodium (Na+) channels
(Ritchie and Rogart, 1977). This distribution of Na+ channels is similar to the axon initial segment and consequently, the node of Ranvier is the site of generation of the action potential in saltatory conduction (Rasband, 2010a). The nodal region is characterised by increased mitochondrial density due to the metabolically active nature of this site (Zhang et al., 2010). Immediately adjacent to the node, is the paranodal region which is 3–5 µm in length (Berthold et al., 2005, Inouye et al.,
2014). At the paranode, microvillar processes from the outermost regions of Schwann cells protrude into the axolemma (Ichimura and Ellisman, 1991). These microvilli contain various matrix proteins that support node formation, maintenance, and function (Corfas et al., 2004, Ogawa et al., 2006). The axolemma at the paranode contains a complex of cell adhesion molecules that includes caspr and contactin, which permits the formation of an axo-glial junction between Schwann cells and axonal membrane (Einheber et al., 1997, Hivert et al., 2016, Menegoz et al., 1997,
Peles et al., 1997, Rios et al., 2000). Continuing underneath the myelin segment, the paranode is followed by the juxtaparanode which extends 10 µm in length (Arroyo et al., 1999, Inouye et al., 2014). The juxtaparanode is characterised by a high density of fast potassium (K+) channels, which may function to dampen nodal excitability after action potential generation to prevent re-excitation or maintain internodal excitability
(Chiu and Ritchie, 1984, Rasband et al., 1998, Wang et al., 1993, Waxman and
Ritchie, 1993). Absence of a normal axo-glial junction results in an accumulation of
K+ channels in the paranode from the adjacent juxtaparanode and an impairment in
17 Literature Review saltatory conduction (Bhat et al., 2001). Thus an important function of the axo-glial junction is to form a barrier between the Na+ channels in the node and K+ channels located in juxtaparanode (Corfas et al., 2004). Finally, adjacent to the juxtaparanode is the internodal region, which constitutes 95% of the length of myelin segments and spans 150–1200 µm (Lascelles and Thomas, 1966, Salzer et al., 2008). Between the internodal axolemma and the inner membrane of the Schwann cell, there is a uniform separation termed the periaxonal space, which spans 15 nm (Salzer et al., 2008). The internodal region is much less densely populated compared to the other domains, however it contains a variety of ion channels, pumps and exchangers involved in the maintenance of membrane potential, and in particular, internodal excitability
(Krishnan et al., 2009a).
Unmyelinated Fibres
In the PNS, there are more thin unmyelinated axons (<1 µm in diameter) than myelinated axons (Feldman et al., 2017, Malik et al., 2011). These axons are known as C fibres and function to carry information for the autonomic nervous system as well as afferent impulses in response to pain (Catala and Kubis, 2013). C fibres are enveloped and grouped together by non-myelinating Schwann cells to form Remak bundles (Feldman et al., 2017). A consequence of the lack of myelin in C fibres is slow impulse conduction (Miller, 2014). Unlike myelinated fibres, the Na+ and K+ channels taking part in action potential generation are distributed uniformly along the axon and instead of saltatory conduction, impulse propagation occurs through local excitation of the neighbouring patch of membrane to generate an action potential
(McCormick, 2013). Injury to these fibres may occur in the early phase of diseases of
18 Literature Review the nervous system and may cause symptoms of burning pain (Lauria et al., 2014,
Malik, 2014). C fibres will be discussed in more detail in the further sections.
Function of Peripheral Axons
The function of peripheral axons is to provide a means of communication for neurons through an action potential. Specifically, the action potential is series of discreet changes in the voltage of the axon produced by the flow of Na+ and K+ across the axolemma (Hodgkin and Huxley, 1952d). Pioneering studies by Hodgkin, Huxley, and Katz on giant squid axons provided the physiological basis of the mathematical model that describes an action potential generation, which was contingent on the permeability of axolemma to Na+ and K+ ions (Hodgkin and Huxley, 1952a, 1952b,
1952c, Hodgkin and Katz, 1949).
Membrane Potential
The sequence of events that underlie the action potential require an actively maintained chemical and electrical concentration gradient across the inner and external surface of the axolemma and is known as the membrane potential. This gradient is fundamental for the action potential and is enabled by the phospholipid bilayer of the membrane that serves as barrier between the intracellular and extracellular environment and is maintained by energy-dependent ion pumps, which are affected in metabolic neuropathies (Cannon, 2014, McCormick, 2013). By convention, the membrane potential is relative to the inside of the axon and can be represented by the equation:
� = ������� ������ �ℎ� ���� (� ) − ������� ������� �ℎ� ���� (� )
19 Literature Review
At rest, axons possess a relative negative charge on the inner surface of the axolemma due the unequal distribution of Na+, K+ and chloride (Cl–) ions across the membrane
(McCormick, 2013). In human axons, the resting membrane potential is –75 to –80 mV and determined by the equilibrium potential for K+ and to a lesser extent, by the equilibrium potential for Na+ and the Na+/K+ pump (Uncini and Kuwabara, 2015). Cl– appears to contribute considerably less to the determination of resting potential of mammalian neurons. If the membrane potential becomes more positive than the resting value, this is termed depolarisation and if the membrane potential becomes more negative than the resting value, this is referred to as hyperpolarisation. While the axolemma is largely impermeable to ions, it contains specialised transmembrane proteins known as ion channels (Cannon, 2014). These ion channels open to form selective permeation pathways that facilitate the passage of specific ions across the phospholipid bilayer through an aqueous pore (Cannon, 2014). Opening and closing of ion channels is said to be voltage-gated as their permeability is regulated by membrane voltage. It is the rapid redistribution of charge through these ions channels that forms the molecular basis of the action potential.
The Action Potential
An action potential is a transient reversal (<1 ms) of the resting membrane potential such that the inside of axolemma becomes positively charged with respect to the outside momentarily (Barnett and Larkman, 2007). The action potential consists of three phases and each phase represents a characteristic change in membrane potential which is driven by the flow of a specific ion across the membrane (Figure 2).
20 Literature Review
+40
Upstroke Downstroke
Threshold Membrane Membrane voltage (mV) depolarisation –75 hyperpolarisation
Resting potential After-hyperpolarisation Time Figure 2. The action potential
Initiation of an action potential may be caused by a chemical or physical stimulus producing a ‘local depolarising response’ or ‘generator potential’ (Barnett and
Larkman, 2007). In motor neurons, which are focus of this thesis, this local depolarisation leads to the opening of a relatively small number of voltage-gated Na+ channels in the axon hillock or node of Ranvier leading to an inward current of Na+ into the axon from the extracellular space (Barnett and Larkman, 2007, Hodgkin and
Katz, 1949). Should depolarisation reach a sufficient level, known as ‘threshold’, an action potential will be triggered. For most neurons, threshold is around –45 to –55 mV. At this membrane potential, the complement of open Na+ channels is sufficient for the inward Na+ current to exceed the combined outward current of K+ and Cl– ions
21 Literature Review that opposes membrane depolarisation (McCormick, 2013). At threshold, there is rapid opening of many more voltage-gated Na+ channels, resulting in an even greater influx of Na+ (McCormick, 2013). This increase of Na+ influx results in further depolarisation, which in turn triggers the opening of more Na+ ion channels in a positive feedback fashion (McCormick, 2013). In this sense, the process of voltage- gated Na+ channel recruitment is termed ‘regenerative’; once threshold is reached, an action potential will always occur, therefore being an ‘all-or-none’ response. Once an action potential has been triggered, the depolarisation can reverse the membrane potential to as high as +40 mV, but not quite reaching the equilibrium potential of Na+
(Barnett and Larkman, 2007). The equilibrium potential of Na+ is the membrane potential at which there is no net movement of Na+ across the axolemma and is approximately +66 mV (Cannon, 2014). Depolarisation of the membrane is limited due to the fast inactivation kinetics of the transient Na+ ion channels involved the action potential (McCormick, 2013).
The second phase of the action potential involves restoration of the membrane potential towards the resting value by what is termed repolarisation. While voltage- gated K+ channels open in this phase allowing the efflux of K+ ions, repolarisation is primarily the result of the inactivation of transient Na+ channels and current leak into the internode (Ritchie, 1995). Voltage-gated K+ channels were found to be responsible for repolarisation of the axolemma in giant squid axons, however they have a negligible contribution in human axons (Schwarz et al., 1995).
The final phase of the action potential is where the effect of voltage-gated K+ channels is primarily seen in human axons. The slow deactivation kinetics of some
22 Literature Review voltage-gated K+ channels enables the prolonged efflux of K+ ions, causing an after- hyperpolarisation beyond the resting membrane potential (McCormick, 2013).
Resting membrane potential is eventually returned after the closure of these ion channels and activity of energy-dependent pumps.
Saltatory Conduction
Saltatory conduction (from the Latin saltare, “to leap”) refers to the propagation of action potentials along myelinated axons from one node of Ranvier to the next
(Huxley and Stämpeli, 1949). Classically, this is thought to be due to the insulator effect of myelin, which enables almost all the current produced by one node of
Ranvier to travel through the internode (without leak) and excite the adjacent node with an internode delay of only about 20 μs (McCormick, 2013, Ramahi and Ruff,
2014). Importantly, the magnitude of current generated at the node is >5 times greater than what is required to generate an action potential at the neighbouring node and is referred to as the safety factor of transmission (Huxley and Stämpeli, 1949, Stämpfli,
1954, Tasaki, 1953). The safety factor ensures the propagating impulse will not diminish and can be calculated as a ratio of current leaving the node of Ranvier to current required for excitation at the next node (Tasaki, 1953). This ratio must be greater than one for current conduction through a node to be successful and it can be adversely affected in demyelinating pathologies (Bowley and Chad, 2019, Kiernan and Kaji, 2013).
Ion Channels, Pumps, and Exchangers
The molecular basis of the action potential is formed by the presence of ion channels that form selective permeation pathways across the phospholipid bilayer of the axonal
23 Literature Review membrane. While the first electrophysiological recordings from individual ion channels were not made until the 1970s, Hodgkin and Huxley predicted the key properties now known to be essential for the action potential: ion selectivity, voltage sensitivity, and importantly, channel closing, which ensures the action potential moves along the axon in one direction (Barnett and Larkman, 2007, Neher and
Sakmann, 1976). Ion channels are transmembrane proteins that have a three- dimensional structure and an aqueous pore (McCormick, 2013). The net flow of ions through these channels is the source of the electrical current that rapidly changes the membrane potential during an action potential (Cannon, 2014). These rapid transients are possible because of the high density of ion channels, the high throughput of an open channel (107ions/s), and the ability of channels to change conformation from open or closed within 1 ms or less (Cannon, 2014). Many of the ion channels involved in the action potential are voltage-gated, as changes in their conformation to open or close are in response to alterations in membrane potential (McCormick, 2013). The movement of ions in or out of the axon is determined by their concentration gradient across the axolemma, which is established by energy-dependent pumps and exchangers (Barnett and Larkman, 2007). The following section outlines the variety of ion channels, pumps and exchangers that present in specific areas of the axon and are essential to impulse conduction and other electrical properties of the axon (Figure
3).
24 Literature Review
Myelin sheath
+ + + Na Na Na Na+/K+ Ca2+ Nap Nat Nat Kf Ih 3Na+ K+ K+ K+ 3Na+ 2K+
2+ Ks Ca Ks 3Na+ Na+/Ca2+ Na+/K+–ATPase exchanger
Node Paranode Juxtaparanode Internode
Figure 3. Distribution of ion channels, pumps, and exchangers in a myelinated axon. Nat: + + transient voltage-gated Na channel; Nap: persistent voltage-gated Na channel; Kf: fast + + voltage-gated K channel; Ks: slow voltage-gated K channel; Ih: inward rectifier
Sodium (Na+) Channels
Action potentials in mammalian axons can be modelled entirely by voltage-gated Na+
(NaV) channel kinetics, therefore making these channels the most important contributor to neurotransmission (Schwarz et al., 1995). Structurally, NaV channels comprise of a single pore-forming a-subunit with one or two auxiliary b-subunits
(Krishnan et al., 2009a, Noda et al., 1984).
The a-subunit is a large, single polypeptide glycoprotein, with a molecular mass of approximately 270 kDa (McCormick, 2013). It consists of four (I-IV) linked homologous transmembrane protein domains, with each domain comprising of six a- helical segments that span the neuronal membrane (S1-6) (Eijkelkamp et al., 2012).
The a-subunit contains specialised regions, which are necessary for activation,
25 Literature Review inactivation and ion permeation. The S4 segment of each domain acts as the membrane ‘voltage sensor’ and has an amino acid sequence that is conserved among the different voltage-gated ion channels (Catterall, 2012, Stühmer et al., 1989).
Importantly, this segment can respond to alterations in membrane potential to mediate the voltage-dependant conformational that occur during channel activation and inactivation (Krishnan et al., 2009a). An additional pore loop between S5 and S6 contributes to the formation of the pore and acts as a selectivity filter to only permit the passage of Na+ from the extra-cellular environment through the aqueous pore
(Krishnan et al., 2009a).
The b1 and b2-subunits are members of the immunoglobulin domain family of cell- adhesion molecules with molecular masses of 39 and 37 kDa, respectively
(McCormick, 2013, Namadurai et al., 2015). They contain a large extracellular domain, a transmembrane region and a small intracellular domain. To date, four b- subunits (NaVb1-4) have been identified (Namadurai et al., 2015). The b subunits interact with the extracellular matrix, intracellular cytoskeleton, and cell adhesion molecules (Catterall et al., 2005). They are involved in channel localisation, modulating channel gating, as well as increasing the amplitude of the inward Na+ current (Catterall et al., 2005, Gurnett and Campbell, 1996, Isom, 2001).
Correct action potential generation and propagation is dependent on the three distinct operational states of NaV channels: closed, open and inactivated. At the resting membrane potential, NaV channels are closed. Membrane depolarisation causes a conformational change of the S4 helix, which in turn opens the channel to permit the influx of Na+ (McCormick, 2013). Positive charges in the S4 region may act as
26 Literature Review voltage sensor such that an increase in the positivity inside the cell results in the conformational change (McCormick, 2013). Sustained depolarisation causes the channel to become inactivated. The channel becomes inactivated within milliseconds of opening and is caused by folding of the intracellular loop that binds domains III-
IV, thereby blocking the channel (Taddese and Bean, 2002). The mechanism of inactivation is hypothesised to be due to a block of the aqueous pore triggered or facilitated as a secondary consequence of activation (McCormick, 2013).
To date, ten Nav a-subunit isoforms have been cloned and functionally characterised
(NaV1.1–1.9 and Nax) (Catterall, 2012, Hiyama et al., 2002). These a-subunit isoforms are expressed in a tissue-specific fashion and have different electrophysiological characteristics, kinetic properties, and toxin sensitivities (Goldin,
2001). There are three isoforms that are primarily present in the peripheral nervous system, NaV1.7, NaV1.8 and NaV1.9 (Bennett, 2014, Catterall, 2012, McDermott et al., 2019). NaV1.1, NaV1.2, and NaV1.6 have been identified in both the peripheral and central nervous systems (Eijkelkamp et al., 2012). NaV1.6 is of particular importance because it is the predominant isoform present at the node of Ranvier in humans and produces the persistent and transient Na+ currents (Caldwell et al., 2000, Herzog et al., 2003).
During the phase of action potential that corresponds with membrane depolarisation,
+ 98% of the Na current is due to the opening of ‘transient’ NaV (Nat) channels, which exhibit the stereotypical rapid gating kinetics (Burke et al., 2001). The remaining 2%
+ of the total Na current is attributable to ‘persistent’ NaV (Nap) channels, which activate at membrane potentials 10-15 mV more negative than transient NaV channels
27 Literature Review and undergoes incomplete voltage-dependent inactivation (Crill, 1996, Kiss, 2008,
Krishnan et al., 2009a). Thus, Nap channels may be activated at resting membrane potential and permit a Na+ conductance over a wider range of membrane potentials compared to transient NaV channels (Baker and Bostock, 1998). Although Nap channels only contributing a small fraction of the total Na+ current, they have an important role in regulating membrane excitability by amplifying subthreshold depolarising inputs, which contributes to repetitive neuronal firing (Eijkelkamp et al.,
2012, Taddese and Bean, 2002). The activation of Nap channels at resting membrane potential suggests they contribute to ectopic symptom generation such as paraesthesia, pain and cramping (Bennett et al., 2019, Bostock and Rothwell, 1997, Kwai et al.,
2013, Misawa et al., 2009, Mogyoros et al., 1997, Tamura et al., 2006). However, it remains is unclear whether the Nat and Nap conductances arise from different NaV channels or if they arise from the same channel type but with different gating kinetics
(Krishnan et al., 2009a).
Disease States
Mutations in the genes encoding for the various a or b-subunits may result in altered gating kinetics, reduced channel density and improper Na+ currents leading to various neurological disorders. There is a clear association between altered NaV channel function and familial forms of epilepsy and pain disorders (Bennett, 2014, Bennett et al., 2019, Bennett and Woods, 2014, Blesneac et al., 2018, Catterall, 2012, Colloca et al., 2017, Eijkelkamp et al., 2012, Themistocleous et al., 2018). Mutations in these genes have also been implicated in inherited forms of autism, ataxia, and migraine
(Eijkelkamp et al., 2012)
28 Literature Review
Potassium (K+) Channels
In both structure and function, K+ channels represent the most diverse family of ion channels (Humphries and Dart, 2015, Kuang et al., 2015). This section will only focus on K+ channels relevant to the mammalian peripheral axon. Mammalian axons express several types of potassium channels including voltage-gated K+ channels
+ + + (KV), inwardly rectifying K channels (KIR), leak K channels, and K channels sensitive to Ca2+ or Na+ (Coetzee et al., 1999, Krishnan et al., 2009a). K+ channels consist of four a-subunits, which may be associated with auxiliary b-subunits
(Krishnan et al., 2009a). The typical structure of K+ channels is a tetramer of four a- subunits arranged around a central pore, with each subunit contributing one transmembrane domain to form the passage (Kuang et al., 2015). The transmembrane pore has an ion selectivity filter that is highly selective and at least 10,000 times more permeant for K+ than Na+ ions (Kuang et al., 2015). The a-subunit consists of a variable number of transmembrane domains depending on the K+ channel type, which is the initial division of the phylogenetic tree for K+ channels (Humphries and Dart,
+ 2+ 2015). KV channels have six transmembrane domains, K channels sensitive to Ca
+ have seven transmembrane domains, while KIR channels and leak K channels have two and four transmembrane domains, respectively (Humphries and Dart, 2015). In neurons, due to equilibrium potential of K+ (approximately –85 mV), the opening of
K+ channels generally mediates outwards currents which serve to dampen cellular excitability (Humphries and Dart, 2015).
+ Voltage-gated K (KV) Channels
There are 40 known human KV channels, which can be separated into 12 subfamilies on the basis of their a-subunit (KVa1–12) (Gutman et al., 2005). Like Nav channels,
29 Literature Review
the positively charged S4 segment make KV channels electrically sensitive and also mediates the eventual opening of the conduction pathway (Kuang et al., 2015, Long et al., 2007). In addition to their a-subunit, KV channels can be classified further according to their encoding gene, channel kinetics, location and toxin sensitivity
(Krishnan et al., 2009a). While the mammalian myelinated axon has a variety of KV channels, the following section will only focus on the two types of KV most relevant to the in vivo assessment of axonal ion channel function, those with fast or slow kinetics.
Fast KV channels (KV1) were first shown to be concentrated in the juxtaparanode, with lower densities present in the internode (Röper and Schwarz, 1989, Wang et al.,
1993). Further evidence suggests they are also enriched at the axon initial segment
(Rasband, 2010b). In the juxtaparanode of myelinated axons, the Kv1 subunits that have been specifically identified are KV1.1, KV1.2, and KV1.4 (Rasband, 2004,
Tsantoulas and McMahon, 2014). Anchoring molecules on the axolemma and
Schwann cell membrane are crucial for the clustering of juxtaparanodal KV channels
(Hivert et al., 2016, Rash et al., 2016, Uncini and Kuwabara, 2015). Fast KV channels are activated between membrane potentials –50 mV and +50 mV and deactivate within a few milliseconds at membrane potentials of –120 mV and – 60 mV (Reid et al., 1999, Safronov et al., 1993, Vogel and Schwarz, 1995). In non-myelinated axons, blockade of fast KV channels leads to action potential prolongation, which suggests these channels may play a role in membrane repolarisation (Bostock et al., 1981,
Sherratt et al., 1980). However, as mentioned above, in mature myelinated axons repolarisation occurs due to the inactivation of transient Na+ channels and current leak into the internode (Ritchie, 1995). In functional terms, fast KV are thought to be
30 Literature Review responsible for limiting the re-excitation of the node following conduction of an action potential and maintaining a stable nodal resting potential (Barrett and Barrett,
1982, Chiu and Ritchie, 1984, Rasband, 2004, Waxman and Ritchie, 1993).
Slow KV channels are formed by the combination of KV7.2 and KV7.3 subunits
(Gutman et al., 2005). Slow KV are present in highest density at the node and axon initial segment and they have slow activation and deactivation kinetics (Devaux et al.,
2004, Röper and Schwarz, 1989, Schwarz et al., 2006). Slow KV channels do not inactivate and generate a steady outward current that stabilises the membrane in response to depolarising currents (Brown and Passmore, 2009). Slow KV channels are active at resting membrane potential and play an important role in its maintenance
(Reid et al., 1999, Schwarz et al., 2006). Due to their slow activation, slow KV channels are not thought to be responsible for membrane repolarisation following action potential conduction, however they do produce a small hyperpolarising afterpotential to reduce neuronal excitability (Baker et al., 1987, Eng et al., 1988).
Slow KV channels are also activated in response to prolonged depolarisation during high frequency activity to prevent inappropriate after-discharge (Schwarz et al.,
2006).
Disease States
In humans, loss of function mutations of KV1.1 are associated with episodic ataxia type 1, an autosomal dominant condition characterised by myokymia and severe contractions of the head and limbs resulting in loss of coordination and balance
(Browne et al., 1994). An accumulating body of research suggests that some human neuropathic pain syndromes are caused by autoimmune antibodies against KV1
31 Literature Review subunits (Bennett and Vincent, 2012, Tsantoulas and McMahon, 2014). Peripheral nerve demyelination results in the redistribution of the fast KV channels into the nodal, paranodal, and internodal regions of the axon due to the loss of structural molecules that usually anchor these channels in the juxtaparanode and results in impaired impulse conduction (Arroyo et al., 1999, Boyle et al., 2001, Rasband et al.,
1998). Mutations in KV7.2 and KV7.3 lead to neonatal epilepsy and myokymia, providing further evidence these channels play an important role in maintaining resting membrane potential and limiting repetitive firing (Dedek et al., 2001,
Humphries and Dart, 2015). Loss of function mutations in KV7.3 have also been associated with autism spectrum disorders (Gilling et al., 2013).
Inwardly Rectifying K+ Channels
+ In terms of structure, KIR channels are the simplest K channels with each subunit being formed by two transmembrane domains separated by a pore-forming region
(Humphries and Dart, 2015). These subunits form homo- or hetero-tetramers to produce functional KIR channels (Bichet et al., 2003, Hibino et al., 2010). KIR channels can be divided into seven subfamilies based on their modulatory mediators and properties of ion conduction (Kuang et al., 2015). The unique feature of KIR channels is that they possess a characteristic asymmetrical K+ conductance whereby
K+ moves into the cell on hyperpolarisation rather than an outward K+ conductance on depolarisation as seen in other K+ channels (Humphries and Dart, 2015, Kuang et al.,
2015). KIR channels also possess a cytosolic domain that regulates the gating of the channel (Kuang et al., 2015). Inward rectification occurs because these channels are blocked by intracellular substances on depolarisation, whereas these blockers are
+ released on hyperpolarisation to permit K influx (Hibino et al., 2010). KIR channels
32 Literature Review tend to be active around the equilibrium potential for K+, and thus help set and maintain the resting membrane potential, but close in response to depolarisation so as not to oppose membrane excitation (Kuang et al., 2015). There are three relevant types of inward rectifiers currents, a K+-selective inward rectification dependent on extracellular K+ concentration, a current mediated by adenosine triphosphate (ATP),
+ + and a mixed cation conductance (Ih) of Na and K through hyperpolarisation- activated cyclic nucleotide-gated (HCN) channels (Humphries and Dart, 2015, Mayer and Westbrook, 1983, Pape, 1996).
The K+-selective channel opens with hyperpolarisation and also exhibits voltage- dependent gating that is influenced by the extracellular K+ concentration, with activation at more depolarised potentials as the extracellular K+ concentration increases (Krishnan et al., 2009a). The conductance depends on the difference between the membrane potential and the K+ equilibrium potential, rather than on membrane potential alone, with an increase in extracellular K+ leading to an influx of
K+ (Birch et al., 1991).
The ATP-sensitive inward rectifier provides a link between cell metabolism and electrical activity (Krishnan et al., 2009a). In the PNS, they are primarily localised in the dorsal root ganglion, where they are thought to play a minor role in setting the basal excitability (Chi et al., 2007, Du et al., 2011, Kawano et al., 2009). These channels are voltage-insensitive however they are inhibited by intracellular ATP (Du and Gamper, 2013, Jonas et al., 1991). Given their location, opening of these channels reduces excitability and pain induced by a range of painful stimuli (Du et al., 2011,
Kawano et al., 2009, Tsantoulas and McMahon, 2014). In situations of low energy
33 Literature Review supply, these channels are open and cause membrane hyperpolarisation, which may protect the axon from the progressive depolarising effects of reduced Na+/K+–ATPase function (Krishnan et al., 2009a, Sun and Feng, 2013). While these channels are mainly found in sensory neurons, agonists of ATP-sensitive channels have been used in animal models of diabetic neuropathy to counter the reduced function of Na+/K+–
ATPase and have demonstrated improvements in motor nerve conduction velocity
(Greene, 1986b, Hohman et al., 2000).
The Ih current flowing through HCN channels plays an important role in the determination of resting membrane potential (Biel et al., 2009, Pape, 1996). HCN channels are voltage-gated channels that belong to a superfamily known as ‘pore-loop cation’ channels (Biel et al., 2009). In mammals, four subunits (HCN1–4) have been identified and their encoding genes are members of the voltage-gated K+ superfamily
(Krarup and Moldovan, 2012). In neurons, mainly HCN1 and HCN2 are expressed
(Chaplan et al., 2003). In vivo, these subunits form homotetramers arranged around a central pore (Biel et al., 2009). Each subunit consists of six transmembrane a-helical segments (S1-S6), with a voltage sensor at S4 and a selectivity filter between S5 and
S6 (Benarroch, 2013, Biel et al., 2009). In the peripheral nervous system, they are mainly localised in the dorsal root ganglion (Biel et al., 2009). While HCN channels are selective for Na+ and K+, the conductance for K+ is greater and an increase in extracellular K+ concentration strongly increases current amplitude (Benarroch, 2013,
Krarup and Moldovan, 2012). HCN channels are activated by hyperpolarisation and directly by cytosolic cAMP (Biel et al., 2009, Krarup and Moldovan, 2012). This conductance, which is thought to originate from the internode, exhibits a slow and complex time course of activation, with an increase in the rate of activation with
34 Literature Review
increasing hyperpolarisation (Baker et al., 1987, Biel et al., 2009). Ih currents begin to activate at –45 mV to –60 mV and activation generally reaches a maximum at –110 mV and does not inactivate (Biel et al., 2009, Pape, 1996). Blockade of this conductance leads to an increase in the input resistance of the cell, suggesting that Ih plays a role in lowering input resistance particularly in situations of membrane hyperpolarisation that would reduce nerve excitability, such as heightened Na+/K+–
ATPase activity during prolonged nerve stimulation (Applegate and Burke, 1989,
Benarroch, 2013). Furthermore, variations in the modulation and expression of HCN channels in sensory axons compared to motor axons has provided some explanation for the functional differences between these neuron types (Howells et al., 2012).
Disease States
Dysfunctional ATP-sensitive inward rectifiers have been associated with generalised seizures after brief hypoxia (Yamada and Inagaki, 2005). Abnormal regulation of
HCN expression or function has been implicated in epilepsy as well as inflammatory and neuropathic pain (Benarroch, 2013, Dibbens et al., 2010, Jiang et al., 2008).
Sodium-Potassium Pump (Na+/K+–ATPase)
+ + The Na /K pump consists of two subunits, a and b arranged in a tetramer (ab)2
(McCormick, 2013). The Na+/K+ pump is located primarily located in the internodal membrane however clusters of this protein has been observed in the nodal and paranodal regions of the axon (Ariyasu et al., 1985, Bostock et al., 1991, Mata et al.,
1991, Wood et al., 1977, Young et al., 2008). The Na+/K+ pump participates in the maintenance of the resting membrane potential by transporting Na+ and K+ against their concentration gradients, with 10% of this maintenance being ascribed to pump
35 Literature Review activity (Thomas, 1972). The Na+/K+ pump is strongly stimulated by increases in intracellular concentration of Na+ and extrudes 3 Na+ ions out of the cell while transporting 2 K+ ions into the cell, achieving this task through the hydrolysis of ATP
(McCormick, 2013, Thomas, 1972). Pump activity also increases with elevations in extracellular K+ and membrane depolarisation and decreases with membrane hyperpolarisation (Rakowski et al., 1989, Rang and Ritchie, 1968). Because of the unequal transport of ions, the operation of this pump generates a hyperpolarizing electrical potential and is said to be electrogenic (McCormick, 2013). Paralysis of the
Na+/K+ pump leads to an excess of positive charge within the axon causing membrane depolarisation. Paralysis also abolishes the direct contribution of the hyperpolarising pump current to the membrane potential, leading to extracellular accumulation of K+ and further depolarisation (Krishnan et al., 2009a). Reduced axonal Na+/K+ pump function due to metabolic changes occurring as a result of hyperglycaemia has been demonstrated in animal models of diabetic neuropathy (Greene et al., 1988, Sima,
1996, Stevens et al., 1993). The intracellular Na+ accumulation that would be expected from pump dysfunction may contribute to the osmotic and structural changes that have been observed in diabetic nerves (Breiner et al., 2017, Brismar,
1993).
Disease States
Mutations in the gene encoding for an a-subunit of the Na+/K+ pump have been documented in some patients with familial hemiplegic migraine as well as benign familial infantile epilepsy, and may underlie the hyperexcitability characteristic of these conditions (De Fusco et al., 2003, Vanmolkot et al., 2003).
36 Literature Review
Na+/Ca2+ Exchanger
The Na+/Ca2+ exchanger is a membrane transporter that maintains intracellular Ca2+ homeostasis and is likely to be located in the node (Craner et al., 2004). It composed of 10 transmembrane helices and extrudes one Ca2+ ion while importing 3 Na+ ions
(Liao et al., 2012, Liao et al., 2016). This process is driven by the electrochemical gradient for Na+ and does not require ATP (Waxman and Ritchie, 1993).
Furthermore, the net direction of transport is determined by both the membrane potential and transmembrane Na+ and Ca2+ gradients (Zhang and David, 2016).
Reverse mode operation of the exchanger, resulting in Ca2+ influx and Na+ efflux, would be favoured by axonal depolarisation and elevated intracellular Na+ concentration, both resulting from energy depletion (Zhang and David, 2016).
Furthermore, reduced function of the Na+/K+ pump due to ATP depletion results in axoplasmic accumulation of Na+, which in turn reverses the Na+/Ca2+ exchanger to remove excess Na+ (Freeman et al., 2016). Consequently, there is Ca2+ accumulation and this may activate calpain, a protease capable of inducing proteolytic cleave of neurofilaments, mitochondrial damage, and Wallerian degeneration (Uncini and
Kuwabara, 2015).
37 Literature Review
Clinical Assessment of Peripheral Nerve Structure and Function
Over the recent years, several new techniques have been developed to investigate peripheral nerve function and structure in health and disease (Gasparotti et al., 2017).
In addition to widening the spectrum of diagnostic tools, these advancements have provided new insights into disease pathophysiology and an understanding of the relationship between structure and function (Frank et al., 2019).
In the assessment of large myelinated fibres, there are various electrophysiological techniques available to investigate nerve function (Shabeeb et al., 2018). At present, the gold standard for assessing nerve function in the clinical setting are nerve conduction studies (NCS). Methodologically, NCS involve maximal surface stimulation of a peripheral nerve and measuring the peak response and conduction velocity at the innervated muscle belly or sensory nerve by placing an active electrode on the skin over the recording site (Tavee, 2019). Measurements obtained from NCS are used to determine the number of functioning nerve fibres and speed of conduction to discern the underlying pathology as either axon loss or demyelination, respectively
(Tavee, 2019). Despite their diagnostic utility, NCS are not able to provide further information on the underlying causes of altered conduction, such as changes in membrane potential or ion channel dysfunction (Kiernan et al., 2005a). Another emerging neurophysiological technique is axonal excitability (also known as nerve excitability) which examines the properties underlying the excitability of the axon
(Kiernan et al., 2020, Tomlinson et al., 2018). While NCS utilise maximal surface stimuli and measure impulse conduction between stimulating and recording sites, axonal excitability studies utilise submaximal surface stimuli and provide insight into the function of ion channels, pumps, and exchangers embedded in the axonal
38 Literature Review membrane at the point of stimulation (Kiernan et al., 2020). The use of submaximal stimulation in axonal excitability studies is also advantageous because it is less painful than nerve conduction studies, which is important for patient comfort and compliance throughout the testing period. In the assessment of large fibre structure, magnetic resonance neurography and ultrasonography have emerged as useful tools to complement electrophysiological examination. For example, nerve ultrasonography provides an indication of the changes in nerve structure (fascicular pattern, epineurium, perineurium), the amount of perineural-endoneurial fluid, and the precise location of pathological processes (Gasparotti et al., 2017). The large fibre assessments relevant to this thesis are axonal excitability studies and nerve ultrasonography.
NCS, axonal excitability studies, and neuroimaging cannot detect the pathological processes occurring in small unmyelinated axons (Gasparotti et al., 2017, Kiernan et al., 2020). Conduction properties of small nerve fibres may be investigated using evoked potentials (laser-evoked, contact heat-evoked, and pain-evoked) and microneurography, however each of these techniques have their own inherent limitations (Colloca et al., 2017, Gasparotti et al., 2017). Methods to assess small fibre structure include skin biopsy, which is the gold standard of small fibre loss, and in vivo confocal corneal microscopy (Colloca et al., 2017). Skin biopsy enables the assessment and reliable quantification of density of intraepidermal and subepidermal nerve fibres, making this technique a useful clinical research tool (Lauria et al., 2004,
Lauria et al., 2011). Corneal confocal microscopy allows the visualisation and quantification of the small nerve fibres of trigeminal origin (Oliveira-Soto and Efron,
39 Literature Review
2001). Of the small nerve fibre assessments, only corneal confocal microscopy is of relevance to this thesis.
40 Literature Review
Axonal Excitability
While some properties of axons were investigated sporadically in the early 1900s, it was not until 1970 that the concept of axonal excitability was thoroughly explored in humans (Bergmans, 1970). Bergmans demonstrated that changes in the minimal voltage required to elicit a response from a single motor unit (a motor neuron and the skeletal muscle fibres innervated by that motor neuron) reflected physiological alterations of the axon (Bergmans, 1970). This voltage was termed ‘threshold’.
Bergmans discovered that by measuring changes in the threshold of axons, induced by artificial polarisation or impulse activity, considerable information about axon physiology, such as membrane potential, could be deduced (Bergmans, 1970, Bostock et al., 1998). However, despite the importance of these findings, the technical difficulty of his method and the continual manual adjustment it required proved to be a major hurdle for routine use of these techniques. Ultimately, this hindered the expansion of nerve excitability to the clinical setting.
Axonal Excitability Studies
The development of a computer assisted axonal excitability assessment software,
Qtrac, renewed interest in axonal excitability as a clinical technique for the investigation of peripheral nerve function in health and disease (Kiernan et al., 2000).
Since the introduction of Qtrac, nerve excitability studies have been utilised in vivo to investigate the pathophysiological mechanisms underlying metabolic neuropathies, immune-mediated neuropathies, hereditary neuropathies, channelopathies, neurodegeneration, neurotoxicity, ataxia, and trauma (Kiernan et al., 2020, Tomlinson et al., 2018).
41 Literature Review
Current axonal excitability study protocols utilise ‘threshold tracking’ as the preferred technique (Bostock et al., 1998). The basic premise of threshold tracking is to measure the change in the stimulus strength required to produce a compound action potential of a defined size. This required stimulus is termed the threshold. Typically, the target compound action potential corresponds to ~40% of the maximum compound action potential. Further, tracking a compound actional potential of defined size allows the study of axons of similar size (Bostock et al., 1998, Burke et al.,
2001). When the test response is smaller than the tracking target, the intensity of the subsequent stimulus is increased and conversely, when the response is larger than the tracking target, the stimulus is reduced (Bostock et al., 1998). Responses are tracked as a percentage change in threshold for normalisation.
While threshold is a measure of excitability and may be used as a biomarker of membrane potential, its interpretation may be confounded by various factors
(Krishnan et al., 2009a). It is therefore necessary to assess multiple threshold parameters across various testing paradigms (Kiernan et al., 2020). The key elements involved in a standard axonal excitability assessment protocol include the initial stimulus–response curve followed by four distinct testing paradigms: strength- duration properties, threshold electrotonus, current-threshold relationship, and the recovery cycle. With the exception of strength-duration properties, the other testing paradigms utilise conditioning stimuli and are designed to investigate how axons behave when membrane potential is changed. Paradigms are summarised in Table 2
(see page 53)
42 Literature Review
Stimulus-Response Curve
First, stimulus–response curves are generated in a dose response pattern by which the stimulus output is increased until the maximal compound muscle action potential is achieved (Figure 4). The stimulus required to reach ~40% of the maximal response, the threshold, is then determined and remains the target for the succeeding test paradigms.
Figure 4. Stimulus-response curve. Curve depicts an increase in compound action potential size with increasing stimulus intensity. Threshold refers to stimulus intensity required to elicit a target response (in this case, 40% of the maximal compound action potential). Membrane depolarisation causes a left-ward shift while hyperpolarisation causes a right-ward shift.
43 Literature Review
Strength-Duration Properties
The relationship between stimulus strength and duration dictates that as the stimulus duration is increased, the stimulus intensity required to produce a target response is reduced and vice versa (Weiss, 1901). It was originally described as an exponential relationship and from this, rheobase was defined as the minimum current (of infinite duration) required to elicit a response from the nerve and chronaxie was defined as the stimulus duration at double the rheobase current (Figure 5) (Lapicque, 1909,
Weiss, 1901). Weiss proposed a simple linear relationship between stimulus charge and stimulus duration to describe strength-duration behaviour: Q = a + bt (Weiss,
1901). This linear relationship has aided the derivation of strength-duration properties in axonal excitability studies, as the rheobase can be calculated from the gradient and the strength-duration time constant (SDTC (τSD), analogous to the chronaxie in human peripheral axons) can be calculated from the x-intercept from only two stimulus durations (Figure 5) (Bostock, 1983, Mogyoros et al., 1996). τSD illustrates the rate at which threshold current increases as stimulation duration is reduced, which increases with depolarisation and decreases with hyperpolarisation (Krishnan et al., 2009a,
Mogyoros et al., 1996). τSD reflects the passive properties of the nodal membrane and is a surrogate marker of Nap function (Bostock, 1983, Bostock and Rothwell, 1997,
Mogyoros et al., 1996). As mentioned previously, Nap channels have an important role in repetitive neuronal firing and ectopic symptom generation, making them a clinically relevant axonal excitability measure (Bostock and Rothwell, 1997).
44 Literature Review
A
Rheobase Stimulus StrengthStimulus (mA)
Stimulus Duration (ms) Chronaxie (τSD) B ) mA.ms ( Slope = Rheobase Charge
τSD Threshold
0 Stimulus Duration (ms)
Figure 5. Strength-duration relationship. (A) Plot of stimulus strength versus stimulus duration, illustrating rheobase as the threshold current required for a stimulus of infinite duration and chronaxie as the stimulus duration at which the threshold current is twice the rheobase. (B) Plot of threshold charge versus stimulus width, illustrating the determination of strength-duration time constant using Weiss’ law. Strength–duration time constant is determined as the negative intercept on the x-axis of the line measured using two stimulus widths.
45 Literature Review
Threshold Electrotonus
Threshold electrotonus (TE) examines properties of the internodal membrane
(Kiernan et al., 2020). Assessment of TE is achieved by measuring the change in threshold at multiple time points during and after prolonged (100 ms) subthreshold depolarising (TEd; +40% of threshold) and hyperpolarising (TEh; –40% of threshold) conditioning currents (Figure 6). These conditioning currents are not sufficient to produce an action potential, however they enable a slow spread of current under the myelin sheath into the internode, consequently altering the membrane potential and activating a number of accommodative internodal conductances (Baker et al., 1987,
Kiernan et al., 2020). Threshold electrotonus is the only clinical method of assessing internodal conductances in vivo. As the internodal region accounts for approximately
99.9% of the axonal membrane, this technique provides valuable information regarding the complex changes in excitability that occur with membrane polarisation
(Salzer et al., 2008). The pattern of threshold change to depolarising and hyperpolarising currents can be broken down into distinct characteristic phases.
46 Literature Review
Figure 6. Threshold electrotonus. Curve illustrates change in threshold during and after 100 ms subthreshold polarising currents (depolarising direction plotted upwards in red and hyperpolarisation direction plotted downward in blue). Black arrows indicate changes in threshold electrotonus with membrane depolarisation (‘fanning-in’) while white arrows depict changes with membrane hyperpolarisation (‘fanning-out’).
In response to subthreshold depolarising currents, an initial depolarising fast phase (F phase) develops that is proportional to the applied conditioning current and reflects depolarisation of the node (Krishnan et al., 2009a). The F phase is followed by a slow depolarisation phase (S1 phase) due to the gradual spread of current into the internode
47 Literature Review
and activation of fast KV channels, which limits responses to depolarising currents
(Baker et al., 1987, Bostock et al., 1998, Kiernan et al., 2020). Subsequently, nodal slow KV channels are activated and produce an accommodative response to depolarisation and slow decay toward control threshold (S2 phase) (Baker et al., 1987,
Bostock and Baker, 1988). When the depolarising condition current ends at 100 ms, threshold rapidly returns to baseline and overshoots controls levels before slowly returning to baseline. This undershoot is attributed to the slow deactivation of slow
KV channels and repolarisation of the nerve (Bostock and Rothwell, 1997).
In response to subthreshold hyperpolarising currents, there is an initial hyperpolarising fast phase (F phase) proportional to the applied conditioning current and reflects nodal hyperpolarisation (Kiernan et al., 2020). This is followed by a slow hyperpolarising phase (S1 phase), which is more pronounced than the S1 phase seen in TEd as internodal KV channels are deactivated by hyperpolarisation and thus do not limit the extent of polarisation (Bostock et al., 1998). Finally, extended hyperpolarisation is eventually tempered by the activation of inwardly rectifying cation current (Ih) (Pape, 1996). With cessation of the hyperpolarising stimuli, there is a slow return of threshold to baseline followed by an overshoot as a result of the slow
+ deactivation of Ih and reactivation of persistent Na currents (Kiernan et al., 2020).
Compared to other nerve excitability measures, there is marked variability in TEh, with the between-subject variability being much greater than the within-subject variability (Howells et al., 2013, Tomlinson et al., 2010). This variation is due to differences in the voltage for half-activation of Ih between individuals (Howells et al.,
2012, Jankelowitz et al., 2007a).
48 Literature Review
Importantly, threshold electrotonus is also a sensitive measure of membrane potential
(Baker and Bostock, 1989, Kiernan and Bostock, 2000). Depolarisation of the membrane potential causes an activation of fast and slow KV channels so that conductance is increased and the threshold electrotonus waveform becomes flatter, resulting in a ‘fanning-in’ appearance (Bostock et al., 1998, Kiernan and Bostock,
2000). Hyperpolarisation of the membrane potential reduces the conductance of the internodal membrane, causing a ‘fanning-out’ of the waveform (Bostock et al., 1998,
Kiernan and Bostock, 2000).
Current-Threshold Relationship
The current-threshold relationship (I/V) enables the assessment of the rectifying properties of the axon by using longer lasting conditioning currents of variable strengths. This is achieved by tracking the change in threshold of a 1 ms test impulse following injection of 200 ms conditioning currents stepped from –100%
(hyperpolarising) to +50% (depolarising) of threshold. Strong depolarising currents result in outward rectification which is achieved by the activation of fast and slow KV channels, while strong hyperpolarising currents result in activation of the inwardly rectifying cation conductance, Ih (Kiernan et al., 2000). The current-threshold relationship results in a characteristic plot analogous to a conventional current-voltage plot and enables quantification of the resting input conductance as well as inwardly and outwardly rectifying conductances (Figure 7) (Kiernan et al., 2020). Resting input conductance is observed from the slope between –10% to +10% current injection and will be affected by channels open at resting membrane potential (Kiernan et al.,
2020). The steepness of the curve in the depolarising and hyperpolarising direction reflects outward and inward rectification respectively; the smaller the change in
49 Literature Review threshold for an injected current, the greater the accommodation (Kiernan et al.,
2020). For example, a steepening in response to hyperpolarising currents indicates greater Ih.
50
Outward rectification
0 Resting I/V slope Current (% threshold)
-50 Inward rectification
-100 -500 0 Threshold Reduction (%)
Figure 7. Current-threshold relationship. Polarising currents are 200 ms in duration but vary in strength. Curve is analogous to a current–voltage plot, with response to depolarising current depicted in the upper right quadrant in red and response to hyperpolarizing current in the lower left quadrant in blue.
Recovery Cycle
The recovery cycle utilises paired pulses with varying interstimulus intervals
(between 2 and 200 ms) to assess the recovery of excitability following impulse conduction. Following impulse conduction myelinated axons undergo a well-
50 Literature Review recognised sequence of changes in excitability (Figure 8).
Figure 8. The recovery cycle. Plot illustrates changes in threshold at various time points (2– 200 ms) following a supramaximal conditioning impulse.
Immediately following an action potential, inactivation of transient Na+ channels produces an absolute refractory period, during which the axon is completely inexcitable and no further action potentials can be generated (Hodgkin and Huxley,
1952d). It is impossible to measure the duration of absolute refractoriness because there is a limitation on the stimulus that can be delivered to a patient (Burke et al.,
2001). As Na+ channels recover from inactivation, there is a relative refractory period, which is characterised by decreased excitability. During the relative refractory period, an action potential may be generated with a greater threshold current than normal. In
51 Literature Review human axons, this period typically lasts for 3 ms (Hess et al., 1979, Maurer et al.,
1977). Relative refractoriness varies with membrane potential due to the effect on Na+ channels. Membrane depolarisation increases the relative refractory period due to the greater extent of inactivated of Na+ channels (Burke et al., 2001). Conversely, hyperpolarising shifts in membrane potential decrease the number of inactivated Na+ channels and decreases relative refractoriness (Burke et al., 2001). Unlike most nerve excitability measures, the relative refractory period is extremely temperature sensitive
(Kiernan et al., 2001). While refractoriness is mainly determined by Na+ channels, paranodal demyelination or structural deficiencies may allow fast KV channels to participate (Garg et al., 2018, Howells et al., 2018, Jankelowitz and Burke, 2013).
Following the refractory period, the axon becomes more easily excitable, termed superexcitability. The basis of superexcitability reflects the size of the depolarising afterpotential produced by a passive capacitive charged stored on the internodal membrane (Barrett and Barrett, 1982, Kiernan et al., 1996). This produces a re- excitation of the node by the back-flow of current from the internodal membrane through low resistance pathways under and through the myelin sheath (Barrett and
Barrett, 1982). Superexcitability peaks between 5–7 ms following impulse conduction.
Superexcitability is followed by another period of reduced excitability, termed subexcitability. Subexcitability is due to axonal hyperpolarisation and subsides with the gradual closure of slow KV channels opened during the conditioning discharge
(Lin et al., 2000, Taylor et al., 1992). Axons begin to become subexcitable at 15-20
52 Literature Review
ms and peak subexcitability occurs at approximately 35-40 ms. Resting level of
excitability returns approximately 100 ms following impulse generation.
Table 2. Nerve excitability testing paradigms Testing Channels Outcome Measures Significance and Pathological States Paradigm Assessed
Strength- • Nap • Rheobase • τSD reflects passive properties of nodal duration • SDTC (τSD) membrane properties • Ectopic symptom generation (↑ τSD)
Threshold • Kf • Various time • Examines internodal conductances electronus • Ks points • Membrane depolarisation (‘fanning-in) vs.
• Ih • Peak threshold membrane hyperpolarisation (fanning out’) reduction • S2 accommodation
Current- • Kf • Resting I/V slope • Assesses outward (Kf and Ks activation) threshold • Ks • Minimum I/V and inward (Ih activation) rectifying relationship • Ih slope properties of axon
Recovery • Nat • RRP • Membrane depolarisation (increase in cycle • Kf • Superexcitability RRP) vs. membrane hyperpolarisation
• Ks • Subexcitability (decrease in RRP) • Paranodal demyelination or structural deficiencies
+ + Nat: transient voltage-gated Na channel; Nap: persistent voltage-gated Na channel; Kf: fast
+ + voltage-gated K channel; Ks: slow voltage-gated K channel; Ih: inward rectifier; SDTC
(τSD): strength-duration time constant; RRP: relative refractory period
Sources of Variability in Axonal Excitability Studies
Investigators have examined the influence of particular demographic variables in
studies of axonal excitability variability in healthy controls. There are small but
significant changes with age, conflicting results regarding sex, and a minimal effect
with BMI (Casanova et al., 2014, Jankelowitz et al., 2007b, McHugh et al., 2011).
53 Literature Review
Studies have not explicitly assessed ethnicity as an independent variable on nerve excitability measures. Other sources of variability are site of stimulation, temperature, serum K+, and certain drugs.
Site of Stimulation
Nerve excitability studies examine properties of the axonal membrane at the point of stimulation. Studies have compared nerve excitability recordings of the median nerve
(to abductor pollicis brevis) and ulnar nerve (to abductor digiti minimi) to assess their interchangeability within healthy subjects (Murray and Jankelowitz, 2011). While the current-threshold relationship and recovery cycle were similar, differences were found in the strength-duration time constant and measurements from threshold electrotonus, which questioned the interchangeability of these sites. Of particular relevance to the development of peripheral neuropathy, studies have demonstrated length-dependent differences in nerve excitability between the upper and lower limbs.
Slow KV conductances are more prominent for median axons than for peroneal axons, suggesting that axons innervating the lower limbs have less protection from depolarising stress and could develop ectopic activity more readily (Kuwabara et al.,
2001). This is supported by further evidence that compared to proximal axons in the lower limb, distal axons are more susceptible to ischaemia (Krishnan et al., 2005a).
Effect of Temperature
While the effect of temperature has effects on nerve conduction velocity and amplitude, its effect on axonal excitability is less pronounced in the range of temperatures typically encountered in a clinical setting (Kiernan et al., 2000, Kiernan et al., 2001). Nevertheless, in axonal excitability studies skin temperature should be
54 Literature Review measured at the site of stimulation, be greater than 32 °C, and kept stable throughout the test (Kiernan et al., 2020). The insensitivity of most nerve excitability measures to temperature change is explicable by their strong dependence on membrane potential which is only minimally affected by temperature (Kiernan et al., 2001). However, it must be noted that the relative refractory period is very sensitive to temperature due to alterations in the kinetics of Na+ channel gating (Kiernan et al., 2000). With respect to the range of temperatures that may be encountered clinically (29–35 °C), Na+ channel inactivation is accelerated with warmer temperatures and the relative refractory period is shortened, while with cooler temperatures the relative refractory period is prolonged (Kiernan et al., 2001). At more extreme temperatures, the effects on axonal excitability are more pronounced. With extreme cooling (20 °C) and hyperthermia
(~40 °C), axons show changes in nerve excitability consistent with alterations in membrane potential and changes in accommodation (Howells et al., 2013, Kovalchuk et al., 2018).
Effect of Serum K+
Both intra- and extracellular K+ concentration are well established determinants of impulse generation and conduction (Hodgkin and Huxley, 1952d). The evidence suggesting that serum K+ has an effect on nerve excitability was observed in end- stage renal disease, in which there was various changes in nerve excitability measures consistent with membrane depolarisation that were attributed to the hyperkalaemia
(Krishnan et al., 2005b, Z'Graggen et al., 2006). Further studies confirmed that it was in fact serum K+ responsible for these changes and not other serum solutes. Arnold and colleagues (2014) applied a serum K+ clamp during haemodialysis while all other solutes were removed and motor excitability studies were performed before, during,
55 Literature Review and after the session (Arnold et al., 2014). During the dialysis session, membrane depolarisation was sustained and only normalised after serum K+ had been dialysed to within normal limits. Changes in axonal excitability due to alterations in serum K+ concentration have since been extended to healthy populations and it was found that the relative refractory period and superexcitability were sensitive to serum K+
(Kuwabara et al., 2007). Boërio and colleagues (2014) demonstrated that nerve excitability measures were affected by serum K+ even within the normal physiological ranges of 3.5–4.5 mmol/L (Boërio et al., 2014). Significant differences were noted between low-normal (3.5–3.9 mmol/L) and high-normal (4.3–4.5 mmol/L) values and that the direction of change in axonal excitability recordings was consistent with patterns reported for abnormally high serum K+ in end-stage renal disease.
Effects of Drugs
A number of drugs have purported effects on ion channel function and may therefore influence axonal excitability recordings. Riluzole, mexiletine, lidocaine, and flupirtine have been demonstrated to affect peripheral nerve excitability (Fleckenstein et al.,
2013, Kuwabara et al., 2005, Moldovan et al., 2014, Vucic et al., 2013). However, the exact effects of this drugs remain unclear.
Limitations of Axonal Excitability
While nerve excitability studies are useful tools in providing information regarding axonal ion channel function, there are limitations which must be considered. A significant limitation is that their application is specific to the assessment of large myelinated fibres. Small, unmyelinated fibres may be damaged in various metabolic, infectious, genetic, immune-mediated, and drug-induced conditions (Themistocleous
56 Literature Review et al., 2014). Axonal excitability studies cannot assess small fibre function in these conditions. However small fibre structure and function may be investigated using thermal thresholds, microneurography, evoked potentials, sudomotor function, laser
Doppler flare, skin biopsy, and in vivo corneal confocal microscopy (Malik, 2020).
Another potential limitation of axonal excitability occurs when there is significant axonal degeneration. As mentioned above, nerve excitability studies typically use
40% of the maximal compound muscle action potential as the threshold for all testing paradigms and this allows the study of axons of similar size (Bostock et al., 1998).
However, the nerve excitability measures obtained at this tracking level are only reflective of the axons recruited at that stimulus intensity. This becomes problematic when attempting to compare diseased axons with healthy axons if electrical recruitment of the diseased axons is unpredictable (Shibuta et al., 2010, Shibuta et al.,
2013). Analyses of nerve excitability recordings made at a single tracking level therefore limits the scope of axonal function that can be determined.
Mathematical Modelling of Axonal Excitability
Nerve excitability properties depend on a complex interaction between nodal and internodal ion channels and ion pumps, passive membrane components, and the intracellular and extracellular constituents of the peripheral nerve (Kiernan et al.,
2020). Because of this complexity, deducing the biophysical basis of abnormal nerve excitability recording or predicting the effects of a pathological change is very difficult. This issue has been partially circumvented through the development and refinement of computer-assisted mathematical modelling of nerve excitability recordings to aid interpretation (Boërio et al., 2014, Bostock et al., 1991, Howells et
57 Literature Review al., 2012, Kiernan et al., 2005b). Since its development, modelling has been implemented in nerve excitability studies of metabolic, immune-mediated, toxic, inflammatory, degenerative, and hereditary neuropathies (Garg et al., 2018, Howells et al., 2018, Jankelowitz and Burke, 2013, Kwai et al., 2016a, Liang et al., 2014, Lin et al., 2008).
Nerve Ultrasonography
Ultrasound is defined as a sound wave (oscillating mechanical pressure wave) with frequency above 20 kHz, which is the upper limit of the acoustic range of normal human hearing (Vlassakov and Sala-Blanch, 2015). Sound waves are generated by transducers that consist of piezoelectric crystals. Transducers function both as emitters (transforming an electrical signal into mechanical vibration, known as the reverse piezoelectric effect) and receivers (converting the returning mechanical waves into an electrical signal, known as the piezoelectric effect) of ultrasound (Vlassakov and Sala-Blanch, 2015).
Ultrasonographic assessment of peripheral nerves (the recurrent laryngeal nerve and nerves of the upper and lower limb) in healthy subjects, patients, and cadavers were reported as early as the 1980s (Fornage, 1988, Solbiati et al., 1985). A few years later,
Buchberger and colleagues (1991) demonstrated the utility of ultrasound for the diagnosis of carpal tunnel syndrome (Buchberger et al., 1991). A decade after the commencement of peripheral nerve ultrasound, improvements in image resolution allowed the correlation of the appearance of sonographs with histological findings as well as accurate morphometric studies (Heinemeyer and Reimers, 1999, Silvestri et al., 1995). Diagnostically, peripheral nerve ultrasonography complements
58 Literature Review electrophysiology studies and clinical examination by providing real-time visualisation of normal and abnormal morphology, mobility, and vascularity of nerves and surrounding tissues (Vlassakov and Sala-Blanch, 2015). Imaging in transverse
(cross-sectional) and longitudinal planes, dynamic examination, and colour doppler are available in aiding accurate diagnosis of nerve pathology (Suk et al., 2013). Major limitations of nerve ultrasonography as a technique is that it is heavily operator- dependant and standardised protocols are evolving (Vlassakov and Sala-Blanch,
2015).
Sonographic Features of a Normal Peripheral Nerve
Axons are enveloped by a thin supporting layer known as the endoneurium. A group of axons and their endoneuriums form a fascicle (Figure 9). The nerve fascicle constitutes a distinct structural and functional unit and is surrounded and protected by a thick cellular layer known as the perineurium (Reina et al., 2015). The nerve fascicles in a peripheral nerve are further enveloped and protected by the epineurium, which composed of fibroadipose tissue (Reina et al., 2015). Fascicles can be isolated or clustered in fascicular groups, with connective tissue between them. The epineurium can be divided into a thicker epineural membrane, which surrounds the entire peripheral nerve, and an interfascicular epineurium between fascicles or fascicular groups. The interfascicular epineurium contains a greater amount of fat, intrinsic blood vessels, and other supportive tissues (Reina et al., 2015).
Most peripheral nerves of the upper and lower limbs, including cervical roots and the brachial plexus, can be easily imaged at high-quality using ultrasound provided the nerve is less than 5–6 cm in depth from the skin (Vlassakov and Sala-Blanch, 2015).
59 Literature Review
On cross-sectional scans, distal peripheral nerves have a typical ‘honeycomb’ appearance, with ovoid hypoechoic nerve fascicles dispersed within a milieu of hyperechoic perineurial connective tissue (Vlassakov and Sala-Blanch, 2015).
Longitudinally, the fascicles appear as linear hypoechoic bands. The epineurium usually appears hyperechoic due to its composition of dense connective tissue with high acoustic impedance. Its distinctive appearance helps to delineate nerves from their surrounding structures.
A validated quantitative method of assessing peripheral nerve sonographs is measuring the nerve cross-sectional area in the transverse plane by tracing the inner margin of the epineurium (Figure 9) (Cartwright et al., 2008, Won et al., 2013).
Peripheral nerve cross-sectional measurements have good intraobserver, interobserver, and side-to-side reliability (Cartwright et al., 2013a, Tagliafico and
Martinoli, 2013). The relationship between cross-sectional area with age, weight, height, and sex has varied across studies (Cartwright et al., 2013b). Cross-sectional area varies with site and is usually greater distally and closer to entrapment sites in healthy individuals (Zaidman et al., 2009).
Sonographic Features of an Injured Peripheral Nerve
The principle ultrasound signs of neuropathology are an increase in cross-sectional area and changes in fascicular architecture and echogenicity (Figure 9). Nerve ultrasound may be used in the study of systemic neuropathies in which the peripheral nervous system is widely affected, and focal neuropathies involving a specific nerve at a specific site which are often due to compression. The most common systemic polyneuropathy sonographic sign is the increased cross-sectional nerve area
60 Literature Review
(Vlassakov and Sala-Blanch, 2015). In situations of nerve compression, the most significant ultrasound signs are proximal thickening of the nerve, an abrupt thinning in the area of compression, distal oedema, loss of nerve ultrastructure (Vlassakov and
Sala-Blanch, 2015).
A B
Figure 9. Sonograph of healthy (A) and neuropathic median nerve (B). Image demonstrates increase in median nerve cross-sectional area (yellow dashed lined) in patient with diabetic neuropathy.
In Vivo Corneal Confocal Microscopy
The cornea is the most densely innervated structure in the human body (Shaheen et al., 2014). 50–450 sensory trigeminal neurons transmit nerve fibres via the ophthalmic division of the trigeminal nerve to enter the cornea as stromal nerves, which form the dense sub-basal plexus, and then converge to an inferior-central location termed the inferior whorl (Al-Aqaba et al., 2019, Müller et al., 2003). Free nerve endings originate from these sub-basal nerves (Müller et al., 2003). Density of corneal nerves is greater centrally than peripherally and the entire cornea is estimated to contain 7000 nerve terminals per square millimetre (Marfurt et al., 2010, Müller et al., 2003). Corneal nerves are responsible for the sensations of touch, pain, and temperature and play an important role in the blink reflex, wound healing, and tear
61 Literature Review production and secretion (Shaheen et al., 2014). Approximately 20% of corneal nerves are thinly myelinated Aδ mechanoreceptors that generate acute pain (Shaheen et al., 2014). Another 70% are polymodal, which fire in response to wide range of noxious stimuli (Al-Aqaba et al., 2019). The remaining 10% are C fibre cold receptors (Shaheen et al., 2014). As in other peripheral innervation, non-myelinating
Schwann cells wrap several C fibres together to form a Remak bundle, which is essential for maintenance and function of unmyelinated axons and nociceptors
(Müller et al., 2003). Corneal nerve fibres are highly vulnerable to degeneration following minimal perturbation due to metabolic, toxic, immune, or inflammatory injury but are also capable of regeneration (Azmi et al., 2019, Jia et al., 2018, Lewis et al., 2017, Petropoulos et al., 2020, Sacchetti and Lambiase, 2017).
Corneal nerves can be visualised and quantified using in vivo corneal confocal microscopy (CCM) which has emerged as a rapid, non-invasive, and reliable clinical tool over the last 20 years (Oliveira-Soto and Efron, 2001). CCM requires the examiner to capture multiple images of the sub-basal nerve plexus and analyse them using a manual or automated software (Dehghani et al., 2014, Petropoulos et al.,
2014). Measures of interest include nerve density, length, branch density, tortuosity, and fractal dimension (geometric complexity) in the central and inferior whorl regions of the cornea to assess small fibre damage (Figure 10) (Ferdousi et al., 2020b). CCM has been shown to be a highly reproducible method between testing occasions, examiners, and instruments and normative values have been established (Hertz et al.,
2011, Kalteniece et al., 2017, Petropoulos et al., 2013, Tavakoli et al., 2015). CCM has shown comparable diagnostic efficiency with intraepidermal nerve fibre density via skin biopsy, which is the gold standard to assess small nerve fibre neuropathy but
62 Literature Review is invasive, requires laboratory processing, and may cause infection (Abhishek and
Khunger, 2015, Alam et al., 2017, Chen et al., 2015). CCM has predominantly been used to assess patients with diabetes mellitus, but the technique has also been used to demonstrate small-fibre damage in chemotherapy-induced neuropathy, human immunodeficiency virus neuropathy, Fabry’s disease, Charcot-Marie-Tooth disease type 1A, inflammatory neuropathies, as well as Parkinson’s disease, multiple sclerosis, and amyotrophic lateral sclerosis (Markoulli et al., 2018, Petropoulos et al.,
2020). While CCM is emerging as a useful tool in the investigation and diagnosis of diabetic neuropathy, specific studies have observed no differences in measures between patients with and without diabetic neuropathy (Andersen et al., 2018a,
Petropoulos et al., 2015). As the cornea is avascular and the nerve fibres lack myelination, corneal nerves are especially susceptible to the metabolic perturbations that occur in diabetes (Petropoulos et al., 2020). Other evidence also points to impaired blood flow to the trigeminal ganglion in animal models of diabetes as a cause of degeneration (Davidson et al., 2012). A major limitation of corneal confocal microscopy as a technique is that only one-thousandth of the total cornea surface area can be imaged at a time (typically 0.16 mm2) (Allgeier et al., 2018). This small field of view is insufficient to reliably characterise corneal nerve morphology, especially in longitudinal studies. To address this, several mosaicking techniques have been developed (Allgeier et al., 2018). Disadvantages of corneal confocal microscopy include the operator-dependant nature of the technique, patient discomfort, and necessity of a topical anaesthetic.
63 Literature Review
A B
C D
Figure 10: Confocal micrograph of healthy (A, B) and neuropathic (C, D) cornea. Figure demonstrates the normal corneal nerve architecture in the central (A) and inferior whorl (B) regions. In contrast, a loss of corneal nerves in the central (C) and inferior whorl (D) regions is observed in a patient with diabetic neuropathy.
64 Literature Review
Diabetic Neuropathy
Diabetic neuropathy is the most prevalent chronic complication of both type 1 and type 2 diabetes and occurs in more than half of patients with diabetes (Pop-Busui et al., 2017, Selvarajah et al., 2019). Diabetic neuropathy may affect different parts of the nervous system and thus present with diverse clinical manifestations (Feldman et al., 2019). The most common type of neuropathy is distal symmetric polyneuropathy
(DSPN) and accounts for approximately 75% of diabetic neuropathies (Albers and
Pop-Busui, 2014, Dyck et al., 2011). Estimates of the incidence of DSPN after the diagnosis of diabetes vary. Evidence from large observational cohort studies, the
Diabetes Control and Complications Trial (DCCT), and Epidemiology of Diabetes
Interventions and Complications (EDIC) Study suggests at least 20% of type 1 diabetes patients develop DSPN after 20 years and 50% of type 2 diabetes patients develop DSPN after 10 years (Pop-Busui et al., 2017). DSPN places a substantial burden on healthcare systems, society, and the affected individual. In the USA, the total annual cost of managing symptomatic DSPN and its complications was estimated to be between US$4.6 billion and $13.7 billion (Feldman et al., 2019,
Selvarajah et al., 2019). DSPN is also the key initiating factor for the development of foot ulceration and most common cause of lower-limb amputation in high-income countries (Armstrong et al., 2017, Boulton et al., 2018). For the individual, DSPN is associated with increased morbidity, distressing neuropathic pain, and significant impairments in quality of life (Ziegler et al., 2014). The major predictors of DSPN are duration of diabetes and poor glycaemic control (Grisold et al., 2017, Tesfaye et al.,
2005). Other risk factors for the development of DSPN in both type 1 and type 2 diabetes are older age, increased height, smoking, alcohol abuse, hyperlipidaemia, hypertension, and presence of other diabetic complications (Callaghan et al., 2015,
65 Literature Review
Ziegler et al., 2014). In type 2 diabetes, obesity and insulin resistance have also been associated with DSPN (Andersen et al., 2018b, Callaghan et al., 2020, Christensen et al., 2020, Han et al., 2015, O'Brien et al., 2017, Ylitalo et al., 2011, Zhou et al., 2020).
Furthermore, there is strong association between the number of metabolic syndrome components and DSPN (Bonadonna et al., 2006, Callaghan et al., 2016a, Callaghan et al., 2016b, Costa et al., 2004, Hanewinckel et al., 2016a). With respect to prevention, rigorous glucose control can decrease the incidence of DSPN in type 1 diabetes but has little effect in type 2 diabetes, which suggests different mechanisms underlie
DSPN in each condition (Callaghan Brian C. et al., 2012, Callaghan B. C. et al., 2012,
Gaede et al., 2008). In type 2 cohorts, insulin-sensitising therapies have been shown to reduce the incidence of neuropathy (Pop-Busui et al., 2013).
Clinical Presentation of DSPN
DSPN manifests with a ‘stocking and glove’ distribution whereby the hands and lower limbs are usually affected (Feldman et al., 2019). Symptoms occur in a length- dependent fashion such that they start distally at the toes, spread proximally, and then to the upper limb digits when the symptoms reach the knees (Feldman et al., 2019).
The most common early symptoms involve small fibres which include pain and dysesthesias (Malik et al., 2011, Pop-Busui et al., 2017). Pain is present in up to 25% of patients with DSPN and may be described as burning, lancinating, tingling, or shooting and is typically worse at night (Pop-Busui et al., 2017). Pain may also be accompanied with paraesthesia, allodynia, and hyperalgesia (Pop-Busui et al., 2017).
Reductions in intra-epidermal and corneal nerve fibres at the inferior whorl, which is the most distal extension of corneal nerve fibres, are among the earliest changes
(Edwards et al., 2017, Ferdousi et al., 2020b, Malik et al., 2011, Petropoulos et al.,
66 Literature Review
2015). Involvement of large fibres, which usually occurs later in the disease course, may cause numbness, tingling without pain, loss of protective sensation (which is particularly indicative of DSPN), abnormal ankle reflexes (which occurs early), reduced vibration sensation, impaired joint position, and diminished touch-pressure sensations (Albers and Pop-Busui, 2014, Pop-Busui et al., 2017). Clinical findings of
DSPN are a loss of sensation to pinprick and cold temperature, which are both mediated by small fibres, and loss of vibration and proprioception, which are both mediated by large fibres, in a ‘stocking and glove’ distribution (Feldman et al., 2019).
With advanced DSPN, weakness of the small muscles of the foot and dorsiflexors is observed and foot drop may also develop (Albers and Pop-Busui, 2014, Feldman et al., 2019).
Nerve conduction studies are the current gold standard for the diagnosis of DSPN
(Selvarajah et al., 2019). Findings in DSPN include reduced amplitudes, decreased conduction velocities, and prolonged F responses (Weisman et al., 2013). Typically, reductions in sensory fibre amplitudes are observed before decreases in the amplitude of motor fibres (Feldman et al., 2019). Abnormalities in nerve conduction (which may be subclinical) appears to be first objective and quantitative indication of DSPN and is necessary in the research setting to confirm the diagnosis (Dyck et al., 2011, Tesfaye et al., 2010). However, it should be noted that nerve conduction studies are not required to meet the definition of DSPN in routine clinical practice, and as such, the diagnosis of DSPN could be held to more rigorous standards. Polyneuropathy due to vitamin deficiencies (e.g. B12), other endocrine and metabolic disorders (e.g. hypothyroidism), or with an infectious, inflammatory, toxin, inherited, vascular, and neoplastic basis must be excluded (Zochodne, 2014). In addition to conventional
67 Literature Review
nerve conduction studies, other electrophysiological measures may have some utility
in the diagnosis of DSPN (Shabeeb et al., 2018).
Pathogenesis and Pathophysiology of DSPN
Research into the field of diabetic neuropathy has largely focused on the metabolic
and/or redox state of the dorsal root ganglion and Schwann cells (Feldman et al.,
2017). Various pathways have been implicated in the pathogenesis of DSPN (Figure
11). Ultimately, these pathways lead to an excess formation of mitochondrial and
cytosolic reactive oxygen species, inflammation, and Na+/K+ pump dysfunction
(Feldman et al., 2017). Indeed, axonal excitability studies have added to our
understanding of DSPN, however these techniques are limited to the study of large
fibres. Nevertheless, investigations using axonal excitability are discussed where
relevant.
Insulin Hypergylcaemia Metabolic Syndrome Resistance/Deficiency
Increased glucose Polyol PKC Hexosamine Dyslipidaemia variability pathway pathway pathway
Increased Increased Increased Mitochondrial Excessive free sorbitol diacylglycerol F-6-P dysfunction fatty acids Neuronal PKC NADPH depletion activation
Na+/K+ pump Oxidative stress Inflammation dysfunction
Impaired Axoglial Axonal Nerve regeneration dysjunction degeneration dysfunction
Neuropathy
Figure 11. Pathogenesis of distal symmetric polyneuropathy. PKC: protein kinase C; F-6-P:
fructose-6-phosphate; NADPH: nicotinamide adenine dinucleotide phosphate.
68 Literature Review
Polyol, Hexosamine, and Protein Kinase C Pathways
Excess glucose in hyperglycaemia is converted to sorbitol by aldose reductase, which results in osmotic imbalance in the cell and compensatory efflux of myoinositol and taurine (Greene et al., 1988). Loss of myoinositol impairs normal nerve physiology as it is an essential component for the Na+/K+ pump (Greene, 1986a, 1986b, Greene and
Lattimer, 1984). Consequently, this would result in an accumulation of intracellular sodium and hence, reduction in the trans-axonal Na+ gradient. Further injury to neurons occurs as aldose reductase activity depletes cellular stores of nicotinamide adenine dinucleotide phosphate (NADPH) which is required for the regeneration of glutathione, an essential antioxidant (Feldman et al., 2017). This leads to the generation of cytoplasmic reactive oxygen species which cause cellular dysfunction
(Oates, 2008). Furthermore, sorbitol is metabolised to fructose which can create further osmotic swelling (Oates, 2008).
Evidence of a reduction in the trans-axonal Na+ gradient was observed in early axonal excitability studies of DSPN. In these studies, hyperglycaemia was associated with excitability changes indicative of reduced nodal Na+ currents, namely reductions in both the SDTC and relative refractory period and an increase in the rheobase (Kitano et al., 2004, Misawa et al., 2006a, Misawa et al., 2005b, Misawa et al., 2004).
Treatment of hyperglycaemia was observed to reverse these observations, suggesting an increase of Na+ currents presumably through restoration of the trans-axonal Na+ gradient (Kitano et al., 2004, Kuwabara and Misawa, 2008). It is important to note changes in the relative refractory period (increase or decrease) have varied between studies, however it tends to increase with the severity of DSPN (see below). Clinical
69 Literature Review trial findings demonstrated that administration of epalrestat, an aldose reductase inhibitor, was found to rapidly increase nodal Na+ currents, as measured by SDTC, and improve nerve conduction in patients with DSPN (Misawa et al., 2006b). Clinical trials of aldose reductase inhibitors have largely failed to treat DSPN but improvements in nerve conduction parameters have been observed in some cases
(Feldman et al., 2017, Sekiguchi et al., 2019). Finally, while a reduction in Na+ currents seems to be implicated in DSPN, it should be noted that an abnormal increase in nodal Na+ currents is associated with neuropathic pain and worsening quality of life measures in patients with diabetes (Kwai et al., 2013, Misawa et al.,
2006a, Misawa et al., 2009).
The hexosamine and protein kinase C (PKC) pathways are also involved in peripheral nerve injury. Increased glycolysis in response to hyperglycaemia results in a glycolysis intermediate, fructose-6-phosphate, entering the hexosamine pathway, which undergoes a series of reactions that ultimately results in damage to peripheral nerves (Feldman et al., 2017). In addition to the polyol pathway, a disruption to
Na+/K+ pump function may occur via the PKC pathway. Increased glycolysis results in the accumulation of diacylglycerol which activates neuronal PKC, and in turn, disrupts function of the Na+/K+ pump and perhaps impairs nerve blood flow (Geraldes and King, 2010).
The possibility that Na+/K+ pump dysfunction may play a role in the development of
DSPN in diabetes has also been supported by axonal excitability studies. Excitability recordings undertaken in patients with DSPN by Krishan and colleagues (2005) demonstrated a ‘fanning-in’ appearance of the threshold electrotonus, consistent with
70 Literature Review axonal depolarisation potentially from impaired pump activity (Krishnan and Kiernan,
2005). Other findings from this study were reductions in the relative refractory period and an increase in superexcitability, consistent with a reduction in nodal Na+ conductance suggesting that a reduction in Na+/K+ pump activity is coupled with decreased nodal Na+ currents (Krishnan and Kiernan, 2005). Impaired Na+/K+ pump function in diabetes is further supported by excitability studies which implemented dynamic manoeuvres such as limb ischaemia and maximal voluntary contraction to demonstrate decreased pump activity in DSPN (Krishnan et al., 2008, Kuwabara et al., 2002). Other axonal excitability studies have demonstrated changes indicative of reductions in nodal Na+ conductances and Na+/K+ pump function in diabetes patients without DSPN or with subclinical DSPN, which worsen with increasing severity of
DSPN. (Bae et al., 2011, Sung et al., 2012). With neuropathy progression there is
‘fanning- in’ of threshold electrotonus, a decrease in subexcitability, and increases in the rheobase, relative refractory period, and superexcitability (Sung et al., 2012).
In animal models of DSPN, intracellular Na+ accumulation and decreased Na+/K+ pump function have shown to disrupt the nodal and paranodal regions through axoglial dysjunction and myelin retraction (Sima and Brismar, 1985, Sima et al.,
1986). It would therefore be expected that the function of nodal and paranodal channels would be affected in patients with diabetes. Indeed, early axonal excitability studies in patients with DSPN demonstrated that hyperglycaemia caused a reduction in nodal and paranodal K+ conductances (Misawa et al., 2005a). These observations were further substantiated with axonal excitability studies undertaken in type 1 diabetes patients without neuropathy demonstrating similar results, suggesting these reductions occur early in the disease course (Arnold et al., 2013a, Kwai et al., 2016a).
71 Literature Review
Mathematical modelling of these recordings indicated there were reductions in a range of Na+ and K+ conductances in the nodal and paranodal regions of the axon
(Kwai et al., 2016a).
Large scale alterations in peripheral nerve structure are also observed in DSPN with ultrasonography (Lee and Dauphinée, 2005). Increase in sorbitol is thought to cause oedema which increases peripheral nerve cross-sectional area (Vlassakov and Sala-
Blanch, 2015). Significant enlargement of cross-sectional area of the median and tibial nerves has been validated as a good parameter in the diagnosis of DSPN (Riazi et al., 2012, Watanabe et al., 2009, Watanabe et al., 2010). Association studies using nerve ultrasound and nerve excitability have demonstrated that while cross-sectional area increases with DSPN in type 1 and type 2 diabetes, cross-sectional area correlates with worsening nerve excitability measures in type 1 diabetes only (Borire et al., 2018b). This differential relationship lends further credence to the notion that different pathophysiological mechanisms may underlie DSPN in type 1 and type 2 diabetes. Further to this notion, the pathophysiological mechanisms of peripheral nerve dysfunction in LADA, which was described as an admixture of type 1 and type
2 diabetes, are yet to be to be characterised. This will be examined in Chapter 4.
Insulin, C-peptide, and Glycaemic Variability
Although insulin does not directly control glucose transport into the peripheral nervous system, it is recognised as a neurotrophic factor and insulin deficiency is associated with the development of diabetic neuropathy (Feldman et al., 2017,
Zaharia et al., 2019). Insulin receptors are expressed on peripheral nerves, intraneural mitochondria, and on Schwann cells which is of significance given the interplay
72 Literature Review between Schwann cells and axons that has recently emerged as a key mediator in the pathogenesis of DSPN (Feldman et al., 2019, Goncalves et al., 2017). In type 1 diabetes, insulin deficiency is accompanied with decreases in C-peptide. C-peptide has been observed to be neuroprotective in experimental studies, possibly through an increase in Na+/K+ pump function, but only marginal benefits were seen clinical supplementation in type 1 diabetes patients with DSPN (Wahren et al., 2016, Wahren and Larsson, 2015). In type 2 diabetes, C-peptide is inversely associated with DSPN, independent of confounding factors (Qiao et al., 2017).
In type 1 diabetes, the mode of insulin delivery has been observed to have an effect on ion channel function. Kwai and colleagues (2015) found that patients receiving CSII had axonal excitability parameters preserved within normal limits at follow-up in contrast to patients receiving MDII, who displayed prominent abnormalities (Kwai et al., 2015). The possibility that the apparent normalisation of nerve function with CSII treatment was due to a greater stability in acute glucose variability was considered.
Subsequent studies demonstrated that greater acute glucose variation, as measured by mean amplitude of glycaemic excursion (which is postulated to cause oxidative stress), was positively associated with worsening axonal excitability measures (Kwai et al., 2016b). The effect of other acute measures of glucose control, namely continuous overall net glycaemic action and time in range, on peripheral nerve structure and function are yet to be explored. These associations will be investigated in Chapter 3.
In type 2 diabetes, insulin-sensitisation treatments have been shown to decrease the incidence of neuropathy compared to insulin-providing treatments (Pop-Busui et al.,
73 Literature Review
2013). Insulin resistance has been observed to develop in experimental models of type
2 diabetes and could play a role in the pathogenesis of DSPN (Grote and Wright,
2016, Kim and Feldman, 2012). Evidence of this concept is lacking in human studies however. While treatment of DSPN primarily focuses on symptomatic management, animal studies have observed neuroprotective effects of commonly used medication used to treat type 2 diabetes (El Mouhayyar et al., 2020, Pop-Busui et al., 2017). The effect of these medications on axonal excitability will be investigated in Chapter 6.
The Metabolic Syndrome
In type 2 diabetes, the components of the metabolic syndrome promote the onset and progression of diabetic neuropathy independent of hyperglycaemia (Callaghan et al.,
2016a, Feldman et al., 2017). Animal models of metabolic syndrome demonstrate there is neuronal dysfunction and oxidative stress, even in normoglycaemia (Cortez et al., 2014). Dyslipidaemia increases the formation of acylcarnitine in peripheral nerves and an excess leads to mitochondrial dysfunction and apoptotic fission (Stino et al.,
2020). Dyslipidaemia can also lead to an increase in oxidized low-density lipoprotein
(LDL) and advanced glycation end-LDL, and both can activate various pro- inflammatory signalling pathways that disrupt mitochondrial function (Eid et al.,
2019, Stino et al., 2020). The excessive free fatty acids catabolised by b-oxidation in response to hyperlipidaemia are harmful to the peripheral nervous system, especially
Schwann cells (Feldman et al., 2019). As mentioned above, there is a strong association between the number of metabolic syndrome components and DSPN, however the underlying mechanisms are yet to be examined. These pathophysiological mechanisms will be studied in Chapter 5.
74 Literature Review
Uraemic Neuropathy
Neurological complications are a major cause of morbidity in CKD and affect both the CNS and PNS causing encephalopathy, stroke, cognitive impairment, restless leg syndrome, autonomic disturbances, and neuromuscular disorders such as peripheral neuropathy (known as uraemic neuropathy) and myopathy (Arnold et al., 2016a,
Baumgaertel et al., 2014, Chillon et al., 2016, Vellanki and Bansal, 2015). Uraemic neuropathy is a distal sensorimotor polyneuropathy caused by uremic toxins. Despite improvements in dialysis care, uraemic neuropathy remains the most common neurological disorder in end-stage kidney disease and occurs in 60-100% of dialysis patients as well as approximately 70% of pre-dialysis patients (Aggarwal et al., 2013,
Karunaratne et al., 2018, Krishnan et al., 2005b, Krishnan et al., 2009b, Laaksonen et al., 2002, Tilki et al., 2009).
Clinical Presentation
Uraemic neuropathy typically presents with sensory symptoms such as paraesthesia, burning, and numbness which progresses slowly in a length-dependent fashion
(Basilio Vagner, 2011, Krishnan and Kiernan, 2009, Said, 2013). Given the length- dependent nature of uraemic neuropathy, there is preferential involvement of distal nerves in the lower limbs than upper limbs (Krishnan and Kiernan, 2007, Said, 2013).
Clinical examination in early stages reveals symptoms and signs confined to the lower limbs, including distal sensory loss to vibration and reduced ankle deep tendon reflexes (Krishnan and Kiernan, 2009, Said, 2013). With more severe disease, upper limb involvement may occur in a ‘stocking and glove’ distribution (Krishnan et al.,
2009b, Said, 2013). In advanced cases, motor nerve involvement occurs leading to muscle atrophy and weakness, which is most prominent distally (Krishnan and
75 Literature Review
Kiernan, 2009). Pain, which is usually absent in the early stages, may become prominent with advanced neuropathy (Said, 2013). Patients presenting with both
CKD and diabetes develop length-dependent neuropathy of greater severity (Krishnan and Kiernan, 2009). Currently, no instrument has been formally validated to assess the severity of peripheral neuropathy in patients with CKD. This will be addressed in
Chapter 1 of this thesis, in which the Total Neuropathy Score will be validated for
CKD patients with and without diabetes. Furthermore, in patients with both CKD and diabetes, the relative contributions of each of these conditions to neuropathy development is yet to be determined. This question will be addressed in Chapter 2.
Nerve conduction studies are the gold standard for the diagnosis of uraemic neuropathy (Krishnan et al., 2009b). Nerve conduction studies in CKD patients with neuropathy demonstrate features of generalised neuropathy of the axonal type with reductions in sensory amplitudes and, to a lesser extent, reduced motor amplitudes with relative preservation of conduction velocities (Krishnan and Kiernan, 2009).
Reduction in the sural sensory amplitude is the most sensitive nerve conduction parameter in the diagnosis of uraemic neuropathy (Laaksonen et al., 2002).
Pathogenesis and Pathophysiology
Observational studies in the 1970s indicated there were lower rates of neuropathy in patients treated with peritoneal dialysis compared with those on haemodialysis (Babb et al., 1971). This led to the theory that accumulation of ‘‘middle molecules,’’ substances with a molecular weight of 300–12000 Da such as b2-microglobulin and parathyroid hormone, were the cause of uraemic neuropathy and the better clearance of these molecules by the peritoneal membrane compared to standard haemodialysis
76 Literature Review membranes explained the reduced neuropathy prevalence with peritoneal dialysis
(Babb et al., 1981, Dhondt et al., 2000). However, the middle molecule hypothesis has been disputed on a number of bases (Kjellstrand et al., 1972). The main criticism was the lack of evidence that any of middle molecules are neurotoxic, with the exception of parathyroid hormone (Kjellstrand, 1981, Vanholder et al., 1994).
However, human studies examining the relationship between parathyroid hormone and nerve conduction velocity yielded conflicting results (Avram et al., 1978, Di
Giulio et al., 1978, Schaefer et al., 1980).
Despite this setback, the hypothesis of a dialysable toxin remained prevalent but the mechanism remained unclear. Many substances have been investigated as a potential uraemic neurotoxin including urea, creatinine, guanidine, methylguanidine, guanidinosuccinic acid, uric acid, oxalic acid, phenols, aromatic hydroxyacids, indican, amines, myo-inositol, amino acids and neurotransmitters, though none of these have yielded compelling results or findings were restricted to in vitro studies
(Anand et al., 2019, Bostock et al., 2004).
The hypothesis that neurotoxic effect of a substance may be caused by an alteration in membrane excitability was proposed by Nielsen based on in vitro studies of muscle and red blood cells obtained from dialysis patients (Bittar, 1967, Nielsen, 1973, Welt et al., 1964). Nielsen hypothesised that one or more of the toxins known to accumulate in uraemia may cause neuropathy by inhibiting the activity of the axonal
Na+/K+ pump (Nielsen, 1973). As the Na+/K+ pump is electrogenic with 3 Na+ ions being extruded for every 2 K+ ions pumped into the axon, inhibition leads to an excess of positive charge on the inner aspect of the axonal membrane and an accumulation of
77 Literature Review extracellular K+, which causes further depolarisation (Kaji and Sumner, 1989).
Disruption of these ion gradients created by the Na+/K+ pump may cause reverse operation of the Na+/Ca2+ exchanger, leading to increased levels of intracellular Ca2+ and axonal loss (Craner et al., 2004).
Kiernan and colleagues (2002) were the first to demonstrate evidence of axonal membrane depolarisation, as assessed by axonal excitability, primarily due to hyperkalaemia in end-stage kidney disease (Kiernan et al., 2002). While clinical features of uraemic neuropathy are predominantly lower limb, these excitability studies were conducted on the median nerve and yet demonstrated significant alterations in excitability prior to dialysis (Kiernan et al., 2002, Krishnan et al.,
2006b). Nerve excitability studies were extended to the lower limb in dialysis patients and findings again indicated membrane depolarisation was due to serum K+ and not other substances such as parathyroid hormone, b2-microglobulin, or urea (Krishnan et al., 2005b). Krishnan and colleagues (2005) noted a correlation between the severity of symptoms and nerve excitability abnormalities, suggesting that the altered membrane potential may be directly related to neuropathic symptoms (Krishnan et al.,
2005b).
Despite the strong evidence of K+ as the cause of membrane depolarisation in end- stage kidney disease, the possibility this was due to impaired Na+/K+ pump function remained. Na+/K+ pump function was subsequently studied in dialysis patients using manoeuvres such as limb tourniquet and forced maximal voluntary contraction
(Krishnan et al., 2006a, Krishnan et al., 2006c). These findings argued against Na+/K+ pump impairment and instead suggested heightened pump activity as a compensatory
78 Literature Review mechanism for membrane potential changes in dialysis patients. Hyperkalaemia as the causative agent of membrane depolarisation in dialysis patients was subsequently confirmed using serum K+ clamp studies (Arnold et al., 2014). In addition to causing alterations in nerve excitability, elevations in serum K+, but not parathyroid hormone
or b2-microglobulin, were also found to correspond with nerve enlargement, which reduced after dialysis (Borire et al., 2017).
While these studies implicated hyperkalaemia in the pathophysiology of uraemic neuropathy in end-stage kidney disease, recent clinical trial evidence has highlighted the role of serum K+ in uraemic neuropathy in the pre-dialysis populations as well
(Arnold et al., 2017). Restriction of dietary potassium has been shown to prevent nerve excitability abnormalities indicative of membrane depolarisation at follow-up in stage 3–4 CKD (Arnold et al., 2017). Furthermore, potassium restriction prevented the progression of neuropathy and resulted in an improvement in gait speed (Arnold et al., 2017).
79 Methodology
Methodology
80 Methodology
Recruitment of Patients and Control Subjects
Studies were approved by the South East Sydney Area Health Service Human
Research Ethics Committee (Northern Section) and the Human Research Ethics
Committee of the University of New South Wales. All participants enrolled provided written informed consent to the procedures in accordance with the Declaration of
Helsinki.
Recruitment of Patients
All patients were recruited from the Diabetes Mellitus Centre and the Kidney Care
Centre at the Prince of Wales Hospital in Sydney, Australia. The entire patient cohort consisted of patients with a diagnosis of either type 1 diabetes, type 2 diabetes, latent autoimmune diabetes of adults, chronic kidney disease, and diabetic kidney disease.
All patients enrolled had been diagnosed with their respective conditions for a minimum of one year. Patients were excluded from participating in studies if they had any of the following: renal transplant, a history of neurotoxic/neuromodulatory treatment, carpal tunnel syndrome, peripheral oedema, an additional condition known to cause neuropathy (such as vitamin B12 deficiency), or a neuromuscular, movement, psychiatric, or developmental disorder. In studies utilising in-vivo corneal confocal microscopy, exclusion criteria included current eye infections, corneal abrasions, history of refractive surgery, trauma to anterior segment, or contact lens wear.
Recruitment of Controls
Control subjects were recruited through Prince of Wales Hospital and the University of New South Wales. All control subjects underwent clinical assessment to exclude neuropathy and underwent testing under the same conditions as patients. In addition
81 Methodology to the exclusion criteria above, control participants were excluded from studies if they had a history of taking medication to treat blood pressure, elevated glucose, or lipids.
Equipment and Materials
Axonal Excitability
• High Performance AC Amplifier (ICP511 AC amplifier, Grass Technologies, West Warwick, US) for amplifying nerve excitability recordings
• Data acquisition device (DAQ PCI-6221); Shielded connector block (BNC- 2110); Cable (SHC-68-68-EPM); (National Instruments, Austin, USA) used to convert recordings to a digital signal
• Isolated linear bipolar constant current stimulator (DS5, Digitimer, Welwyn Garden City, UK)
• Hum Bug 50/60 Hz Noise Eliminator (Quest Scientific Instruments, North Vancouver, Canada) used to cancel background electrical noise
• Automated threshold tracking and analysis software - QTRAC (Institute of Neurology, Queen Square, London) with TROND axonal excitability protocol
• Conventional non-polarizable ECG electrodes (Unilect 7831Q; Unomedical, Stonehouse, Great Britain) to provide surface stimulation/reference for surface recordings
• Electrosurgical neutral electrode (Unilect 2406M, Unomedical, Stonehouse, Great Britain) used as ground electrodes
• Red Dot Trace Prep (2236, 3M Canada) for abrading skin prior to electrode placement
82 Methodology
• Thermistor thermometer (5831-A, Omega Engineering, Manchester, UK) used to measure skin temperature
Nerve Ultrasonography
• MyLabOne system with a 10–18 MHz linear probe (Esaote, Genoa, Italy) used for ultrasound scans • Aquasonic Clear ultrasound gel (Parker Laboratories, New Jersey, USA) to provide a conductive medium
In-vivo Corneal Confocal Microscopy
• Heidelberg Retinal Tomograph III Rostock Cornea Module (Heidelberg Engineering GmbH, Heidelberg, Germany) for corneal confocal microscopy
• ACCMetrics Corneal Nerve Fibre Analyser (University of Manchester, Manchester, UK) for confocal image analysis
Nerve Conduction Studies
• Medelec Synergy EMG system (Oxford Instruments, Old Woking, UK) used for nerve conduction studies
• Surface EMG recording electrodes (Kendall Soft-E, H69P, Tyco Healthcare, Gosport, UK).
Neuropathy Assessment
• Neurotip disposable needles (Owen Mumford Ltd., Oxford, UK) for assessing pinprick sensation.
• 128 Hz tuning fork used to measure vibration sense
• Clinical reflex hammer to test tendon reflexes
83 Methodology
Glucose Variability
• iPro CGM System (Medtronic, CA, USA) for blinded continuous glucose recording
• EnliteTM sensor (Medtronic, CA, USA) for measurement of interstitial glucose
• ACCU-CHEK Performa Glucometer (Roche Diagnostics, Mannheim, Germany) for measuring on the spot blood glucose levels obtained by finger- stick.
• Gylcemic Variability Analyzer Program (Óbuda University, Budapest, Hungary)
Axonal Excitability Assessment and Mathematical Modelling
Axonal excitability studies provide important insights into the biophysical properties of human axons in health and disease. Excitability studies enable the rapid acquisition of multiple excitability variables by measuring the change in stimulus required to elicit a target response (~40% of maximal compound action potential) following various patterns of conditioning stimuli. The key elements involved in a standard axonal excitability assessment protocol include the initial stimulus–response curve followed by four distinct testing paradigms: strength-duration properties, threshold electrotonus, current-threshold relationship, and the recovery cycle. With the exception of strength-duration properties, the other testing paradigms utilise conditioning stimuli and are designed to investigate how axons behave when membrane potential is changed. The information gained from these paradigms provides indirect measures of axonal membrane potential and the activity of ion
84 Methodology channels, energy dependent pumps and ion exchange processes involved in impulse conduction (see Axonal Excitability in Literature Review, page 41).
All nerve excitability studies in this thesis were undertaken on the median motor nerve. To reduce skin impedance 3M Red Dot Trace Prep (2236, 3M Canada) was used to prepare both the stimulating and recording sites. Non-polarisable electrodes
(Unilect 7831Q; Unomedical, Stonehouse, Great Britain) were utilised to stimulate the median nerve. The cathode was placed over the optimal site of median nerve stimulation (site of least resistance) and the anode placed 10 cm proximal over the radial bone. The recording electrode was positioned over the motor point of the APB muscle, the reference electrode was placed ~4 cm distally on the proximal phalanx and the earth plate (Unilect 2406M, Unomedical, Stonehouse, Great Britain) was placed on the palm. Temperature was monitored close to the site of stimulation and maintained at >32 °C throughout the study (Figure 12).
85 Methodology
Computer QTRAC
Stimulator
Recording electrodes
Earth Stimulating electrodes
Figure 12. Median motor nerve excitability set-up. Image demonstrates electrode placement and equipment for the computerised QTRAC system with stimulation delivered through an isolated bipolar constant current stimulator.
Stimulation was controlled by a computerised threshold tracking system, QTRAC
(Institute of Neurology, Queen Square, UK), converted to current, and delivered by an isolated linear bipolar constant current stimulator (DS5, Digitimer, Welwyn Garden
City, UK). Compound motor action potentials were recorded by the QTRAC software following amplification (ICP511 AC amplifier, Grass Technologies, West Warwick,
US) and digitisation (DAQ PCI-6221; Shielded connector block BNC-2110; Cable
SHC-68-68-EPM; National Instruments, Austin, USA). A Hum Bug 50/60 Hz Noise
Eliminator (Quest Scientific Instruments, North Vancouver, Canada) was used to remove excess electronic noise. QTRAC threshold tracking software utilised a proportional tracking system that altered the stimulus intensity according to the
86 Methodology difference between actual and target responses, i.e. stimulus intensity was automatically increased or decreased to achieve an action potential of a defined size.
Responses were tracked as a percentage change in threshold for normalisation.
Stimulus-Response Curve
To begin, stimulus–response curves were generated by an incremental (1 mA) increase in the stimulus intensity of a 1 ms impulse until a maximal response was obtained. Reversing this process, the QTRAC software automatically decreases the stimulus until the response returns to zero. At each level of current decrease, the average of three responses are taken. A target response of the steepest point on the stimulus-response curve (~40% of the maximal compound action potential, see Figure
4, page 43) was automatically chosen. The stimulus required to elicit this target response (or change there-of), termed the threshold, was the quantitative value measured for the remainder of the protocol. Test pulses for threshold were 1 ms for all paradigms.
Strength-Duration Properties
Strength-duration properties were assessed by measuring the change in stimulus required to elicit threshold utilising four different stimulus widths: 0.2 ms, 0.4 ms, 0.8 ms and 1 ms. The value reflecting the strength-duration time constant was calculated from the x-axis intercept of the linear relationship between stimulus intensity and stimulus duration derived using Weiss’ formula (see Figure 5, page 45) (Bostock,
1983, Mogyoros et al., 1996).
87 Methodology
Threshold Electrotonus
Threshold electrotonus (TE) was determined by plotting the percentage of threshold change when test pulses were applied during and after 100 ms of subthreshold depolarising (TEd; +40% of threshold) and hyperpolarising (TEh; –40% of threshold) conditioning currents. A total of 26 time points were assessed and more than 10 measures are acquired from this paradigm. By convention TE is plotted as threshold reduction on the y-axis and time on the x-axis (see Figure 6, page 47).
In response to TEd and TEh, measures are determined from the average percentage threshold change between 10–20 ms, 20–40 ms, and 90–100 ms. After cessation of conditioning currents, TEd(undershoot) and TEh(overshoot) are calculated as the mean value of maximal threshold reduction in the 20 ms following cessation of the conditioning stimulus. Additional parameters calculated from responses to TEd include peak threshold reduction and S2 accommodation, which is defined as the difference between peak threshold reduction and TEd(90–100 ms).
Current-Threshold Relationship
The current-threshold (I/V) relationship also utilised subthreshold conditioning currents. However, unlike TE, these currents are longer in duration (200 ms) and their intensity is altered in 10% increments from –100% (hyperpolarising) to +50%
(depolarising) of threshold. The change in excitability induced by these condition currents was assessed at a single time-point following the 200 ms polarising current.
Given that the slope of the I/V relationship is a threshold analogue of input conductance, all I/V variables are expressed as the slope of this relationship between various polarising stimulus intensities (see Figure 7, page 50). There are three
88 Methodology variables obtained from the I/V protocol: resting I/V slope (calculated from the polarising currents –10% to +10%), minimum I/V slope (calculated as the minimum of the curve obtained by fitting a straight line to each three adjacent points in turn), and hyperpolarising I/V slope (calculated from the three most hyperpolarised intervals).
Recovery Cycle
The recovery cycle assessed the restoration of axonal excitability following a supramaximal stimulus. The change in threshold was measured at 18 conditioning-test intervals from 2 to 200 ms after a supramaximal stimulus was delivered. To avoid contamination during the short conditioning-test intervals, the response from a single supramaximal conditioning stimulus was subtracted from the conditioning + test response.
The recovery cycle of excitability has a well-defined and reproducible series of events from which three variables are measured (see Figure 8, page 51). First, the relative refractory period, measured in milliseconds, is the time taken to return to baseline threshold (or to cross the x–axis). Following this, a period of enhanced excitability, termed superexcitability, was calculated as the maximal percentage threshold reduction, averaged from 3 adjacent points. The final phase, subexcitability, was calculated as the maximal percentage threshold increase after 10 ms, averaged from 3 adjacent points.
89 Methodology
Mathematical Modelling
Mathematical Modelling of axonal excitability recordings were undertaken in select studies to investigate the pathological basis of axonal dysfunction. Axonal excitability recordings were analysed using the using the Bostock model of axonal excitability, which is a validated model of the human axon based on a single node and internode connected by paranodal pathways through and under the myelin sheath (Bostock et al., 1991, Jankelowitz et al., 2007a, Kiernan et al., 2005b). The model assists in the interpretation of excitability findings between control and disease data by providing an indication of the underlying changes in and around the axonal membrane in the disease state. This includes changes in the maximal conductance and permeabilities of different types of Na+ and K+ ion channels, alterations in pump currents, biophysical properties, and surrounding ionic concentrations (Figure 13). An approximation of the actual distribution of voltage-gated K+ channels in myelinated axons is achieved by the inclusion of slow and fast K+ channels in both compartments, with slow K+ channels predominantly located at the node, and fast K+ channels predominantly at the internode. Transient and persistent voltage-gated Na+ channels were modelled only at the nodal membrane and the hyperpolarisation-activated cation current was restricted to the internodal axolemma. Leak conductances, Na+/K+–ATPase pump currents and axolemmal capacitances were modelled in both axonal compartments.
The Barrett-Barrett conductance, which represents current flow through and underneath the myelin sheath between the node and internode, was also investigated.
90 Methodology
RestingResting membraneMembrane Pump potentialpotential current
Permeability/ Conductance
Capacitance
Ion concentrations
Figure 13. Mathematical model parameters. Image depicts that changes in maximal conductance and permeabilities of different types of Na+ and K+ ion channels, alterations in pump currents, various biophysical properties, and surrounding ionic concentrations may be investigated.
The model was first adjusted to fit the mean nerve excitability data obtained from the
control group before fitting the mean data of the disease group. Modelling analyses
involved changes in a single or a combination of parameters in a hypothesis-driven
and iterative fashion to objectively fit simulated excitability data with the mean
recorded data as closely as possible using a least squares approach. The overall
discrepancy was assessed and minimised using the weighted sum of the squares of the
error terms between the control and disease group data of the four excitability
91 Methodology paradigms: strength-duration behaviour, threshold electrotonus, current-threshold relationship, and recovery cycle. Weighting factors of these paradigms were 0.5, 1, 1, and 2, respectively and were kept constant for each analysis. Minimum interstimulus interval for the recovery cycle was set at 3 ms. Analyses were run in unclamped mode to permit secondary changes in resting membrane potential caused by changes in conductances or pump currents.
Nerve Ultrasonography
Median nerve ultrasound was performed prior to electrophysiological testing. Imaging was performed with a MyLabOne system with an adjustable 10–18 MHz linear array transducer (Esaote, Italy) using the ‘Musculoskeletal’ factory preset (acoustic power
100%, line density set at medium, dynamic range set at 14, persistence set at 1) and other settings (focus, gain, and depth) were kept constant for each examination. Nerve ultrasound was conducted while participants were sitting comfortably with their forearm fully supinated and fingers semi-extended. The forearm was supported on an armrest to ensure the elbow was flexed at 90°. The median nerve was first identified in the transverse plane at the carpal tunnel inlet at the level of the pisiform bone and then tracked proximally between the superficial (flexor digitorum superficialis) and deep (flexor pollicis longus and flexor digitorum profundus) muscles until the junction of the middle and distal third of the forearm (Figure 14). The median nerve cross-sectional area was then measured on the screen (in mm2) with a stylus using the continuous trace method by outlining the inner margin of the epineurium, which is a validated measure in nerve ultrasound (Cartwright et al., 2008, Won et al., 2013).
92 Methodology
A B
Figure 14. Median nerve ultrasonography. (A) Image depicts median nerve was tracked proximally from the carpal tunnel to the junction of the middle and distal third of the forearm. (B) Image demonstrates measurement of median nerve cross-sectional area.
In-vivo Corneal Confocal Microscopy
Participants were scanned bilaterally with a corneal confocal microscope to visualise corneal nerve morphology (Heidelberg Engineering GmbH, Heidelberg, Germany)
(Figure 15). Prior to image acquisition, topical anaesthesia was applied to the cornea of both eyes and the instrument was positioned onto the centre of the cornea. During the examination, the participant was instructed to fixate on a roving black target on a computer screen placed 35 cm perpendicular to their contralateral eye in order to image nerves in the central cornea and nerves at the inferior whorl. The microscope was set on ‘sequence mode’ and 100 contiguous images of the sub-basal nerve plexus were captured over 40 seconds. For capturing images of the inferior whorl, the roving target was shifted higher to an estimated superonasal area of participant’s contralateral eye. The inferior whorl was subsequently located inferior and slightly nasal to the corneal apex. 100 contiguous images were captured over 40 seconds.
Eight central and three to four inferior whorl images, not overlapping by more than
20%, from both eyes of each participant were selected for quantification. For the central cornea, images with bright and vertically arranged fibres were selected, while
93 Methodology for the inferior whorl, images that captured the entire region were chosen
(Petropoulos et al., 2015, Vagenas et al., 2012). Images were analysed using a validated and fully automated nerve analysis software (Corneal Nerve Fiber Analyzer
V.2, ACCMetrics, University of Manchester, Manchester, United Kingdom) to quantify corneal nerve fibre length (total length of main nerves and nerve branches per square millimeter), density (number of main nerves per square millimeter), branch density (total number of main nerve branches per square millimeter), and fractal dimension (a measure of structural complexity with a lower value indicating less corneal nerve intricacy) in the central cornea (Dehghani et al., 2014, Petropoulos et al., 2014). Corneal nerve fibre length and fractal dimension were also quantified in the inferior whorl. Corneal nerve measures are presented as an average of both eyes.
Microneuromas were defined as nerve abnormalities that present as irregularly shaped, terminal enlargements of nerve endings with variable hyper-reflectivity and poorly defined margins, which were singular or in clusters of two or three.
Microneuromas were counted manually and the same microneuroma imaged in more than one frame was considered as a count of one.
Figure 15. Corneal confocal microscopy
94 Methodology
Nerve Conduction Studies
Nerve conduction studies were performed as part of the clinical assessments of neuropathy in all patients. The tibial and sural nerves were assessed as per standard neurophysiology protocols to determine the compound muscle action potential and sensory nerve action potential upon maximal stimulation, respectively (Liveson and
Ma, 1992). Conduction velocities were also assessed. All studies were performed using a Medelec Synergy EMG system (Oxford Instruments, Old Woking, UK) and surface EMG recording electrodes (Kendall Soft-E, H69P, Tyco Healthcare, Gosport,
UK). The tibial nerve was assessed by orthodromic stimulation between the medial malleolus and calcaneal tendon and recording from abductor hallucis. The sural nerve was assessed by antidromic stimulation approximately at the junction of the middle and lower thirds of the leg (10–16 cm proximal to the lateral malleolus), just lateral to the midline, and recording between the lateral malleolus and calcaneal tendon
(Liveson and Ma, 1992).
Clinical Assessment of Neuropathy
Total Neuropathy Score
Neuropathy presence and severity was determined using the Total Neuropathy Score
(TNS) (Figure 16). The TNS is comprised of eight items which are each scored from
0–4. The scores from the eight items of the TNS were summed to give a total score from 0–32, with a higher score indicating more severe neuropathy and zero indicating an absence of neuropathy. The TNS was administered in a scripted fashion.
95 Methodology
Total Neuropathy Score (TNS)
Figure 16. Total Neuropathy Score. Scoring system for the eight items of the Total Neuropathy Score are shown.
Participants were first asked if they had experienced any abnormal sensations in their
limbs such as feelings of numbness, pins and needles or prickling. If the subject
responded positively, a score ranging from 1–4 was assigned based on how far these
sensations extended proximally: in the fingers or toes (1), up to the ankles or wrists
(2), up to knees or elbows (3) or above (4). Patients were then asked if they felt weak
in their arms or legs. If the subject acknowledged a deficit, this weakness was graded
from 1–4 based on the amount difficulty experienced by the participant: slight (1),
moderate (2), requiring assistance (3) or paralysis (4).
96 Methodology
Pinprick sensibility was assessed using a Neurotip (Owen Mumford, U.K.) The sharp and blunt ends of the Neurotip were first pressed against the participants’ upper limb to allow them to distinguish between the two ends. After the subject had closed their eyes, the sharp and the blunt ends of the Neurotip were then pressed on the most distal aspect of each toe and finger, gradually more proximally over the surface of the foot and hand, and eventually to the knee and elbow until the participant reported the object to be sharp. A reduction in pinprick sensibility in the fingers or toes was scored as (1), up to the wrist or ankle was scored as (2), up to the elbow or knee was scored as (3) or even more proximally was scored as (4). Vibration sensibility was assessed using a 128-Hz tuning fork. The vibratory stimulus was first introduced on the upper limb and subjects were asked if they could feel the vibration or “buzz”. After the subject had closed their eyes, the tuning fork was placed over the most distal aspect of each toe and finger, gradually more proximally over the medial/lateral malleoli and medial/lateral wrist, and eventually at the medial/lateral knee and elbow, until the subject reported they felt the vibration. Vibration sensibility was graded in a similar manner to pinprick sensibility.
Strength was graded using manual muscle tests of ankle dorsiflexion as outlined in
Medical Research Council (MRC) Guidelines (Medical Research Council of the
United Kingdom, 1976). Mild weakness (MRC grade 4) was assigned a score of (1), moderate weakness (MRC grade 3) was assigned a score of (2), severe weakness
(MRC grade 2) was assigned a score of (3) and paralysis (MRC grade 1–0) was assigned a score of (4). Assessments of deep tendon reflexes were first examined at the ankle and then at the knee. If reinforcement was required to induce the ankle reflex this was scored as (1) and if the ankle reflex was completely absent, this was
97 Methodology scored as (2). If reinforcement was required to induce the patellar reflex this was scored as (3) or (4) if this was completely absent too.
Nerve conduction studies, as described above, were then undertaken to assess the compound muscle action potential and sensory nerve action potential upon maximal stimulation of the tibial and sural nerve, respectively. A score graded from 1–4 was attributed depending on how far the amplitude fell below the lower limit of the normal range, determined from an internal normative data set.
Modified Toronto Clinical Neuropathy Score
In select studies where sensory symptoms required heavy weighting, neuropathy severity and presence were assessed using the modified Toronto Clinical Neuropathy
Score. The modified Toronto Clinical Neuropathy Score is validated for the assessment of diabetic distal symmetric polyneuropathy and is comprised of 11 items
(Bril et al., 2009). Symptoms (foot pain, numbness, tingling, weakness, ataxia, and upper limb involvement) were graded from 0–3 depending on their presence and interference with the participant’s activities of daily living. When present, symptoms without interference with sense of well‐being or activities of daily living are graded as
(1), those which interfere with sense of well‐being, but not with activities of daily living as (2), or those which interfere with both as (3). Sensory tests (pinprick, light touch, temperature, vibration, and position sense) were graded from 0–3 depending on the degree of deficit distally. If abnormal, each test was scored as (1) if the deficit was at toes only, (2) if the deficit was between toes and ankle, or (3) if the deficit was above the ankle. Pinprick was assessed using a Neurotip (Owen Mumford, U.K.) and temperature was assessed using a cold tuning fork. Nerve conduction studies, as
98 Methodology outlined above, were also undertaken in the tibial and sural nerves. Presence of neuropathy was defined using the Toronto Diabetic Neuropathy Expert Group definition of confirmed diabetic neuropathy: the presence of nerve conduction abnormality and a sign or symptom of neuropathy (Tesfaye et al., 2010).
Glucose Variability
In select studies where glucose variability was of interest, participants with type 1 diabetes underwent six days of blinded continuous glucose monitoring (iPro,
Medtronic, CA, USA; sensor: EnliteTM, Medtronic, CA, USA). Glucose recordings were analysed using Glycemic Variability Analyzer Program, which has been validated for the assessment of glucose variability (Figure 17) (Marics et al., 2015).
Variables calculated were time in and above range, expressed as a percentage of total monitoring time, and continuous overall net glycaemic action (CONGA), which can assess intra-day and inter-day glucose variability. Time in range was considered when glucose was between 3.9–10.0 mmol/L (70–180 mg/dL) and time above range was defined when glucose values exceeded 10.0 mmol/L (180 mg/dL) (Battelino et al.,
2019). CONGA is defined as the standard deviation of the summed differences between the current glucose observation and an observation n hours prior, with a higher CONGA indicating greater glycaemic variation (McDonnell et al., 2005). The interval, n, was 1 hour. CONGA was calculated for each day of monitoring and averaged. Mathematically, CONGA is calculated using the formula:
∗ ∑ (� − � ) ����� = �∗ − 1
99 Methodology
Where
Dt = glucose readingt – glucose readingt-m
∗ ∑ � � = �∗ k* = number of observations where there is an observation n x 60 min ago. m = n x 60
For example, for an n of 1 and blood glucose (BG) monitoring beginning at 0800, the calculation would be: BG at 0900 minus BG at 0800; BG at 0905 minus BG at 0805;
BG at 0910 minus BG at 0810 and so on. CONGA is the standard deviation of these differences.
Normal range
Figure 17. Analysis of continuous glucose monitoring recordings. Dashed red lines indicate target glucose range, which allows the percentage time spent in (or outside) target range to be calculated. The software also computes CONGA for each day of monitoring, which is then averaged.
100 Methodology
Metabolic Syndrome Components
Metabolic syndrome was defined using the updated National Cholesterol Education
Program Adult Treatment Panel III criteria (Grundy et al., 2005). Study participants were considered to have metabolic syndrome if they fulfilled any 3 of the 5 criteria: increased waist circumference, elevated triglycerides, reduced HDL, elevated blood pressure, and elevated fasting glucose (Grundy et al., 2005). All type 2 diabetes participants had an elevated fasting glucose but were considered to have each of the following other components if:
1. Waist circumference was ≥102 cm (≥40 inches) for males or ≥88 cm (≥35
inches) for females,
2. Triglycerides ≥ 1.7 mmol/L (≥150 mg/dL),
3. HDL <1.03 mmol/L (<40 mg/dL) in males or <1.3 mmol/L (<50 mg/dL) in
females,
4. Systolic blood pressure ≥ 130 mmHg or diastolic blood pressure ≥ 85 mmHg.
Waist circumference was estimated retrospectively from BMI using a predictive linear regression model that accounts for age, sex, and ethnicity and is validated for overweight and obese individuals (Bozeman et al., 2012).
For males:
�� = �! + �"��� + �#��� + �$����� + �%���� + �
101 Methodology
For females, different models were applied depending if the patient was 35 years or older (one constant for age < 35 years and a separate intercept and slope for age ≥ 35 years):
�� = �! + �"��� + �#�{��� ≥ 35} + �$��� × �{��� ≥ 35} + �%����� + �&���� + �
Statistical Analyses
Data were analysed using SPSS Statistics Version 25.0 for Windows (IBM Corp,
New York, USA). Participants were de-identified and analysis of the relevant independent variables for each study were undertaken in a blinded manner using a coded system. Normality of data was first assessed using Shapiro–Wilk tests.
Normally distributed data is expressed as mean ± standard deviation while non- normally distributed data is written as median and quartile 1 to quartile 3. By convention, nerve excitability data is expressed as mean ± standard error. Where appropriate, and with post-hoc corrections if applicable, a one-way analysis of variance, Kruskal-Wallis tests independent t-tests, Mann–Whitney U tests, Pearson chi-square analyses were applied to compare the means of demographic data, clinical measures, nerve excitability variables, median nerve cross-sectional area, corneal confocal microscopy measures between groups. Depending on normality, Pearson or
Spearman correlations were applied to determine relationships between clinical variables and outcome measures. Linear or logistic regression analyses were undertaken to investigate the effects of demographic or clinical variables on nerve excitability outcomes. Paired sample t-or Wilcoxon signed-rank tests were used to compare means in prospective studies. For all tests, statistical significance was considered when p < 0.05.
102 Chapter 1
Chapter 1 – Validation of the Total Neuropathy Score as a means to assess peripheral neuropathy in chronic kidney disease
103 Chapter 1
Summary and Link to Thesis
In the Literature Review, I discussed uraemic neuropathy as a complication of chronic kidney disease and described the clinical presentation. Despite being a prevalent complication, there is no formally validated outcome measure to evaluate the severity of peripheral neuropathy in chronic kidney disease. In the following chapter, I formally validate the Total Neuropathy Score as a means to assess the severity of peripheral neuropathy in stages 3–5 chronic kidney disease with and without diabetes.
By demonstrating the Total Neuropathy Score to have construct validity and good internal reliability in patients with chronic kidney disease, its use for subsequent chapters in this thesis can be justified.
This work has been published:
Issar, T., Arnold, R., Kwai, N. C. G., Pussell, B. A., Endre, Z. H., Poynten, A. M., . .
. Krishnan, A. V. (2018). The utility of the Total Neuropathy Score as an
instrument to assess neuropathy severity in chronic kidney disease: A
validation study. Clinical Neurophysiology, 129(5), 889-894. doi:
10.1016/j.clinph.2018.02.120
TI was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition.
104 Chapter 1
Abstract
Objective: To demonstrate construct validity of the Total Neuropathy Score (TNS) in assessing peripheral neuropathy in subjects with chronic kidney disease (CKD).
Methods: 113 subjects with CKD and 40 matched controls were assessed for peripheral neuropathy using the TNS. An exploratory factor analysis was conducted and internal consistency of the scale was evaluated using Cronbach’s alpha. Construct validity of the TNS was tested by comparing scores between case and control groups.
Results: Factor analysis revealed valid item correlations and internal consistency of the TNS was good with a Cronbach’s alpha of 0.897. As a whole group, subjects with
CKD scored significantly higher on the TNS (CKD: median, 6, interquartile range, 1–
13; controls: median, 0, interquartile range, 0–1; p<0.001). Subgroup analysis revealed construct validity was maintained for subjects with stages 3–5 CKD with and without diabetes.
Conclusions: The TNS is a valid measure of peripheral neuropathy in patients with
CKD. The TNS is the first neuropathy scale to be formally validated in patients with
CKD.
105 Chapter 1
1. Introduction
Chronic kidney disease (CKD) is a condition characterised by progressive loss of renal function over time. The leading causes of CKD are diabetes, hypertension and glomerulonephritis (Couser et al., 2011). Peripheral neuropathy in CKD, known as uraemic neuropathy, is the most common neurological complication of CKD and affects ~70% of pre-dialysis and ~90% of dialysis patients (Aggarwal et al., 2013,
Krishnan and Kiernan, 2009). Peripheral neuropathy initially presents with distal sensory loss to pinprick and vibration as well as reduced or absent tendon reflexes in the lower limbs (Krishnan and Kiernan, 2009). In severe cases, motor nerves may be affected resulting in weakness and atrophy in the lower limb (Krishnan and Kiernan,
2009).
There is no formally validated outcome measure to assess the severity of neuropathy in CKD (Hanewinckel et al., 2016b). Improvement of neuropathy is the objective of various interventions such as dietary modulation and enhanced dialysis strategies
(Australian New Zealand Clinical Trials Registry: ACTRN12610000538044,
ACTRN12609000615280) as well as renal transplantation (Arnold et al., 2017). Thus, a valid outcome measure is essential. The Total Neuropathy Score (TNS) was developed as a composite measure of peripheral nerve impairment (Table 1.1) and has high inter-rater and intra-rater reliability (Cornblath et al., 1999). The eight items that constitute the TNS provide a comprehensive examination of different nerve fibres and provide a single score, with a higher score indicating more severe neuropathy. The
TNS and its modified forms have been validated as a clinical tool to assess the severity of systemic neuropathies such as diabetic neuropathy and chemotherapy- induced peripheral neuropathy (Binda et al., 2015, Cavaletti et al., 2013, Cavaletti et
106 Chapter 1 al., 2007, Cornblath et al., 1999, Gilchrist and Tanner, 2013). A number of studies investigating neuropathy in CKD have also utilised a modified TNS (Arnold et al.,
2014, Arnold et al., 2016b, Arnold et al., 2013b, Borire et al., 2017, Borire et al.,
2018a). However, there has been no formal validation of the TNS in patients with
CKD, which could potentially represent a barrier for its future use in clinical trials.
Table 1.1. Components of the Total Neuropathy Score Item Assessment Nerve fibres examined
Sensory symptoms Scripted interview Sensory – multiple fibre types
Motor symptoms Scripted interview Motor – a, b motor fibres
Pinprick sensation Sharp sensation using Neurotip Sensory – type III (A∂), type IV (C) fibres
Vibration sensation 128-Hz tuning fork Sensory – type II (Ab) fibres
Strength Manual ankle dorsiflexion Motor – a motor fibres
Deep tendon reflexes Ankle and knee stretch reflex Sensory – type Ia (Aa) and II (Ab) fibres Motor – a and g fibres
Tibial CMAP Nerve conduction study Motor
Sural SNAP Nerve conduction study Sensory
CMAP: compound muscle action potential; SNAP: sensory neuron action potential
The objective of this study was to investigate the validity of the TNS as an instrument to assess the presence and severity of neuropathy in patients with CKD. Construct validity was assessed by testing the hypothesis that the TNS could distinguish groups of patients clinically diagnosed with CKD from healthy controls matched for age, sex and BMI for the presence of neuropathy.
107 Chapter 1
2. Methods
This study was approved by the South East Sydney Area Health Service Human
Research Ethics Committee (Northern Section) and the Human Research Ethics
Committee of the University of New South Wales. All patients provided written informed consent to the procedures in accordance with the Declaration of Helsinki.
Participants were recruited from the Kidney Care Centre at the Prince of Wales
Hospital, Sydney, Australia.
English-speaking subjects with clinically confirmed stage 3–5 CKD aged 18–85 able to give informed consent were invited to participate in the study. Subjects were excluded from the study if they had any of the following: peripheral oedema or joint injury, a history of neurotoxic treatment, had undergone a kidney transplant, other conditions known to cause neuropathy (with the exception of diabetes) and neuromuscular, movement, psychiatric or developmental disorders. A total of 113 patients (CKDT) were included in the final analysis. For comparison, 40 healthy age-, sex- and BMI-matched controls were recruited. Subjects were excluded if they had: peripheral oedema or joint injury, a history of neurotoxic treatment, conditions known to cause neuropathy and neuromuscular, movement, psychiatric or developmental disorders. Neurophysiological data was collected on the same day that participants had blood drawn for assessment of serum biochemistry. Stage of CKD and assignment of subjects for subgroup analysis was established using estimated glomerular filtration rate (eGFR, mL/min/1.73m2) (National Kidney Foundation,
2002). Subjects were grouped into stage 3/4 if their eGFR was between 15–59
(CKD3/4). Subjects with an eGFR below 15 were classified as stage 5 (CKD5).
108 Chapter 1
Participants with a clinically confirmed absence of diabetes were designated as
CKDNoDM.
Neuropathy presence and severity was determined using the TNS by trained administrators. A single examiner administered the TNS in a scripted fashion for each participant. Participants were first asked if they had experienced any abnormal sensations in their limbs such as feelings of numbness, pins and needles or prickling.
If the subject responded positively, a score ranging from 1–4 was assigned based on how far these sensations extended proximally: in the fingers or toes (1), up to the ankles or wrists (2), up to knees or elbows (3) or above (4). Patients were then asked if they felt weak in their arms or legs. If the subject acknowledged a deficit, this weakness was graded from 1–4 based on the amount difficulty experienced by the participant: slight (1), moderate (2), requiring assistance (3) or paralysis (4).
Pinprick sensibility was assessed using a Neurotip (Owen Mumford, U.K.) The sharp and blunt ends of the Neurotip were first pressed against the participants’ upper limb to allow them to distinguish between the two ends. After the subject had closed their eyes, the sharp and the blunt ends of the Neurotip were then pressed on the most distal aspect of each toe and finger, gradually more proximally over the surface of the foot and hand, and eventually to the knee and elbow until the participant reported the object to be sharp. A reduction in pinprick sensibility in the fingers or toes was scored as (1), up to the wrist or ankle was scored as (2), up to the elbow or knee was scored as (3) or even more proximally was scored as (4). Vibration sensibility was assessed using a 128-Hz tuning fork. The vibratory stimulus was first introduced on the upper limb and subjects were asked if they could feel the vibration or “buzz”. After the
109 Chapter 1 subject had closed their eyes, the tuning fork was placed over the most distal aspect of each toe and finger, gradually more proximally over the medial/lateral malleoli and medial/lateral wrist, and eventually at the medial/lateral knee and elbow, until the subject reported they felt the vibration. Vibration sensibility was graded in a similar manner to pinprick sensibility.
Strength was graded using manual muscle tests of ankle dorsiflexion as outlined in
Medical Research Council (MRC) Guidelines (Medical Research Council of the
United Kingdom, 1976). Mild weakness (MRC grade 4) was assigned a score of (1), moderate weakness (MRC grade 3) was assigned a score of (2), severe weakness
(MRC grade 2) was assigned a score of (3) and paralysis (MRC grade 1–0) was assigned a score of (4). Assessments of deep tendon reflexes were first examined at the ankle and then at the knee. If reinforcement was required to induce the ankle reflex this was scored as (1) and if the ankle reflex was completely absent, this was scored as (2). If reinforcement was required to induce the patellar reflex this was scored as (3) or (4) if this was completely absent too.
A nerve conduction study (Medelec Synergy, Oxford Instruments, UK) was then undertaken to assess the compound muscle action potential (CMAP) and sensory nerve action potential (SNAP) upon maximal stimulation of the tibial and sural nerve respectively, as outlined in previous studies that have utilised the TNS in CKD
(Arnold et al., 2014, Arnold et al., 2016b, Arnold et al., 2013b, Borire et al., 2017,
Borire et al., 2018a, Liveson and Ma, 1992). A score graded from 1–4 was attributed depending on how far the amplitude fell below the lower limit of the normal range.
110 Chapter 1
The scores from the eight items of the TNS were then summed to give a total score from 0–32, with a higher score indicating more severe neuropathy.
All results were analysed using SPSS version 24 (IBM Corporation, U.S.). An exploratory factor analysis was first conducted with the principle component method to evaluate the validity of the items constituting the TNS. Internal consistency of the
TNS was assessed using Cronbach’s alpha and item-total score correlations. As per guidelines, a threshold of >0.7 for Cronbach’s alpha was set to assess internal consistency and a threshold of >0.3 was set for item-total score correlations
(Campbell, 1991). All data was then assessed for normality using a Shapiro-Wilk test.
Variables that were normally distributed are expressed as mean and standard deviation (SD) and variables that were non-normal in distribution are expressed as median and interquartile range (IQR). Construct validity of the TNS was assessed by comparing the TNS results between all subjects with CKD and healthy controls with a
Mann-Whitney U test. Differences in scores for individual items of the TNS between groups were evaluated with a Mann-Whitney U test. Subgroup analyses comparing total neuropathy scores and individual item scores were then conducted using Mann-
Whitney U tests to determine if construct validity was maintained for patients with stage 3, stage 4, stage 5 CKD or CKD without diabetes.
3. Results
3.1. Subject Demographics
Subject demographics are summarised in Table 1.2. 113 subjects with CKD and 40 healthy age-, sex- and BMI-matched controls were recruited into this study.
Participants with CKD ranged from 21 to 82 years of age while control participants
111 Chapter 1 ranged from 35 to 78 years of age. No significant differences were found in age or
BMI between subjects with CKD and controls and sex ratios were maintained. Of the
113 subjects with CKD, 59 (52%) had diabetes and in 52 of these cases, diabetes was established as the primary cause of CKD. In subjects without diabetes, CKD was mainly attributed to hypertension and glomerulonephritis. In terms of CKD severity, there were 70 subjects in the CKD3/4 subgroup and 43 in the CKD5 subgroup, all of which were receiving dialysis. Importantly, there were no significant differences between subgroups and controls for age or BMI and similar sex ratios were preserved.
There were 54 patients in the CKDNoDM subgroup. No significant differences were found in age or BMI between CKDNoDM and controls and sex ratios were maintained
Table 1.2. Subject demographics
CKDT CKD3/4 CKD5 CKDNoDM Controls (n=113) (n=70) (n=43) (n=54) (n=40) Age (years; median, 65 (53-70) 66 (56-69) 61±13 60 (49-69) 61 (51-64) IQR or mean ± SD) Sex (% male) 62 57 70 57 60 BMI (kg/m2; mean ± 31±6 30±5 35±6 28±5 28±4 SD) Cause of CKD (%) N/A Diabetes 46 50 40 N/A Hypertension 14 11 19 30 Glomerulonephritis 9 6 14 15 PKD 6 6 7 11 IgA nephropathy 6 4 7 7 Reflux nephropathy 5 4 5 9 Other 14 19 8 28
BMI: body mass index; IgA: Immunoglobulin A; PKD: polycystic kidney disease
112 Chapter 1
3.2 Factor Analysis
Exploratory factor analysis revealed the sample size to be adequate (Kaiser-Meyer-
Olkin value, 0.894), with a significant Bartlett’s test of sphericity (p<0.001) confirming that at least two items of the TNS correlated significantly with each other and the prerequisite for further analysis was satisfied. All eight items on the TNS demonstrated communality extractions >0.4, which indicates that all items correlated sufficiently with each other (sensory symptoms, 0.494; motor symptoms, 0.447; pinprick sensation, 0.684; vibration sensation, 0.695; strength, 0.516; deep tendon reflexes, 0.616; tibial CMAP, 0.660; sural SNAP, 0.771) (Costello, 2005). Based on the Eigen-values and Scree Plot generated, only 1 factor (which accounted for 61.05% of the variance) could be extracted, which implies that the instrument measured one construct. All eight items of the TNS had loading values greater than 0.6 with this factor (sensory symptoms, 0.703; motor symptoms, 0.668; pinprick sensation, 0.827; vibration sensation, 0.834; strength, 0.718; deep tendon reflexes, 0.785; tibial CMAP,
0.813; sural SNAP, 0.878), which highlights that the domains of the TNS had a good association with the construct (Comrey, 1973).
3.3 Internal Consistency of the TNS
The TNS demonstrated good consistency, as measured by a Cronbach’s alpha of
0.897 (Table 1.3). All items of the TNS demonstrated good corrected-item total correlations with no items scoring less than 0.3 (Campbell, 1991). Deletion of any item did not improve Cronbach’s alpha but it remained 0.897 if motor symptoms were to be removed. The sural SNAP component of the TNS had the greatest measure of reliability (corrected item-total correlation, 0.838) and motor symptoms had the least reliability (corrected item-total correlation, 0.573).
113 Chapter 1
Table 1.3. Internal consistency of the TNS Cronbach’s alpha 0.897 TNS item Corrected item-total correlation Cronbach’s alpha if item deleted Sensory symptoms 0.607 0.890 Motor symptoms 0.573 0.897 Pinprick sensation 0.776 0.875 Vibration sensation 0.775 0.874 Strength 0.634 0.895 Deep tendon reflexes 0.714 0.880 Tibial CMAP 0.736 0.879 Sural SNAP 0.838 0.871
CMAP: compound muscle action potential; SNAP: sensory neuron action potential
3.4 Construct Validity
Subjects with CKD scored significantly higher on the TNS compared to controls
(CKDT: median, 6; IQR, 1–13; controls: median, 0; IQR, 0–1; p<0.001; Figure 1.1).
None of the participants received the highest possible score of 32. Of the 113 subjects with CKD, 19 had a total neuropathy score of 0, 15 of which had stage 3 or 4 CKD.
Significant differences between subjects with CKD and controls were found for all eight domains on the TNS (Table 1.4). Further subgroup analysis revealed that all subjects with CKD scored significantly higher on the TNS compared to controls regardless of stage (CKD3/4: median, 5.5; IQR, 1–10; p<0.001; CKD5: median, 7.0;
IQR, 1–17; p<0.001). Subjects without diabetes (CKDNoDM) also had significantly higher total neuropathy scores compared to controls (CKDNoDM: median, 2, IQR, 0–5; p<0.001). As shown in Table 1.5, comparison of item sub-scores between subgroups and controls revealed that subjects with stage 3/4 or 5 CKD scored significantly higher on all components of the TNS, as did subjects without diabetes (CKDNoDM).
114 Chapter 1
Figure 1.1. Histogram of TNS results for all patients with CKD compared with controls.
Table 1.4. Comparison of TNS item scores between groups
TNS item CKDT (n=113) Controls (n=40) % With deficit Mean % With Deficit Mean Sensory symptoms 44 0.82*** 8 0.08 Motor symptoms 18 0.29** 0 0.00 Pinprick sensation 59 1.13*** 5 0.05 Vibration sensation 57 1.39*** 3 0.03 Strength 31 0.38*** 0 0.00 Deep tendon reflexes 57 1.16*** 13 0.15 Tibial CMAP 29 0.94*** 0 0.00 Sural SNAP 55 1.71*** 5 0.05
CMAP: compound muscle action potential; SNAP: sensory neuron action potential. Mann- Whitney U tests were used for each comparison. **p<0.01, ***p<0.001
115 Chapter 1
Table 1.5. Comparison of TNS item scores between subgroups and controls.
TNS item CKD3/4 CKD5 CKDNoDM Controls (n=70) (n=43) (n=54) (n=40) % With Mean % With Mean % With Mean % With Mean deficit deficit deficit deficit Sensory 44 0.89*** 42 0.72*** 32 0.54** 8 0.08 symptoms Motor 14 0.24* 23 0.37** 9 0.13* 0 0.00 symptoms Pinprick 49 0.90*** 77 1.51*** 46 0.74*** 5 0.05 sensation Vibration 51 1.16*** 65 1.77*** 43 0.98*** 3 0.02 sensation Strength 23 0.29** 44 0.54*** 19 0.19** 0 0.00 Deep tendon 57 1.19*** 56 1.12*** 32 0.50* 13 0.15 reflexes Tibial 26 0.74** 35 1.26*** 9 0.32* 0 0.00 CMAP Sural SNAP 53 1.56*** 58 1.95*** 35 0.88*** 5 0.05
Mann-Whitney U tests were used to compare item scores for CKD3/4, CKD5, and CKDnoDM to Controls. *p<0.05, **p<0.01, ***p<0.001
4. Discussion
The present study sought to demonstrate construct validity of the TNS in the assessment of peripheral neuropathy in subjects with CKD. Despite its use in previous investigations of CKD, there had been no formal validation procedures. This study has provided novel findings in that is has confirmed construct validity of the TNS in
CKD through formal statistical procedures. An exploratory factor analysis was first conducted to investigate the correlation among the eight items constituting the TNS and their suitability as a group in measuring a single factor. It was established that the items constituting the TNS appropriately correlated with a single factor and the association between each item and the single factor was strong. A reliability
116 Chapter 1 assessment of the association between the TNS items and the single factor extracted was then computed using Cronbach’s alpha. It was determined that the TNS had good internal consistency and removal of any item did not improve internal consistency of the TNS, highlighting that all eight items on the TNS contributed positively to the overall score. This implies that each item on the TNS is an appropriate measure of neuropathy and supports the assertion that the TNS is a suitable outcome measure to assess nerve impairment in peripheral neuropathy in patients with CKD. Removal of motor symptoms, which had the lowest item-total correlation, did not change
Cronbach’s alpha indicating that it contributes slightly to the overall measurement.
However, it is an essential component of the TNS as motor impairment becomes apparent in the late stages of CKD and should remain in the instrument (Krishnan and
Kiernan, 2009). CKD patients scored significantly higher in total neuropathy and for all components of the TNS compared to healthy controls. A sufficient range of scores was displayed in the group with CKD demonstrating that the TNS can assess the various severities of neuropathy.
Subgroup analyses revealed that construct validity was maintained over various stages of CKD. Subjects with stage 3/4 or 5 CKD had significantly higher total neuropathy scores and sub-scores for the individual items of the TNS when compared to controls.
An increase in total neuropathy score was observed with the progression of CKD, which further supports the validity of the TNS as an instrument to assess the severity of peripheral neuropathy. Subjects with CKD but without diabetes had significantly higher total neuropathy scores and sub-scores compared to controls. This was maintained in those with stage 3/4 or 5 CKD without diabetes (data not shown), indicating that the TNS can be utilised across different stages of CKD in patients
117 Chapter 1 without diabetes. Patients with stage 2 CKD (eGFR ranging from 60–89 mL/min/1.73m2) were not included in this study as signs of neuropathy are not present with an eGFR above 60 mL/min/1.73m2 (Hanewinckel et al., 2017).
Currently no composite clinical tool to assess peripheral neuropathy in patients with
CKD has undergone formal validation. Laaksonen et al. (2002) determined that two of the most sensitive measures for the diagnosis of peripheral neuropathy were vibration detection from the feet and the sural SNAP, both of which are included in the TNS (Laaksonen et al., 2002). In keeping with these findings, reductions in vibration sensibility and sural SNAP were the highest scored items on the TNS (Table
1.4). The Sural SNAP component of the TNS demonstrated the strongest item-total correlation and would have reduced Cronbach’s alpha the most if removed (Table
1.3). In contrast, reports of motor symptoms, strength and tibial CMAP were the lowest scoring items on the TNS (Table 1.4).
The TNS in its original and modified forms has been employed in studies assessing differing types of neuropathy (Cavaletti et al., 2013, Cavaletti et al., 2007, Cornblath et al., 1999, Gilchrist and Tanner, 2013). In its original form, the TNS includes nerve conduction studies of the sural and peroneal nerve. However, the peroneal nerve is vulnerable to compression injury in kidney disease (Galassi et al., 1998). Therefore, in the interest of avoiding this confounding issue and keeping with methodologies of other studies of peripheral neuropathy in CKD, the tibial nerve was examined instead
(Arnold et al., 2014, Arnold et al., 2016b, Arnold et al., 2013b, Borire et al., 2017,
Borire et al., 2018a). Despite its use as an assessment tool in previous studies, there was no formal validation undertaken and the current study was therefore performed to
118 Chapter 1 rectify this oversight. This study has confirmed the validity of the TNS and therefore continued application is supported and favoured over other scales. Finally, while it would have been favourable to conduct an inter-rater reliability assessment in this study, high inter-rater reliability has previously been established for the TNS
(Cornblath et al., 1999).
It has been demonstrated that the TNS is a valid assessment for the severity of peripheral neuropathy in patients with CKD. The TNS is the first instrument to have undergone formal validation in patients with CKD. A comprehensive scale that can assess a variety of nerve fibres is an ideal clinical tool to identify and characterise neuropathy in patients with CKD. As a valid outcome measure, the TNS will be especially useful in assessing the efficacy of various strategies to treat peripheral neuropathy in future clinical trials. Inter-rater and intra-observer reliability as well as the performance of the TNS in measuring changes over time in longitudinal studies should be examined in future studies.
119 Chapter 2
Chapter 2 – Investigation of the effect of diabetes and chronic kidney disease on axonal pathophysiology
120 Chapter 2
Summary and Link to Thesis
In Chapter 1, I demonstrated that the Total Neuropathy Score is a validated outcome measure to assess the severity of peripheral neuropathy in stages 3–5 chronic kidney disease with and without diabetes. In the following chapter, I explore the pathophysiological mechanisms underlying axonal dysfunction in chronic kidney disease caused by diabetes, which is known as diabetic kidney disease. Features of peripheral neuropathy between chronic kidney disease, type 2 diabetes, and diabetic kidney disease were first compared using the Total Neuropathy Score. It was established that patients with diabetic kidney disease present with an especially severe neuropathy phenotype. Axonal excitability studies were then utilised to investigate the relative contributions of type 2 diabetes and chronic kidney disease to axonal dysfunction in diabetic kidney disease. It was found that chronic kidney disease, and not type 2 diabetes, underlies nerve dysfunction in patients with diabetic kidney disease. This was evidenced by the shared features of ion channel dysfunction between diabetic kidney disease and chronic kidney disease, as well as an observed association with elevated potassium.
This work has been published:
Issar, T., Arnold, R., Kwai, N. C. G., Walker, S., Yan, A., Borire, A. A., . . .
Krishnan, A. V. (2019). Relative contributions of diabetes and chronic kidney
disease to neuropathy development in diabetic nephropathy patients. Clinical
Neurophysiology, 130(11), 2088-2095. doi: 10.1016/j.clinph.2019.08.005
TI was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition.
121 Chapter 2
Abstract
Objective: Chronic kidney disease (CKD) caused by diabetes is known as diabetic kidney disease (DKD). The present study aimed to examine the underlying mechanisms of axonal dysfunction and features of neuropathy in DKD compared to
CKD and type 2 diabetes (T2DM) alone.
Methods: Patients with DKD (n = 30), CKD (n = 28) or T2DM (n = 40) and healthy controls (n = 41) underwent nerve excitability assessments to examine axonal function. Clinical neuropathy severity was assessed using the Total Neuropathy
Score. A validated mathematical model of human axons was utilised to provide an indication of the underlying causes of nerve pathophysiology.
Results: Total neuropathy score was significantly higher in patients with DKD compared to those with either CKD or T2DM (p < 0.05). In DKD, nerve excitability measures (S2 accommodation and superexcitability, p < 0.05) were more severely affected compared to both CKD and T2DM and worsened with increasing serum K+
(p < 0.01). Mathematical modelling indicated the basis for nerve dysfunction in DKD was an elevation of extracellular K+ and reductions in Na+ permeability and the hyperpolarisation-activated cation current, which was similar to CKD.
Conclusions: Patients with DKD manifested a more severe neuropathy phenotype and shared features of nerve dysfunction to that of CKD. The CKD, and not diabetes component, appears to underlie axonal pathophysiology in DKD.
122 Chapter 2
1. Introduction
Diabetic kidney disease (DKD) is the chronic loss of renal function due to diabetes.
The leading cause of chronic kidney disease (CKD) is type 2 diabetes (T2DM) and the growing incidence of T2DM has led to an increase in DKD worldwide (Tuttle et al., 2014). Peripheral neuropathy is one of the most common complications of both
T2DM and CKD (Aggarwal et al., 2013, Feldman et al., 2017). The most frequent clinical manifestation of peripheral neuropathy in both conditions is a distal symmetric polyneuropathy, which is characterised by initial sensory loss and impaired reflexes in the lower limbs with subsequent muscle weakness and atrophy (Krishnan and Kiernan, 2009, Pop-Busui et al., 2017). Peripheral neuropathy in diabetes is known as diabetic neuropathy and affects at least 50% of patients (Feldman et al.,
2017). In CKD, peripheral neuropathy affects approximately 70% of pre-dialysis patients (Hanewinckel et al., 2017). The overlap of diabetes and CKD is thought to contribute to a more rapid onset of nerve injury in patients who have both conditions, however the relative effects of each condition in DKD has not been explored (Pop-
Busui et al., 2010b).
The aim of the study was to assess the difference in neuropathy phenotype in DKD compared to T2DM and CKD alone and evaluate the relative contributions of T2DM and CKD to nerve dysfunction in DKD. Peripheral nerve pathophysiology in each condition was assessed using nerve excitability techniques, which provide an insight into the activity of ion channels, pumps, and exchangers of the axonal membrane. For each cohort, nerve excitability values were evaluated in a mathematical model of the human motor nerve to provide precise data on the exact location of the pathological change in the peripheral nerves of each disease group.
123 Chapter 2
2. Methods
2.1. Subjects
This study was approved by the South East Sydney Area Health Service Human
Research Ethics Committee (Northern Section) and the Human Research Ethics
Committee of the University of New South Wales. All subjects enrolled provided written informed consent to the procedures in accordance with the Declaration of
Helsinki. A total of 98 patients with T2DM, CKD, or DKD were consecutively recruited from the Diabetes Mellitus Centre and Kidney Care Centre at the Prince of
Wales Hospital in Sydney, Australia. Required sample size was calculated based on nerve excitability measures that have previously been demonstrated to be abnormal in in T2DM and CKD (superexcitability and S2 accommodation) with 80% power
(Arnold et al., 2017, Kwai et al., 2013). Allocation into the T2DM group was predicated on a clinical diagnosis of T2DM for at least one year and an estimated glomerular filtration rate (eGFR) greater than 80 mL/min/1.73m2. Inclusion into the
CKD group was based on a clinical diagnosis of stage 3 or 4 CKD (eGFR between
15–59 ml/min/1.73m2) (National Kidney Foundation, 2002) and an absence of T2DM.
Assignment into the DKD group required both a clinical diagnosis of T2DM and stage 3 or 4 CKD. Patients were excluded from participating in the study if they had any of the following: renal transplant, neurotoxic/neuromodulatory treatment, carpal tunnel syndrome, peripheral oedema, an additional condition known to cause neuropathy, or a neuromuscular, movement, psychiatric, or developmental disorder.
41 healthy controls were recruited for comparison.
Patient demographic, serum biochemistry, and clinical data of interest included age,
+ sex, glycated haemoglobin A1c (HbA1c), eGFR, body mass index (BMI), serum K ,
124 Chapter 2 urea, creatinine, and neuropathy severity, which was evaluated using the Total
Neuropathy Score (TNS). The TNS is a validated instrument to evaluate peripheral neuropathy in diabetes, CKD, and DKD (Cornblath et al., 1999, Issar et al., 2018).
The TNS is comprised of eight domains that assess sensory and motor peripheral nerve function and each domain is scored from 0–4. These eight domains are: sensory and motor symptoms, vibration (128 Hz tuning fork) and pinprick sensation
(NeurotipTM, Owen Mumford, United Kingdom), deep tendon reflexes, manual muscle strength, and sural sensory nerve (SNAP) and tibial motor nerve amplitude
(CMAP) (Medelec Synergy, Oxford Instruments, UK). Total neuropathy scores range from 0–32 and a higher score is indicative of more severe neuropathy while a score of zero indicates an absence of neuropathy.
2.2. Nerve Excitability Studies
Nerve excitability was assessed at the median nerve using the TROND protocol as applied by Qtrac software (Digitimer, London, United Kingdom). The median nerve was orthodromically stimulated through surface electrodes (Ambu, Sydney,
Australia), proximal to the wrist at the site of least resistance using a DS5 Isolated
Bipolar Current Stimulator (Digitimer, London, United Kingdom). Compound muscle action potentials (CMAP) were recorded from the abductor pollicis brevis muscle.
Skin temperature at the point of stimulation was kept above 32°C (Burke et al., 1999,
Kiernan et al., 2001). The excitability assessment consisted of four distinct testing paradigms which provide indirect information regarding the activity of voltage-gated sodium (Na+) and potassium (K+) ion channels, energy-dependent pumps, and exchangers embedded within the axon membrane at the point of stimulation.
Stimulus-response curves were first generated using a 1 ms test pulse to obtain the
125 Chapter 2 maximal CMAP amplitude and a target response of 40% of maximum was calculated.
The current required to elicit this target response, known as ‘threshold’, was tracked in the four excitability testing paradigms: strength–duration behaviour, threshold electrotonus, current–threshold relationship, and recovery cycle.
1. Strength–duration behaviour was examined by plotting the relationship
between the stimulus intensity required to reach threshold from four different
stimulus durations (0.2, 0.4, 0.8, and 1.0 ms). Weiss’ Law was then applied to
calculate the strength–duration time constant, which is a marker of persistent
Na+ conductance in the Node of Ranvier (Krarup and Moldovan, 2009).
2. Threshold electrotonus examines the changes in threshold in response to
depolarising (stimulating) or hyperpolarising (inhibiting) conditioning
currents. These changes were determined after a 1 ms test pulse is applied
during or after 100 ms of a subthreshold conditioning current of +40%
(depolarising) or –40% (hyperpolarising) of control threshold, established
from the initial stimulus-response curve. Percentage change in threshold was
plotted at 10 ms intervals and key excitability measures (threshold change
between 10-20 ms and S2 accommodation) were extracted from the plot. S2
accommodation is the specific phase of depolarising threshold electrotonus in
which threshold reduction is limited and begins to return to control level.
Threshold electrotonus assesses various nodal and internodal conductances,
which contribute to the excitability of peripheral nerves and the ionic
maintenance of their intracellular and extracellular environment (Bostock et
al., 1998, Kiernan et al., 2000).
3. The current–threshold relationship quantifies rectification properties of the
internode in response to long-lasting polarising currents (Bostock et al., 1998,
126 Chapter 2
Kiernan et al., 2000). This relationship was determined by plotting the
percentage threshold change of 1 ms test pulses following 200 ms polarising
currents ranging from +50% (depolarising) to –100% (hyperpolarising) of the
control threshold. The minimum I/V slope represents the minimum slope
calculated by fitting a straight line between three adjacent points.
4. The recovery cycle examines the changes in threshold following a nerve
impulse. Percentage change in threshold was plotted for a range of
conditioning–test intervals following a 1 ms supramaximal conditioning
stimulus. The relative refractory period, superexcitability, and subexcitability
measures were subsequently determined. The recovery cycle assesses the
activity of nodal Na+ channels and fast K+ ion channels (Bostock et al., 1998).
2.3. Mathematical Modelling
To investigate the pathological basis for axonal dysfunction in each disease, nerve excitability recordings obtained from groups were further analysed using the Bostock model of axonal excitability, which is a validated model of the human axon based on a single node and internode connected by pathways through and under the myelin sheath (Bostock et al., 1991, Jankelowitz et al., 2007a, Kiernan et al., 2005b). The model assists in the interpretation of excitability findings between control and disease data by providing an indication of the underlying changes in and around the axonal membrane in the disease state. This includes changes in the maximal conductance and permeabilities of different types of Na+ and K+ ion channels, alterations in pump currents, biophysical properties, and surrounding ionic concentrations (Figure 2.1).
The model was first adjusted to fit the mean nerve excitability data obtained from the control group before fitting the mean data of the T2DM, CKD, or DKD group.
127 Chapter 2
Modelling of each disease was approached with a hypothesis based on factors previously implicated in nerve dysfunction. Transient and persistent Na+ channels were modelled at the node whereas fast and slow K+ channels were modelled in the node and internode. Currents through these Na+ and K+ channels were modelled as permeabilities, in accordance with the Constant-Field Theory (Boërio et al., 2014).
Pump and leak currents and axolemmal capacitances were assessed in both
+ compartments. The hyperpolarisation-activated cation current (Ih), which carries Na and K+ ions, was modelled only in the internode. Extracellular K+ and the Barrett-
Barrett conductance, which represents current flow between the node and internode through and underneath the myelin sheath, were also investigated (Howells et al.,
2012). To better account for the effect of extracellular K+, the permeabilities of fast and slow K+ channels were made proportional to extracellular K+ (Boërio et al.,
2014). Modelling analyses involved changes in a single or a combination of parameters in an iterative fashion to objectively fit simulated excitability data with the mean recorded data as closely as possible using a least squares approach. The
‘discrepancy’ between the simulated and recorded data was obtained by the comparing the error between simulated and recorded data of the four excitability paradigms: strength-duration behaviour, threshold electrotonus, current-threshold relationship, and recovery cycle. Weighting factors of these paradigms were 0.5, 1, 1, and 2, respectively and were kept constant for the modelling of each disease.
Minimum interstimulus interval for the recovery cycle was set at 3 ms. Analyses were run in unclamped mode to permit secondary changes in resting membrane potential caused by changes in conductances or pump currents.
128 Chapter 2
Myelin#
+ + Na # Na # Na+/K+#
Nap# Nat# Kf# Kf# Ih#
K+# K+# K+# K+# 2K+#
Ks# Ks# 3Na+# Na+/K+/ATPase# GBB#
Node# Internode#
Figure 2.1. Schematic of the peripheral nerve membrane highlighting the two compartments of the axon and the parameters that were modelled. Parameters investigated included: + voltage-gated sodium (Na ) channels (transient, Nat and persistent, Nap), voltage gated- potassium (K+) ion channels (fast and slow), sodium-potassium pump (Na+/K+-ATPase), the hyperpolarisation-activated activated cation current (Ih) and the Barrett-Barrett conductance (GBB), which is the flow of current between the node and internode through and under the myelin sheath.
2.4. Statistical Analyses
Results were analysed using SPSS Statistics Version 25.0 for Windows (IBM Corp,
New York, USA). Shapiro–Wilk tests were undertaken to determine the normality of data. Where appropriate and with post-hoc tests if required, a one-way analysis of variance, Kruskal-Wallis tests, independent t-tests, Mann–Whitney U tests, Pearson chi-square analyses were applied to compare means of demographic data, clinical measures and extracted nerve excitability variables between groups. Relationships between demographic and clinical data with nerve excitability variables were investigated using bivariate correlations in the form of a Pearson correlation coefficient (r) or a Spearman’s rank correlation coefficient (ρ). Statistical significance
129 Chapter 2 was considered when p < 0.05.
3. Results
3.1. Subject Demographics
Patient demographics were representative of the various disease cohorts (summarised in Table 2.1). The DKD group was matched for age and sex with T2DM and CKD groups. No difference was observed in diabetes duration, percentage of patients receiving insulin, glycaemic control and BMI between diabetes cohorts. Kidney filtration, urea, and creatinine were matched between DKD and CKD groups. All disease groups were matched for serum K+. No significant differences were found in male to female ratio between disease cohorts and controls (p = 0.194).
130 Chapter 2
Table 2.1. Subject demographics DKD T2DM CKD Control p value (n = 30) (n = 40) (n = 28) (n = 41) DKD DKD DKD vs. vs. vs. T2DM CKD Control Age 66 59 59 55 0.113 0.374 <0.001 (years) (61–69) ± 10 (50–71) (54–59) Sex 50 51 63 73 - - - (% male)
BMI 32 30 27 25 0.203 0.004 <0.001 (kg/m2) (27–36) ± 5 ± 5 (23–27)
HbA1c 7.6 7.7 - - 0.753 - - (%) (7.3–8.9) (6.8–9.4) Diabetes 15 11 - - 0.126 - - duration (years) ± 8 ± 7 Patients on 67 65 - - - - - Insulin (%) eGFR 39 90 33 - <0.001 0.097 - (mL/min/1.73m2) ± 10 (87–90) ± 11 K+ 4.6 4.6 4.7 - 0.826 0.581 - (mmol/L) ± 0.5 (4.4–4.9) ± 0.4 Urea 11.2 5.1 11.7 - <0.001 0.539 - (mmol/L) (9.3–15.6) (3.5–6.0) (9.8–13.1) Creatinine 155 70 151 - <0.001 0.522 - (mmol/L) (125–187) (62.1–80) (137–178)
Normally distributed data is expressed as mean ± SD while non-normally distributed data is expressed as median and quartile 1 to quartile 3. Normally distributed data was compared using a one-way ANOVA (with Bonferroni corrections) while non-normally distributed data was compared using a Kruskal-Wallis test (with Dunn-Bonferroni corrections).
3.2 Neuropathy Scoring
Assessment of the total neuropathy scores and subscores for the eight items of the
TNS established that patients with patients with DKD had a significantly higher total
neuropathy score compared to patients with T2DM (p = 0.019) and CKD (p = 0.020;
Table 2.2). Significantly higher subscores in DKD were observed in the TNS domains
of motor nerve conduction amplitude (Tibial CMAP: vs. T2DM, p = 0.008; vs. CKD,
p = 0.002) and sensory nerve conduction amplitude (Sural SNAP: vs. T2DM, p =
131 Chapter 2
0.003; vs. CKD, p = 0.011). In DKD, the greatest impairment was seen in sensory nerve conduction while motor symptoms and manual muscle strength items had the lowest subscore values.
Table 2.2. Total neuropathy score and subscore comparison DKD T2DM CKD Control p value (n = 30) (n = 40) (n = 28) (n = 41) DKD DKD vs. T2DM vs. CKD Total neuropathy 9.2 ± 1.4 5.3 ± 1.0 4.9 ± 1.1 0 0.019 0.020 score Sensory symptoms 1.0 ± 0.2 1.0 ± 0.2 0.9 ± 0.2 0 0.691 0.513 Motor symptoms 0.4 ± 0.2 0.3 ± 0.1 0.4 ± 0.1 0 0.286 0.811 Pinprick sensation 0.9 ± 0.2 0.5 ± 0.1 0.8 ± 0.2 0 0.108 0.678 Vibration sensation 1.4 ± 0.3 0.8 ± 0.2 0.8 ± 0.2 0 0.070 0.086 Strength 0.4 ± 0.2 0.3 ± 0.1 0.1 ± 0.1 0 0.353 0.103 Tendon reflexes 1.6 ± 0.3 1.3 ± 0.2 0.9 ± 0.2 0 0.302 0.043 Tibial CMAP 1.3 ± 0.3 0.4 ± 0.2 0.2 ± 0.2 0 0.008 0.002 Sural SNAP 2.0 ± 0.3 0.8 ± 0.2 0.8 ± 0.2 0 0.003 0.011
All data is expressed as mean ± SEM. Subscores range from 0–4, with a higher value indicating a greater severity of impairment. Total neuropathy scores range from 0–32, with a higher score indicative of greater neuropathy severity. CMAP: compound muscle action potential; SNAP: sensory nerve action potential. Data was compared using a Kruskal-Wallis test (with Dunn- Bonferroni corrections).
3.3 Nerve Excitability
Nerve excitability findings are presented in Figure 2.2 and extracted variables are summarised in Table 2.3. Patients with DKD exhibited severe abnormalities in excitability measures compared to patients with either T2DM or CKD. In DKD, there was a significant reduction in S2 accommodation (p = 0.007) and minimum I/V slope
(p = 0.001) when compared to T2DM. Further, there was marked increase in superexcitability (p = 0.003) and decrease in subexcitability (p = 0.003) in the
132 Chapter 2 recovery cycle. Compared to patients with only CKD, patients with DKD demonstrated significant increases in superexcitability (p = 0.030) and the strength- duration time constant (p = 0.025) and a reduction in S2 accommodation (p = 0.025)
Across all disease groups, no other differences in depolarising or hyperpolarising threshold electrotonus or in the current-threshold relationship were found. No correlation was found between total neuropathy score and excitability indices.
Compared to controls, patients with DKD demonstrated significantly worse changes in all nerve excitability parameters, with the exception of the strength-duration time constant.
Figure 2.2. Mean excitability data for patients with T2DM, CKD or DKD and healthy controls for the threshold electrotonus (A), recovery cycle (B) and current-threshold paradigms (C).
133 Chapter 2
Table 2.3. Nerve excitability findings DKD T2DM CKD Control p value (n = 30) (n = 40) (n = 28) (n = 41) DKD DKD DKD CKD T2DM vs. T2DM vs. CKD vs. Control vs. Control vs. Control SDTC 0.46 0.48 0.41 0.44 0.499 0.025 0.368 0.197 0.052 (ms) ± 0.01 ± 0.02 ± 0.02 ± 0.01 TEd (10–20 ms) 61.7 65.4 64.0 67.9 0.075 0.350 0.001 0.015 0.079 (%) ± 1.4 ± 1.1 ± 1.1 ± 0.9 TEd (S2) 18.7 21.1 21.0 22.9 0.007 0.025 <0.001 0.033 0.015 (%) ± 0.7 ± 0.5 ± 0.7 ± 0.5
Minimum 0.22 0.26 0.24 0.26 0.001 0.271 0.001 0.084 0.999 I/V slope ± 0.01 ± 0.01 ± 0.01 ± 0.01
Superexcitability –15.4 –19.9 –19.7 –23.6 0.003 0.030 <0.001 0.019 0.003 (%) ± 1.2 ± 0.9 ± 1.5 ± 0.9 Subexcitability 9.4 11.9 11.0 14.6 0.003 0.080 <0.001 <0.001 0.001 (%) ± 0.7 ± 0.5 ± 0.6 ± 0.6 RRP 3.7 3.3 3.5 3.2 0.091 0.545 0.006 0.018 0.125 (ms) ± 0.2 ± 0.1 ± 0.1 ± 0.1
By convention, all excitability data is expressed as mean ± SEM with the exception of strength- duration time constant, minimum I/V slope and RRP. SDTC: strength-duration time constant; TEd: depolarising threshold electrotonus; S2: S2 accommodation phase; I/V: current-threshold; RRP: relative refractory period. Normally distributed data was compared using a one-way ANOVA (with Bonferroni corrections) while non-normally distributed data was compared using a Kruskal-Wallis test (with Dunn-Bonferroni corrections).
Nerve excitability measures in the CKD and T2DM cohorts were also found to be
abnormal when compared to the control group (Table 2.3). Patients with CKD or
T2DM exhibited decreases in S2 accommodation (CKD: p = 0.033; T2DM: p =
0.015) and subexcitability (CKD: p < 0.001; T2DM: p = 0.001) and increases in
superexcitability (CKD: p = 0.019; T2DM: p = 0.003) and in the undershoot phase of
depolarising threshold electrotonus (CKD: p = 0.011; T2DM: p = 0.005).
Additionally, depolarising threshold electrotonus at 10–20 ms was reduced (p = 0.015
and the relatively refractory period was increased (p = 0.018) in CKD but not T2DM.
134 Chapter 2
In DKD and CKD, increasing serum K+ significantly correlated with more severe changes in nerve excitability variables. In DKD, increase in serum K+ was associated with more severe alterations in various phases of depolarising threshold electrotonus
(10–20 ms: r = –0.680, p < 0.001; S2 accommodation: r = –0.526, p = 0.003, undershoot: r = 0.423, p = 0.022) and recovery cycle (superexcitability: r = 0.660, p <
0.001; relative refractory period: ρ = 0.409, p = 0.034). In CKD, higher serum K+ correlated with worsening depolarising threshold electrotonus at 10–20 ms (r = –
0.390, p = 0.044) and superexcitability (r = 0.464, p = 0.017). In contrast, no relationship was found between serum K+ and abnormal excitability measures in
T2DM. In DKD and CKD, no association was found between urea or creatinine levels and nerve excitability variables. No relationship was observed between HbA1c and excitability parameters in DKD and T2DM.
3.4 Mathematical Modelling
Nerve excitability findings were modelled by varying membrane properties previously implicated in nerve dysfunction in each condition and findings are summarised in Figure 2.3 and Table 2.4. Analysis of the DKD findings indicated that alterations in nerve excitability recordings were due to an elevation of extracellular
K+ concentration (Control: 3.6; DKD: 4.4, mmol/L) and a 20% reduction in Na+ permeability (Control: 4.8; DKD: 3.85, cm3s-1 x 10–9). A considerable improvement in the model could be achieved by a 26% reduction in Ih conductance (Control: 5.5;
DKD: 4.05, nanosiemens). Together, these alterations in the model reduced the discrepancy between the Control group and patients with DKD by 86%.
135 Chapter 2
Modelling of the CKD data was similar to the DKD data, particularly in relation to
+ Na permeability and Ih conductance. It revealed the simplest explanation for nerve excitability changes was a combination of a 15% decrease in Na+ permeability
3 -1 –9 (Control: 4.8; CKD: 4.1, cm s x 10 ) and a 28% reduction in Ih conductance
(Control: 5.5; CKD: 3.95, nanosiemens). These variations could account for 85% of the discrepancy between the two groups, which was considerably more than any other two-parameter combination.
136 Chapter 2
(A)#
Myelin%
+ + Na # Na # Na+/K+#
Nap# Nat# Ih#
+ !#Na #permeability# !#Ih#conductance#
"#Extracellular#K+#
Node% Internode%
(B)#
Myelin%
+ + Na # Na # Na+/K+#
Nap# Nat# Ih#
+ !#Na #permeability# !#Ih#conductance#
Node% Internode%
(C)#
Myelin%
Na+# Na+#
Nap# Nat# Kf# Kf# "#Nodal#and#internodal#fast#K+# !#Na+#permeability# K+# K+# permeability# + "#Persistent#Na #channels# + K # K+# "#Internodal#slow#K+# !#Nodal#slow#K+#permeability# !#BarreABBarreA#conductance# permeability#
Ks# Ks#
GBB%
Node% Internode%
Figure 2.3. Summary of mathematical modelling of nerve excitability data for diabetic kidney disease (A), chronic kidney disease (B), and type 2 diabetes (C). Pathophysiological mechanisms underlying nerve dysfunction in diabetic kidney disease are similar to those of chronic kidney disease.
137 Chapter 2
Table 2.4. Modelled parameters between DKD, CKD, T2DM, and control cohorts Parameter DKD CKD T2DM Control
Permeability of Na+ channels at node (cm3s-1 x 10–9) 3.85 4.1 4.05 4.8
% of Na+ channels that are persistent (%) 0.835 0.69
Permeability of nodal slow K+ channels (cm3s-1 x 10–9) 0.48 0.56
Permeability of internodal slow K+ channels (cm3s-1 x 10–9) 0.0064 0.006
Permeability of nodal fast K+ channels (cm3s-1 x 10–9) 0.25 0.2
Permeability of internodal fast K+ channels (cm3s-1 x 10–9) 1.53 1.35
Barrett-Barrett conductance (nS) 33 33.9
Hyperpolarisation-activated cation current (nS) 4.05 3.95 5.5 Extracellular K+ (mmol/L) 4.4 3.6 Key parameter changes applied in the mathematical model for disease cohorts from control values. Conductance values are expressed in nanosiemens (nS).
Mathematical modelling of the T2DM data was quite different to both DKD and
CKD. It demonstrated that the changes in T2DM were best explained by altering a range of conductances and permeabilities through ion channels in the node and the internode, as previously demonstrated in studies of type 1 diabetes (Kwai et al.,
2016a). Specifically, the changes were best accounted for by a decrease in Na+ permeability and slow nodal K+ channel permeability as well as an increase in the percentage of persistent Na+ channels and fast K+ channel permeability. Together, these changes accounted for 79% of the discrepancy between the Control and T2DM groups.
4. Discussion
This study has provided evidence that DKD patients exhibit more severe neuropathy and greater nerve dysfunction when compared to patients with either CKD or T2DM
138 Chapter 2 alone. Patients with DKD had a higher total neuropathy score and exhibited greater reductions in motor and sensory nerve conduction amplitudes when compared to either T2DM or CKD alone, which suggests that the combined pathological effects of diabetic and uraemic neuropathy results in more severe nerve injury (Table 2.2).
Consistent with these findings, patients with DKD demonstrated greater nerve dysfunction, as assessed by nerve excitability techniques, when compared to either
T2DM or CKD (Table 2.3). Specifically, more severe alterations in S2 accommodation and superexcitability were observed in DKD. The magnitude of change was similar to previously reported measures in type 2 diabetes patients with severe diabetic neuropathy (total neuropathy scores ranging from 24–32) and patients with end-stage kidney disease (Borire et al., 2018a, Kwai et al., 2013). A significantly larger, and apparent normalisation of, the strength-duration time constant was present in DKD compared to CKD. It is hypothesised that this is due the increase of persistent
Na+ current that occurs in diabetic neuropathy (Kwai et al., 2013, Misawa et al.,
2009). Significant differences in subexcitability and minimum I/V slope were also evident between DKD and T2DM but not CKD. Nerve dysfunction was also evident in CKD and T2DM alone when compared to healthy controls. While serum K+ was within normal range in both DKD and CKD, increasing K+ nevertheless correlated with greater abnormalities in nerve function. This was not the case in T2DM. Further, no correlation was found between HbA1c and excitability measures in DKD and
T2DM. Mathematical modelling suggested the underlying causes of peripheral nerve dysfunction in DKD mirrored those previously implicated in CKD, rather than
T2DM.
139 Chapter 2
Mathematical modelling of the DKD nerve excitability indicated that the underlying basis for nerve dysfunction was an elevation in extracellular K+ to 4.4 mmol/L in conjunction with reductions in Na+ permeability and the hyperpolarisation-activated cation current. The finding that axonal dysfunction was partially driven by an increase in K+ from both the modelling and correlation analyses suggests that the CKD component has a greater role in nerve dysfunction in DKD. This is further supported by clinical trial data which have demonstrated that dietary potassium restriction, which is thought to reduce nerve injury due to CKD, appears to have an equally beneficial role in CKD patients with diabetes as CKD patients without diabetes
(Arnold et al., 2017).
In the CKD cohort, abnormalities in nerve excitability were present even though the mean serum K+ concentration of 4.7 mmol/L fell within normal range. Increasing K+ correlated with greater nerve dysfunction, which is consistent with previous findings from a randomised controlled trial of dietary K+ restriction in CKD in which it was demonstrated that the maintenance of serum K+ ≤ 4.5 mmol/L is neuroprotective
(Arnold et al., 2017). The results of the current investigation suggest that peripheral nerves in CKD patients are especially susceptible to the development of abnormal excitability properties even within the high normal range for serum K+. This argument is further supported by previous excitability studies in patients with end-stage renal failure without diabetes, in which an elevation of K+ instead of other suspected uraemic toxins was implicated in axonal dysfunction (Arnold et al., 2014, Krishnan et al., 2005b). Modelling analysis of the CKD data indicated that the changes in nerve excitability were due a decrease in the permeability of Na+ at the node and conductance of Ih in the internodal compartment of the axon, which is underneath the
140 Chapter 2 myelin sheath. The decrease in Na+ permeability is consistent with findings in experimental uraemic neuropathy (Brismar and Tegnèr, 1984). Alteration in Ih may be a reflection of structural changes in Schwann cells, such as swelling, that occurs with submyelinic accumulation of K+ or the impaired buffering of internodal K+ by these cells (Baker, 2002, Brazhe et al., 2011, Said, 2013).
Mathematical modelling of the excitability data in the T2DM cohort were consistent with previous studies of specific nodal and internodal ion channels in clinical and experimental T2DM (Table 2.4). The pattern of a decrease in Na+ permeability and increase in the percentage of persistent Na+ has been observed in human and animal studies (Brismar et al., 1987, Hong and Wiley, 2006, Krishnan and Kiernan, 2005,
Misawa et al., 2009). Peripheral nerve biopsies from neuropathic T2DM patients and
T2DM animal models have also demonstrated there is diffuse redistribution of fast K+ channels from their juxtaparanodal position, which may explain the increase in permeability of these channels in both compartments of the axon (Zenker et al.,
2012). The alterations in the nodal and internodal conductances and permeabilities determined from the modelling analysis of the T2DM excitability recordings are similar to findings in type 1 diabetes, with the exception of the Barrett-Barrett conductance (Kwai et al., 2016a). In contrast to T2DM, it was observed that the
Barrett-Barrett conductance (current flow between the node and internode through and underneath the myelin sheath) is increased in type 1 diabetes. It should however be noted there were differences in neuropathy severity between the diabetes groups of these two studies, as patients with type 1 diabetes in Kwai et al. (2016) were neuropathy-free and patients with T2DM in the present study had established neuropathy. Further, potassium channel function was modelled as a conductance in
141 Chapter 2
Kwai et al. (2016) as opposed to permeability, which has been shown to be more accurate (Boërio et al., 2014).
While we have provided evidence that CKD primarily underlies nerve pathophysiology in DKD, given that the DKD cohort exhibited more severe neuropathy than either the CKD or T2DM groups, the pathological effects of T2DM are likely to be involved in some way. Studies of insulin signalling in DKD have led to the proposal that insulin resistance is a key factor in the decline of renal function
(Karalliedde and Gnudi, 2016). Greater insulin resistance is independently associated with the development of microalbuminuria and is thought to cause an escalation in oxidative stress and pro-inflammatory mediators in the kidney (Karalliedde and
Gnudi, 2016, Parvanova et al., 2006). Consequently, this would initiate nerve injury via pathways implicated in uraemic neuropathy. Interestingly, insulin has also been shown to be a potent neurotrophic factor and insulin receptors are abundantly expressed on peripheral nerves (Feldman et al., 2017). The greater severity of neuropathy and nerve dysfunction observed in the DKD compared to T2DM may therefore be in part due to greater insulin resistance in DKD, which may result in an inability of peripheral nerves to respond to trophic support.
In conclusion, patients with DKD manifest a more severe neuropathy phenotype and greater nerve dysfunction than patients with either T2DM or CKD alone. Analysis of nerve excitability findings in DKD suggests the CKD component of the condition has a major role in causing axonal dysfunction. Future clinical studies of diabetic neuropathy should examine and report renal status of patients as a complicating pathophysiological factor.
142 Chapter 3
Chapter 3 – Association between acute glucose control and axonal function and structure in type 1 diabetes
143 Chapter 3
Summary and Link to Thesis
In the previous chapter, it was found that chronic kidney disease, and not type 2 diabetes, underlies axonal dysfunction in diabetic kidney disease. The following two chapters explore the pathophysiological mechanisms of peripheral nerve dysfunction in autoimmune diabetes. In the Literature Review, I discussed the emergence of acute measures of glucose control in the investigation of diabetic neuropathy in type 1 diabetes. In this chapter, I explore the association between acute measures of glucose control and peripheral nerve function and structure in type 1 diabetes. Continuous glucose monitoring was utilised to calculate continuous overlapping net glycaemic action (a measure of glucose variability) as well as percentage time in (and above) target range. Following glucose monitoring, peripheral nerve function was assessed using axonal excitability whilst peripheral nerve structure was investigated using corneal confocal microscopy. It was found that poorer acute glucose control was associated with worsening axonal excitability outcomes and increased the likelihood of abnormal nerve function. Poorer acute glucose control was also associated with greater alterations in corneal nerve morphology.
This work has been published:
Issar, T., Tummanapalli, S. S., Kwai, N. C. G., Chiang, J. C. B., Arnold, R., Poynten,
A. M., . . . Krishnan, A. V. (2020). Associations between acute glucose control
and peripheral nerve structure and function in type 1 diabetes. Diabetic
Medicine. doi: 10.1111/dme.14306
TI was responsible for the study design, recruitment, data collection (with exception of in-vivo corneal confocal microscopy – SST, JCBC), data interpretation, and the manuscript composition.
144 Chapter 3
Abstract
Objective: To examine the associations between of continuous overlapping net glycaemic action (CONGA), percentage time in hyperglycaemia (%HG) or normoglycaemia (%NG) and peripheral nerve structure and function in type 1 diabetes
Methods: Twenty-seven participants with type 1 diabetes underwent continuous glucose monitoring followed by corneal confocal microscopy and nerve excitability assessments. CONGA, %HG (>10.0 mmol/L), and %NG (3.9–10.0 mmol/L) were correlated against corneal nerve fibre length and density in the central cornea and inferior whorl region, corneal micro-neuromas, and a nerve excitability score while controlling for age, sex, diabetes duration, and HbA1c.
Results: An increase in CONGA (median 2.5 [2.0–3.1] mmol/L) or %HG (mean 46 ±
18%) was associated with a worse nerve excitability score (r = –0.433, p = 0.036 and r = –0.670, p = 0.0012, respectively). In contrast, greater %NG (51 ± 17%) correlated with better nerve excitability scores (r = 0.672, p = 0.0011). Logistic regression revealed that increasing %HG increased the likelihood of abnormal nerve function
(OR: 1.11, 95% CI: 1.01–1.23, p = 0.037). An increase in CONGA and %HG were associated with worsening nerve conduction measures, while longer %NG correlated with improved nerve conduction variables. CONGA and %HG were associated with inferior whorl corneal nerve fibre length (r = 0.483, p = 0.034 and r = 0.591, p =
0.021, respectively) and number of micro-neuromas (r = 0.433, p = 0.047 and r =
0.516, p = 0.020, respectively).
145 Chapter 3
Conclusions: Short-term measures of glucose control are associated with impaired nerve function and alterations in corneal nerve morphology.
146 Chapter 3
1. Introduction
Diabetic neuropathy is a common microvascular complication of diabetes (Pop-Busui et al., 2017). Findings from the Diabetes Control and Complications Trial and
Epidemiology of Diabetes Interventions and Complications study have indicated that despite strict HbA1c control, there is still a high frequency of neuropathy development in type 1 diabetes (Pop-Busui et al., 2010a). While HbA1c is an important clinical marker of glucose control, it fails to capture acute glucose excursions, the magnitude and frequency of intra- and inter-day glucose variations, and at times, an accurate mean glucose (Battelino et al., 2019, Beck et al., 2017). These measures however can be otherwise determined from continuous glucose monitoring. Recently, there has been a shift towards focusing on time in (or outside) optimal glucose range and glucose variability as more accurate predictors of the development of diabetic complications (Battelino et al., 2019, Beck et al., 2019b, Ceriello et al., 2019, Lu et al., 2018). Given the inaccuracies of HbA1c, monitoring of these emerging measures of glucose control may be more effective for the prevention of diabetic complications
(Beck et al., 2019a).
The aim of the present study was to determine the effects of time outside of range and glucose variability on peripheral nerve structure and function in type 1 diabetes.
Peripheral nerve structure was assessed using corneal confocal microscopy while nerve function was assessed using nerve excitability studies, which provide insight into axonal ion channel dysfunction in peripheral nerve disorders (Chen et al., 2015,
Kiernan et al., 2020, Kwai et al., 2016a, Tavakoli et al., 2013). These outcomes were correlated with measures obtained from continuous monitoring, namely continuous
147 Chapter 3 overall net glycaemic action (CONGA) and percentage time in (and above) target range, to capture the degree and duration of glucose variation.
2. Methods
2.1. Participants
This study was approved by the Human Research Ethics Committee of the University of New South Wales (Sydney, Australia) and the South East Sydney Area Health
Service Human Research Ethics (reference number: 14/012). All participants recruited into the study provided written informed consent to the procedures in accordance with the Declaration of Helsinki. Twenty-seven individuals with established type 1 diabetes for at least five years were consecutively enrolled from the
Diabetes Mellitus Centre of Prince of Wales Hospital (Sydney, Australia). Sample size calculations were based on detecting differences in nerve function between individuals with type 1 diabetes and people without diabetes, and was determined from the known reduction in subexcitability in type 1 diabetes, as demonstrated by
Kwai et al. (type 1 diabetes: 10.4 ± 4.7; people without diabetes: 15.4 ± 3.7 (Kwai et al., 2016a). The power calculation was based on obtaining 80% power with an alpha error = 0.05. Participant demographic data collected was age, sex, duration of diabetes, insulin delivery modality (multiple daily insulin injections or continuous subcutaneous insulin infusion), HbA1c, lipid profile, systolic blood pressure and BMI.
Exclusion criteria for prospective participants of the study included any of the following: carpal tunnel syndrome, peripheral oedema, previous neurotoxic therapy, current treatment with neuropathic pain medications, chronic kidney disease, any other condition known to cause neuropathy, current eye infections or corneal abrasions, previous refractive surgery, trauma to anterior segment, or contact lens
148 Chapter 3 wear, or if they had a neuromuscular, movement, psychiatric or developmental disorder. A cohort of 35 healthy people without diabetes similar for age, BMI and sex proportion were recruited from the University of New South Wales (Sydney,
Australia) and served to provide reference values for peripheral nerve structure and function.
2.2. Neuropathy Assessment
Neuropathy status was assessed using the Total Neuropathy Score, which is validated for the assessment of diabetic neuropathy (Cornblath et al., 1999). The Total
Neuropathy Score is comprised of eight items: sensory and motor symptoms, sensation to vibration (128 Hz tuning fork), and pinprick (NeurotipTM, Owen
Mumford, Oxford, United Kingdom), deep tendon reflexes, manual muscle strength, and sural (sensory nerve amplitude and velocity) and tibial (motor nerve amplitude and latency) nerve conduction studies (Natus, Middleton, USA), as per standard clinical neurophysiological protocols (Cornblath et al., 1999, Liveson and Ma, 1992).
Each item of the Total Neuropathy Score is scored from 0 to 4 and summed to give a total score ranging from 0 to 32.
2.3. Continuous Glucose Monitoring and Assessment of Glucose Control
All participants with diabetes underwent six days of blinded continuous glucose monitoring (iPro, Medtronic, CA, USA; sensor: EnliteTM, Medtronic, CA, USA) to assess glucose control prior to testing. Glucose recordings were analysed using
Glycemic Variability Analyzer Program, which has been validated for the assessment of glucose variability (Marics et al., 2015). Variables calculated were time in and above range, expressed as a percentage of total monitoring time, and continuous
149 Chapter 3 overall net glycaemic action (CONGA), which can assess intra-day and inter-day variability (Beck et al., 2019b, Frontoni et al., 2013, McDonnell et al., 2005). Time in range (%NG) was considered when glucose was between 3.9–10.0 mmol/L and time above range (%HG) was defined when glucose values exceeded 10.0 mmol/L
(Battelino et al., 2019). CONGA is defined as the standard deviation of the summed differences between the current glucose observation and an observation n hours prior, with a higher CONGA indicating greater glycaemic variation (McDonnell et al.,
2005). The interval, n, was 1 hour. CONGA was calculated for each day of monitoring and averaged. The formula for CONGA is included in the Supplementary
Material.
2.4. Nerve Excitability Studies and Receiver Operating Characteristic Curve Analysis
To assess peripheral nerve function, motor nerve excitability assessments were conducted on the median nerve (as in previous studies of diabetes utilising nerve excitability) following continuous glucose monitoring (Kiernan et al., 2020). The nerve excitability assessment consisted of two testing paradigms that provide indirect information regarding the function of voltage-gated sodium (Na+) and potassium (K+) ion channels within the axon membrane at the point of stimulation (Issar et al., 2019).
First, stimulus-response curves were generated using a 1 millisecond test pulse to obtain the maximal compound muscle action potential amplitude from abductor pollicis brevis. A target response of 40% of this maximum was subsequently calculated. The current required to elicit this target response, known as ‘threshold’, was tracked in two testing paradigms: threshold electrotonus and recovery cycle (Issar et al., 2019).
150 Chapter 3
1. Threshold electrotonus examines the changes in threshold in response to
depolarising (stimulating) or hyperpolarising (inhibiting) conditioning currents to
assess a variety of nodal and internodal ion channels (Bostock et al., 1998). These
changes were determined after a 1 millisecond test pulse was applied during or
after 100 milliseconds of a subthreshold conditioning current of +40%
(depolarizing) or –40% (hyperpolarising) of control threshold, established from
the initial stimulus-response curve. Percentage change in threshold was plotted at
10 millisecond intervals (Issar et al., 2019).
2. The recovery cycle examines the changes in threshold following a supramaximal
nerve impulse and assesses the activity of nodal Na+ channels and fast K+ channels
(Bostock et al., 1998). Percentage change in threshold was plotted for a range of
conditioning–test intervals (2–200 millisecond) following a 1 millisecond
supramaximal conditioning stimulus (Issar et al., 2019).
As utilised in previous studies, a composite score of nerve function was created from the nerve excitability parameters to reflect multiple aspects of nerve physiology
(Arnold et al., 2017, Park et al., 2009). The excitability measures contributing to the composite score were depolarising threshold electrotonus between 40–60 ms and subexcitability of the recovery cycle, which reflect internodal and nodal properties of the axon, respectively, and are adversely affected in type 1 diabetes (Kwai et al.,
2016a). A lower score indicates worse function. Nerve excitability scores, which reflect the overall function of ion channels embedded in the axon membrane, were defined as being ‘normal’ or ‘abnormal’ based on cut-offs determined from all participant data using a receiver operating characteristic curve. Cut-offs that maximised the sum of sensitivity and specificity (Youden’s index) were chosen.
151 Chapter 3
2.5. Corneal Confocal Microscopy
All participants were scanned bilaterally with a corneal confocal microscope to visualise corneal nerve morphology and assess the presence of micro-neuromas
(Heidelberg Engineering GmbH, Heidelberg, Germany). Eight central and three to four inferior whorl images of the cornea from both eyes of each participant were selected for quantification. Images selected were not overlapping by more than 20%.
Corneal micrographs were analysed using a validated and fully automated nerve analysis software to quantify corneal nerve fibre density (defined as the number of main branches per square millimetre, no./mm2) and length (defined as the average length of nerve fibres in millimetre per square millimetre, mm/mm2) in the central cornea. Corneal nerve fibre length was also obtained in the inferior whorl, which is the region that contains the most distal aspect of corneal nerves and is sensitive to changes in diabetes (Corneal Nerve Fiber Analyzer V.2, ACCMetrics, University of
Manchester, Manchester, United Kingdom)(Chen et al., 2017, Kalteniece et al., 2018,
Petropoulos et al., 2015). Corneal nerve measures are presented as an average of both eyes. Micro-neuromas, which have been observed in instances of regeneration, were defined as nerve abnormalities that present as irregularly shaped, terminal enlargements of nerve endings with variable hyper-reflectivity and poorly defined margins, which were singular or in clusters of 2-3 (Aggarwal et al., 2019). These were counted manually and the same micro-neuroma imaged in more than one frame was considered as a count of 1. Results are expressed as the total number of micro- neuromas from both eyes.
152 Chapter 3
2.6. Statistical Analysis
Data were analysed using SPSS Statistics Version 25.0 for Windows (IBM Corp,
New York, USA). The normality of the data was first determined using Shapiro–Wilk tests. Data that was normally distributed is expressed as mean ± standard deviation while data that was non-normally distributed is represented as median and quartile 1 to quartile 3. Independent t-tests, Mann–Whitney U tests, Pearson chi-square analyses were used to compare the means of demographic data, clinical measures, nerve excitability scores, and corneal confocal measurements between participants with diabetes and people without diabetes. The relationship between %NG, %HG,
CONGA, nerve excitability and corneal confocal microscopy was investigated using partial correlations adjusted for age, sex, diabetes duration, and HbA1c. Spearman’s rank correlation coefficient was used to examine the associations between %NG,
%HG, CONGA, and HbA1c. Logistic regression was used to predict abnormal nerve function from %HG, %NG, or CONGA. For all tests, statistical significance was considered when p < 0.05.
3. Results
3.1. Participant Demographics
Participant demographics, neuropathy scores, and neurophysiological measurements are summarised in Table 3.1. Diabetes participants and people without diabetes were similar for age, sex proportion, and BMI. Overall, participants had long-standing diabetes, were above target HbA1c, and there was approximately even distribution between participants treated with multiple daily insulin injections and those using insulin pump therapy.
153 Chapter 3
Table 3.1. Participant demographics T1DM People without p value (n = 27) diabetes (n = 35) Age 36 (24–56) 35 (30–52) 0.504 (years) Sex 18:9; 67% 20:15; 57% 0.445 (M:F; % men) BMI 25 (24–29) 24 (23–26) 0.106 (kg/m2) HbA 1c 67 (58–78) - - (mmol/mol) HbA 1c 8.3 (7.5–9.3) - - (%) Diabetes duration 20 ± 8 - - (years) Insulin delivery 15:12 - - (MDII:CSII; % MDII) Triglycerides 0.95 (0.68–1.23) - - (mmol/L) LDL 2.22 ± 0.73 - - (mmol/L) HDL 1.58 ± 0.52 - - (mmol/L) Systolic BP 120 (113–131) - - Total neuropathy 3 (0–5) 0 < 0.001 score Sural amplitude 8.8 (3.3–16.4) 20.3 (12.8–28.0) < 0.001 (μV) Sural NCV 41 (31–49) 50 (41–56) < 0.05 (m/s) Tibial amplitude 7.5 ± 5.6 14.0 ± 5.6 < 0.001 (mV) Tibial latency 3.9 (3.4–4.2) 3.4 (3.2–3.9) 0.237 (ms) Normally distributed data is expressed as mean ± SD while non-normally distributed data is expressed as median and quartile 1 to quartile 3. T1DM, type 1 diabetes; MDII, multiple daily insulin injections; CSII, continuous subcutaneous insulin infusion; LDL, low density lipoprotein; HDL, high density lipoprotein; NCV, nerve conduction velocity. Normally distributed data was compared using an independent t-test while non-normally distributed data was compared using a Mann-Whitney U test.
3.2. Continuous Glucose Monitoring
All participants with diabetes completed blinded continuous glucose monitoring with no complications. Average %HG was 46 ± 18% and mean %NG was 51 ± 17%,
154 Chapter 3 demonstrating that participants were largely either above or within recommended range. Median CONGA was 2.5 mmol/L (2.0–3.1 mmol/L). Percentage time in hyperglycaemia correlated with HbA1c (r = 0.616, p = 0.0008) and CONGA (r =
0.430, p = 0.025) but no association was found between HbA1c and CONGA (r =
0.126, p = 0.541). No relationship was observed between the demographic variables age, sex, BMI, diabetes duration, mode of insulin delivery, lipid profile, systolic blood pressure or total neuropathy score (including sub-scores) to short-term measures of glucose control. According to the Tesfaye definition of confirmed neuropathy (presence of a nerve conduction abnormality and a sign or symptom), 12 of the 27 participants would be classified as having peripheral neuropathy.
3.3. Nerve Conduction Studies
Partial correlations between glycaemic metrics and nerve conduction measures adjusted for age, sex, diabetes duration, and HbA1c are listed in Table 3.2. CONGA was associated with tibial latency (r = 0.527, p = 0.018). Percentage time in hyperglycaemia correlated with tibial nerve amplitude (r = –0.504, p = 0.023) and tibial latency (r = 0.509, p = 0.022). A relationship between time in normoglycaemia and tibial nerve amplitude (r = 0.429, p = 0.049) as well as tibial latency (r = –0.544, p = 0.015) was also observed.
155 Chapter 3
Table 3.2. Partial correlations between acute glycaemic metrics and measures of peripheral nerve structure and function Sural Sural Tibial Tibial Excitability IWL Micro-neuroma amplitude (μV) NCV (m/s) amplitude (mV) latency (ms) score (mm/mm2) count CONGA –0.047 –0.241 –0.114 0.527* –0.433* 0.483* 0.433* %HG –0.021 –0.209 –0.504* 0.509* –0.670** 0.591* 0.516* %NG 0.080 0.211 0.429* –0.544* 0.672** –0.542* –0.545*
Correlations have been controlled for age, sex, diabetes duration, and HbA1c. CONGA, continuous net overall glycaemic action; %HG, percentage time above target range; %NG, percentage time in target range; NCV, nerve conduction velocity; IWL; inferior whorl corneal fibre length. *P < 0.05, **P < 0.01
3.4. Nerve Excitability Assessments and Scores
Participants with diabetes had a significantly lower mean composite nerve excitability
score compared to people without diabetes (diabetes: 61.0 ± 6.3; people without
diabetes: 67.6 ± 5.0, p < 0.0001) indicative of impaired peripheral nerve function.
Receiver operating characteristic analysis established that a composite score above
65.6 (area under curve = 0.84) was indicative of normal nerve function (Figure. 3.1).
Time spent in hyperglycaemia or CONGA was associated with greater abnormalities
in the nerve excitability score (r = –0.670, p = 0.0012; r = –0.433, p = 0.036,
respectively) (Table 3.2). Conversely, longer time within target range correlated with
a better score (r = 0.672 p = 0.0011) (Table 3.2). These associations were controlled
for age, sex, diabetes duration, and HbA1c. Subsequent logistic regression revealed
that a greater percentage time in hyperglycaemia predicted an abnormal nerve
excitability score (OR: 1.11; 95% CI: 1.01–1.23, p = 0.037; beta coefficient, 0.106;
standard error, 0.051) however no such relationship for percentage time in
normoglycaemia or CONGA was observed. No correlation was found between the
demographic variables age, sex, BMI, diabetes duration, mode of insulin delivery,
lipids, or systolic blood pressure and composite nerve excitability scores. No
association was found between HbA1c and nerve excitability (r = –0.292, p = 0.147).
156 Chapter 3
1 True positive rate (sensitivity) rate positive True
0 0 1 False positive rate (1-specificity)
Figure 3.1. Receiver operating characteristic analysis of composite nerve excitability score obtained from participants with diabetes and people without diabetes. Area under the curve is equal to 0.84.
3.5. Corneal Confocal Microscopy
Participants with diabetes had a marked reduction in mean inferior whorl length
(diabetes: 12.4 ± 4.2 mm/mm2; people without diabetes: 18.2 ± 2.7 mm/mm2, p <
0.0001) and a greater number of micro-neuromas, which has been observed in instances of corneal nerve regeneration (diabetes: 4 (2–7); people without diabetes: 0
(0–2), p = 0.00011) (Figure. 3.2A and 3.2B) (Aggarwal et al., 2019). Partial correlations between corneal confocal microscopy and short-term measures of glucose control are listed in Table 3.2. A significant association was observed between
157 Chapter 3 inferior whorl corneal nerve length and CONGA (r = 0.483, p = 0.034), above range
(r = 0.591, p = 0.021), as well as within range (r = –0.542, p = 0.034) (Table 3.2).
Increases in glucose variability, as measured by CONGA, and time spent in hyperglycaemia were accompanied with a larger number of micro-neuromas (r =
0.433, p = 0.047; r = 0.516, p = 0.020, respectively). Less micro-neuromas were observed with greater time in target range (r = –0.545, p = 0.014). In the central cornea, marked reductions in mean corneal nerve fibre density (diabetes: 21.3 ± 7.4 no./mm2; people without diabetes: 30.3 ± 4.4 no./mm2, p < 0.0001) and length
(diabetes: 13.1 ± 3.3 mm/mm2; people without diabetes: 17.1 ± 1.9 mm/mm2; p <
0.0001) were observed in participants with diabetes, but these measures did not correlate with CONGA or percentage time in or above target range. No association was found between inferior whorl corneal nerve length and the demographic variables age, sex, BMI, diabetes duration, mode of insulin delivery, lipids and systolic blood pressure. No significant relationship between inferior whorl length and HbA1c was observed (r = –0.374, p = 0.153). Neuropathy severity was associated with reductions in central cornea nerve density (r = –0.729, p < 0.001) and length (r = –0.772, p <
0.001) as well a decreased inferior whorl corneal nerve length (r = –0.765, p < 0.001).
158 Chapter 3
Figure 3.2. Corneal confocal image highlighting a micro-neuroma of a corneal nerve fibre (A) and representative image of the inferior whorl from a participant with type 1 diabetes (B). Micro-neuromas (indicated by the red arrow) are characterised by irregularly shaped, terminal enlargements of nerve endings with variable hyper-reflectivity and poorly defined margins.
159 Chapter 3
4. Discussion
The current study has demonstrated that acute glucose variations affect peripheral nerve structure and function in type 1 diabetes. Given the current emphasis on the need to understand the relationship between measures of short-term glucose control and diabetic complications, we assessed how percentage time in, or above target range and glucose variability relate to peripheral nerve pathophysiology independent of age, sex, diabetes duration, and HbA1c. Despite being relatively asymptomatic, participants with diabetes demonstrated clear differences in peripheral nerve structure and function in comparison to people without diabetes. The findings of this study indicate that a greater percentage of time spent above target range as well as increased glucose variability, as measured by CONGA, correlate with worsening peripheral nerve function, as measured by a composite nerve excitability score reflective of overall ion channel function. In support of this finding, longer durations within target range were associated with improvements in nerve function. Logistic regression analysis indicated that single percentage increases in the percentage time above range increased the likelihood of abnormal nerve function by 11%. In regard to nerve morphology, corneal confocal microscopy demonstrated that increases in glucose variability and time above range were associated with structural alterations in the inferior whorl as well as an increased presence of micro-neuromas (enlarged nerve terminal spouts). As utilised in the current study, CONGA was designed specifically for continuous glucose monitoring and may be used as a marker of intra-day or inter- day glucose variability (Frontoni et al., 2013, Standl et al., 2011). Unlike other measures of glucose variability, CONGA does not require arbitrary glucose cut-offs or definitions of meaningful excursions, logarithmic transformations, selection of peaks or nadirs, or defined meal times, all which complicate interpretation
160 Chapter 3
(McDonnell et al., 2005). Neither structural nor functional measures correlated with
HbA1c, highlighting that short-term measures of glucose might have more clinical utility in the monitoring of neuropathy.
The importance of percentage time in range is an emerging concept in the study of diabetic complications and has not been examined in the context of diabetic neuropathy (Battelino et al., 2019). The results of the current study indicate that percentage time above target range is strongly associated with worsening peripheral nerve function and our findings also suggest that glucose variability may have an additional role. Nevertheless, the potentially damaging effect of glucose variability is consistent with previous studies, which have observed a relationship between increasing glucose variability, measured over 5 or 6 days, and nerve dysfunction as well as clinical signs of neuropathy in type 1 diabetes (Akaza et al., 2018, Kwai et al.,
2016b). Large fluctuations in glucose in type 1 diabetes are associated with oxidative stress and endothelial dysfunction, which may explain how glucose variability causes nerve injury and impair nerve function (Ceriello et al., 2012, Meng et al., 2015).
Interestingly, animal studies have observed that glucose fluctuation in hyperglycaemia disrupts peripheral nerve structure and nerve conduction velocities, increases pro-inflammatory cytokines, and reduces expression of anti-oxidants, more so than sustained hyperglycaemia alone after 4 months (Yang et al., 2019).
Human studies assessing the effect of acute glucose control on corneal nerve morphology are limited. It has previously been observed that higher glucose variability is associated with greater corneal nerve branching (Mahelkova et al.,
2018). Animal experiments have observed a trend between poor short-term glucose
161 Chapter 3 control and longer corneal nerve lengths (Yorek et al., 2014). In the current study, while corneal nerve length was significantly reduced in diabetes participants overall, increasing glucose variability and time above range correlated with greater nerve length in the inferior whorl. The inferior whorl is an area which contains the most distal aspect of corneal nerves and is highly sensitive to the effects of diabetes
(Kalteniece et al., 2018). Given the length-dependent nature of nerve injury in diabetes, it is understandable that corneal nerves in the inferior whorl may exhibit prompt alterations in morphology. It is possible this increase in corneal nerve length occurred as a regenerative response, given the positive association between glucose variability or time above range and the micro-neuroma count. In previous studies, micro-neuromas have been observed to increase following corneal nerve injury and in corneal neuropathic pain, with decreases noted after treatment (Aggarwal et al., 2019,
Cruzat et al., 2017). The finding in this study that micro-neuromas are associated with poorer measures of glucose control may indicate that a reduction in corneal nerves is a stimulus for regeneration and hence an indicator of prior nerve injury. Since corneal nerves are classified as small fibres and therefore lack the support and protection from myelin, they may be especially susceptible to nerve injury (Feldman et al., 2017).
Intraepidermal nerve fibres have also been observed to regenerate and increase in the number of distal swellings in small fibre neuropathy, which provides further support that this change may have been regenerative in nature (Cheng et al., 2013). It is not known if the corneal nerves that regenerate are indeed healthy or aberrant, although this may be answered from prospective studies assessing the relationship between micro-neuromas, corneal sensitivity, and corneal nerve tear neuromediators such as substance P.
162 Chapter 3
In the current study, no correlation between HbA1c and worsening nerve function or alterations in nerve morphology was observed. The lack of correlation between HbA1c and corneal confocal microscopy is consistent with other studies (Ferdousi et al.,
2020a). While HbA1c is recognised as the key clinical marker for the development of diabetes complications, it fails to account for individual patterns of glucose control, magnitude and frequency of glucose excursions, differences in various ethnic groups, and some comorbidities (Beck et al., 2017, Lu et al., 2018). Interestingly, greater variability of HbA1c itself has been shown to increase the risk of diabetic microvascular complications, except neuropathy (Virk et al., 2016). This further emphasises the requirement to find a better marker of glucose control in the study and monitoring of diabetic neuropathy.
Limitations of the current investigation include the cross-sectional design of the study, lack of an oxidative stress biomarker, pain assessment, and quantitative sensory testing. In particular, the relatively small sample size of the study is to be emphasised.
It was also not powered to assess differences between cohorts treated with insulin injections compared to those managed with insulin pumps, which has important cost and quality of life associations (Wan et al., 2018). Future studies may examine the association between time spent in hypoglycaemia and peripheral nerve structure and function, which could not be assessed in this study. Future investigations may also assess the effects of unblended or closed loop continuous glucose monitoring insulin delivery systems on microvascular complications (Bergenstal et al., 2011). In terms of the generalisability of our findings, our participants were recruited from a large metropolitan public clinic, and were similar for diabetes duration, HbA1c, and BMI to the cohort followed up in the Diabetes Control and Complications Trial (Diabetes
163 Chapter 3
Control and Complications Trial/Epidemiology of Diabetes Interventions and
Complications Study Research Group, 2016).
In conclusion, we have provided evidence that short-term measures of glucose control are associated with impaired nerve function and alterations nerve structure in type 1 diabetes. Future studies should examine the benefit of sustained improvement of acute glycaemia on the development and progression of peripheral neuropathy.
164 Chapter 4
Chapter 4 – Axonal pathophysiology in Latent Autoimmune Diabetes of Adulthood
165 Chapter 4
Summary and Link to Thesis
In Chapter 3, I explored the association between acute glucose control and axonal structure and function in type 1 diabetes. It was found that poor acute glucose control was associated with impaired peripheral nerve function and altered corneal nerve morphology. In the following chapter, I will investigate the pathophysiological mechanisms underlying peripheral nerve dysfunction in another type of autoimmune diabetes, latent autoimmune diabetes of adulthood. Nerve ultrasonography and axonal excitability studies were undertaken in patients with latent autoimmune diabetes of adulthood and findings were compared to patients with type 1 or type 2 diabetes. It was established that patients with latent autoimmune diabetes of adulthood exhibited more severe changes in nerve ultrasound and axonal excitability outcome measures when compared to type 1 or type 2 diabetes cohorts. It was also revealed that the basis of nerve dysfunction in latent autoimmune diabetes of adulthood, as indicated by mathematical modelling of nerve excitability recordings, is different to that observed in type 1 and type 2 diabetes.
This work has been published:
Issar, T., Yan, A., Kwai, N. C. G., Poynten, A. M., Borire, A. A., Arnold, R., &
Krishnan, A. V. (2020). Altered peripheral nerve structure and function in
latent autoimmune diabetes in adults. Diabetes/Metabolism Research and
Reviews, 36(3), e3260. doi: 10.1002/dmrr.3260
TI was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition.
166 Chapter 4
Abstract
Objective: The present study was undertaken to investigate mechanisms of peripheral nerve dysfunction in latent autoimmune diabetes in adults (LADA)
Methods: Participants with LADA (n = 15) underwent median nerve ultrasonography and nerve excitability to examine axonal structure and function, in comparison to cohorts of type 1 diabetes (n = 15), type 2 diabetes (n = 23) and healthy controls (n =
26). The LADA group was matched for diabetes duration, glycaemic control, and clinical neuropathy severity with the type 1 and type 2 diabetes groups. A validated mathematical model of the human axon was utilised to investigate the pathophysiological basis of nerve dysfunction.
Results: The most severe changes in nerve structure and function were noted in the
LADA group. The LADA cohort demonstrated a significant increase in nerve cross- sectional area compared to type 1 participants and controls. Compared to type 1 and 2 diabetes, measures of threshold electrotonus, which assesses nodal and internodal conductances, were significantly worse in LADA in response to both depolarising currents and hyperpolarising currents. In the recovery cycle, participants with LADA had a significant increase in the relative refractory period. Mathematical modelling of excitability recordings indicated the basis of nerve dysfunction in LADA was different to type 1 and 2 diabetes.
Conclusions: Participants with LADA exhibited more severe changes in nerve excitability measures and different underlying pathophysiological mechanisms compared to participants with type 1 or 2 diabetes. Intensive management of risk factors to delay the progression of neuropathy in LADA may be required.
167 Chapter 4
1. Introduction
Latent autoimmune diabetes in adults (LADA) is a subtype of autoimmune diabetes
(Redondo, 2013). Autoimmune diabetes is characterised by the presence of one or more-islet specific autoantibodies that has a causal role in the destruction of pancreatic islet cells (Atkinson and Maclaren, 1994). These include islet cell autoantibodies and autoantibodies targeting three major islet autoantigens: glutamic acid decarboxylase (GAD), protein tyrosine phosphatase, and insulin (Guglielmi et al., 2012). LADA is similar in clinical presentation to type 1 diabetes but manifests with a later onset in life and a slower progression towards insulin dependence
(Redondo, 2013). LADA also shares metabolic and genetic features with type 2 diabetes, although recent evidence suggests LADA is genetically closer to type 1 diabetes (Mishra et al., 2017, Pozzilli and Pieralice, 2018, Redondo, 2013).
Diabetic peripheral neuropathy is the most common microvascular complication of diabetes. The present study was undertaken to investigate the pathophysiology of peripheral nerve disease in LADA compared to type 1 and 2 diabetes. Nerve function was assessed using nerve excitability techniques, which provides insight into the function of the ion channels, pumps, and exchangers of the axonal membrane and have previously demonstrated changes in both type 1 and type 2 diabetes (Arnold et al., 2013a, Kwai et al., 2015, Kwai et al., 2016a, Kwai et al., 2013, Sung et al., 2012,
Sung et al., 2017). A healthy control cohort was also recruited to examine underlying causes of nerve dysfunction in LADA. To accompany nerve electrophysiology, nerve structure was investigated using high-resolution ultrasound to assess cross-sectional fascicular area (Borire et al., 2018b).
168 Chapter 4
2. Methods
2.1. Participants
This study was approved by the South East Sydney Area Health Service Human
Research Ethics Committee (Northern Section) and the Human Research Ethics
Committee of the University of New South Wales. All participants enrolled provided written informed consent to the procedures in accordance with the Declaration of
Helsinki. Thirty individuals with established autoimmune diabetes of at least seven years were consecutively recruited from the Diabetes Centre of Prince of Wales
Hospital in Sydney, Australia. Participants were grouped into having LADA (n = 15,
12 men and 3 women) or type 1 diabetes (n = 15, 7 males and 8 females).
Subgrouping of participants into LADA was based on the criteria suggested by the
Immunology of Diabetes Society: age of onset after 30 years of age, the presence of at least one circulating autoantibody mentioned above, and insulin independence for the first 6 months after disease onset, as seen in other studies of LADA (Alam et al.,
2019, Lu et al., 2015). A type 2 diabetes cohort (n = 23, 11 males, 12 females) and a cohort of age-matched healthy control participants (n = 26, 11 males, 15 females) were also recruited. Required cohort sizes were determined from changes in nerve excitability parameters observed in previous studies of participants with diabetes and were calculated at 80% power, with an a = 0.05 (Krishnan and Kiernan, 2005).
Individuals were excluded from the study if they had any of the following: carpal tunnel syndrome, peripheral oedema, a history of neurotoxic treatment, current treatment with neuropathic pain agents, chronic kidney disease, any other condition known to cause neuropathy, or if they had a neuromuscular, movement, psychiatric or developmental disorder. Participant demographic data and medical history of interest included age, sex, diabetes duration, mode of insulin delivery (multiple daily insulin
169 Chapter 4
injects or continuous subcutaneous insulin infusion), HbA1c, lipid profile, and BMI.
The LADA cohort was matched to the type 1 and type 2 diabetes groups for diabetes duration, HbA1c and neuropathy severity, which was determined using the Total
Neuropathy Score (TNS) (Cornblath et al., 1999). The TNS is comprised of sub- scores in eight domains: sensory and motor symptoms, sensation to vibration (128 Hz tuning fork), and pinprick (NeurotipTM, Owen Mumford, Oxford, United Kingdom), tendon reflexes, strength, and sural (sensory) and tibial (motor) nerve conduction studies (Natus, Middleton, USA), as per standard protocols (Liveson and Ma, 1992).
Scores range from 0–32, with a higher score indicating more severe neuropathy and zero indicating an absence of neuropathy.
2.2. Nerve Ultrasound
All participants underwent ultrasonography of the median nerve at a non-entrapment site using a 10–18 MHz linear probe. The machine (MyLabTMOne, Esaote, Genoa,
Italy) was operated under the ‘Musculoskeletal’ factory preset (acoustic power 100%, line density set at medium, dynamic range set at 14, and persistence set at 1) and other settings such as depth, gain, and focus were kept constant for each examination. The median nerve was chosen due to the high reproducibility of the cross-sectional area
(CSA) measurement that can be obtained (Borire et al., 2018b, Yan et al., 2020).
Participants were evaluated while sitting comfortably with their forearm fully supinated and fingers semi-extended. The forearm was supported on an armrest to ensure the elbow was flexed at 90°. The median nerve was first identified in the transverse plane at the carpal tunnel inlet at the level of the pisiform bone and then traced proximally between the superficial (flexor digitorum superficialis) and deep
(flexor pollicis longus and flexor digitorum profundus) muscles until the junction of
170 Chapter 4 the middle and distal third of the forearm. The median nerve CSA was then measured on the screen (in mm2) with a stylus using the continuous trace method by outlining the inner margin of the epineurium (Cartwright et al., 2012). Each CSA measurement was completed by two observers and then averaged.
2.3. Nerve excitability and Mathematical Modelling
Motor nerve excitability assessment was then conducted on the median nerve for each participant using the TROND protocol and Qtrac software (Digitimer, London,
United Kingdom). The median nerve was stimulated through non-polarisable surface electrodes (Ambu, Sydney, Australia), proximal to the wrist at the site of least resistance using a DS5 Isolated Bipolar Current Stimulator (Digitimer, London,
United Kingdom). Skin temperature at the point of stimulation was maintained above
32°C (Kiernan et al., 2001). Compound muscle action potentials (CMAP) were recorded from the abductor pollicis brevis muscle. The assessment consisted of four different testing paradigms, each providing indirect information regarding current flow (conductance) through voltage-gated sodium (Na+) and potassium (K+) ion channels as well as an indication of the activity of energy-dependent pumps (Na+/K+-
ATPase) and exchangers embedded within the axon membrane at the point of stimulation. Stimulus-response curves were first generated using a 1 ms test pulse to obtain the maximal CMAP amplitude and a target response of 40% of maximum was calculated. The current required to elicit this target response (threshold) was tracked in the four paradigms: strength–duration behaviour, threshold electrotonus, current– threshold relationship, and recovery cycle.
1. Strength–duration behaviour was examined by plotting the relationship
between stimulus intensity required to reach threshold from four different
171 Chapter 4
stimulus durations (0.2, 0.4, 0.8, and 1.0 ms). Weiss’ Law was applied to
calculate the strength–duration time constant, which is a marker of persistent
Na+ conductance in the Node of Ranvier (Burke et al., 2001).
2. Threshold electrotonus assesses the change in threshold after depolarising
(stimulating) or hyperpolarising (inhibiting) conditioning currents. These
changes were determined after a 1 ms test pulses was applied during or after
100 ms subthreshold conditioning currents of +40% (depolarising) or –40%
(hyperpolarising) of control threshold, which was determined from the initial
stimulus-response curve. Percentage change in threshold was plotted at 10 ms
intervals and key excitability measures (depolarising threshold change
between 10–20 ms, S2 accommodation, hyperpolarising threshold change
between 10–20 ms, 20–40 ms, and 90–100 ms) were determined. S2
accommodation is the specific phase of depolarising threshold electrotonus in
which threshold reduction is limited and begins to return to control level
(Burke et al., 2001). Threshold electrotonus examines various conductances at
the node and internode (the region of axonal membrane underneath the myelin
sheath) that contribute to the excitability of peripheral nerves (Bostock et al.,
1998).
3. The current–threshold relationship assesses rectification properties of the
internode in response to long lasting polarising currents (Bostock et al., 1998).
This relationship was determined by plotting the percentage threshold change
of 1 ms test pulses following 200 ms polarising currents ranging from +50%
(depolarising) to –100% (hyperpolarising) of the control threshold.
4. The recovery cycle studies the changes in threshold during the recovery
period after a nerve impulse. Percentage change in threshold was plotted over
172 Chapter 4
a range of conditioning–test intervals following a 1 ms supramaximal
conditioning stimulus (Burke et al., 2001). The relative refractory period is the
initial period of the recovery cycle in which there is decreased excitability
following the inactivation of nodal Na+ channels after a supramaximal
conditional stimulus. Superexcitability is the subsequent phase in which there
is increased excitability due to a depolarising afterpotential (Burke et al.,
2001). Subexcitability is the final period of the recovery cycle in which there
is decreased excitability due to slowly activating K+ channels located at the
node (Bostock et al., 1998).
To investigate the pathological basis of nerve dysfunction in each type of diabetes, nerve excitability recordings obtained from the diabetes groups were further analysed using the Bostock model of axonal excitability (Bostock et al., 1991). The model assists in the interpretation of nerve excitability findings by providing an indication of the underlying changes in the axonal membrane in the disease state. This includes changes in maximum conductance of different types of Na+ and K+ ion channels as well as the sodium-potassium pump. The model was adjusted to fit the mean excitability data of the healthy control participants before fitting the mean excitability data of the LADA, type 1 diabetes, and type 2 diabetes groups. Modelling analyses involves iterative changes in a single or combination of parameters in the control group to reproduce the mean recorded data of each disease group as closely as possible. This simulated excitability data is objectively fit using a least squares approach, in which the overall discrepancy was assessed and minimised using the weighted sum of the squares of the error terms between the control and disease group data. Persistent and transient Na+ conductances were modelled only at the node while
173 Chapter 4 fast and slow K+ conductances were modelled at the node and internode. Current from
Na+/K+–ATPase activity was modelled in both axonal compartments. The Barrett-
Barrett conductance, which represents current flow through and underneath the myelin sheath between the node and internode, was also investigated.
2.4. Statistical analyses
All results were analysed using IBM SPSS Statistics version 25.0 for Windows (IBM
Corp, Armonk, NY, USA). Shapiro–Wilk tests were undertaken to determine the normality of data. Where appropriate and with post-hoc corrections if applicable, an analysis of variance, Kruskal-Wallis test, independent t-tests, Mann–Whitney U tests or Pearson chi-square analyses were applied to compare means of demographic data, extracted excitability parameters, and median nerve CSA between groups. Similarly, either Pearson or Spearman bivariate correlations were applied to determine relationships between demographic, excitability, and ultrasound data. Linear regression analyses were undertaken to investigate the effects of demographic variables on nerve excitability and ultrasound measures. Statistical significance was considered when P < 0.05. Normally and non-normally distributed clinical data is expressed as mean ± standard deviation and median (quartile 1–quartile 3) (with the exception of TNS), respectively. All nerve excitability data is expressed as mean ± standard error. Figures were generated using GraphPad Prism version 7.00 for
Windows (GraphPad Software, La Jolla, CA, USA).
174 Chapter 4
3. Results
3.1. Participant Demographics
Participant demographic data is summarised in Table 4.1. Groups were matched to ensure that there were no differences in glycaemic control, diabetes duration, baseline clinical neuropathy severity, and male to female ratio between the LADA participants and type 1 or type 2 cohorts. As expected, the LADA and Type 1 cohorts differed on age (P < 0.001), and the LADA group had a lower BMI compared to the participants with type 2 diabetes (P = 0.003).
175 Chapter 4
Table 4.1. Participant Characteristics
P value LADA T1DM T2DM Control LADA LADA LADA (n = 15) (n = 15) (n = 23) (n = 26) vs. T1DM vs. T2DM vs. Control
Age (years) 58 ± 15 40 ± 8 60 ± 9 57 ± 8 <0.001 0.593 0.710
Sex (M:F) 12:3 7:8 11:12 11:15 0.128 0.088 0.030
Diabetes duration (years) 16 ± 8 21 ± 10 13 ± 6 0.300 0.838 -
HbA1c (%) 8.7 ± 1.5 8.4 ± 1.8 8.6 ± 1.4. 0.555 0.704 -
HbA1c (mmol/mol) 72 ± 16 69 ± 20 70 ± 15 0.555 0.704 -
Insulin (%; mode; 100 100 61 MDII, CSII) - - - (14 MDII, 1 CSII) (8 MDII, 7 CSII) (all MDII)
LDL (mmol/L) 2.0 (1.5–2.5) 2.5 (1.9–3.3) 2.8 (2.0–3.5) 0.186 0.043 -
HDL (mmol/L) 1.6 (1.4–2.0) 1.4 (1.1–2.0) 1.2 (1.1–1.5) 0.270 0.001 -
Triglycerides (mmol/L) 0.9 (0.7–1.1) 1.1 (0.7–1.6) 1.4 (1.1–2.6) 0.418 0.001 -
BMI (kg/m2) 25 (22–26) 26 (22–30) 28 (25–35) 25 (22–26) 1 0.003 1
TNS 4 ± 5 3 ± 4 4 ± 3 0 0.447 0.769 <0.001
Sural SNAP (µV) 11.8 ± 3.5 12.5 ± 1.4 12.7 ± 1.3 16.2 ± 2.7 0.824 0.759 0.018
Sural NCV (m/s) 43.3 ± 8.8 42.7 ± 6.3 45.2 ± 8.0 47.1 ± 5.5 0.830 0.521 0.180
Tibial CMAP (mV) 8.7 ± 1.8 9.7 ± 1.1 7.2 ± 0.7 12.6 ± 2.0 0.539 0.380 0.015
Tibial latency (ms) 4.1 (3.6–4.5) 3.7 (3.1–4.2) 4.0 (3.5–4.7) 3.6 (3.2–3.9) 0.134 0.769 0.110
Normally distributed data is expressed as mean ± SD while non-normally distributed data is expressed as median and quartile 1 to quartile 3 (with the exception of TNS data). T1DM, type 1
diabetes; T2DM, type 2 diabetes; HbA1c, glycated haemoglobin; MDII, multiple daily insulin injections; CSII, continuous subcutaneous insulin infusion; LDL, low-density lipoprotein; HDL, high-density lipoprotein; BMI, body mass index; TNS, total neuropathy score; SNAP, sensory nerve action potential; CMAP, compound muscle action potential. Normally distributed data was compared using a one-way ANOVA (with Bonferroni corrections) while non-normally distributed data was compared using a Kruskal-Wallis test (with Dunn-Bonferroni corrections).
176 Chapter 4
3.2. Nerve Excitability and Ultrasound Between Groups
There were significant changes noted in nerve cross-sectional area in LADA participants (Figure. 4.1). Median nerve CSA was significantly larger in the LADA group (8.18 ± 1.41 mm2) compared to healthy controls (7.20 ± 1.33 mm2, P < 0.05).
A significant difference in median CSA was also found between LADA and type 1 diabetes (7.10 ± 1.03 mm2) groups, with greater nerve size noted in LADA participants (P < 0.05). In contrast, no significant differences in median nerve CSA were observed between type 1 or type 2 diabetes (7.55 ± 1.61 mm2) and the control group (P = 0.368 and P = 0.759, respectively).
Figure 4.1. Median nerve cross-sectional area measurements. Cross-sectional area of the median nerve (outlined) at the junction between the middle and distal third of the forearm. (A) Sonograph from a LADA participant (10.70 mm2) compared to (B) age-matched healthy control participant (6.18 mm2).
The greater changes in nerve cross-sectional area in LADA were accompanied by more severe changes in nerve function, assessed using nerve excitability techniques.
Prominent abnormalities in nerve excitability parameters from the threshold electrotonus and recovery cycle paradigms were observed in the LADA cohort
177 Chapter 4 compared to the control group (Figure. 4.2 and Figure. 4.3). In threshold electrotonus, the percentage threshold change was significantly reduced in response to depolarising currents (10–20 ms and S2 accommodation, both P < 0.01) and increased with hyperpolarising currents (10–20 ms, P < 0.01; 20–40 ms, P < 0.01; 90–100 ms, P <
0.05). In the recovery cycle, a significant increase in the relative refractory period and superexcitability (LADA: –19.1 ± 1.8; control: –24.0 ± 1.1) as well as a decrease in subexcitability (LADA: 11.4 ± 1.3; control: 14.6 ± 0.7) was observed in individuals with LADA (P < 0.001, P < 0.05, and P < 0.001 respectively).
178 Chapter 4
(A) (B) ** ** 90 40 * * * P = 0.184 80 30
70 20
60 10 S2 Accommodation S2 (% change in threshold) 10–20 ms (depolarizing) 50 (% change in threshold) 0
LADA Type 2 Type 1 Control LADA Type 2 Type 1 Control
(C) ** (D) * ** -50 -40 * * * -60 -60
-70 -80 -80 -100 -90 -120 (% change in threshold) -100 (% change in threshold) 10–20 ms (hyperpolarizing) 20–40 ms (hyperpolarizing)
LADA Type 2 Type 1 Control LADA Type 2 Type 1 Control
(E) (F) * 0 *** * 8 P = 0.115 ** *** -50 6
-100 4
-150 Period (ms) 2 Relative Refractory
(% change in threshold) -200
90–100 ms (hyperpolarizing) 0
LADA Type 2 Type 1 Control LADA Type 2 Type 1 Control
Figure 4.2. Group comparison of nerve excitability parameters. Group comparison of nerve excitability measures highlighting that participants with LADA demonstrated the greatest abnormalities in axonal properties compared to participants with other forms of diabetes and healthy controls. (A–E) are parameters of the threshold electrotonus paradigm and (F) is a parameter of the recovery cycle. All values given as mean (± SEM). By convention, variables of threshold electronus and recovery cycle (except relative refractory period) are expressed as percentage change in threshold. Type 1, type 1 diabetes; Type 2, type 2 diabetes. TEd, depolarising threshold electrotonus; TEh, hyperpolarising threshold electrotonus. Normally distributed data was compared using an Independent t-test while non-normally distributed data was compared using a Mann-Whitney U test. *P < 0.05, **P < 0.01, ***P < 0.001
179 Chapter 4
S2 accommodation A 80 B
70 60 Depolarising threshold electrotonus LADA 40 Control 10–20 ms 50 20 Type 1 DM Type 2 DM
0 30 Subexcitability
-20 Relative 10–20 ms refractory period -40 10 Threshold change (%)
Threshold reduction (%) -60 20–40 ms 90–100 ms -80 Hyperpolarising -10 threshold electrotonus
-100 Superexcitability
-120 -30 0 100 200 2 20 200 Delay (ms) Interstimulus Interval (ms) Figure 4.3. Group comparison of nerve excitability recordings. Mean nerve excitability data for participants with LADA, type 1 diabetes, or type 2 diabetes and healthy controls for threshold electrotonus (A) and recovery cycle (B) testing paradigms. Type 1 DM, type 1 diabetes; Type 2 DM, type 2 diabetes.
When compared to type 1 and type 2 diabetes cohorts, LADA participants again
demonstrated more severe changes in neurophysiological parameters (Figure. 4.2 and
Figure. 4.3). Compared to type 1 and type 2 diabetes, threshold change at various time
points in threshold electrotonus was significantly reduced after depolarising currents
(LADA vs. type 1: 10–20 ms and S2 accommodation, both P < 0.05; LADA vs. type
2: 10–20 ms, P < 0.05, S2 accommodation, P = 0.184). Threshold change after
hyperpolarising currents was increased compared to type 1 and type 2 diabetes
(LADA vs. type 1: 10–20 ms, 20–40 ms, 90–100 ms, all P < 0.05; LADA vs. type 2:
10–20 ms and 20–40 ms, P < 0.05, 90–100 ms, P = 0.115). In the recovery cycle, a
significant increase in the relative refractory period was observed in individuals with
LADA (vs. type 1: P < 0.01; vs. type 2: P < 0.001) but no differences were observed
in superexcitability (LADA: –19.1 ± 1.8, type 1: –22.8 ± 1.9, P = 0.168; type 2: –21.0
180 Chapter 4
± 1.1, P = 0.342) or subexcitability (LADA: 11.4 ± 1.3, type 1: 10.4 ± 0.8, P = 0.901; type 2: 12.0 ± 0.7, P = 0.132).
An increase in median nerve CSA was associated with more severe changes in nerve function in LADA and type 1 diabetes. Enlargement of CSA significantly correlated with depolarising threshold electrotonus at 10–20 ms (LADA: r = –0.461, type 1: r =
–0.522, P < 0.05 for both), 40–60 ms (LADA: r = –0.572, P < 0.05; type 1: r = –
0.787, P < 0.01), peak (LADA: r = –0.565, type 1: r = –0.638, P < 0.05 for both), 90–
100 ms (LADA: r = –0.481, P < 0.05; type 1: r = –0.627, P < 0.01). In contrast, no correlation was found between structural and functional measures in type 2 diabetes.
Linear regression analyses determined that age or BMI had no effect on nerve excitability measures or median nerve CSA. Further, mode by which insulin was delivered had no effect on nerve excitability and ultrasound outcomes.
3.3. Mathematical Modelling
Mathematical modelling indicated that the pathological changes in nerve function observed the LADA group were due to different mechanisms than those of type 1 and type 2 diabetes (Table 4.2). Nerve excitability recordings from the LADA cohort were best accounted for by relatively large alterations in two conductances. Specifically, a combination of a 62.8% reduction in the percentage of nodal persistent Na+ channels and a 62.3% reduction of internodal fast K+ conductance could account for 78% of the discrepancy. In contrast, modelling of excitability findings between the type 1 or type
2 and control data indicated that a different pattern of changes reduced the discrepancy between these diabetes cohorts and control recordings. In these cohorts, a combination of reduced nodal Na+ permeability and conductance though nodal and
181 Chapter 4 internodal K+ channels was observed. A decrease in the Barrett-Barrett conductance was also noted, which reflects a decrease of current through and underneath the myelin sheath between the node and internode.
Table 4.2. Modelled parameters for LADA, type 1 diabetes, type 2 diabetes and control cohorts Parameter LADA Type 1 DM Type 2 DM Control Permeability of Na+ channels at node 3.5 3.6 4.1 (cm3s-1 x 10–9) % of Na+ channels that are persistent 0.29 0.78 0.78 0.78 (%) Max conductance of nodal slow K+ 35 37.6 43.1 channels (nS) Max conductance of internodal 0.3 0.29 0.3 slow K+ channels (nS) Max conductance of nodal fast K+ 17.4 17.4 17.4 channels (nS) Max conductance of internodal 46 105 112 122 fast K+ channels (nS) Barrett-Barrett conductance 33.7 32.4 35.2 (nS) Na+/K+-ATPase pump current 0.8 2.6 0 (pA) Key parameter changes applied in the mathematical model for diabetes and control groups. Conductance values are expressed in nanosiemens (nS) and pump current is expressed as picoamperes (pA). The Barrett-Barrett conductance represents current flow through and underneath the myelin sheath between the nodal and internodal compartments of the axon. Type 1 DM, type 1 diabetes; Type 2 DM, type 2 diabetes
4. Discussion
The present study has demonstrated that LADA participants have more severe changes in nerve structure and function, when compared to groups of participants with type 1 and type 2 diabetes matched for diabetes duration, glycaemic control, and neuropathy severity. Nerve ultrasonography revealed that participants with LADA had a significant increase in nerve CSA compared to type 1 diabetes participants and the control group. Although diabetic neuropathy commences with sensory symptoms,
182 Chapter 4 motor nerve excitability was assessed due to its previously demonstrated utility as an early marker of nerve dysfunction in diabetes (Arnold et al., 2013a, Krishnan and
Kiernan, 2005, Krishnan et al., 2008, Kwai et al., 2016a). Furthermore, previous studies have shown that changes in motor nerve excitability correlate with sensory symptoms and neuropathy-related quality of life in people with diabetes (Erdoğan et al., 2011, Kwai et al., 2013). Motor nerve excitability demonstrated more severe abnormalities in the LADA cohort compared to both type 1 and type 2 diabetes cohorts. A correlation between nerve cross-sectional area and function was found in
LADA and type 1 diabetes (Borire et al., 2018b). Limitations of the present study include cross-sectional nature of the investigation (especially considering that nerve excitability testing only reflects peripheral nerve properties at the site of stimulation at a point in time), the age difference between the LADA and type 1 groups, the lack of an insulin sensitivity assessment such as a hyperinsulinemic-euglycemic clamp, and the small sample size, which may have concealed differences in diabetes duration between groups.
Mathematical modelling of the nerve excitability results indicated that the more severe changes in nerve function observed in LADA were due to distinct and relatively large alterations nodal Na+ and internodal K+ conductances. In contrast, it was found that the functional changes observed in the type 1 and type 2 diabetes cohorts were due to a different combination of reduced conductances (Brismar et al.,
1987, Hong and Wiley, 2006). It is hypothesised that the mechanisms underlying nerve dysfunction in LADA may result in an altered progression of neuropathy in comparison to type 1 and type 2 diabetes. As the LADA group was matched for diabetes duration, glycaemic control and neuropathy severity with the type 1 and type
183 Chapter 4
2 cohorts, it is unlikely that the greater severity of nerve dysfunction observed in
LADA is simply a function of metabolic or clinical features. It is more likely that other factors such as immune-mediated injury and an increase in insulin resistance within the nervous system are involved in the accelerated progression of nerve dysfunction.
Investigations of diabetic peripheral neuropathy in LADA are limited in number.
Studies have reported a similar prevalence of neuropathy between LADA and type 2 diabetes but significantly higher rates in LADA than type 1 diabetes despite a shared autoimmune basis (Arikan et al., 2005, Isomaa et al., 1999, Wang et al., 2015). The findings of the latter studies are consistent with the patterns of change in nerve pathophysiology observed in the current investigation. In keeping with our findings that patients with LADA demonstrated the greatest nerve dysfunction, recent results have also highlighted there is greater small nerve fibre pathology in LADA compared to type 2 diabetes (Alam et al., 2019). No differences in nerve conduction parameters between the LADA and type 2 diabetes groups were observed, which is similar to
Wang et al. (2015) when diabetes duration of the cohorts is considered (Wang et al.,
2015). Of note, a correlation between nerve structure and function has previously been observed in type 1, but not type 2, diabetes (Borire et al., 2018b). This is consistent with the findings of the current study as median nerve enlargement correlated with greater dysfunction in the autoimmune diabetes cohorts, suggesting similar mechanisms are involved in the disease pathophysiology. While previous studies have shown that an increase in the relative refractory period occurs with longer disease duration in type 1 diabetes patients without neuropathy, no such
184 Chapter 4 relationship was observed in any of the diabetes groups in the current study, all of which had established neuropathy (Arnold et al., 2013a).
The indication that the changes in nerve function observed in type 1 and type 2 diabetes were due to a variety of altered nodal and internodal conductances is consistent with previous modelling investigations of nerve dysfunction in preclinical diabetic neuropathy (Kwai et al., 2016a). Some differences in specific conductances are present between the studies, however it should be noted participants in the current investigation had established neuropathy. In contrast, the finding that nerve dysfunction in LADA is more severe and occurs due to relatively large reductions in
Na+ and K+ conductances is indicative of advanced pathological changes in these participants. A combination of immune-mediated mechanisms of nerve injury in autoimmune diabetes and an increase in insulin resistance within the nervous system may underlie these observations. Small studies have shown an association between anti-GAD antibodies and more severe changes in peripheral nerve function in type 1 diabetes (Hoeldtke et al., 2000, Louraki et al., 2016). Other studies have observed raised antinuclear antibodies and a presence of anti-insulin antibodies in individuals with type 1 and 2 diabetes with nerve dysfunction (Janahi et al., 2015, Morano et al.,
1999). Secondly, insulin resistance within the peripheral nervous system has been reported in animal models of type 2 diabetes (Kim and Feldman, 2012). Insulin is a potent neurotrophic factor that supports axonal growth as well as Schwann cell physiology and receptors for insulin are abundantly expressed on peripheral nerves, at the node of Ranvier, and on Schwann cell membranes (Grote and Wright, 2016,
Rachana et al., 2018, Waldbillig and LeRoith, 1987). Schwann cell dysfunction has recently been recognised as an integral factor in the development of diabetic
185 Chapter 4 neuropathy as these cells envelope the internode and export metabolites necessary for energy production within the axon (Goncalves et al., 2017). Interestingly, the degree of systemic insulin resistance in people with LADA has been found to be similar to that of individuals with type 2 diabetes and long-term type 1 diabetes (Behme et al.,
2003, Chiu et al., 2007). Earlier development of insulin resistance in the peripheral nervous system (tissue-specific) would result in an inability to respond to the neurotrophic properties of insulin and impaired Schwann cell function, which may amount to increased peripheral nerve injury. Taken together, these factors may account for the severe changes in nerve function observed in the LADA cohort.
The implication of this study is that there is evidence to suggest that the pathophysiological mechanisms underlying nerve dysfunction in LADA are different to those of type 1 or 2 diabetes. Frequent assessment of neuropathy biomarkers and more rigorous monitoring may be required for people with LADA. Longitudinal studies examining epidemiological and clinical variables contributing to complication development in LADA cohorts would also be beneficial in providing further pathological insights.
186 Chapter 5
Chapter 5 – Pathophysiological mechanisms underlying altered axonal function and structure in type 2 diabetes with metabolic syndrome
187 Chapter 5
Summary and Link to Thesis
In the previous chapter, I investigated the pathophysiological mechanisms underlying peripheral nerve dysfunction in latent autoimmune diabetes of adulthood. It was found that these patients have more severe changes in nerve ultrasound and axonal excitability measures compared to patients with type 1 or type 2 diabetes. Of particular importance, it was established that ion channel dysfunction in latent autoimmune diabetes of adulthood was different to that observed in type 1 or type 2 diabetes. The following chapter will explore the pathophysiological mechanisms underlying axonal dysfunction in type 2 diabetes with and without the metabolic syndrome. In the Literature Review, I discussed how the metabolic syndrome may promote the onset and progression of peripheral neuropathy. In the following study, it was determined that patients with type 2 diabetes and the metabolic syndrome demonstrate more severe changes in nerve ultrasound, corneal confocal microscopy, and axonal excitability when compared to patients with type 2 diabetes alone.
Mathematical modelling of axonal excitability recordings indicated that the more severe changes in nerve function was due to a combined reduction in nodal sodium channel permeability as well a reduction in the activity of the Na+/K+ pump. In contrast, in type 2 diabetes alone, only nodal sodium permeability was reduced.
This work has been submitted to the European Journal of Neurology and is under review. TI was responsible for the study design, recruitment, data collection (with exception of in-vivo corneal confocal microscopy – SST), data interpretation, and the manuscript composition.
188 Chapter 5
Abstract
Objective: There is a strong association between the metabolic syndrome in diabetes and the development of peripheral neuropathy, however the pathophysiological mechanisms remain unknown.
Methods: Participants with type 2 diabetes and metabolic syndrome (T2DM/MetS, n
= 89) and type 2 diabetes alone (T2DM, n = 59) underwent median nerve ultrasound and excitability studies to assess peripheral nerve structure and function. A subset of
T2DM/MetS (n = 24) and T2DM (n = 22) participants underwent confocal microscopy to assess central and inferior whorl corneal nerve structure. Clinical neuropathy severity was assessed using the modified Toronto Clinical Neuropathy
Score (mTCNS). Diabetes groups were similar for age, sex distribution, diabetes duration, HbA1c, and insulin treatment. 60 healthy controls similar for age and sex distribution were recruited for comparison.
Results: Participants with T2DM/MetS manifested with a greater mTCNS compared to T2DM alone (p < 0.05). Median nerve cross-sectional area was larger in the
T2DM/MetS group compared to the T2DM cohort (p < 0.05). Participants with
T2DM/MetS had reductions in both central (all p < 0.01) and inferior whorl (all p <
0.05) nerve measures. Compared to T2DM, the T2DM/MetS group demonstrated more severe changes in nerve excitability measures, which was due to reduced sodium channel permeability and sodium-potassium pump function. In comparison, only sodium channel permeability was reduced in T2DM.
189 Chapter 5
Conclusions: Compared to participants with type 2 diabetes alone, those with diabetes and metabolic syndrome manifested greater alterations in peripheral nerve structure and function, which may be due to reduced function of the sodium-potassium pump.
190 Chapter 5
1. Introduction
Peripheral neuropathy is a common complication of diabetes and is a major cause of disability (Feldman et al., 2017). The condition is characterised by sensory loss, pain, weakness, ulceration, and amputation. Multiple studies have reported an association between the metabolic syndrome and neuropathy in type 2 diabetes and pre-diabetes
(Callaghan et al., 2018, Callaghan et al., 2016a, Callaghan et al., 2016b, Hanewinckel et al., 2016a, Smith and Singleton, 2013, Stino and Smith, 2017).
A diagnosis of metabolic syndrome is based on three of any of the following criteria: elevated waist circumference, hypertriglyceridemia, reduced HDL cholesterol, hypertension, and elevated fasting glucose (Grundy et al., 2005). Obesity, in particular, has been recognised as the key component associated with neuropathy
(Callaghan et al., 2018, Callaghan et al., 2016a, Callaghan et al., 2016b, Hanewinckel et al., 2016a). In patients with type 2 diabetes, risk and prevalence of neuropathy has been observed to be greater as the number of metabolic syndrome factors increases, however the effect of each component contributing to this association remain unclear
(Callaghan et al., 2016a).
The present study was undertaken to investigate the pathophysiological mechanisms underlying peripheral neuropathy in type 2 diabetes with metabolic syndrome.
Mechanisms of peripheral nerve function were examined using nerve excitability studies, which provide insight into the function of ion channels and the sodium- potassium pump embedded in the axonal membrane of peripheral nerves. In addition, nerve morphology was assessed using nerve ultrasound and corneal confocal microscopy.
191 Chapter 5
2. Methods
2.1. Subjects
This study was approved by the South East Sydney Area Health Service Human
Research Ethics Committee and the Human Research Ethics Committee (HREC) of the University of New South Wales (HREC number: 14/012). All subjects enrolled provided written informed consent in accordance with the Declaration of Helsinki. A total of 148 individuals with type 2 diabetes were consecutively recruited from the
Diabetes Centre of Prince of Wales Hospital in Sydney, Australia. Demographic data of interest included age, sex, diabetes duration, insulin treatment, HbA1c, lipid profile, blood pressure, and waist circumference. Waist circumference was estimated retrospectively from BMI using a predictive model that accounts for age, sex, and ethnicity and is validated for overweight and obese individuals (Bozeman et al.,
2012). Exclusion criteria for the study included any of the following: carpal tunnel syndrome, peripheral edema, a BMI > 40 kg/m2 (since waist circumference could not be estimated accurately using the predictive model), a history of neurotoxic treatment, current treatment with neuropathic pain agents, chronic kidney disease, any other condition known to cause neuropathy, current eye infections, corneal abrasions, history of refractive surgery, trauma to anterior segment, or contact lens wear, or if they had a neuromuscular, movement, psychiatric or developmental disorder
(Bozeman et al., 2012). 60 healthy controls similar in age and sex distribution were recruited from the University of New South Wales for comparison. Control participants were excluded from the study if their BMI exceeded 27 kg/m2 (as per previous studies) and if they had a history of taking medication to treat blood pressure, elevated glucose, or lipids (Callaghan et al., 2016b).
192 Chapter 5
2.2. Metabolic Syndrome and Participant Grouping
Metabolic syndrome was defined using the updated National Cholesterol Education
Program Adult Treatment Panel III criteria (Grundy et al., 2005). Required sample size for each group was based on previous studies utilising nerve excitability and confocal microscopy in to type 2 diabetes (Yan et al., 2020). Type 2 diabetes participants were considered to have metabolic syndrome (T2DM/MetS) if they fulfilled any 3 of the 5 criteria: increased waist circumference, elevated triglycerides, reduced HDL, elevated blood pressure, and elevated fasting glucose (Grundy et al.,
2005). All type 2 diabetes participants had an elevated fasting glucose but were considered to have each of the following other components if:
1. Waist circumference was ≥102 cm (≥40 inches) for males or ≥88 cm (≥35
inches) for females,
2. Triglycerides ≥ 1.7 mmol/L (≥150 mg/dL),
3. HDL <1.03 mmol/L (<40 mg/dL) in males or <1.3 mmol/L (<50 mg/dL) in
females,
4. Systolic blood pressure ≥ 130 mmHg or diastolic blood pressure ≥ 85 mmHg.
Participants that did not meet the criteria for metabolic syndrome were allocated into a type 2 diabetes only group (T2DM).
2.3. Neuropathy Assessment
Neuropathy severity was assessed using the validated modified Toronto Clinical
Neuropathy Score (mTCNS), which is comprised of 11 items (Bril et al., 2009).
Symptoms (foot pain, numbness, tingling, weakness, ataxia, and upper limb involvement) were graded from 0–3 depending on their presence and interference
193 Chapter 5 with the participant’s activities of daily living. Sensory tests (pinprick, light touch, temperature, vibration, and position sense) were graded from 0–3 depending on the degree of deficit distally (Bril et al., 2009). Each item was summed to give a total score ranging from 0–33, with a higher score indicating more severe neuropathy.
Nerve conduction studies were also undertaken on the sural (sensory nerve amplitude and velocity) and tibial (compound muscle action potential and latency) nerves
(Liveson and Ma, 1992). Presence of neuropathy was defined using the Toronto
Diabetic Neuropathy Expert Group definition of confirmed diabetic neuropathy: the presence of nerve conduction abnormality and a sign or symptom of neuropathy
(Tesfaye et al., 2010).
2.4. Nerve Excitability Assessment
Motor nerve excitability studies were conducted on the median nerve to assess nerve function. Motor nerve excitability was selected due to its utility as an early marker of nerve dysfunction in diabetes (Kiernan et al., 2020). First, stimulus-response curves were generated using a 1 ms test pulse to obtain the maximal CMAP amplitude from abductor pollicis brevis. A target response of 40% of this maximum was subsequently calculated. The current required to elicit this target response, known as ‘threshold’, was tracked in four testing paradigms which each provide indirect information regarding the function of voltage-gated sodium (Na+) and potassium (K+) ion channels within the axon membrane (comprised of the nodal and internodal regions).
1. Strength–duration behavior: the relationship between the stimulus intensity
required to reach threshold from four different stimulus durations (0.2, 0.4,
0.8, and 1.0 ms) was used to calculate the strength–duration time constant,
194 Chapter 5
which is a marker of persistent Na+ conductance in the Node of Ranvier
(Bostock et al., 1998, Kiernan et al., 2020).
2. Threshold electrotonus: Changes in threshold were determined after a 1 ms
test pulse was applied during or after 100 ms of a subthreshold conditioning
current of +40% (depolarising) or –40% (hyperpolarising) of control threshold
(established from the initial stimulus-response curve) to assess the activity of a
variety of nodal and internodal ion channels (Bostock et al., 1998, Kiernan et
al., 2020).
3. The current–threshold relationship: threshold change of 1 ms test pulses
following 200 ms conditioning currents ranging from +50% (depolarizing) to
–100% (hyperpolarising) of the control threshold was used to assess
rectification properties of the internode (Bostock et al., 1998, Kiernan et al.,
2020).
4. Recovery cycle: changes in threshold following a 1 ms supramaximal
conditioning stimulus were obtained to assesses the activity of nodal Na+
channels and fast K+ channels (Bostock et al., 1998, Kiernan et al., 2020).
2.5. Mathematical Modeling
To investigate the pathological basis for nerve dysfunction, nerve excitability recordings were analysed using the Bostock model of axonal excitability, which is a validated model of the human axon and has been used in the interpretation of nerve excitability for metabolic, toxic, inflammatory, and hereditary neuropathies (Garg et al., 2018, Jankelowitz and Burke, 2013, Kiernan et al., 2020, Liang et al., 2014). The model assists in interpretation by providing an indication of the underlying changes in and around the axonal membrane in the disease state. This includes changes in the
195 Chapter 5 maximal conductance and permeabilities of different types of Na+ and K+ ion channels in the nodal and internodal membrane, alterations in the sodium-potassium pump current, and other biophysical properties. The model was first adjusted to fit the mean nerve excitability data obtained from the Control group before fitting the mean data of the T2DM and T2DM/MetS groups. Modeling of nerve dysfunction was based on the known electrophysiological changes implicated in diabetic neuropathy
(Brismar et al., 1987, Krishnan et al., 2008, Scarpini et al., 1993). Modeling analyses involved changes in a single or a combination of membrane parameters in an iterative fashion to objectively fit simulated excitability data with the mean recorded data as closely as possible using a least squares approach.
2.6. Nerve Ultrasound
All participants underwent ultrasonography of the median nerve at a non-entrapment site using a 10–18 MHz linear probe. The median nerve was chosen due to the high reproducibility of the cross-sectional area (CSA) measurement that can be obtained
(Yan et al., 2020). The machine (MyLabTMOne, Esaote, Genoa, Italy) was operated under the ‘Musculoskeletal’ factory preset and all settings such as depth, gain, and focus were kept constant for each examination. Ultrasound was conducted while participants were sitting comfortably with their forearm fully supinated and fingers semi-extended. The forearm was supported on an armrest to ensure the elbow was flexed at 90°. The median nerve was first identified in the transverse plane at the carpal tunnel inlet and then traced proximally between the superficial (flexor digitorum superficialis) and deep (flexor pollicis longus and flexor digitorum profundus) muscles until the junction of the middle and distal third of the forearm.
The nerve CSA was then measured on the screen (in mm2) by outlining the inner
196 Chapter 5 margin of the epineurium (Yan et al., 2020). Each CSA measurement was completed by two observers and then averaged.
2.7. Corneal Confocal Microscopy
Participants enrolled after 2018 were scanned bilaterally with a confocal microscope to assess corneal nerve structure (Heidelberg Engineering GmbH, Heidelberg,
Germany). Eight central and four inferior whorl images, not overlapping by more than
20%, from both eyes of each participant were selected for quantification. Images were analysed using a validated and fully automated nerve analysis software (Corneal
Nerve Fiber Analyzer V.2, ACCMetrics, University of Manchester, Manchester,
United Kingdom) to quantify corneal nerve fibre length (total length of main nerves and nerve branches per square millimeter), density (number of main nerves per square millimeter), branch density (total number of main nerve branches per square millimeter), and fractal dimension (a measure of structural complexity with a lower value indicating less corneal nerve intricacy) in the central cornea. Corneal nerve fibre length and fractal dimension were also quantified in the inferior whorl, which is the region that contains the most distal aspect of corneal nerves (Chen et al., 2017,
Petropoulos et al., 2015). Corneal measures are presented as an average of both eyes.
2.8. Statistical Analyses
Data were analysed using SPSS Statistics Version 25.0 for Windows (IBM Corp,
New York, USA). Normality of data was assessed using Shapiro–Wilk tests.
Normally distributed data is expressed as mean ± standard deviation while non- normally distributed data is written as median and quartile 1 to quartile 3. Nerve excitability data is expressed as mean ± standard error. Where appropriate, and with
197 Chapter 5 post-hoc corrections if applicable, independent t-tests, Mann–Whitney U tests, or
Pearson chi-square analyses were applied to compare the means of demographic data, clinical variables, and outcome measures between groups. Pearson or Spearman correlations were applied to determine relationships between clinical variables and outcome measures. Partial correlations were used to control for age, diabetes duration, and HbA1c. A logistic regression analysis was used to determine the relationship between metabolic syndrome components and the presence of confirmed neuropathy, controlling for age, diabetes duration, and HbA1c. Statistical significance was considered when P < 0.05.
3. Results
3.1. Participant Demographics
In total, 148 participants with type 2 diabetes were recruited. Subjects were grouped according to the presence (T2M/MetS, n = 89) or absence of metabolic syndrome
(T2DM, n = 59). Demographics are summarised in Table 5.1. Diabetes groups were similar for disease duration, proportion of participants treated with insulin, LDL, and
HbA1c. As expected, participants with T2DM/MetS differed significantly from subjects with T2DM in terms of waist circumference, triglycerides, HDL, and systolic blood pressure. All groups were similar for age and sex distribution.
198 Chapter 5
Table 5.1. Participant demographics
T2DM/MetS T2DM Control T2DM/MetS T2DM/MetS (n = 89) (n = 59) (n = 60) vs. T2DM vs. Control Age (years) 62 ± 9 60 ± 11 57 ± 7 0.751 0.060
Sex (% male) 53 53 56 0.975 0.717
Diabetes duration (years) 13 (8–18) 10 (4–18) 0.078
Participants on insulin (%) 63 51 0.123
HbA1c (%) 7.7 (6.8–8.7) 7.4 (6.7–9.0) 0.549
Waist circumference 110.1 100.4 0.001 (cm) (102.4–116.3) (92.3–113.5)
Triglycerides (mmol/L) 1.8 (1.3–2.4) 1.0 (0.8–1.4) <0.001
LDL (mmol/L) 2.1 (1.6–2.9) 2.4 (1.5–2.6) 0.985
HDL (mmol/L) 1.0 (0.9–1.3) 1.5 (1.3–1.6) <0.001
SBP (mmHg) 130 120 <0.001 (124–145) (118–128) DBP (mmHg) 78 (70–83) 80 (70–80) 0.929 mTCNS 4 (0–8) 1 (0–5) 0 0.026 <0.001
Confirmed neuropathy (%) 38 20 0 <0.001 <0.001
Sural SNAP (μV) 9.4 11.8 15.3 0.084 0.003 (3.1 –13.9) (6.5–16.3) (12.4–27.8) Sural NCV (m/s) 40.9 43.2 47.0 0.561 <0.05 (32.3–45.6) (36.4–47.6) (39.3–45.6) Tibial CMAP 6.3 8.1 15.5 (mV) 0.063 <0.001 (3.2 –10.2) (4.8–10.9) (12.1–19.1)
Tibial latency (ms) 4.2 4.2 3.4 0.851 <0.001 (3.6–4.8) (3.6–4.7) (3.1–3.8) Normally distributed data is expressed as mean ± SD and non-normally distributed data is expressed as median and quartile 1 to quartile 3. Waist circumference was calculated retrospectively using a predictive model that accounts for age, sex, and ethnicity and is validated for overweight and obese individuals. SBP: systolic blood pressure; DBP: diastolic blood pressure; mTCNS: modified Toronto Clinical Neuropathy Score; SNAP: sensory neuron action potential; NCV; nerve conduction velocity; CMAP: compound muscle action potential. As per the Toronto Diabetic Neuropathy Expert Group definition, confirmed neuropathy was based on the presence of nerve conduction abnormality and a sign or symptom of neuropathy. Normally distributed data was compared using a one-way ANOVA (with LSD corrections) while non-
199 Chapter 5 normally distributed data was compared using a Kruskal-Wallis test (with Dunn-Bonferroni corrections).
3.2. Neuropathy Assessment
Based on the Toronto Diabetic Neuropathy Expert Group definition of confirmed diabetic neuropathy, neuropathy was confirmed to be present in 38% of participants with T2DM/MetS compared to 20% in the T2DM group (p < 0.001). Evaluation of the mTCNS findings determined that participants with T2DM/MetS had significantly worse neuropathy compared to participants with T2DM alone (p < 0.05) (Table 5.1).
Further analysis revealed that subjects with T2DM/MetS scored significantly worse for individual items foot pain (p < 0.01), numbness (p < 0.05), ataxia (p < 0.01), and upper limb symptoms (p < 0.05) but not tingling, weakness, pinprick, temperature, light touch, vibration, and position sense. Considering both diabetes groups together, sural and tibial amplitudes reduced with waist circumference (sural: r = –0.373, p <
0.001; tibial: r = –0.356, p < 0.001) and increased with HDL (sural: r = 0.261, p <
0.01) after controlling for age, HbA1c, and disease duration. Neuropathy severity increased with the number of metabolic syndrome components (r = 0.348, p < 0.001), independent of age, HbA1c and diabetes duration. In particular, a higher neuropathy severity score was found with increasing waist circumference (r = 0.395, p < 0.001) and triglycerides (r = 0.472, p < 0.001). Logistic regression analysis revealed that waist circumference, when analysed with other components of metabolic syndrome, was a significant predictor of neuropathy presence when controlling for age, diabetes duration and HbA1c (OR: 1.12; 95% CI: 1.05–1.19, p < 0.001).
3.3. Nerve Ultrasound and Corneal Confocal Microscopy
200 Chapter 5
Nerve ultrasound and corneal confocal microscopy were used to assess peripheral nerve structure, with significant alterations being observed in participants with
T2DM/MetS. Subjects with T2DM/MetS demonstrated significant enlargement in median nerve CSA (8.3 ± 1.7 mm2) when compared to T2DM alone (7.6 ± 1.9 mm2, p
< 0.05) and the Control group (7.2 ± 1.3 mm2, p < 0.05) (Figure 5.1). When controlling for age, diabetes duration, HbA1c and neuropathy severity, analysis revealed that waist circumference corresponded with median nerve CSA enlargement
(r = 0.504, p < 0.01). Corneal confocal microscopy, undertaken in a subset of participants matched for age, male to female proportion, HbA1c, diabetes duration, proportion of subjects dependent on insulin, and neuropathy severity, demonstrated there were significant structural differences between the T2DM/MetS and T2M groups (Table 5.2). All corneal nerve measurements were significantly reduced in
T2DM/MetS compared to T2DM (Table 5.2 and Figure 5.2). After accounting for age, diabetes duration, HbA1c, and neuropathy severity, increases in waist circumference and triglycerides were associated with reductions in structural measures in both the central and inferior whorl regions of the cornea (Table 5.3).
Figure 5.1. Median nerve ultrasonography highlighting differences in median nerve cross- sectional area between participants with (A) type 2 diabetes and metabolic syndrome (11.74 mm2), (B) type 2 diabetes alone (8.84 mm2), and (C) healthy controls (6.99 mm2).
201 Chapter 5
Table 5.2. Corneal confocal microscopy cohorts T2DM/MetS (n = 24) T2DM (n = 22) T2DM/MetS vs. T2DM Age (years) 65 ± 11 60 ± 13 0.160
Sex (% male) 64 54 0.515
Diabetes duration (years) 15 ± 7 11 ± 8 0.060
Participants on insulin (%) 67 50 0.251
HbA1c (%) 7.8 ± 1.3 7.6 ± 1.7 0.574 mTCNS 4 (0–9) 1 (0–5) 0.090
CNFL(mm/mm2) 11.7 ± 3.9 14.8 ± 3.1 <0.01
CNFD(no./mm2) 18.9 ± 7.4 25.0 ± 5.7 <0.01
CNBD(no./mm2) 25.2 ± 13.2 37.7 ± 16.6 <0.01
CNFrD 1.45 (1.41–1.48) 1.49 (1.45–1.50) <0.01
IWL (mm/mm2) 11.0 (7.0–14.6) 14.5 (11.3–16.6) <0.05
IWFrD 1.44 (1.41–1.48) 1.48 (1.45–1.50) <0.05
Corneal confocal microscopy was undertaken in a subset of participants. Normally distributed data is expressed as mean ± SD and non-normally distributed data is expressed as median and quartile 1 to quartile 3. CNFL: central nerve fibre length; CNFD: central nerve fibre density; CNBD: central nerve branch density; CNFrD: central nerve fractal dimension; IWL: inferior whorl fibre length; IWFrD: inferior whorl fractal dimension. Normally distributed data was compared using an independent t-test while non-normally distributed data was compared using a Mann-Whitney U test.
202 Chapter 5
Figure 5.2. Representative corneal confocal micrographs comparing differences in nerve structure in the central and corresponding inferior whorl region of the cornea for participants with type 2 diabetes and metabolic syndrome (A, B), type 2 diabetes alone (C, D), and healthy controls (E, F).
203 Chapter 5
Table 5.3. Corneal confocal microscopy correlations CNFL CNFD CNBD CNFrD IWL IWFrD Waist –0.400** –0.330* –0.439** –0.371** –0.468*** –0.403** circumference Triglycerides –0.488* –0.383* –0.383* –0.519** –0.617*** –0.682***
Association between waist circumference or triglycerides and corneal confocal measurements in the central or inferior whorl regions of the cornea. Partial correlations have been controlled for age, diabetes duration, HbA1c, and neuropathy severity. CNFL: central nerve fibre length; CNFD: central nerve fibre density; CNBD: central nerve branch density; CNFrD: central nerve fractal dimension; IWL: inferior whorl fibre length; IWFD: inferior whorl fibre density; IWBD: inferior whorl branch density; IWFrD: inferior whorl fractal dimension. *P < 0.05, **P < 0.01, ***P < 0.001
3.4. Nerve Excitability and Mathematical Modeling
To provide pathological insights into the underlying causes of the greater neuropathy severity observed in the T2DM/MetS group, nerve excitability studies were undertaken. Participants with T2DM/MetS demonstrated more severe changes in nerve function in comparison to the T2DM and Control groups (Kiernan et al., 2020).
In the depolarising threshold electrotonus testing paradigm, subjects with
T2DM/MetS had a less percentage threshold reduction between 40–60 ms (49.6 ± 0.5
%) and accommodation half-time (40.4 ± 0.5 ms) when compared to T2DM alone
(51.3 ± 0.6 %, p < 0.05 and 42.5 ± 0.6 ms, p < 0.01, respectively) and less percentage threshold reduction between 40–60 ms when compared to the Control group (52.7 ±
0.6 %, p <0.001). Superexcitability was significantly increased in the T2DM/MetS group (–18.3 ± 0.8 %) in comparison to T2DM (–20.6 ± 0.6 %, p < 0.05) and Control
(–24.7 ± 0.7 %, p < 0.001) participants.
Nerve excitability recordings from the T2DM and T2DM/MetS participants were subsequently modeled against the Control group to provide insight into the cause of
204 Chapter 5 altered membrane function in each cohort. Analysis indicated that the nerve excitability findings observed in individuals with T2DM alone could be explained best by a reduction in nodal Na+ permeability (T2DM: 3.6; Control: 4.2, cm3s-1 x 10–
9). This single change reduced the discrepancy between the T2DM and Control groups by 80%. In contrast, sole decreases in nodal Na+ permeability in the T2DM/MetS participants could not adequately explain the impaired nerve function observed in this group, as this only minimised the discrepancy by 67% (T2DM/MetS: 3.55; Control:
4.2, cm3s-1 x 10–9). However, when a decrease in nodal Na+ permeability
(T2DM/MetS: 3.55; Control: 4.2, cm3s-1 x 10–9) was combined with a reduction in the sodium-potassium pump current (T2DM/MetS: –4.4; Control: 0, picoamperes), the discrepancy between the T2DM/MetS and Control cohorts was reduced by 79%. A sole decrease in the sodium-potassium pump current could only minimise the discrepancy between these groups by 23%, suggesting that a decrease in Na+ permeability and sodium-potassium pump activity were occurring simultaneously in
T2DM/MetS.
4. Discussion
The present study has provided insights regarding the difference in pathophysiology underlying nerve dysfunction in people with type 2 diabetes, with and without the metabolic syndrome. Nerve excitability techniques were applied in both groups, with data analysed using a validated mathematical model of the human motor axon. In type
2 diabetes without metabolic syndrome, a reduction in sodium permeability is the dominant abnormality underlying nerve dysfunction while in people with metabolic syndrome, there is involvement of both a decrease in sodium permeability and reduced function of the sodium-potassium pump. From a clinical perspective,
205 Chapter 5 participants with type 2 diabetes with metabolic syndrome had more severe clinical neuropathy and greater abnormalities in a range of surrogate markers of nerve structure and function. Limitations of the current study include the retrospective estimation of waist circumference from BMI, the lack of sensory nerve excitability studies, and cross-sectional nature of the study design. Discussion will now focus on the mechanisms whereby altered sodium-potassium pump function in metabolic syndrome may drive greater axonal loss and on the potential avenues of therapy that may be useful in attenuating the degree of nerve injury.
Participants with diabetes and metabolic syndrome demonstrated more severe changes in nerve function, and these changes were attributed to decreased nodal sodium permeability as well as a reduction in the sodium-potassium pump current. Both of these features have been observed in experimental studies on the pathophysiology of diabetic neuropathy (Brismar et al., 1987, Scarpini et al., 1993). The reduction in sodium-potassium pump activity was not observed in participants with diabetes alone, which suggests that impairment in the pump may underlie the more severe changes in nerve function observed in diabetes with metabolic syndrome, especially considering the sodium-potassium pump is critical for normal nerve physiology (Feldman et al.,
2017). Reduced function of the sodium-potassium pump may be explained by decreased mitochondrial ATP production caused by obesity-mediated increases in long chain fatty acids (O'Brien et al., 2017). This may be particularly significant given the observed link between mitochondrial dysfunction in metabolic syndrome and neuron degeneration as well as peripheral neuropathy (Fernyhough and McGavock,
2014, Jha et al., 2017, Rumora et al., 2018, Rumora et al., 2019). In keeping with our finding regarding the reduced function of the sodium-potassium pump, other studies
206 Chapter 5 have observed a link between the metabolic syndrome and decreased pump function in other cells and tissues (Gianfrancesco et al., 2019, Lin et al., 2014, Namazi et al.,
2019, Ng and Hockaday, 1986). Insulin, which is important for the maintenance of the sodium-potassium pump, has also been recognised as a potent neurotrophic factor that supports peripheral nerve growth (Feldman et al., 2017, Sima, 2003). While diabetes cohorts were matched for the proportion of participants that were treated with insulin, another explanation of our findings is that that more severe impairments in nerve function observed in the participants with metabolic syndrome may be due to greater insulin resistance in this cohort. This is supported by recent findings that resistance to insulin has been shown to be a significant factor in the development of peripheral neuropathy in individuals with metabolic syndrome (Han et al., 2015). Of note, GLP-
1 receptor agonists, which are currently used in the treatment of type 2 diabetes, are known to improve insulin sensitivity and further evidence exists they may improve
Na+ currents and sodium-potassium pump function in animal and cell models
(Hinnen, 2017, Liu et al., 2011, Luciani et al., 2010). Considering the pathophysiological mechanisms implicated in nerve dysfunction in the current study,
GLP-1 receptors agonists may mitigate nerve dysfunction and further investigations into their therapeutic potential for diabetic neuropathy are warranted.
Participants with type 2 diabetes and metabolic syndrome also exhibited greater increases in median nerve cross-sectional area than those with diabetes alone. As a consequence of an impaired sodium-potassium pump, intracellular sodium would be retained which would lead to nodal swelling (Sima and Brismar, 1985). Such structural disruption to the peripheral nerve has also been observed to correlate with the reduction in sodium permeability in animal models of diabetic neuropathy (Sima
207 Chapter 5 et al., 1986). Alternatively, enlargement of median nerve cross-sectional area, which was found to correspond with increasing waist circumference independent of neuropathy severity and other confounding variables, may be related to obesity- induced inflammation of peripheral nerves (O'Brien et al., 2017). In animal models of metabolic syndrome, adipose-derived proinflammatory mediators, such as interleukin-
6, interleukin-1b, and TNF-a, have been shown to be involved in peripheral nerve injury and inflammation (Zhang et al., 2018).
Greater small fibre structural deterioration, visualised using corneal confocal microscopy, was also observed in the group with type 2 diabetes and metabolic syndrome. Previous investigations have shown that insulin increases the activity of the sodium-potassium pump in the corneal endothelium, which is important for normal corneal thickness and hydration, however a direct effect on the sodium- potassium pump of corneal nerves is yet to be seen (Hatou et al., 2010). Interestingly, omega-3 fatty acids have been shown to improve pump function in animal models of diabetic neuropathy and dietary supplementation has improved corneal measures in patients with diabetes (Gerbi et al., 1998, Lewis et al., 2017). Insulin has also been observed to have a positive effect on corneal nerve fibres and given that insulin resistance is an important factor in the development of peripheral neuropathy in metabolic syndrome, insulin insensitivity may also explain the significant reduction in corneal nerve measures (Han et al., 2015, Mahelkova et al., 2018). While corneal nerves are especially susceptible to injury, probably due to their lack of myelin, regeneration of these small fibres in the cornea has been observed in type 2 diabetes
(Jia et al., 2018). The regenerative capacity of small fibres in the epidermis decreases
208 Chapter 5 in metabolic syndrome, which may explain the marked reductions in corneal measures observed in the T2M/MetS group (Singleton et al., 2015).
In conclusion, we have provided evidence that individuals with type 2 diabetes and metabolic syndrome manifest greater alterations in nerve structure and function in comparison to individuals with type 2 diabetes alone. A reduction in sodium- potassium pump activity may be the pathophysiological basis for these observed differences. Longitudinal studies to confirm these findings are required.
209 Chapter 6
Chapter 6 – Effect of exenatide on axonal function in type 2 diabetes
210 Chapter 6
Summary and Link to Thesis
In Chapter 5, I explored the pathophysiological mechanisms underlying axonal function and structure in type 2 diabetes with and without the metabolic syndrome. It was found that patients with type 2 diabetes and the metabolic syndrome demonstrated more severe changes in nerve function and structure. The basis of altered nerve function was due to a combined reduction in nodal sodium channel permeability and reduced activity of Na+/K+–ATPase. In type 2 diabetes, altered nerve function was due solely to a reduction in nodal sodium channel permeability.
Interestingly, evidence from animal studies suggest that glucagon-like peptide-1 agonists may improve Na+ currents and Na+/K+–ATPase function. In search of potential neuroprotective options, the final chapter of this thesis will examine the effect of glucose-lowering medication on peripheral nerve function in type 2 diabetes.
Specifically, the effect of exenatide, dipeptidyl peptidase-4 inhibitors, and sodium- glucose co-transporter 2 inhibitors on measures of axonal excitability were assessed.
It was found that prominent abnormalities in axonal excitability remained in patients receiving dipeptidyl peptidase-4 inhibitor or sodium-glucose co-transporter 2 inhibitor therapy. In contrast, patients receiving exenatide demonstrated no differences in nerve excitability compared to healthy controls. The effect of exenatide on axonal excitability was then assessed prospectively. Following exenatide therapy, improvements in measures of nerve excitability were observed.
This work has been submitted to Diabetes Care and is under review. TI was responsible for the study design, recruitment, data collection, data interpretation, and the manuscript composition.
211 Chapter 6
Abstract
Objective: To assess the effect of exenatide, DPP-4 inhibitors, and SGLT2 inhibitors on measures of peripheral nerve function in patients with type 2 diabetes
Methods: Patients receiving either exenatide (n = 32), a DPP-4 inhibitor (n = 31), or a
SGLT2 inhibitor (n = 27) underwent nerve excitability assessments to assess the activity of voltage-gated ion channels and the sodium-potassium pump. Groups were matched for age, sex, HbA1c, diabetes duration, lipids, and clinical neuropathy severity. An additional 10 subjects were assessed prospectively, before the commencement of exenatide and 3 months later. Neuropathy severity was assessed using the Total Neuropathy Score. A cohort of healthy controls (n = 32) were recruited for comparison.
Results: Compared to controls, patients receiving a DPP-4 or SGLT2 inhibitor demonstrated prominent abnormalities in peak threshold reduction (p < 0.05 for both),
S2 accommodation (p < 0.001 and p < 0.01, respectively), superexcitability (p < 0.001 and p < 0.05, respectively), and subexcitability (p < 0.001 for both). In contrast, no differences in nerve excitability measures were observed between the exenatide and control groups. In the prospective arm, patients demonstrated an improvement in S2 accommodation (p < 0.05), superexcitability (p < 0.05), and subexcitability (p < 0.05) at follow-up compared to baseline recordings. These changes were independent of the significant reduction in HbA1c following exenatide treatment.
Conclusions: Exenatide was associated with an improvement in nerve excitability measures in patients with type 2 diabetes.
212 Chapter 6
1. Introduction
Diabetic neuropathy is a highly prevalent complication of diabetes and is a major cause of disability (Pop-Busui et al., 2017). The most common phenotype is a distal symmetric polyneuropathy, in which there is bilateral damage to peripheral nerves in a length-dependent manner resulting in sensory loss, weakness, and pain (Feldman et al., 2019). Unfortunately, disease-modifying treatments for neuropathy remain elusive and management primarily involves the alleviation of painful symptoms (Callaghan et al., 2020, Feldman et al., 2019).
In animal models of diabetic neuropathy, glucagon-like peptide-1 (GLP-1) receptor agonists, dipeptidyl peptidase-4 (DPP-4) inhibitors, and sodium-glucose co- transporter 2 (SGLT2) inhibitors have been associated with improvement in nerve function (El Mouhayyar et al., 2020). In these studies, drug administration was followed by improvements in sensation, alleviation of hypersensitivity, correction in sodium-potassium pump activity, or amelioration in nerve conduction (Bianchi et al.,
2012, Davidson et al., 2011, Gong et al., 2014, Himeno et al., 2011, Jolivalt et al.,
2011, Kan et al., 2012, Lee et al., 2018, Liu et al., 2011, Ma et al., 2018, Moustafa et al., 2018, Shekunova et al., 2019, Takakura et al., 2016). Of particular interest, GLP-1 receptors are known to exist in the peripheral nervous system and some synthetic
GLP-1 agonists are capable of crossing the blood-nerve barrier (Liu et al., 2011,
Muscogiuri et al., 2017). In addition to their clinical utility in managing hyperglycaemia in type 2 diabetes, these drugs may have a role in the prevention or treatment of neuropathy in a clinical setting. However, their effect on measures of axonal function has not been assessed in the human setting.
213 Chapter 6
The present investigation was undertaken to examine the effect of exenatide, on measures of peripheral nerve function in individuals with type 2 diabetes. Studies were also conducted in two disease control groups of type 2 diabetic patients treated
DPP-4 inhibitors or SGLT2 inhibitors. The study utilised nerve excitability techniques which provide insight into the activity of membrane embedded ion channels and energy dependent pumps that contribute to impulse conduction in peripheral nerve axons (Kiernan et al., 2020). Previous studies using these techniques in patients with diabetes have reported nerve excitability abnormalities indicative of altered sodium and potassium ion channel function and reductions in the activity of the sodium- potassium pump (Krishnan and Kiernan, 2005, Krishnan et al., 2008, Kwai et al.,
2013, Sung et al., 2012). Studies of nerve excitability in patients with neuropathy due to diabetes and other aetiologies have demonstrated that this technique is acutely responsive to pharmacological interventions (Arnold et al., 2017, Boërio et al., 2010,
Kamel et al., 2020, Krishnan et al., 2005b, Kwai et al., 2015, Misawa et al., 2006b).
2. Methods
2.1. Participants
This study was approved by the South East Sydney Area Health Service Human
Research Ethics Committee (Northern Section) and the Human Research Ethics
Committee of the University of New South Wales. All subjects enrolled provided written informed consent to the procedures in accordance with the Declaration of
Helsinki. Required sample size for each group was calculated based on the change in
‘subexcitability’, a nerve excitability parameter that is known to reduce in patients with type 2 diabetes (Kwai et al., 2013). The power calculation was based on obtaining 80% power with an alpha error = 0.05.
214 Chapter 6
A total of 90 participants diagnosed with type 2 diabetes for at least 1 year were consecutively recruited from the Diabetes Mellitus Centre at the Prince of Wales
Hospital (Sydney, Australia) for cross-sectional assessment. Exclusion criteria for potential participants in the current study included any of the following: carpal tunnel syndrome, previous neurotoxic therapy, current treatment with neuropathic pain medications, peripheral oedema, stage 3b or greater chronic kidney disease (eGFR <
45) or renal transplantation, any other condition known to cause neuropathy (such as vitamin B12 deficiency), or neuromuscular, movement, psychiatric or developmental disorders.
For cross-sectional studies, participants were allocated into groups depending on their pharmacological treatment of diabetes with exenatide (n = 32), a DPP-4 inhibitor (n =
31; sitagliptin = 17, linagliptin = 14), or a SGLT2 inhibitor (n = 27; dapagliflozin =
14, empagliflozin = 13). All participants with type 2 diabetes had completed at least 1 month of therapy with these medications prior to testing. A cohort of 32 healthy controls similar in age and sex distribution was recruited from the University of New
South Wales (Sydney, Australia) for comparison. All healthy controls had a normal neurological examination and nerve conduction studies as well as no symptoms of peripheral neuropathy. Control participants were excluded from the study if they had a history of taking medication to treat blood pressure, elevated glucose, cholesterol or lipids. In a second series of studies, prospective evaluation was undertaken in 10 subjects with type 2 diabetes who were assessed before and 3 months after the commencement of exenatide. There was no change to any oral therapy for the
215 Chapter 6 duration of study enrolment. In one patient, insulin was ceased following initiation of exenatide.
Patient demographic, serum biochemistry, and clinical data collected included age, sex, diabetes duration, HbA1c, additional diabetes medication (insulin, sulfonylureas, biguanides), lipid profile (triglycerides, LDL, HDL), systolic blood pressure, BMI, and neuropathy severity, which was evaluated using the Total Neuropathy Score
(TNS). The TNS is a validated instrument to evaluate peripheral neuropathy in diabetes (Cornblath et al., 1999). The instrument is comprised of eight items that assess sensory and motor peripheral nerve function and each item is scored from 0–4.
These eight items are: sensory symptoms, motor symptoms, vibration sensation (128
Hz tuning fork), pinprick sensation (NeurotipTM, Owen Mumford, United Kingdom), deep tendon reflexes, manual muscle strength, sural sensory nerve amplitude, and tibial motor nerve amplitude (Medelec Synergy, Oxford Instruments, UK). Total neuropathy scores range from 0–32. A higher score is indicative of more severe neuropathy while a score of zero indicates an absence of neuropathy.
2.1. Nerve Excitability Assessment
In all participants, motor nerve excitability studies were conducted on the median nerve to assess the function of voltage-gated sodium and voltage-gated potassium ion channels as well as the sodium-potassium pumps. Motor nerve excitability was selected due to its utility as an early marker of nerve dysfunction in diabetes (Kwai et al., 2013). The nerve excitability assessment consisted of four testing paradigms, each providing indirect information regarding the function of voltage-gated ion channels or sodium-potassium pumps within the axon membrane (which is comprised of
216 Chapter 6 alternating nodes of Ranvier and internodal regions covered by myelin sheath) at the point of stimulation (Kiernan et al., 2020). Prior to the assessment, stimulus-response curves were generated using a 1 ms test pulse to obtain the maximal CMAP amplitude from abductor pollicis brevis. A target response of 40% of this maximum was subsequently calculated. The current required to elicit this target response, known as
‘threshold’, was tracked in the four testing paradigms: ‘strength-duration behavior’,
‘threshold electrotonus’, ‘current-threshold relationship’, and ‘recovery cycle’.
1. Strength–duration behavior was examined by plotting the relationship between
the stimulus intensity required to reach threshold from four different stimulus
durations (0.2, 0.4, 0.8, and 1.0 ms). Weiss’ Law was subsequently applied to
calculate the strength–duration time constant, which is a marker of persistent
Na+ channel conductance in the Node of Ranvier (Bostock et al., 1998).
2. Threshold electrotonus examines the changes in threshold in response to
depolarising (stimulating) or hyperpolarising (inhibiting) conditioning currents
to assess the activity of a variety of nodal and internodal ion channels that
contribute to the excitability of peripheral nerves and the ionic maintenance of
their intracellular and extracellular environment (Bostock et al., 1998, Kiernan
et al., 2000). Changes in threshold were determined after a 1 ms test pulse was
applied during or after 100 ms of a subthreshold conditioning current of +40%
(depolarising) or –40% (hyperpolarising) of control threshold, established
from the initial stimulus-response curve. Percentage change in threshold was
plotted at 10 ms intervals and key measures were extracted from the plot (peak
depolarising threshold reduction and S2 accommodation). S2 accommodation
is the specific phase of depolarising threshold electrotonus in which threshold
reduction is limited and begins to return to control level due to the activation
217 Chapter 6
of nodal slow potassium channels (Bostock et al., 1998). S2 accommodation is
calculated as the difference between peak threshold reduction and threshold
reduction between 90–100 ms.
3. The current–threshold relationship quantifies rectification properties of the
internode in response to long-lasting polarising currents (Bostock et al., 1998).
Current-threshold relationship was determined by plotting the percentage
threshold change of 1 ms test pulses following 200 ms polarising currents
ranging from +50% (depolarising) to –100% (hyperpolarising) of the control
threshold.
4. The recovery cycle examines the changes in threshold following a
supramaximal nerve impulse and assesses the activity of nodal sodium
channels and fast potassium channels (Bostock et al., 1998). Percentage
change in threshold was plotted for a range of conditioning–test intervals (2–
200 ms) following a 1 ms supramaximal conditioning stimulus. The key
measures of the recovery cycle (superexcitability and subexcitability) were
subsequently determined. Superexcitability is the phase of the recovery cycle
in which there is a decrease in threshold due to charge storage on the
internodal membrane. Subexcitability is the period of the recovery cycle in
which there is an increase in threshold due to hyperpolarisation caused by the
delayed inactivation of slow K+ ion channels.
Analysis of nerve excitability measures obtained from the protocol was restricted to those that have previously been shown to be abnormal in studies of type 2 diabetes, namely peak threshold reduction, S2 accommodation, superexcitability, and subexcitability (Krishnan and Kiernan, 2005, Krishnan et al., 2008, Kwai et al., 2013,
Sung et al., 2012).
218 Chapter 6
2.2. Statistical Analyses
Data were analysed using SPSS Statistics Version 25.0 for Windows (IBM Corp,
New York, USA). Normality of data was first assessed using Shapiro–Wilk tests.
Normally distributed data is expressed as mean ± standard deviation while non- normally distributed data is written as median and quartile 1 to quartile 3. By convention, nerve excitability data is expressed as mean ± standard error. Where appropriate, and with post-hoc corrections if applicable, independent t-tests, Mann–
Whitney U tests, and Pearson chi-square analyses were applied to compare the means of demographic data, clinical measures, and nerve excitability measures between exenatide, DPP-4 and SGLT2 groups. Paired sample t-or Wilcoxon signed-rank tests were used to compare means before and after exenatide therapy. Pearson or Spearman correlations were applied to determine relationships between demographic or clinical variables and nerve excitability measures. Linear regression analyses were undertaken to investigate the effects of demographic or clinical variables on nerve excitability outcomes. Figures were generated using GraphPad Prism version 7.00 for Windows
(GraphPad Software, La Jolla, CA, USA). For all tests, statistical significance was considered when p < 0.05.
3. Results
3.1. Participant demographics
Demographic and clinical measures for exenatide, DPP-4, and SGLT2 groups are summarised in Table 6.1. Diabetes groups were similar in age, sex distribution, diabetes duration, HbA1c, lipid profile, systolic blood pressure, BMI, and neuropathy severity. Overall, the severity of neuropathy in all diabetes groups was mild.
219 Chapter 6
Table 6.1. Subject demographics and clinical measures
p value Exenatide (A) DPP-4 (B) SGLT2 (C) Control (n = 32) (n =31) (n = 27) (n = 32) A vs Overall B vs C Age (years) 59 ± 9 63 ± 12 62 ± 11 58 ± 5 0.448 0.231
Sex (% male) 53 65 60 59 0.655 0.838
BMI (kg/m2) 34 (29–36) 30 (27–34) 29 (26–34) 27 (25–31) 0.119 <0.01
Diabetes duration 11 ± 8 12 ± 7 14 ± 7 0.468 (years)
HbA1c (%) 8.1 ± 1.8 8.2 ± 1.7 8.2 ± 1.6 0.933
Insulin (% cohort) 41 52 59 0.353
Sulfonylurea (% cohort) 31 42 15 0.077
Biguanide (% cohort) 84 55 78 0.013
Triglycerides (mmol/L) 1.7 (1.2–2.5) 1.8 (1.3–3.0) 1.8 (0.9–2.4) 0.630
LDL (mmol/L) 2.2 ± 0.9 2.5 ± 1.1 2.1 ± 0.8 0.387
HDL (mmol/L) 1.2 ± 0.3 1.2 ± 0.3 1.2 ± 0.4 0.977
SBP (mmHG) 130 (120–140) 130 (120–140) 128 (118–140) 0.824
TNS 3 (1–8) 4 (1–7) 2 (0–9) 0 0.805 <0.001
Sural SNAP (µV) 13.6 (6.5–16.2) 9.3 (4.5–17.7) 12.1 (4.5–16.3) 17.5 (11.6–26.8) 0.633 0.078
Tibial CMAP (mV) 7.4 (3.8–9.7) 7.4 (3.7–11.2) 7.9 (3.8–10.9) 13 (9.8–17.9) 0.936 0.032
Normally distributed data is expressed as mean ± SD and non-normally distributed data is expressed as median and quartile 1 to quartile 3. BMI: body mass index; LDL: low-density lipoprotein; HDL: high-density lipoprotein; SBP: systolic blood pressure; TNS: total neuropathy score; SNAP: sensory nerve action potential; CMAP: compound muscle action potential. Normally distributed data was compared using a one-way ANOVA while non-normally distributed data was compared using a Kruskal-Wallis test.
220 Chapter 6
3.2. Nerve Excitability Findings
Analysis of nerve excitability parameters was undertaken in a hypothesis driven manner, focussing on those measures that have been shown to be abnormal in previous studies of type 2 diabetes (Krishnan and Kiernan, 2005, Krishnan et al.,
2008, Kwai et al., 2013, Sung et al., 2012). In the cross-sectional study, linear regression analyses determined the demographic and clinical variables listed in Table
6.1 had no significant effect on the nerve excitability outcomes described below.
Mean nerve excitability measures for the exenatide, DPP-4, and SGLT2 groups are compared in Figure 6.1. Patients treated with a DPP-4 or SGLT2 inhibitor demonstrated prominent abnormalities in numerous nerve excitability parameters when compared to healthy controls. The pattern of change in these measures is consistent with previous studies of nerve excitability in type 2 diabetes (Krishnan and
Kiernan, 2005, Krishnan et al., 2008, Kwai et al., 2013, Sung et al., 2012). Analysis of the nerve excitability recordings obtained from the DPP-4 group revealed significant abnormalities in threshold change in response to depolarising currents during the threshold electrotonus paradigm (peak depolarising threshold reduction, p
< 0.05, S2 accommodation, p < 0.001). In the recovery cycle, a significant increase in superexcitability (p < 0.001) and decrease in subexcitability (p < 0.001) was also observed. In the SGLT2 group, there were significant reductions in peak depolarising threshold reduction (p < 0.05), S2 accommodation (p < 0.01), and subexcitability (p <
0.001) as well as a marked increase in superexcitability (p < 0.05).
221 Chapter 6
* ** * 40 *** 90 p = 0.613 p = 0.794
80 30
70 20
Peak 60 10 50 S2 Accommodation S2
(% threshold reduction) 0 (% threshold reduction) 40
DPP-4 DPP-4 SGLT2 Control SGLT2 Control Exenatide Exenatide * *** 0 *** 30 *** p = 0.293 p = 0.090 -10 20 -20
-30 10
-40 Subexcitability Superexcitability (% change in threshold) -50 (% change in threshold) 0
DPP-4 DPP-4 Control SGLT2 Control SGLT2 Exenatide Exenatide
Figure 6.1. Group comparison of nerve excitability measures. Patients receiving DPP-4 inhibitor (n = 31; sitagliptin = 17, linagliptin = 14) or SGLT2 inhibitor therapy (n = 27; dapagliflozin = 14, empagliflozin = 13) demonstrated prominent abnormalities in nerve excitability measures known to be altered in type 2 diabetes. In contrast, no differences in nerve excitability variables were found between patients receiving exenatide and healthy controls. All values are expressed as mean ± SEM. Normally distributed data was compared using an independent t-test while non-normally distributed data was compared using a Mann- Whitney U test. *p < 0.05, **p < 0.01, ***p < 0.001
The major focus of this study was to assess whether treatment with exenatide could restore nerve function towards normal values. In contrast to data obtained from the
DPP-4 and SGLT 2 groups, there were no excitability abnormalities noted in subjects receiving exenatide, when compared to healthy controls (Figure 6.1). Specifically, there were no changes noted in the threshold electrotonus (peak depolarising
222 Chapter 6 reduction, p = 0.613, S2 accommodation, p = 0.794) or recovery cycle measures
(superexcitability, p = 0.293, subexcitability, p = 0.090).
To further assess the effect of exenatide treatment, prospective studies were undertaken in a small group of subjects, who were assessed before and 3 months after the commencement of exenatide. Data obtained from prospective studies are summarised in Table 6.2. Figure 6.2 demonstrates the pattern of nerve excitability changes of a representative patient following exenatide therapy. At follow-up assessment, there was an increase in S2 accommodation (pre-exenatide: 22.6 ± 0.9, post-exenatide: 24.9 ± 1.1, p < 0.05) and subexcitability (pre-exenatide: 13.0 ± 1.0, post exenatide: 14.7 ± 1.3, p < 0.05) as well as a decrease in superexcitability (pre- exenatide: –22.1 ± 2.1, post-exenatide: –24.0 ± 2.2, p < 0.05) compared to baseline recordings. The change in these three nerve excitability measures was in a direction towards control values and is consistent with an improvement in peripheral nerve function. While there was a significant reduction in HbA1c as well as a trending improvement in systolic blood pressure following the commencement of exenatide, there was no correlation between the percentage change in HbA1c, systolic blood pressure and percentage change in nerve excitability parameters.
223 Chapter 6
Table 6.2. Prospective exenatide cohort clinical measures at baseline and at 3-month follow-up
Pre-exenatide Post-exenatide p value (n = 10) (n = 10) Age (years) 57 ± 12 Sex (% male) 60 Diabetes Duration 8 ± 7 (years)
1c HbA (%) 7.8 ± 1.7 7.3 ± 1.7 <0.05 BMI (kg/m2) 35 (31–38) 35 (30–38) 0.114 Insulin (% cohort) 40 30 0.343
Sulfonylurea (% cohort) 30 30 1.000
Biguanide (% cohort) 90 90 1.000
Triglycerides (mmol/L) 2.1 (1.3–2.6) 1.8 (1.1–3.0) 0.414
LDL (mmol/L) 2.6 ± 0.9 2.6 ± 1.2 0.500 HDL (mmol/L) 1.1 ± 0.3 1.2 ± 0.4 0.437 SBP (mmHG) 134 ± 13 128 ± 11 0.051 TNS 1 (0–3) 1 (0-2) 0.751
Sural SNAP (µV) 14.0 ± 8.4 15.3 ± 7.4 0.360
Tibial CMAP (mV) 11.5 ± 4.6 11.8 ± 5.0 0.642
Normally distributed data is expressed as mean ± SD and non-normally distributed data is expressed as median and quartile 1 to quartile 3. BMI: body mass index; LDL: low-density lipoprotein; HDL: high-density lipoprotein; SBP: systolic blood pressure; SNAP: sensory nerve action potential; CMAP: compound muscle action potential. Normally distributed data was compared using a paired sample test while non-normally distributed data was compared using a Wilcoxon signed-rank test.
224 Chapter 6
Depolarising (A) threshold electrotonus 75
50 S2 accommodation
Peak threshold reduction 25
0
-25
-50 Post-exenatide -75 Pre-exenatide Threshold reduction (%) reduction Threshold -100
Hyperpolarising -125 threshold electrotonus
-150 0 100 200 Delay (ms)
(B) 75
50
Subexcitability 25
0 Threshold change (%) change Threshold
-25
Superexcitability
-50 2 20 200 Interstimulus interval (ms)
Figure 6.2. Nerve excitability recordings showing threshold electrotonus (A) and the recovery cycle (B) of a 72-year-old male patient with mild neuropathy (TNS = 3). Patient was treated with exenatide and metformin. Figure demonstrates changes in nerve excitability measures before and 3 months after commencement of exenatide. Dashed line represents mean nerve excitability recording of the healthy control group.
225 Chapter 6
4. Discussion
In the present study, it was observed that exenatide treatment was associated with improved peripheral nerve function in individuals with type 2 diabetes, as assessed using nerve excitability techniques. In cross sectional studies, nerve excitability measures obtained from participants receiving exenatide demonstrated no significant differences to values obtained from the healthy control cohort. In the small prospective cohort, improvements in nerve function were noted following the commencement of exenatide therapy. In contrast to the results obtained in the exenatide group, prominent abnormalities in numerous nerve excitability measures were observed in participants receiving SGLT2 inhibitor or DPP-4 inhibitor therapy when compared to the control group.
The direction of change in nerve excitability measures following exenatide commencement is consistent with improved nerve function (Kwai et al., 2013, Sung et al., 2012). A possible explanation for the improvement in these measures may be due to activation of neuroprotective pathways following receptor agonism. This is supported by the fact that GLP-1 receptors are known to exist in the peripheral nervous system (Liu et al., 2011, Muscogiuri et al., 2017). In animal models of diabetes, it has been suggested that GLP-1 receptor agonism in the peripheral nervous system initiates downstream signalling pathways that correct sodium-potassium pump function, prevent oxidative stress, or activate anti-inflammatory pathways that ultimately prevent neurodegeneration (Liu et al., 2011, Ma et al., 2018, Moustafa et al., 2018). Liu and colleagues reported that cAMP, a stimulator of the sodium- potassium pump, was increased in the sciatic nerve following GLP-1 receptor
226 Chapter 6 agonism in diabetic rats and was postulated as a potential mechanism of neuroprotection (Liu et al., 2011). In keeping with those findings, the changes in nerve excitability measures following exenatide treatment in the current study are consistent with increased function of the sodium-potassium pump (Kiernan and
Bostock, 2000, Krishnan et al., 2008). It is also possible that GLP-1 receptor agonism improves nerve function through an increase in microvascular perfusion, which has been observed in recent clinical studies (Aung et al., 2019). It is unlikely the improvements in nerve function observed following commencement of exenatide were due to reductions in HbA1c, and systolic blood pressure since the percentage change in these variables did not correlate with nerve excitability measures. This suggests that the improvement in nerve function was independent of reductions in
HbA1c and systolic blood pressure.
Despite these apparent benefits of GLP-1 receptor agonism on peripheral nerves, 18 months of exenatide therapy in a randomised trial had no significant effect on clinical neuropathy measures in participants with mild to moderate neuropathy (Jaiswal et al.,
2015). The investigators postulated that this lack of response might have been due to trial participants having a neuropathy severity that was too advanced to be amenable to treatment. In the current study, neuropathy severity in the exenatide cohort was mild and while neuropathy scores did not improve following treatment, acute effects on nerve excitability measures (which may change more readily than clinical neuropathy measures) were still observed. A prolonged period of observation is needed to assess whether acute benefits afforded by exenatide treatment translate into improved long-term clinical outcomes in terms of neuropathy severity and patient reported outcomes.
227 Chapter 6
While numerous animal studies have shown that DPP-4 or SGLT2 inhibitor therapy improves measures of peripheral nerve function, nerve excitability outcomes in the current study remained abnormal in both of these groups (Bianchi et al., 2012,
Davidson et al., 2011, Lee et al., 2018, Takakura et al., 2016). Though the mechanism of action of DPP-4 inhibitors is to increase the level of endogenous GLP-1, it is possible that this is not as beneficial as direct administration of a GLP-1 analogue which may stimulate the GLP-1 receptor directly and are generally considered to be more potent than DPP-4 inhibitors (Brunton, 2014). However, it should be noted that animal studies have provided evidence that DPP-4 inhibitors may ameliorate neuropathy via improvements in sodium-potassium pump activity, which is similar to the proposed mechanism of GLP-1 receptor agonists (Bianchi et al., 2012).
Improvements in nerve function following SGLT2 inhibition observed in animal studies were likely due to an improvement in glycaemic control rather than a direct effect on nerve function, as these transporters are found in the kidney tubules
(Takakura et al., 2016). In the clinical context, investigations into the benefit of DPP-
4 inhibition have shown a decreased incidence of neuropathy in retrospective studies
(Kolaczynski et al., 2016).
Limitations of the current investigation include the lack of SGLT2 inhibitor and DPP-
4 inhibitor prospective cohorts and the small sample size of the study, particularly in the prospective exenatide cohort. Furthermore, longitudinal studies with prolonged follow-up are required to determine the extent of benefit on clinical outcomes for patients receiving this treatment. Future trials examining the potential of GLP-1 receptor agonism on ameliorating neuropathy severity should aim to enrol patients
228 Chapter 6 with minor neuropathy, rather than patients with mild to moderate, or even severe neuropathy (Jaiswal et al., 2015). Consideration of the type of GLP-1 receptor agonist administered should also be based on the previously identified predictive factors of achieving efficacy with a GLP-1 receptor agonist (Monami et al., 2017, Simioni et al.,
2018). Benefit of GLP-1 receptor agonism may also prove more efficacious if treatment is administered prior to the development of severe neuropathy.
In conclusion, it was observed that exenatide alters acute peripheral nerve function in individuals with type 2 diabetes. Longitudinal studies are required to determine the effect of exenatide therapy on long-term outcomes of peripheral nerve function and clinical measures of neuropathy.
229 Summary and Future Directions
Summary and Future Directions
230 Summary and Future Directions
Summary
The studies that comprise this thesis have explored the pathophysiological mechanisms of axonal dysfunction in diabetes and chronic kidney disease using axonal excitability studies, nerve ultrasonography and in-vivo confocal microscopy.
Studies were undertaken in patients with type 1 diabetes, type 2 diabetes, latent autoimmune diabetes of adulthood (LADA), the metabolic syndrome, chronic kidney disease (CKD), and diabetic kidney disease (DKD). The knowledge gained from these studies provides novel insights into the mechanisms and contributing factors underlying diabetic and uraemic neuropathy.
As this thesis investigated uraemic neuropathy, an instrument capable of assessing the severity of neuropathy in chronic kidney disease was essential. Therefore, in Chapter
1, the Total Neuropathy Score was formally validated for patients with CKD. The scale was first established to be measuring a single construct and the eight components of the Total Neuropathy Score were found to have good association with this single construct. The instrument was then demonstrated to have good internal reliability and that the sural sensory nerve action potential amplitude was the most reliable of the eight items. Finally, the scale was then shown to have construct validity for patients with stages 3–5 CKD with and without diabetes. Of particular importance, the Total Neuropathy Score is the first instrument to be formally validated in patients with CKD.
Having validated the Total Neuropathy Score in CKD cohorts, in Chapter 2 the pathophysiological mechanisms underlying neuropathy in patients with DKD were then investigated. The major aim of this study was to determine the relative
231 Summary and Future Directions contributions of CKD and type 2 diabetes to nerve dysfunction in DKD. Patients with
DKD, CKD, and type 2 diabetes were assessed using the Total Neuropathy Score and subsequently underwent axonal excitability assessments. In addition to having a worse neuropathy phenotype, patients with DKD demonstrated more severe changes in nerve excitability measures compared to patients with CKD or type 2 diabetes.
Importantly, increasing serum K+ was associated with more severe alterations in axonal excitability measures in DKD and CKD cohorts. Nerve excitability recordings were then analysed using a validated mathematical model of the human axon. In
DKD, modelling indicated that changes in nerve excitability were due to an elevation
+ + of extracellular K and reductions in nodal Na permeability and internodal Ih conductance. Similarly, modelling suggested that altered nerve excitability in CKD
+ was due to decreased nodal Na permeability and internodal Ih conductance. In contrast, a variety of nodal and internodal conductances and permeabilities were affected in type 2 diabetes. Together, these findings indicated that CKD, and not type
2 diabetes, underlie axonal dysfunction in DKD.
Investigations were then conducted into autoimmune diabetes. In Chapter 3, the impact of acute glucose control on peripheral nerve structure and function was assessed in type 1 diabetes. This was particularly important given the emerging clinical interest of the role of short-term glucose variation in the development of diabetic complications. Patients with type 1 diabetes underwent six days of blinded continuous glucose monitoring. Measures of short-term glucose control, namely continuous overall net glycaemic action (CONGA) and percentage time in (and above) target range were calculated. At the conclusion of glucose monitoring, patients underwent axonal excitability assessments and corneal confocal microscopy. It was
232 Summary and Future Directions found that greater time spent in hyperglycaemia and higher CONGA were associated with greater abnormalities in nerve excitability. Furthermore, greater time above target range increased the likelihood of abnormal axonal excitability. Higher CONGA and greater time spent in hyperglycaemia were also associated with longer inferior whorl corneal nerve length, which may have occurred as a regenerative response, evidenced by the increasing number of microneuromas that were present with worse acute glucose control.
Studies into autoimmune diabetes were continued. In Chapter 4, the pathophysiological mechanisms underlying axonal dysfunction in LADA were examined. Patients with LADA underwent axonal excitability assessments and nerve ultrasonography and findings were compared to participants with type 1 or type 2 diabetes. Compared to type 1 diabetes patients, individuals with LADA demonstrated an increase in median nerve cross-sectional area. When compared to patients with type 1 or type 2 diabetes, patients with LADA demonstrated more severe changes in axonal excitability, specifically changes in threshold electrotonus and the recovery cycle. Subsequent modelling of nerve excitability recordings obtained from the
LADA group indicated the mechanisms underlying nerve dysfunction was different to type 1 and type 2 diabetes. While a range of nodal and internodal conductances in type 1 and type 2 diabetes were affected, patients with LADA had a reduction in persistent nodal Na+ channels and a decreased conductance through fast K+ channels.
This finding indicates that different pathophysiological mechanisms to those seen in type 1 and type 2 diabetes are responsible for the more severe changes in axonal function observed in the LADA group.
233 Summary and Future Directions
Investigations into type 2 diabetes were then undertaken. The prevalence of the metabolic syndrome is increasing worldwide, which is important given that the metabolic syndrome is associated with the development and progression of peripheral neuropathy. In Chapter 5, the effect of the metabolic syndrome in type 2 diabetes was examined. Type 2 diabetes participants with and without the metabolic syndrome underwent axonal excitability assessments, nerve ultrasonography, and corneal confocal microscopy. With respect to nerve structure, individuals with both type 2 diabetes and the metabolic syndrome demonstrated larger median nerve cross- sectional area and reductions in all corneal confocal measures when compared to those with type 2 diabetes alone. An effect on nerve function was also observed as patients with both type 2 diabetes and the metabolic syndrome demonstrated more severe changes in nerve excitability measures. Mathematical modelling indicated the more severe changes in nerve excitability in type 2 diabetes patients with the metabolic syndrome was due to a reduction in nodal Na+ permeability and sodium- potassium pump function. In contrast, in type 2 diabetes alone, only nodal Na+ permeability was reduced. This suggests that the metabolic syndrome has an additional effect on the pathophysiology of axonal dysfunction in which impaired function of the sodium-potassium is involved.
The final study of this thesis sought to investigate the neuroprotective potential of glucose-lowering medication by examining their effect on peripheral nerve function.
In Chapter 6, type 2 diabetes patients receiving either exenatide, a DPP-4 inhibitor, or a SGTL2 inhibitor underwent axonal excitability assessments. When compared to healthy controls, patients on DPP-4 or SGLT2 inhibitor treatment demonstrated prominent abnormalities in nerve excitability measures. In contrast, no differences
234 Summary and Future Directions between axonal excitability measures were observed between subjects receiving exenatide therapy and healthy controls. An additional cohort of prospective exenatide participants were recruited. Nerve excitability assessments were conducted prior to the commencement of exenatide and 3 months later. Changes in excitability measures consistent with an improvement in nerve function were observed following exenatide therapy. Importantly, these changes were independent of improvements in HbA1c and systolic blood pressure following exenatide therapy.
Future Directions
Findings from the investigations undertaken in this thesis have opened further avenues of research into the field of peripheral neuropathy in diabetes and CKD.
Future studies that build upon the pathophysiological mechanisms found to be involved in nerve dysfunction should be undertaken. The effect of potential neuroprotective treatments should also be assessed.
In Chapter 2, it was observed that the CKD component of DKD underlies axonal dysfunction and these patients present with an especially severe neuropathy phenotype. Since previous clinical trial evidence indicates that dietary potassium restriction is neuroprotective in CKD, a similar study could be replicated in a larger
DKD population to determine if the progression to the severe neuropathy phenotype observed in DKD can be prevented or slowed. Prospective studies should include
DKD patients with varying stages of neuropathy to investigate the efficacy of potassium restriction at each of these stages.
235 Summary and Future Directions
In Chapter 3, it was observed that LADA patients exhibit more severe changes in nerve excitability than type 1 and type 2 diabetes patients. To build upon these findings, longitudinal studies examining epidemiological and clinical variables that may contribute to this observation, such as autoantibody titre and degree of insulin resistance, are required. Furthermore, as LADA is heterogeneous in clinical presentation, future studies could categorise patient profile based on demographic and clinical measures.
In Chapter 4, continuous glucose monitoring studies demonstrated that greater percentage time spent in hyperglycaemia and higher CONGA were associated with impaired nerve function and altered corneal nerve morphology. Further studies should examine patients longitudinally to confirm these findings and assess if these associations can be corrected with improvements in short-term glucose control.
Markers of oxidative stress should also be assessed to determine their association with measures of short-term glucose control. Future studies could also compare the effect of glucose fluctuation when in hyperglycaemia compared to normoglycaemia or hypoglycaemia.
In Chapter 5, patients with type 2 diabetes and the metabolic syndrome were found have more severe changes in peripheral nerve structure and function than patients with type 2 diabetes alone. These changes were possibly due to a reduction in the function of the sodium-potassium pump that was observed in type 2 diabetes patients with metabolic syndrome. As insulin resistance has been found to be a significant factor implicated in the development of peripheral neuropathy in metabolic syndrome, future studies should determine the degree of insulin sensitivity in patients. Further
236 Summary and Future Directions studies in type 2 diabetes could also examine the effect of insulin-sensitising treatment on peripheral nerve structure and function longitudinally. Other factors that may contribute to peripheral neuropathy, such as advanced glycation end products should also be investigated.
In Chapter 6 it was observed that exenatide, a glucagon-like peptide-1 (GLP-1) agonist, improved nerve excitability measures. Future studies should compare the effectiveness of the different GLP-1 agonists currently available, as they may differ in the properties that affect nerve function. These include the ability to enter the peripheral nervous system and their maximal therapeutic effect upon agonism of the
GLP-1 receptor on peripheral nerves. While exenatide did not appear to improve clinical measures of neuropathy in previous a clinical trial, it is possible that the recruited patients were too advanced in neuropathic status to observe a therapeutic effect. As it was demonstrated that exenatide does affect peripheral nerve function assessed using nerve excitability, future clinical trials investigating the neuroprotective potential of exenatide (or other GLP-1 agonists) may wish to address the difference in response between patients with mild and severe neuropathy.
237 References
References
238 References
Abhishek K, Khunger N. Complications of skin biopsy. Journal of Cutaneous and Aesthetic Surgery 2015;8(4):239-241.
Aggarwal HK, Sood S, Jain D, Kaverappa V, Yadav S. Evaluation of spectrum of peripheral neuropathy in predialysis patients with chronic kidney disease. Renal Failure 2013;35(10):1323-1329.
Aggarwal S, Colon C, Kheirkhah A, Hamrah P. Efficacy of autologous serum tears for treatment of neuropathic corneal pain. Ocular Surface 2019;17(3):532-539.
Akaza M, Akaza I, Kanouchi T, Sasano T, Sumi Y, Yokota T. Nerve conduction study of the association between glycemic variability and diabetes neuropathy. Diabetology & Metabolic Syndrome 2018;10(1):69.
Al-Aqaba MA, Dhillon VK, Mohammed I, Said DG, Dua HS. Corneal nerves in health and disease. Progress in Retinal and Eye Research 2019;73:100762.
Alam U, Jeziorska M, Petropoulos IN, Asghar O, Fadavi H, Ponirakis G, . . . Malik RA. Diagnostic utility of corneal confocal microscopy and intra-epidermal nerve fibre density in diabetic neuropathy. PLoS One 2017;12(7).
Alam U, Jeziorska M, Petropoulos IN, Pritchard N, Edwards K, Dehghani C, . . . Malik RA. Latent autoimmune diabetes of adulthood (LADA) is associated with small fibre neuropathy. Diabetic Medicine 2019;36(9):1118-1124.
Albers JW, Pop-Busui R. Diabetic neuropathy: mechanisms, emerging treatments, and subtypes. Current Neurology and Neuroscience Reports 2014;14(8):473-473.
Alicic RZ, Rooney MT, Tuttle KR. Diabetic Kidney Disease: Challenges, Progress, and Possibilities. Clinical journal of the American Society of Nephrology 2017;12(12):2032-2045.
Allgeier S, Bartschat A, Bohn S, Peschel S, Reichert K-M, Sperlich K, . . . Köhler B. 3D confocal laser-scanning microscopy for large-area imaging of the corneal subbasal nerve plexus. Scientific Reports 2018;8(1):7468.
American Diabetes Association. 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes—2020. Diabetes Care 2020;43(Supplement 1):S14-S31.
Anand U, Korchev Y, Anand P. The role of urea in neuronal degeneration and sensitization: An in vitro model of uremic neuropathy. Molecular pain 2019;15.
Andersen ST, Grosen K, Tankisi H, Charles M, Andersen NT, Andersen H, . . . Karlsson P. Corneal confocal microscopy as a tool for detecting diabetic polyneuropathy in a cohort with screen-detected type 2 diabetes: ADDITION- Denmark. Journal of diabetes and its complications 2018a;32(12):1153-1159.
Andersen ST, Witte DR, Dalsgaard EM, Andersen H, Nawroth P, Fleming T, . . . Charles M. Risk factors for incident diabetic polyneuropathy in a cohort with screen-
239 References detected type 2 diabetes followed for 13 years: Addition-Denmark. Diabetes Care 2018b;41(5):1068-1075.
Applegate C, Burke D. Changes in excitability of human cutaneous afferents following prolonged high-frequency stimulation. Brain 1989;112(1):147-164.
Arikan E, Sabuncu T, Ozer EM, Hatemi H. The clinical characteristics of latent autoimmune diabetes in adults and its relation with chronic complications in metabolically poor controlled Turkish patients with Type 2 diabetes mellitus. Journal of diabetes and its complications 2005;19(5):254-258.
Ariyasu RG, Nichol JA, Ellisman MH. Localization of sodium/potassium adenosine triphosphatase in multiple cell types of the murine nervous system with antibodies raised against the enzyme from kidney. Journal of Neuroscience 1985;5(10):2581- 2596.
Armstrong DG, Boulton AJM, Bus SA. Diabetic foot ulcers and their recurrence. The New England journal of medicine 2017;376(24):2367-2375.
Arnold R, Issar T, Krishnan AV, Pussell BA. Neurological complications in chronic kidney disease. JRSM Cardiovascular Disease 2016a;5:2048004016677687.
Arnold R, Kwai N, Lin CS, Poynten AM, Kiernan MC, Krishnan AV. Axonal dysfunction prior to neuropathy onset in type 1 diabetes. Diabetes/metabolism research and reviews 2013a;29(1):53-9.
Arnold R, Pianta TJ, Pussell BA, Kirby A, O'Brien K, Sullivan K, . . . Krishnan AV. Randomized, Controlled Trial of the Effect of Dietary Potassium Restriction on Nerve Function in CKD. Clinical journal of the American Society of Nephrology 2017;12(10):1569-1577.
Arnold R, Pussell BA, Howells J, Grinius V, Kiernan MC, Lin CS, Krishnan AV. Evidence for a causal relationship between hyperkalaemia and axonal dysfunction in end-stage kidney disease. Clinical Neurophysiology 2014;125(1):179-85.
Arnold R, Pussell BA, Kiernan MC, Krishnan AV. Comparative study to evaluate the effects of peritoneal and hemodialysis on peripheral nerve function. Muscle and Nerve 2016b;54(1):58-64.
Arnold R, Pussell BA, Pianta TJ, Grinius V, Lin CS, Kiernan MC, . . . Krishnan AV. Effects of hemodiafiltration and high flux hemodialysis on nerve excitability in end- stage kidney disease. PLoS One 2013b;8(3):e59055.
Arroyo EJ, Xu YTA, Zhou L, Messing A, Peles E, Chiu SY, Scherer SS. Myelinating Schwann cells determine the internodal localization of Kv1.1, Kv1.2, Kvβ2, and Caspr. Journal of Neurocytology 1999;28(4-5):333-347.
Atkinson MA, Maclaren NK. The Pathogenesis of Insulin-Dependent Diabetes Mellitus. The New England journal of medicine 1994;331(21):1428-1436.
Aung MM, Slade K, Freeman LAR, Kos K, Whatmore JL, Shore AC, Gooding KM. Locally delivered GLP-1 analogues liraglutide and exenatide enhance microvascular
240 References perfusion in individuals with and without type 2 diabetes. Diabetologia 2019;62(9):1701-1711.
Avram MM, Feinfeld DA, Huatuco AH. Search for the uremic toxin. Decreased motor-nerve conduction velocity and elevated parathyroid hormone in uremia. The New England journal of medicine 1978;298(18):1000-3.
Azmi S, Jeziorska M, Ferdousi M, Petropoulos IN, Ponirakis G, Marshall A, . . . Malik RA. Early nerve fibre regeneration in individuals with type 1 diabetes after simultaneous pancreas and kidney transplantation. Diabetologia 2019.
Babb AL, Ahmad S, Bergström J, Scribner BH. The middle molecule hypothesis in perspective. American Journal of Kidney Diseases 1981;1(1):46-50.
Babb AL, Popovich RP, Christopher TG, Scribner BH. The genesis of the square meter-hour hypothesis. Transactions - American Society for Artificial Internal Organs 1971;17:81-91.
Bae JS, Kim OK, Kim JM. Altered nerve excitability in subclinical/early diabetic neuropathy: evidence for early neurovascular process in diabetes mellitus? Diabetes Research and Clinical Practice 2011;91(2):183-9.
Baker M, Bostock H. Depolarization changes the mechanism of accommodation in rat and human motor axons. The Journal of Physiology 1989;411(1):545-561.
Baker M, Bostock H, Grafe P, Martius P. Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. The Journal of Physiology 1987;383(1):45-67.
Baker MD. Electrophysiology of mammalian Schwann cells. Progress in biophysics and molecular biology 2002;78(2):83-103.
Baker MD, Bostock H. Inactivation of macroscopic late Na+ current and characteristics of unitary late Na+ currents in sensory neurons. Journal of Neurophysiology 1998;80(5):2538-2549.
Barnett MW, Larkman PM. The action potential. Practical Neurology 2007;7(3):192- 197.
Barrett EF, Barrett JN. Intracellular recording from vertebrate myelinated axons: mechanism of the depolarizing afterpotential. The Journal of Physiology 1982;323(1):117-144.
Basilio Vagner R. Uraemic neuropathy: A review. International Journal of Genetics and Molecular Biology 2011;3(11):155-160.
Battelino T, Danne T, Bergenstal RM, Amiel SA, Beck R, Biester T, . . . Phillip M. Clinical Targets for Continuous Glucose Monitoring Data Interpretation: Recommendations From the International Consensus on Time in Range. Diabetes Care 2019;42(8):1593-1603.
241 References
Baumgaertel MW, Kraemer M, Berlit P. Neurologic complications of acute and chronic renal disease. Handbook of Clinical Neurology 2014;119:383-93.
Bear MF, Connors BW, Paradiso MA. Neuroscience : exploring the brain. 4th ed: Philadelphia : Wolters Kluwer, 2016.
Beck RW, Bergenstal RM, Cheng P, Kollman C, Carlson AL, Johnson ML, Rodbard D. The Relationships Between Time in Range, Hyperglycemia Metrics, and HbA1c. Journal of Diabetes Science and Technology 2019a;13(4):614-626.
Beck RW, Bergenstal RM, Riddlesworth TD, Kollman C, Li Z, Brown AS, Close KL. Validation of time in range as an outcome measure for diabetes clinical trials. Diabetes Care 2019b;42(3):400-405.
Beck RW, Connor CG, Mullen DM, Wesley DM, Bergenstal RM. The Fallacy of Average: How Using HbA1c Alone to Assess Glycemic Control Can Be Misleading. Diabetes Care 2017;40(8):994-999.
Behme MT, Dupre J, Harris SB, Hramiak IM, Mahon JL. Insulin resistance in latent autoimmune diabetes of adulthood. Annals of the New York Academy of Sciences 2003;1005:374-7.
Beirowski B, Babetto E, Golden JP, Chen YJ, Yang K, Gross RW, . . . Milbrandt J. Metabolic regulator LKB1 is crucial for Schwann cell-mediated axon maintenance. Nature Neuroscience 2014;17(10):1351-1361.
Benarroch EE. HCN channels: function and clinical implications. Neurology 2013;80(3):304-10.
Bennett DLH. Voltage-gated sodium channel mutations and painful neuropathy: Nav1.9 joins the family. Brain 2014;137(6):1574-1576.
Bennett DLH, Clark AJ, Huang J, Waxman SG, Dib-Hajj SD. The Role of Voltage- Gated Sodium Channels in Pain Signaling. Physiology reviews 2019;99(2):1079- 1151.
Bennett DLH, Vincent A. Autoimmune pain; An emerging concept. Neurology 2012;79(11):1080-1081.
Bennett DLH, Woods CG. Painful and painless channelopathies. The Lancet Neurology 2014;13(6):587-599.
Bergenstal RM, Tamborlane WV, Ahmann A, Buse JB, Dailey G, Davis SN, . . . Wood MA. Sensor-Augmented Pump Therapy for A1C Reduction (STAR 3) Study: Results from the 6-month continuation phase. Diabetes Care 2011;34(11):2403-5.
Bergmans J. The physiology of single human nerve fibres. Vander, Belgium: University of Louvain; 1970.
Berthold CH, Fraher JP, King RHM, Rydmark M. Chapter 3 - Microscopic Anatomy of the Peripheral Nervous System. In: Dyck PJ, Thomas PK, editors. Peripheral Neuropathy (Fourth Edition). Philadelphia: W.B. Saunders; 2005. p. 35-91.
242 References
Bhat MA, Rios JC, Lu Y, Garcia-Fresco GP, Ching W, St Martin M, . . . Bellen HJ. Axon-glia interactions and the domain organization of myelinated axons requires neurexin IV/Caspr/Paranodin. Neuron 2001;30(2):369-83.
Bianchi R, Cervellini I, Porretta-Serapiglia C, Oggioni N, Burkey B, Ghezzi P, . . . Lauria G. Beneficial effects of PKF275-055, a novel, selective, orally bioavailable, long-acting dipeptidyl peptidase IV inhibitor in streptozotocin-induced diabetic peripheral neuropathy. The Journal of pharmacology and experimental therapeutics 2012;340(1):64-72.
Bichet D, Haass FA, Jan LY. Merging functional studies with structures of inward- rectifier K(+) channels. Nature reviews Neuroscience 2003;4(12):957-67.
Biel M, Wahl-Schott C, Michalakis S, Zong X. Hyperpolarization-Activated Cation Channels: From Genes to Function. Physiological Reviews 2009;89(3):847-885.
Bikbov B, Purcell CA, Levey AS, Smith M, Abdoli A, Abebe M, . . . Vos T. Global, regional, and national burden of chronic kidney disease, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet 2020;395(10225):709-733.
Binda D, Cavaletti G, Cornblath DR, Merkies IS. Rasch-Transformed Total Neuropathy Score clinical version (RT-TNSc((c)) ) in patients with chemotherapy- induced peripheral neuropathy. Journal of the Peripheral Nervous System 2015;20(3):328-32.
Birch BD, Kocsis JD, Di Gregorio F, Bhisitkul RB, Waxman SG. A voltage- and time-dependent rectification in rat dorsal spinal root axons. Journal of Neurophysiology 1991;66(3):719-728.
Bittar EE. Maia muscle fibre as a model for the study of uraemic toxicity. Nature 1967;214(5085):310-2.
Blesneac I, Themistocleous AC, Fratter C, Conrad LJ, Ramirez JD, Cox JJ, . . . Bennett DLH. Rare NaV1.7 variants associated with painful diabetic peripheral neuropathy. Pain 2018;159(3):469-480.
Boërio D, Bostock H, Spescha R, Z'Graggen WJ. Potassium and the excitability properties of normal human motor axons in vivo. PLoS One 2014;9(6):e98262.
Boërio D, Créange A, Hogrel J-Y, Guéguen A, Bertrand D, Lefaucheur J-P. Nerve excitability changes after intravenous immunoglobulin infusions in multifocal motor neuropathy and chronic inflammatory demyelinating neuropathy. Journal of the Neurological Sciences 2010;292(1):63-71.
Bonadonna RC, Cucinotta D, Fedele D, Riccardi G, Tiengo A. The metabolic syndrome is a risk indicator of microvascular and macrovascular complications in diabetes: Results from Metascreen, a multicenter diabetes clinic-based survey. Diabetes Care 2006;29(12):2701-2707.
243 References
Borire AA, Arnold R, Pussell BA, Kwai NC, Visser LH, Padua L, . . . Krishnan AV. Haemodialysis alters peripheral nerve morphology in end-stage kidney disease. Clinical Neurophysiology 2017;128(1):281-286.
Borire AA, Arnold R, Pussell BA, Kwai NC, Visser LH, Simon NG, . . . Krishnan AV. Effects of hemodialysis on intraneural blood flow in end-stage kidney disease. Muscle Nerve 2018a;57(2):287-293.
Borire AA, Issar T, Kwai NC, Visser LH, Simon NG, Poynten AM, . . . Krishnan AV. Correlation between markers of peripheral nerve function and structure in type 1 diabetes. Diabetes/metabolism research and reviews 2018b;34(7):e3028.
Bostock H. The strength-duration relationship for excitation of myelinated nerve: computed dependence on membrane parameters. The Journal of Physiology 1983;341(1):59-74.
Bostock H, Baker M. Evidence for two types of potassium channel in human motor axons in vivo. Brain Research 1988;462(2):354-358.
Bostock H, Baker M, Reid G. Changes in excitability of human motor axons underlying post-ischaemic fasciculations: evidence for two stable states. The Journal of Physiology 1991;441(1):537-557.
Bostock H, Cikurel K, Burke D. Threshold tracking techniques in the study of human peripheral nerve. Muscle & Nerve 1998;21(2):137-58.
Bostock H, Rothwell JC. Latent addition in motor and sensory fibres of human peripheral nerve. The Journal of Physiology 1997;498(1):277-294.
Bostock H, Sears TA, Sherratt RM. The effects of 4‐aminopyridine and tetraethylammonium ions on normal and demyelinated mammalian nerve fibres. The Journal of Physiology 1981;313(1):301-315.
Bostock H, Walters RJL, Andersen KV, Murray NMF, Taube D, Kiernan MC. Has potassium been prematurely discarded as a contributing factor to the development of uraemic neuropathy? Nephrology Dialysis Transplantation 2004;19(5):1054-1057.
Boulton AJM, Armstrong DG, Kirsner RS. Diagnosis and management of diabetic foot complications. American Diabetes Association 2018.
Bowley MP, Chad DA. Chapter 15 - Clinical neurophysiology of demyelinating polyneuropathy. In: Levin KH, Chauvel P, editors. Handbook of Clinical Neurology. 161: Elsevier; 2019. p. 241-268.
Bowling CB, Inker LA, Gutierrez OM, Allman RM, Warnock DG, McClellan W, Muntner P. Age-specific associations of reduced estimated glomerular filtration rate with concurrent chronic kidney disease complications. Clinical journal of the American Society of Nephrology 2011;6(12):2822-8.
Boyle MET, Berglund EO, Murai KK, Weber L, Peles E, Ranscht B. Contactin orchestrates assembly of the septate-like junctions at the paranode in myelinated peripheral nerve. Neuron 2001;30(2):385-397.
244 References
Bozeman SR, Hoaglin DC, Burton TM, Pashos CL, Ben-Joseph RH, Hollenbeak CS. Predicting waist circumference from body mass index. BMC Medical Research Methodology 2012;12(1):115.
Brazhe AR, Maksimov GV, Mosekilde E, Sosnovtseva OV. Excitation block in a nerve fibre model owing to potassium-dependent changes in myelin resistance. Interface Focus 2011;1(1):86-100.
Breiner A, Qrimli M, Ebadi H, Alabdali M, Lovblom LE, Abraham A, . . . Bril V. Peripheral nerve high-resolution ultrasound in diabetes. Muscle and Nerve 2017;55(2):171-178.
Bril V, Tomioka S, Buchanan RA, Perkins BA, m TSG. Reliability and validity of the modified Toronto Clinical Neuropathy Score in diabetic sensorimotor polyneuropathy. Diabetic Medicine 2009;26(3):240-246.
Brismar T. Abnormal Na‐Currents in Diabetic Rat Nerve Nodal Membrane. Diabetic Medicine 1993;10(2 S):110S-112S.
Brismar T, Sima AA, Greene DA. Reversible and irreversible nodal dysfunction in diabetic neuropathy. Annals of neurology 1987;21(5):504-7.
Brismar T, Tegnèr R. Experimental uremic neuropathy: Part 2. Sodium permeability decrease and inactivation in potential clamped nerve fibers. Journal of the neurological sciences 1984;65(1):37-45.
Brown DA, Passmore GM. Neural KCNQ (Kv7) channels. British journal of pharmacology 2009;156(8):1185-95.
Browne DL, Gancher ST, Nutt JG, Brunt ER, Smith EA, Kramer P, Litt M. Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nature Genetics 1994;8(2):136-40.
Brunton S. GLP-1 receptor agonists vs. DPP-4 inhibitors for type 2 diabetes: is one approach more successful or preferable than the other? International Journal of Clinical Practice 2014;68(5):557-567.
Buchberger W, Schön G, Strasser K, Jungwirth W. High-resolution ultrasonography of the carpal tunnel. Journal of Ultrasound in Medicine 1991;10(10):531-537.
Burke D, Kiernan MC, Bostock H. Excitability of human axons. Clinical Neurophysiology 2001;112(9):1575-1585.
Burke D, Mogyoros I, Vagg R, Kiernan MC. Temperature dependence of excitability indices of human cutaneous afferents. Muscle Nerve 1999;22(1):51-60.
Buzzetti R, Zampetti S, Maddaloni E. Adult-onset autoimmune diabetes: current knowledge and implications for management. Nature Reviews Endocrinology 2017;13(11):674-686.
245 References
Caldwell JH, Schaller KL, Lasher RS, Peles E, Levinson SR. Sodium channel Nav1.6 is localized at nodes of Ranvier, dendrites, and synapses. Proceedings of the National Academy of Sciences of the United States of America 2000;97(10):5616-5620.
Callaghan BC, Gallagher G, Fridman V, Feldman EL. Diabetic neuropathy: what does the future hold? Diabetologia 2020;63(5):891-897.
Callaghan BC, Gao L, Li Y, Zhou X, Reynolds E, Banerjee M, . . . Ji L. Diabetes and obesity are the main metabolic drivers of peripheral neuropathy. Annals of clinical and translational neurology 2018;5(4):397-405.
Callaghan BC, Hur J, Feldman EL. Diabetic neuropathy: one disease or two? Current Opinion in Neurology 2012;25(5):536-541.
Callaghan BC, Little AA, Feldman EL, Hughes RA. Enhanced glucose control for preventing and treating diabetic neuropathy. Cochrane database of systematic reviews (Online) 2012;6.
Callaghan BC, Price RS, Feldman EL. Distal Symmetric Polyneuropathy: A Review. Journal of the American Medical Association 2015;314(20):2172-2181.
Callaghan BC, Xia R, Banerjee M, de Rekeneire N, Harris TB, Newman AB, . . . Strotmeyer ES. Metabolic Syndrome Components Are Associated With Symptomatic Polyneuropathy Independent of Glycemic Status. Diabetes Care 2016a;39(5):801-7.
Callaghan BC, Xia R, Reynolds E, Banerjee M, Rothberg AE, Burant CF, . . . Feldman EL. Association Between Metabolic Syndrome Components and Polyneuropathy in an Obese Population. JAMA Neurology 2016b;73(12):1468-1476.
Campbell MJ. Health measurement scales. A practical guide to their development and use. David L. Streiner, Geoffrey R. Norman, Oxford University Press, Oxford. Statistics in Medicine 1991;10(9):1478-1479.
Cannon SC. Ion Channels, Overview. In: Aminoff MJ, Daroff RB, editors. Encyclopedia of the Neurological Sciences (Second Edition). Oxford: Academic Press; 2014. p. 747-751.
Cartwright MS, Demar S, Griffin LP, Balakrishnan N, Harris JM, Walker FO. Validity and reliability of nerve and muscle ultrasound. Muscle and Nerve 2013a;47(4):515-521.
Cartwright MS, Hobson-Webb LD, Boon AJ, Alter KE, Hunt CH, Flores VH, . . . Walker FO. Evidence-based guideline: Neuromuscular ultrasound for the diagnosis of carpal tunnel syndrome. Muscle Nerve 2012;46(2):287-293.
Cartwright MS, Mayans DR, Gillson NA, Griffin LP, Walker FO. Nerve cross- sectional area in extremes of age. Muscle and Nerve 2013b;47(6):890-893.
Cartwright MS, Passmore LV, Yoon JS, Brown ME, Caress JB, Walker FO. Cross- sectional area reference values for nerve ultrasonography. Muscle and Nerve 2008;37(5):566-571.
246 References
Casanova I, Diaz A, Pinto S, de Carvalho M. Motor excitability measurements: the influence of gender, body mass index, age and temperature in healthy controls. Neurophysiologie clinique 2014;44(2):213-8.
Catala M, Kubis N. Chapter 3 - Gross anatomy and development of the peripheral nervous system. In: Said G, Krarup C, editors. Handbook of Clinical Neurology. 115: Elsevier; 2013. p. 29-41.
Catterall WA. Voltage-gated sodium channels at 60: structure, function and pathophysiology. The Journal of Physiology 2012;590(11):2577-89.
Catterall WA, Goldin AL, Waxman SG. International Union of Pharmacology. XLVII. Nomenclature and structure-function relationships of voltage-gated sodium channels. Pharmacological Reviews 2005;57(4):397-409.
Cavaletti G, Cornblath DR, Merkies IS, Postma TJ, Rossi E, Frigeni B, . . . Group CI- P. The chemotherapy-induced peripheral neuropathy outcome measures standardization study: from consensus to the first validity and reliability findings. Annals of oncology 2013;24(2):454-62.
Cavaletti G, Frigeni B, Lanzani F, Piatti M, Rota S, Briani C, . . . Italian NG. The Total Neuropathy Score as an assessment tool for grading the course of chemotherapy-induced peripheral neurotoxicity: comparison with the National Cancer Institute-Common Toxicity Scale. Journal of the Peripheral Nervous System 2007;12(3):210-5.
Ceriello A, Monnier L, Owens D. Glycaemic variability in diabetes: clinical and therapeutic implications. The Lancet Diabetes & Endocrinology 2019;7(3):221-230.
Ceriello A, Novials A, Ortega E, La Sala L, Pujadas G, Testa R, . . . Giugliano D. Evidence that hyperglycemia after recovery from hypoglycemia worsens endothelial function and increases oxidative stress and inflammation in healthy control subjects and subjects with type 1 diabetes. Diabetes 2012;61(11):2993-7.
Cervin C, Lyssenko V, Bakhtadze E, Lindholm E, Nilsson P, Tuomi T, . . . Groop L. Genetic similarities between latent autoimmune diabetes in adults, type 1 diabetes, and type 2 diabetes. Diabetes 2008;57(5):1433-7.
Chadban SJ, Ierino FL. Welcome to the era of CKD and the eGFR. The Medical journal of Australia 2005;183(3):117-8.
Chaplan SR, Guo H-Q, Lee DH, Luo L, Liu C, Kuei C, . . . Dubin AE. Neuronal Hyperpolarization-Activated Pacemaker Channels Drive Neuropathic Pain. The Journal of Neuroscience 2003;23(4):1169-1178.
Chen X, Graham J, Dabbah MA, Petropoulos IN, Ponirakis G, Asghar O, . . . Malik RA. Small nerve fiber quantification in the diagnosis of diabetic sensorimotor polyneuropathy: Comparing corneal confocal microscopy with intraepidermal nerve fiber density. Diabetes Care 2015;38(6):1138-1144.
247 References
Chen X, Graham J, Dabbah MA, Petropoulos IN, Tavakoli M, Malik RA. An Automatic Tool for Quantification of Nerve Fibers in Corneal Confocal Microscopy Images. IEEE transactions on bio-medical engineering 2017;64(4):786-794.
Cheng HT, Dauch JR, Porzio MT, Yanik BM, Hsieh W, Smith AG, . . . Feldman EL. Increased Axonal Regeneration and Swellings in Intraepidermal Nerve Fibers Characterize Painful Phenotypes of Diabetic Neuropathy. The Journal of Pain 2013;14(9):941-947.
Chi XX, Jiang X, Nicol GD. ATP-sensitive potassium currents reduce the PGE2- mediated enhancement of excitability in adult rat sensory neurons. Brain research 2007;1145:28-40.
Chiang JL, Maahs DM, Garvey KC, Hood KK, Laffel LM, Weinzimer SA, . . . Schatz D. Type 1 Diabetes in Children and Adolescents: A Position Statement by the American Diabetes Association. Diabetes Care 2018;41(9):2026-2044.
Chillon JM, Massy ZA, Stengel B. Neurological complications in chronic kidney disease patients. Nephrology Dialysis Transplantation 2016;31(10):1606-14.
Chiu HK, Tsai EC, Juneja R, Stoever J, Brooks-Worrell B, Goel A, Palmer JP. Equivalent insulin resistance in latent autoimmune diabetes in adults (LADA) and type 2 diabetic patients. Diabetes Research and Clinical Practice 2007;77(2):237-44.
Chiu SY, Ritchie JM. On the Physiological Role of Internodal Potassium Channels and the Security of Conduction in Myelinated Nerve Fibres. Proceedings of the Royal Society of London Series B, Biological Sciences 1984;220(1221):415-422.
Christensen DH, Knudsen ST, Gylfadottir SS, Christensen LB, Nielsen JS, Beck- Nielsen H, . . . Thomsen RW. Metabolic Factors, Lifestyle Habits, and Possible Polyneuropathy in Early Type 2 Diabetes: A Nationwide Study of 5,249 Patients in the Danish Centre for Strategic Research in Type 2 Diabetes (DD2) Cohort. Diabetes Care 2020.
Coetzee WA, Amarillo Y, Chiu J, Chow A, Lau D, McCormack T, . . . Rudy B. Molecular diversity of K+ channels. Annals of the New York Academy of Sciences 1999;868:233-85.
Colloca L, Ludman T, Bouhassira D, Baron R, Dickenson AH, Yarnitsky D, . . . Raja SN. Neuropathic pain. Nature Reviews Disease Primers 2017;3(1):17002.
Comrey AL. A First Course in Factor Analysis: Academic Press, 1973.
Coresh J. Update on the Burden of CKD. Journal of the American Society of Nephrology 2017;28(4):1020-1022.
Corfas G, Velardez MO, Ko C-P, Ratner N, Peles E. Mechanisms and Roles of Axon- Schwann Cell Interactions. The Journal of Neuroscience 2004;24(42):9250-9260.
Cornblath DR, Chaudhry V, Carter K, Lee D, Seysedadr M, Miernicki M, Joh T. Total neuropathy score: validation and reliability study. Neurology 1999;53(8):1660- 4.
248 References
Cortez M, Singleton JR, Smith AG. Chapter 9 - Glucose intolerance, metabolic syndrome, and neuropathy. In: Zochodne DW, Malik RA, editors. Handbook of Clinical Neurology. 126: Elsevier; 2014. p. 109-122.
Costa LA, Canani LH, Lisbôa HRK, Tres GS, Gross JL. Aggregation of features of the metabolic syndrome is associated with increased prevalence of chronic complications in Type 2 diabetes. Diabetic Medicine 2004;21(3):252-255.
Costello AB, Osbourne, J. Best practices in exploratory factor analysis: four recommendations for getting the most from your analysis. Practical Assessment Research & Evaluation 2005;10(7).
Couser WG, Remuzzi G, Mendis S, Tonelli M. The contribution of chronic kidney disease to the global burden of major noncommunicable diseases. Kidney International 2011;80(12):1258-70.
Craig ME, Twigg SM, Donaghue KC, Cheung NW, Cameron FJ, Conn J, . . . Silink M. National evidence‐based clinical care guidelines for type 1 diabetes in children, adolescents and adults. Canberra: Australian Government Department of Health and Ageing, 2011.
Craner MJ, Hains BC, Lo AC, Black JA, Waxman SG. Co-localization of sodium channel Nav1.6 and the sodium-calcium exchanger at sites of axonal injury in the spinal cord in EAE. Brain 2004;127(2):294-303.
Crill WE. Persistent sodium current in mammalian central neurons. Annual Review of Physiology 1996;58:349-362.
Cruzat A, Qazi Y, Hamrah P. In Vivo Confocal Microscopy of Corneal Nerves in Health and Disease. The Ocular Surface 2017;15(1):15-47.
Davidson EP, Coppey LJ, Dake B, Yorek MA. Treatment of streptozotocin-induced diabetic rats with alogliptin: effect on vascular and neural complications. Experimental Diabetes Research 2011;2011:810469.
Davidson EP, Coppey LJ, Holmes A, Yorek MA. Changes in Corneal Innervation and Sensitivity and Acetylcholine-Mediated Vascular Relaxation of the Posterior Ciliary Artery in a Type 2 Diabetic Rat. Investigative Ophthalmology & Visual Science 2012;53(3):1182-1187.
Davies MJ, D’Alessio DA, Fradkin J, Kernan WN, Mathieu C, Mingrone G, . . . Buse JB. Management of Hyperglycemia in Type 2 Diabetes, 2018. A Consensus Report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2018;41(12):2669-2701.
Davoudi S, Sobrin L. Novel Genetic Actors of Diabetes-Associated Microvascular Complications: Retinopathy, Kidney Disease and Neuropathy. The review of diabetic studies 2015;12(3-4):243-59.
De Fusco M, Marconi R, Silvestri L, Atorino L, Rampoldi L, Morgante L, . . . Casari G. Haploinsufficiency of ATP1A2 encoding the Na+/K+ pump α2 subunit associated with familial hemiplegic migraine type 2. Nature Genetics 2003;33(2):192-196.
249 References
Dedek K, Kunath B, Kananura C, Reuner U, Jentsch TJ, Steinlein OK. Myokymia and neonatal epilepsy caused by a mutation in the voltage sensor of the KCNQ2 K+ channel. Proceedings of the National Academy of Sciences of the United States of America 2001;98(21):12272-12277.
Dehghani C, Pritchard N, Edwards K, Russell AW, Malik RA, Efron N. Fully automated, semiautomated, and manual morphometric analysis of corneal subbasal nerve plexus in individuals with and without diabetes. Cornea 2014;33(7):696-702.
Devaux JJ, Kleopa KA, Cooper EC, Scherer SS. KCNQ2 is a nodal K+ channel. The Journal of Neuroscience 2004;24(5):1236-44.
Dhondt A, Vanholder R, Van Biesen W, Lameire N. The removal of uremic toxins. Kidney International 2000;58:S47-S59.
Di Giulio S, Chkoff N, Lhoste F, Zingraff J, Drüeke T. Parathormone as a nerve poison in uremia. The New England journal of medicine 1978;299(20):1134-5.
Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study Research Group. Intensive Diabetes Treatment and Cardiovascular Outcomes in Type 1 Diabetes: The DCCT/EDIC Study 30-Year Follow-up. Diabetes Care 2016;39(5):686-693.
Dibbens LM, Reid CA, Hodgson B, Thomas EA, Phillips AM, Gazina E, . . . Petrou S. Augmented currents of an HCN2 variant in patients with febrile seizure syndromes. Annals of neurology 2010;67(4):542-6.
Du X, Gamper N. Potassium channels in peripheral pain pathways: expression, function and therapeutic potential. Current neuropharmacology 2013;11(6):621-640.
Du X, Wang C, Zhang H. Activation of ATP-sensitive potassium channels antagonize nociceptive behavior and hyperexcitability of DRG neurons from rats. Molecular pain 2011;7:35-35.
Dyck PJ, Albers JW, Andersen H, Arezzo JC, Biessels G-J, Bril V, . . . Russell JW. Diabetic polyneuropathies: update on research definition, diagnostic criteria and estimation of severity. Diabetes/Metabolism Research and Reviews 2011;27(7):620- 628.
Edwards K, Pritchard N, Dehghani C, Vagenas D, Russell A, Malik RA, Efron N. Corneal confocal microscopy best identifies the development and progression of neuropathy in patients with type 1 diabetes. Journal of diabetes and its complications 2017;31(8):1325-1327.
Eid S, Sas KM, Abcouwer SF, Feldman EL, Gardner TW, Pennathur S, Fort PE. New insights into the mechanisms of diabetic complications: role of lipids and lipid metabolism. Diabetologia 2019;62(9):1539-1549.
Eijkelkamp N, Linley JE, Baker MD, Minett MS, Cregg R, Werdehausen R, . . . Wood JN. Neurological perspectives on voltage-gated sodium channels. Brain 2012;135(Pt 9):2585-612.
250 References
Einheber S, Zanazzi G, Ching W, Scherer S, Milner TA, Peles E, Salzer JL. The Axonal Membrane Protein Caspr, a Homologue of Neurexin IV, Is a Component of the Septate-like Paranodal Junctions That Assemble during Myelination. Journal of Cell Biology 1997;139(6):1495-1506.
El Mouhayyar C, Riachy R, Khalil AB, Eid A, Azar S. SGLT2 Inhibitors, GLP-1 Agonists, and DPP-4 Inhibitors in Diabetes and Microvascular Complications: A Review. International Journal of Endocrinology 2020;2020:1-11.
Eng DL, Gordon TR, Kocsis JD, Waxman SG. Development of 4-AP and TEA sensitivities in mammalian myelinated nerve fibers. Journal of Neurophysiology 1988;60(6):2168-2179.
Erdoğan Ç, Yücel M, Değirmenci E, Öz O, Akgün H, Odabaşı Z. Nerve excitability properties in early preclinical diabetic neuropathy. Diabetes Research and Clinical Practice 2011;94(1):100-104.
Feldman EL, Callaghan BC, Pop-Busui R, Zochodne DW, Wright DE, Bennett DLH, . . . Viswanathan V. Diabetic neuropathy. Nature Reviews Disease Primers 2019;5(1):41.
Feldman EL, Nave KA, Jensen TS, Bennett DLH. New Horizons in Diabetic Neuropathy: Mechanisms, Bioenergetics, and Pain. Neuron 2017;93(6):1296-1313.
Ferdousi M, Kalteniece A, Azmi S, Petropoulos IN, Ponirakis G, Alam U, . . . Malik RA. Diagnosis of Neuropathy and Risk Factors for Corneal Nerve Loss in Type 1 and Type 2 Diabetes: A Corneal Confocal Microscopy Study. Diabetes Care 2020a.
Ferdousi M, Kalteniece A, Petropoulos I, Azmi S, Dhage S, Marshall A, . . . Malik RA. Diabetic Neuropathy Is Characterized by Progressive Corneal Nerve Fiber Loss in the Central and Inferior Whorl Regions. Investigative Ophthalmology and Visual Science 2020b;61(3):48.
Fernyhough P, McGavock J. Chapter 25 - Mechanisms of disease: Mitochondrial dysfunction in sensory neuropathy and other complications in diabetes. In: Zochodne DW, Malik RA, editors. Handbook of Clinical Neurology. 126: Elsevier; 2014. p. 353-377.
Fitzhugh R. Computation of Impulse Initiation and Saltatory Conduction in a Myelinated Nerve Fiber. Biophysical Journal 1962;2(1):11-21.
Fleckenstein J, Sittl R, Averbeck B, Lang PM, Irnich D, Carr RW. Activation of axonal Kv7 channels in human peripheral nerve by flupirtine but not placebo - therapeutic potential for peripheral neuropathies: Results of a randomised controlled trial. Journal of Translational Medicine 2013;11(1):34.
Forbes JM, Coughlan MT, Cooper ME. Oxidative stress as a major culprit in kidney disease in diabetes. Diabetes 2008;57(6):1446-54.
Fornage BD. Peripheral nerves of the extremities: imaging with US. Radiology 1988;167(1):179-82.
251 References
Fourlanos S, Dotta F, Greenbaum CJ, Palmer JP, Rolandsson O, Colman PG, Harrison LC. Latent autoimmune diabetes in adults (LADA) should be less latent. Diabetologia 2005;48(11):2206-12.
Frank T, Nawroth P, Kuner R. Structure-function relationships in peripheral nerve contributions to diabetic peripheral neuropathy. Pain 2019;160 Suppl 1:S29-S36.
Fraser SD, Blakeman T. Chronic kidney disease: identification and management in primary care. Pragmatic and observational research 2016;7:21-32.
Freeman SA, Desmazieres A, Fricker D, Lubetzki C, Sol-Foulon N. Mechanisms of sodium channel clustering and its influence on axonal impulse conduction. Cellular and molecular life sciences 2016;73(4):723-35.
Frontoni S, Di Bartolo P, Avogaro A, Bosi E, Paolisso G, Ceriello A. Glucose variability: An emerging target for the treatment of diabetes mellitus. Diabetes Research and Clinical Practice 2013;102(2):86-95.
Gaede P, Lund-Andersen H, Parving HH, Pedersen O. Effect of a multifactorial intervention on mortality in type 2 diabetes. The New England journal of medicine 2008;358(6):580-91.
Galassi G, Ferrari S, Cobelli M, Rizzuto N. Neuromuscular complications of kidney diseases. Nephrology Dialysis Transplantation 1998;13(suppl_7):41-47.
Garg N, Park SB, Howells J, Noto YI, Vucic S, Yiannikas C, . . . Kiernan MC. Anti- MAG neuropathy: Role of IgM antibodies, the paranodal junction and juxtaparanodal potassium channels. Clinical Neurophysiology 2018;129(10):2162-2169.
Gasparotti R, Padua L, Briani C, Lauria G. New technologies for the assessment of neuropathies. Nature Reviews Neurology 2017;13(4):203-216.
Geraldes P, King GL. Activation of protein kinase C isoforms and its impact on diabetic complications. Circulation Research 2010;106(8):1319-1331.
Gerbi A, Maixent JM, Barbey O, Jamme I, Pierlovisi M, Coste T, . . . Raccah D. Alterations of Na,K-ATPase isoenzymes in the rat diabetic neuropathy: protective effect of dietary supplementation with n-3 fatty acids. Journal of neurochemistry 1998;71(2):732-40.
Gianfrancesco MA, Dehairs J, L'Homme L, Herinckx G, Esser N, Jansen O, . . . Legrand-Poels S. Saturated fatty acids induce NLRP3 activation in human macrophages through K(+) efflux resulting from phospholipid saturation and Na, K- ATPase disruption. Biochimica et biophysica acta Molecular and cell biology of lipids 2019;1864(7):1017-1030.
Gilchrist LS, Tanner L. The pediatric-modified total neuropathy score: a reliable and valid measure of chemotherapy-induced peripheral neuropathy in children with non- CNS cancers. Supportive care in cancer 2013;21(3):847-56.
252 References
Gilling M, Rasmussen HB, Calloe K, Sequeira AF, Baretto M, Oliveira G, . . . Tommerup N. Dysfunction of the Heteromeric KV7.3/KV7.5 Potassium Channel is Associated with Autism Spectrum Disorders. Frontiers in Genetics 2013;4:54.
Goldin AL. Resurgence of sodium channel research. Annual Review of Physiology 2001;63:871-894.
Goncalves NP, Vaegter CB, Andersen H, Ostergaard L, Calcutt NA, Jensen TS. Schwann cell interactions with axons and microvessels in diabetic neuropathy. Nature Reviews Neurology 2017;13(3):135-147.
Gong N, Xiao Q, Zhu B, Zhang CY, Wang YC, Fan H, . . . Wang YX. Activation of spinal glucagon-like peptide-1 receptors specifically suppresses pain hypersensitivity. The Journal of Neuroscience 2014;34(15):5322-34.
Greene DA. A sodium-pump defect in diabetic peripheral nerve corrected by sorbinil administration: Relationship to myo-inositol metabolism and nerve conduction slowing. Metabolism 1986a;35(4 SUPPL. 1):60-65.
Greene DA. Sorbitol, Myo-Inositol and Sodium-Potassium ATPase in Diabetic Peripheral Nerve. Drugs 1986b;32(2):6-14.
Greene DA, Lattimer SA. Impaired energy utilization and Na-K-ATPase in diabetic peripheral nerve. American Journal of Physiology - Endocrinology and Metabolism 1984;9(4):E311-E318.
Greene DA, Lattimer SA, Sima AAF. Are disturbances of sorbitol, phosphoinositide, and Na+-K+-ATPase regulation involved in pathogenesis of diabetic neuropathy? Diabetes 1988;37(6):688-693.
Grisold A, Callaghan BC, Feldman EL. Mediators of diabetic neuropathy: is hyperglycemia the only culprit? Current opinion in endocrinology, diabetes, and obesity 2017;24(2):103-111.
Grote CW, Wright DE. A Role for Insulin in Diabetic Neuropathy. Frontiers in cellular neuroscience 2016;10:581.
Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, . . . Blood I. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005;112(17):2735-52.
Guglielmi C, Palermo A, Pozzilli P. Latent autoimmune diabetes in the adults (LADA) in Asia: from pathogenesis and epidemiology to therapy. Diabetes/metabolism research and reviews 2012;28 Suppl 2:40-6.
Gurnett CA, Campbell KP. Transmembrane auxiliary subunits of voltage-dependent ion channels. Journal of Biological Chemistry 1996;271(45):27975-27978.
Gutman GA, Chandy KG, Grissmer S, Lazdunski M, McKinnon D, Pardo LA, . . . Wang X. International Union of Pharmacology. LIII. Nomenclature and molecular
253 References relationships of voltage-gated potassium channels. Pharmacological Reviews 2005;57(4):473-508.
Hallan SI, Dahl K, Oien CM, Grootendorst DC, Aasberg A, Holmen J, Dekker FW. Screening strategies for chronic kidney disease in the general population: follow-up of cross sectional health survey. British medical journal 2006;333(7577):1047.
Haller MJ, Atkinson MA, Schatz D. Type 1 diabetes mellitus: etiology, presentation, and management. Pediatric clinics of North America 2005;52(6):1553-78.
Han L, Ji L, Chang J, Wen J, Zhao W, Shi H, . . . Lu B. Peripheral neuropathy is associated with insulin resistance independent of metabolic syndrome. Diabetology and Metabolic Syndrome 2015;7(1):14.
Hanewinckel R, Drenthen J, Ligthart S, Dehghan A, Franco OH, Hofman A, . . . van Doorn PA. Metabolic syndrome is related to polyneuropathy and impaired peripheral nerve function: a prospective population-based cohort study. Journal of neurology, neurosurgery, and psychiatry 2016a;87(12):1336-1342.
Hanewinckel R, Ikram MA, Franco OH, Hofman A, Drenthen J, van Doorn PA. High body mass and kidney dysfunction relate to worse nerve function, even in adults without neuropathy. Journal of the Peripheral Nervous System 2017;22(2):112-120.
Hanewinckel R, Ikram MA, van Doorn PA. Assessment scales for the diagnosis of polyneuropathy. Journal of the Peripheral Nervous System 2016b;21(2):61-73.
Hatou S, Yamada M, Akune Y, Mochizuki H, Shiraishi A, Joko T, . . . Tsubota K. Role of Insulin in Regulation of Na+-/K+-Dependent ATPase Activity and Pump Function in Corneal Endothelial Cells. Investigative Ophthalmology & Visual Science 2010;51(8):3935-3942.
Heinemeyer O, Reimers CD. Ultrasound of radial, ulnar, median, and sciatic nerves in healthy subjects and patients with hereditary motor and sensory neuropathies. Ultrasound in Medicine and Biology 1999;25(3):481-485.
Hertz P, Bril V, Orszag A, Ahmed A, Ng E, Nwe P, . . . Perkins BA. Reproducibility of in vivo corneal confocal microscopy as a novel screening test for early diabetic sensorimotor polyneuropathy. Diabetic Medicine 2011;28(10):1253-60.
Herzog RI, Cummins TR, Ghassemi F, Dib-Hajj SD, Waxman SG. Distinct repriming and closed-state inactivation kinetics of Nav1.6 and Nav1.7 sodium channels in mouse spinal sensory neurons. The Journal of Physiology 2003;551(3):741-750.
Hess K, Eames RA, Darveniza P, Gilliatt RW. Acute ischaemic neuropathy in the rabbit. Journal of the Neurological Sciences 1979;44(1):19-43.
Hibino H, Inanobe A, Furutani K, Murakami S, Findlay I, Kurachi Y. Inwardly Rectifying Potassium Channels: Their Structure, Function, and Physiological Roles. Physiological Reviews 2010;90(1):291-366.
254 References
Hill NR, Fatoba ST, Oke JL, Hirst JA, O'Callaghan CA, Lasserson DS, Hobbs FDR. Global Prevalence of Chronic Kidney Disease - A Systematic Review and Meta- Analysis. PLoS One 2016;11(7):e0158765-e0158765.
Himeno T, Kamiya H, Naruse K, Harada N, Ozaki N, Seino Y, . . . Nakamura J. Beneficial effects of exendin-4 on experimental polyneuropathy in diabetic mice. Diabetes 2011;60(9):2397-406.
Hinnen D. Glucagon-Like Peptide 1 Receptor Agonists for Type 2 Diabetes. Diabetes Spectrum 2017;30(3):202-210.
Hivert B, Pinatel D, Labasque M, Tricaud N, Goutebroze L, Faivre-Sarrailh C. Assembly of juxtaparanodes in myelinating DRG culture: Differential clustering of the Kv1/Caspr2 complex and scaffolding protein 4.1B. Glia 2016;64(5):840-52.
Hiyama TY, Watanabe E, Ono K, Inenaga K, Tamkun MM, Yoshida S, Noda M. Nax channel involved in CNS sodium-level sensing. Nature Neuroscience 2002;5(6):511- 512.
Hodgkin AL, Huxley AF. The components of membrane conductance in the giant axon of Loligo. The Journal of Physiology 1952a;116(4):473-496.
Hodgkin AL, Huxley AF. Currents carried by sodium and potassium ions through the membrane of the giant axon of Loligo. The Journal of Physiology 1952b;116(4):449- 472.
Hodgkin AL, Huxley AF. The dual effect of membrane potential on sodium conductance in the giant axon of Loligo. The Journal of Physiology 1952c;116(4):497-506.
Hodgkin AL, Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. The Journal of Physiology 1952d;117(4):500-544.
Hodgkin AL, Katz B. The effect of sodium ions on the electrical activity of the giant axon of the squid. The Journal of Physiology 1949;108(1):37-77.
Hoeldtke RD, Bryner KD, Hobbs GR, Horvath GG, Riggs JE, Christie I, . . . Lernmark A. Antibodies to glutamic acid decarboxylase and peripheral nerve function in type 1 diabetes. Journal of Clinical Endocrinology and Metabolism 2000;85(9):3297-3308.
Hohman TC, Cotter MA, Cameron NE. ATP-sensitive K+ channel effects on nerve function, Na+, K+ ATPase, and glutathione in diabetic rats. European journal of pharmacology 2000;397(2-3):335-341.
Holley JL. Age, eGFR, and CKD complications. Clinical journal of the American Society of Nephrology 2011;6(12):2729-31.
Holst JJ. The physiology of glucagon-like peptide 1. Physiology reviews 2007;87(4):1409-39.
255 References
Hong S, Wiley JW. Altered expression and function of sodium channels in large DRG neurons and myelinated A-fibers in early diabetic neuropathy in the rat. Biochemical and biophysical research communications 2006;339(2):652-60.
Howells J, Czesnik D, Trevillion L, Burke D. Excitability and the safety margin in human axons during hyperthermia. The Journal of Physiology 2013;591(12):3063- 3080.
Howells J, Matamala JM, Park SB, Garg N, Vucic S, Bostock H, . . . Kiernan MC. In vivo evidence for reduced ion channel expression in motor axons of patients with amyotrophic lateral sclerosis. The Journal of Physiology 2018;596(22):5379-5396.
Howells J, Trevillion L, Bostock H, Burke D. The voltage dependence of I(h) in human myelinated axons. The Journal of Physiology 2012;590(7):1625-40.
Humphries ES, Dart C. Neuronal and Cardiovascular Potassium Channels as Therapeutic Drug Targets: Promise and Pitfalls. Journal of biomolecular screening 2015;20(9):1055-73.
Hussain S, Chand Jamali M, Habib A, Hussain MS, Akhtar M, Najmi AK. Diabetic kidney disease: An overview of prevalence, risk factors, and biomarkers. Clinical Epidemiology and Global Health 2020.
Huxley AF, Stämpeli R. Evidence for saltatory conduction in peripheral myelinated nerve fibres. The Journal of Physiology 1949;108(3):315-339.
Ichimura T, Ellisman MH. Three-dimensional fine structure of cytoskeletal- membrane interactions at nodes of Ranvier. Journal of Neurocytology 1991;20(8):667-681.
Inker LA, Coresh J, Levey AS, Tonelli M, Muntner P. Estimated GFR, albuminuria, and complications of chronic kidney disease. Journal of the American Society of Nephrology 2011;22(12):2322-31.
Inouye H, Liu J, Makowski L, Palmisano M, Burghammer M, Riekel C, Kirschner DA. Myelin Organization in the Nodal, Paranodal, and Juxtaparanodal Regions Revealed by Scanning X-Ray Microdiffraction. PLoS One 2014;9(7):e100592.
International Diabetes Federation. IDF diabetes atlas. 7th Edition ed. Brussels: International Diabetes Federation, 2015.
Isom LL. Sodium channel β subunits: Anything but auxiliary. Neuroscientist 2001;7(1):42-54.
Isomaa B, Almgren P, Henricsson M, Taskinen M-R, et al. Chronic complications in patients with slowly progressing autoimmune type 1 diabetes (LADA). Diabetes Care 1999;22(8):1347-53.
Issar T, Arnold R, Kwai NCG, Pussell BA, Endre ZH, Poynten AM, . . . Krishnan AV. The utility of the Total Neuropathy Score as an instrument to assess neuropathy severity in chronic kidney disease: A validation study. Clinical Neurophysiology 2018;129(5):889-894.
256 References
Issar T, Arnold R, Kwai NCG, Walker S, Yan A, Borire AA, . . . Krishnan AV. Relative contributions of diabetes and chronic kidney disease to neuropathy development in diabetic nephropathy patients. Clinical Neurophysiology 2019;130(11):2088-2095.
Jaiswal M, Martin CL, Brown MB, Callaghan B, Albers JW, Feldman EL, Pop-Busui R. Effects of exenatide on measures of diabetic neuropathy in subjects with type 2 diabetes: results from an 18-month proof-of-concept open-label randomized study. Journal of diabetes and its complications 2015;29(8):1287-94.
Janahi NM, Santos D, Blyth C, Bakhiet M, Ellis M. Diabetic peripheral neuropathy, is it an autoimmune disease? Immunol Lett 2015;168(1):73-79.
Jankelowitz SK, Burke D. Pathophysiology of HNPP explored using axonal excitability. Journal of Neurology, Neurosurgery, and Psychiatry 2013;84(7):806-12.
Jankelowitz SK, Howells J, Burke D. Plasticity of inwardly rectifying conductances following a corticospinal lesion in human subjects. The Journal of Physiology 2007a;581(Pt 3):927-40.
Jankelowitz SK, McNulty PA, Burke D. Changes in measures of motor axon excitability with age. Clinical Neurophysiology 2007b;118(6):1397-404.
Jha SK, Jha NK, Kumar D, Ambasta RK, Kumar P. Linking mitochondrial dysfunction, metabolic syndrome and stress signaling in Neurodegeneration. Biochim Biophys Acta Molecular Basis of Disease 2017;1863(5):1132-1146.
Jia X, Wang X, Wang X, Pan Q, Xian T, Yu X, Guo L. In Vivo Corneal Confocal Microscopy Detects Improvement of Corneal Nerve Parameters following Glycemic Control in Patients with Type 2 Diabetes. Journal of diabetes research 2018;2018:8516276.
Jiang G, Luk AOY, Tam CHT, Xie F, Carstensen B, Lau ESH, . . . Hong Kong Diabetes Register TRSSG. Progression of diabetic kidney disease and trajectory of kidney function decline in Chinese patients with Type 2 diabetes. Kidney International 2019;95(1):178-187.
Jiang Y-Q, Sun Q, Tu H-Y, Wan Y. Characteristics of HCN Channels and Their Participation in Neuropathic Pain. Neurochemical Research 2008;33(10):1979-1989.
Jolivalt CG, Fineman M, Deacon CF, Carr RD, Calcutt NA. GLP-1 signals via ERK in peripheral nerve and prevents nerve dysfunction in diabetic mice. Diabetes, obesity & metabolism 2011;13(11):990-1000.
Jonas P, Koh DS, Kampe K, Hermsteiner M, Vogel W. ATP-sensitive and Ca- activated K channels in vertebrate axons: novel links between metabolism and excitability. Pflügers Archiv European Journal of Physiology 1991;418(1-2):68-73.
Kaji R, Sumner AJ. Ouabain reverses conduction disturbances in single demyelinated nerve fibers. Neurology 1989;39(10):1364-1364.
257 References
Kalteniece A, Ferdousi M, Adam S, Schofield J, Azmi S, Petropoulos I, . . . Malik RA. Corneal confocal microscopy is a rapid reproducible ophthalmic technique for quantifying corneal nerve abnormalities. PLoS One 2017;12(8).
Kalteniece A, Ferdousi M, Petropoulos I, Azmi S, Adam S, Fadavi H, . . . Malik RA. Greater corneal nerve loss at the inferior whorl is related to the presence of diabetic neuropathy and painful diabetic neuropathy. Scientific Reports 2018;8(1):3283.
Kamel J, Loh M, Cook M, MacIsaac RJ, Roberts LJ. Reducing glucose variability with continuous subcutaneous insulin infusion is associated with reversal of axonal dysfunction in type 1 diabetes mellitus. Muscle & Nerve 2020;61(1):44-51.
Kan M, Guo G, Singh B, Singh V, Zochodne DW. Glucagon-like peptide 1, insulin, sensory neurons, and diabetic neuropathy. Journal of Neuropathology and Experimental Neurology 2012;71(6):494-510.
Karalliedde J, Gnudi L. Diabetes mellitus, a complex and heterogeneous disease, and the role of insulin resistance as a determinant of diabetic kidney disease. Nephrology Dialysis Transplantation 2016;31(2):206-13.
Karunaratne K, Taube D, Khalil N, Perry R, Malhotra PA. Neurological complications of renal dialysis and transplantation. Practical Neurology 2018;18(2):115-125.
Kawano T, Zoga V, Kimura M, Liang M-Y, Wu H-E, Gemes G, . . . Sarantopoulos CD. Nitric oxide activates ATP-sensitive potassium channels in mammalian sensory neurons: action by direct S-nitrosylation. Molecular pain 2009;5:12-12.
Kiernan MC, Bostock H. Effects of membrane polarization and ischaemia on the excitability properties of human motor axons. Brain 2000;123(12):2542-2551.
Kiernan MC, Bostock H, Park SB, Kaji R, Krarup C, Krishnan AV, . . . Burke D. Measurement of axonal excitability: Consensus guidelines. Clinical Neurophysiology 2020;131(1):308-323.
Kiernan MC, Burke D, Andersen KV, Bostock H. Multiple measures of axonal excitability: a new approach in clinical testing. Muscle Nerve 2000;23(3):399-409.
Kiernan MC, Burke D, Bostock H. Chapter 5 - Nerve Excitability Measures: Biophysical Basis and Use in the Investigation of Peripheral Nerve Disease. In: Dyck PJ, Thomas PK, editors. Peripheral Neuropathy (Fourth Edition). Philadelphia: W.B. Saunders; 2005a. p. 113-129.
Kiernan MC, Cikurel K, Bostock H. Effects of temperature on the excitability properties of human motor axons. Brain 2001;124(Pt 4):816-25.
Kiernan MC, Isbister GK, Lin CS, Burke D, Bostock H. Acute tetrodotoxin-induced neurotoxicity after ingestion of puffer fish. Annals of neurology 2005b;57(3):339-48.
Kiernan MC, Kaji R. Chapter 4 - Physiology and pathophysiology of myelinated nerve fibers. In: Said G, Krarup C, editors. Handbook of Clinical Neurology. 115: Elsevier; 2013. p. 43-53.
258 References
Kiernan MC, Mogyoros I, Burke D. Differences in the recovery of excitability in sensory and motor axons of human median nerve. Brain 1996;119 (Pt 4):1099-105.
Kiernan MC, Walters RJ, Andersen KV, Taube D, Murray NM, Bostock H. Nerve excitability changes in chronic renal failure indicate membrane depolarization due to hyperkalaemia. Brain 2002;125(Pt 6):1366-78.
Kim B, Feldman EL. Insulin resistance in the nervous system. Trends in Endocrinology and Metabolism 2012;23(3):133-141.
Kiss T. Persistent Na-channels: origin and function. A review. Acta biologica Hungarica 2008;59 Suppl:1-12.
Kitano Y, Kuwabara S, Misawa S, Ogawara K, Kanai K, Kikkawa Y, . . . Hattori T. The acute effects of glycemic control on axonal excitability in human diabetics. Annals of Neurology 2004;56(4):462-467.
Kjellstrand CM. Do middle molecules cause uremic intoxication? (Con). American Journal of Kidney Diseases 1981;1(1):51-6.
Kjellstrand CM, Evans RL, Petersen RJ, Rust LW, Shideman J, Buselmeier TJ, Rozelle LT. Considerations of the middle molecule hypothesis. Proceedings of the Clinical Dialysis and Transplant Forum 1972;2:127-32.
Kolaczynski WM, Hankins M, Ong SH, Richter H, Clemens A, Toussi M. Microvascular Outcomes in Patients with Type 2 Diabetes Treated with Vildagliptin vs. Sulfonylurea: A Retrospective Study Using German Electronic Medical Records. Diabetes Therapy 2016;7(3):483-96.
Kovalchuk MO, Franssen H, Van Schelven LJ, Sleutjes B. Comparing excitability at 37 degrees C versus at 20 degrees C: Differences between motor and sensory axons. Muscle Nerve 2018;57(4):574-580.
Krarup C, Moldovan M. Nerve conduction and excitability studies in peripheral nerve disorders. Current Opinion in Neurology 2009;22(5):460-6.
Krarup C, Moldovan M. Reappraising I(h:) do myelinated motor and sensory axons of human peripheral nerves operate at different resting membrane potentials? The Journal of Physiology 2012;590(7):1515-1516.
Krishnan AV, Kiernan MC. Altered nerve excitability properties in established diabetic neuropathy. Brain 2005;128(Pt 5):1178-87.
Krishnan AV, Kiernan MC. Uremic neuropathy: clinical features and new pathophysiological insights. Muscle Nerve 2007;35(3):273-90.
Krishnan AV, Kiernan MC. Neurological complications of chronic kidney disease. Nature Reviews Neurology 2009;5(10):542-51.
Krishnan AV, Lin CS, Kiernan MC. Excitability differences in lower-limb motor axons during and after ischemia. Muscle Nerve 2005a;31(2):205-13.
259 References
Krishnan AV, Lin CS, Park SB, Kiernan MC. Axonal ion channels from bench to bedside: a translational neuroscience perspective. Progress in Neurobiology 2009a;89(3):288-313.
Krishnan AV, Lin CSY, Kiernan MC. Activity-dependent excitability changes suggest Na+/K+ pump dysfunction in diabetic neuropathy. Brain 2008;131(5):1209- 1216.
Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bostock H, Kiernan MC. Altered motor nerve excitability in end-stage kidney disease. Brain 2005b;128(Pt 9):2164-74.
Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Bostock H, Kiernan MC. Neuropathy, axonal Na+/K+ pump function and activity-dependent excitability changes in end-stage kidney disease. Clinical Neurophysiology 2006a;117(5):992-9.
Krishnan AV, Phoon RK, Pussell BA, Charlesworth JA, Kiernan MC. Sensory nerve excitability and neuropathy in end stage kidney disease. Journal of Neurology, Neurosurgery, and Psychiatry 2006b;77(4):548-51.
Krishnan AV, Phoon RKS, Pussell BA, Charlesworth JA, Bostock H, Kiernan MC. Ischaemia induces paradoxical changes in axonal excitability in end-stage kidney disease. Brain 2006c;129(6):1585-1592.
Krishnan AV, Pussell BA, Kiernan MC. Neuromuscular disease in the dialysis patient: an update for the nephrologist. Seminars in Dialysis 2009b;22(3):267-78.
Kuang Q, Purhonen P, Hebert H. Structure of potassium channels. Cellular and molecular life sciences 2015;72(19):3677-93.
Kuriscák E, Trojan S, Wünsch Z. Model of spike propagation reliability along the myelinated axon corrupted by axonal intrinsic noise sources. Physiological research 2002;51(3):205-215.
Kursula P. Myelin. In: Aminoff MJ, Daroff RB, editors. Encyclopedia of the Neurological Sciences (Second Edition). Oxford: Academic Press; 2014. p. 240-241.
Kuwabara S, Cappelen-Smith C, Lin CSY, Mogyoros I, Burke D. Differences in accommodative properties of median and peroneal motor axons. Journal of Neurology Neurosurgery and Psychiatry 2001;70(3):372-376.
Kuwabara S, Misawa S. Pharmacologic intervention in axonal excitability: in vivo assessment of nodal persistent sodium currents in human neuropathies. Current molecular pharmacology 2008;1(1):61-67.
Kuwabara S, Misawa S, Kanai K, Tamura N, Nakata M, Sawai S, Hattori T. The effects of physiological fluctuation of serum potassium levels on excitability properties in healthy human motor axons. Clinical Neurophysiology 2007;118(2):278- 282.
260 References
Kuwabara S, Misawa S, Tamura N, Kanai K, Hiraga A, Ogawara K, . . . Hattori T. The effects of mexiletine on excitability properties of human median motor axons. Clinical Neurophysiology 2005;116(2):284-289.
Kuwabara S, Ogawara K, Harrori T, Suzuki Y, Hashimoto N. The acute effects of glycemic control on axonal excitability in human diabetic nerves. Internal Medicine 2002;41(5):360-5.
Kwai N, Arnold R, Poynten AM, Lin CS, Kiernan MC, Krishnan AV. Continuous subcutaneous insulin infusion preserves axonal function in type 1 diabetes mellitus. Diabetes/metabolism research and reviews 2015;31(2):175-82.
Kwai NC, Arnold R, Poynten AM, Howells J, Kiernan MC, Lin CS, Krishnan AV. In vivo evidence of reduced nodal and paranodal conductances in type 1 diabetes. Clinical Neurophysiology 2016a;127(2):1700-6.
Kwai NC, Arnold R, Poynten AM, Krishnan AV. Association between glycemic variability and peripheral nerve dysfunction in type 1 diabetes. Muscle & Nerve 2016b;54(5):967-969.
Kwai NC, Arnold R, Wickremaarachchi C, Lin CS, Poynten AM, Kiernan MC, Krishnan AV. Effects of axonal ion channel dysfunction on quality of life in type 2 diabetes. Diabetes Care 2013;36(5):1272-7.
Laaksonen S, Metsärinne K, Voipio-Pulkki L-M, Falck B. Neurophysiologic parameters and symptoms in chronic renal failure. Muscle & Nerve 2002;25(6):884- 890.
Lapicque L. Définition experimentale de l'excitabilité. 1909.
Lascelles RG, Thomas PK. Changes due to age in internodal length in the sural nerve in man. Journal of neurology, neurosurgery, and psychiatry 1966;29(1):40-44.
Laugesen E, Ostergaard JA, Leslie RD, Danish Diabetes Academy W, Workshop S. Latent autoimmune diabetes of the adult: current knowledge and uncertainty. Diabetic Medicine 2015;32(7):843-52.
Lauria G, Borgna M, Morbin M, Lombardi R, Mazzoleni G, Sghirlanzoni A, Pareyson D. Tubule and neurofilament immunoreactivity in human hairy skin: markers for intraepidermal nerve fibers. Muscle Nerve 2004;30(3):310-6.
Lauria G, Cazzato D, Porretta-Serapiglia C, Casanova-Molla J, Taiana M, Penza P, . . . Merkies IS. Morphometry of dermal nerve fibers in human skin. Neurology 2011;77(3):242-9.
Lauria G, Merkies ISJ, Waxman SG, Faber CG. Epidermal Nerve Fibers. In: Aminoff MJ, Daroff RB, editors. Encyclopedia of the Neurological Sciences (Second Edition). Oxford: Academic Press; 2014. p. 76-79.
Lee CM, Colagiuri R, Magliano DJ, Cameron AJ, Shaw J, Zimmet P, Colagiuri S. The cost of diabetes in adults in Australia. Diabetes Research and Clinical Practice 2013;99(3):385-90.
261 References
Lee D, Dauphinée DM. Morphological and Functional Changes in the Diabetic Peripheral Nerve: Using Diagnostic Ultrasound and Neurosensory Testing to Select Candidates for Nerve Decompression. Journal of the American Podiatric Medical Association 2005;95(5):433-437.
Lee KA, Jin HY, Lee NY, Kim YJ, Park TS. Effect of Empagliflozin, a Selective Sodium-Glucose Cotransporter 2 Inhibitor, on Kidney and Peripheral Nerves in Streptozotocin-Induced Diabetic Rats. Diabetes & Metabolism Journal 2018;42(4):338-342.
Leterrier C. The Axon Initial Segment: An Updated Viewpoint. The Journal of Neuroscience 2018;38(9):2135-2145.
Levey AS, Eckardt KU, Tsukamoto Y, Levin A, Coresh J, Rossert J, . . . Eknoyan G. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney International 2005;67(6):2089-100.
Lewis EJH, Perkins BA, Lovblom LE, Bazinet RP, Wolever TMS, Bril V. Effect of omega-3 supplementation on neuropathy in type 1 diabetes: A 12-month pilot trial. Neurology 2017;88(24):2294-2301.
Liang C, Howells J, Kennerson M, Nicholson GA, Burke D, Ng K. Axonal excitability in X-linked dominant Charcot Marie Tooth disease. Clinical Neurophysiology 2014;125(6):1261-1269.
Liao J, Li H, Zeng W, Sauer DB, Belmares R, Jiang Y. Structural insight into the ion- exchange mechanism of the sodium/calcium exchanger. Science 2012;335(6069):686- 690.
Liao J, Marinelli F, Lee C, Huang Y, Faraldo-Gómez JD, Jiang Y. Mechanism of extracellular ion exchange and binding-site occlusion in a sodium/calcium exchanger. Nature Structural and Molecular Biology 2016;23(6):590-599.
Lin CS-Y, Krishnan AV, Lee M-J, Zagami AS, You H-L, Yang C-C, . . . Kiernan MC. Nerve function and dysfunction in acute intermittent porphyria. Brain 2008;131(9):2510-2519.
Lin CSY, Mogyoros I, Burke D. Recovery of excitability of cutaneous afferents in the median and sural nerves following activity. Muscle and Nerve 2000;23(5):763-770.
Lin YK, Chen YC, Kao YH, Tsai CF, Yeh YH, Huang JL, . . . Chen YJ. A monounsaturated fatty acid (oleic acid) modulates electrical activity in atrial myocytes with calcium and sodium dysregulation. International journal of cardiology 2014;176(1):191-8.
Liu L, Li X, Xiang Y, Huang G, Lin J, Yang L, . . . Zhou Z. Latent Autoimmune Diabetes in Adults With Low-Titer GAD Antibodies: Similar Disease Progression With Type 2 Diabetes. A Nationwide, Multicenter Prospective Study (LADA China Study 3) 2015;38(1):16-21.
262 References
Liu WJ, Jin HY, Lee KA, Xie SH, Baek HS, Park TS. Neuroprotective effect of the glucagon-like peptide-1 receptor agonist, synthetic exendin-4, in streptozotocin- induced diabetic rats. British journal of pharmacology 2011;164(5):1410-20.
Liveson JA, Ma DM. Laboratory Reference for Clinical Neurophysiology. New York: Oxford University Press, 1992.
Long SB, Tao X, Campbell EB, MacKinnon R. Atomic structure of a voltage- dependent K+ channel in a lipid membrane-like environment. Nature 2007;450(7168):376-82.
Louraki M, Katsalouli M, Kanaka-Gantenbein C, Kafassi N, Critselis E, Kallinikou D, . . . Karavanaki K. The prevalence of early subclinical somatic neuropathy in children and adolescents with Type 1 diabetes mellitus and its association with the persistence of autoantibodies to glutamic acid decarboxylase (GAD) and islet antigen- 2 (IA-2). Diabetes Research and Clinical Practice 2016;117:82-90.
Lu J, Hou X, Zhang L, Hu C, Zhou J, Pang C, . . . Jia W. Associations between clinical characteristics and chronic complications in latent autoimmune diabetes in adults and type 2 diabetes. Diabetes/metabolism research and reviews 2015;31(4):411-20.
Lu J, Ma X, Zhou J, Zhang L, Mo Y, Ying L, . . . Jia W. Association of Time in Range, as Assessed by Continuous Glucose Monitoring, With Diabetic Retinopathy in Type 2 Diabetes. Diabetes Care 2018;41(11):2370-2376.
Luciani P, Deledda C, Benvenuti S, Cellai I, Squecco R, Monici M, . . . Peri A. Differentiating effects of the glucagon-like peptide-1 analogue exendin-4 in a human neuronal cell model. Cellular and molecular life sciences 2010;67(21):3711-23.
Ma J, Shi M, Zhang X, Liu X, Chen J, Zhang R, . . . Zhang H. GLP1R agonists ameliorate peripheral nerve dysfunction and inflammation via p38 MAPK/NFkappaB signaling pathways in streptozotocininduced diabetic rats. International journal of molecular medicine 2018;41(5):2977-2985.
Mahelkova G, Burdova MC, Mala S, Hoskovcova L, Dotrelova D, Stechova K. Higher Total Insulin Dose Has Positive Effect on Corneal Nerve Fibers in DM1 Patients. Investigative Ophthalmology and Visual Science 2018;59(10):3800-3807.
Malik RA. Chapter 18 - Pathology of human diabetic neuropathy. In: Zochodne DW, Malik RA, editors. Handbook of Clinical Neurology. 126: Elsevier; 2014. p. 249-259.
Malik RA. Diabetic neuropathy: A focus on small fibres. Diabetes/Metabolism Research and Reviews 2020;36(S1):e3255.
Malik RA, Veves A, Tesfaye S, Smith G, Cameron N, Zochodne D, Lauria G. Small fibre neuropathy: Role in the diagnosis of diabetic sensorimotor polyneuropathy. Diabetes/Metabolism Research and Reviews 2011;27(7):678-684.
Marfurt CF, Cox J, Deek S, Dvorscak L. Anatomy of the human corneal innervation. Experimental eye research 2010;90(4):478-92.
263 References
Marics G, Lendvai Z, Lódi C, Koncz L, Zakariás D, Schuster G, . . . Tóth-Heyn P. Evaluation of an open access software for calculating glucose variability parameters of a continuous glucose monitoring system applied at pediatric intensive care unit. Biomedical engineering online 2015;14:37-37.
Markoulli M, Flanagan J, Tummanapalli SS, Wu J, Willcox M. The impact of diabetes on corneal nerve morphology and ocular surface integrity. The Ocular Surface 2018;16(1):45-57.
Mata M, Fink DJ, Ernst SA, Siegel GJ. Immunocytochemical Demonstration of Na+,K+‐ATPase in Internodal Axolemma of Myelinated Fibers of Rat Sciatic and Optic Nerves. Journal of neurochemistry 1991;57(1):184-192.
Maurer K, Hopf HC, Lowitzsch K. Hypokalemia shortens relative refractory period of peripheral sensory nerves in man. Journal of Neurology 1977;216(1):67-71.
Mayer ML, Westbrook GL. A voltage‐clamp analysis of inward (anomalous) rectification in mouse spinal sensory ganglion neurones. The Journal of Physiology 1983;340(1):19-45.
McCormick DA. Chapter 5 - Membrane Potential and Action Potential. In: Squire LR, Berg D, Bloom FE, du Lac S, Ghosh A, Spitzer NC, editors. Fundamental Neuroscience (Fourth Edition). San Diego: Academic Press; 2013. p. 93-116.
McDermott LA, Weir GA, Themistocleous AC, Segerdahl AR, Blesneac I, Baskozos G, . . . Bennett DLH. Defining the Functional Role of NaV1.7 in Human Nociception. Neuron 2019;101(5):905-919.e8.
McDonnell CM, Donath SM, Vidmar SI, Werther GA, Cameron FJ. A Novel Approach to Continuous Glucose Analysis Utilizing Glycemic Variation. Diabetes Technology & Therapeutics 2005;7(2):253-263.
McHugh JC, Reilly RB, Connolly S. Examining the effects of age, sex, and body mass index on normative median motor nerve excitability measurements. Clinical Neurophysiology 2011;122(10):2081-8.
Medical Research Council of the United Kingdom. Aids to the examination of the peripheral nervous system: H.M.S.O, London, 1976.
Menegoz M, Gaspar P, Le Bert M, Galvez T, Burgaya F, Palfrey C, . . . Girault J-A. Paranodin, a Glycoprotein of Neuronal Paranodal Membranes. Neuron 1997;19(2):319-331.
Meng X, Gong C, Cao B, Peng X, Wu D, Gu Y, . . . Su C. Glucose fluctuations in association with oxidative stress among children with T1DM: comparison of different phases. Journal of Clinical Endocrinology and Metabolism 2015;100(5):1828-36.
Miller KE. Axons. In: Aminoff MJ, Daroff RB, editors. Encyclopedia of the Neurological Sciences (Second Edition). Oxford: Academic Press; 2014. p. 361-362.
264 References
Misawa S, Kuwabara S, Kanai K, Tamura N, Hiraga A, Nakata M, . . . Hattori T. Axonal potassium conductance and glycemic control in human diabetic nerves. Clinical Neurophysiology 2005a;116(5):1181-7.
Misawa S, Kuwabara S, Kanai K, Tamura N, Nakata M, Ogawara K, . . . Hattori T. Nodal persistent Na+ currents in human diabetic nerves estimated by the technique of latent addition. Clinical Neurophysiology 2006a;117(4):815-820.
Misawa S, Kuwabara S, Kanai K, Tamura N, Nakata M, Sawai S, . . . Hattori T. Aldose reductase inhibition alters nodal Na+ currents and nerve conduction in human diabetics. Neurology 2006b;66(10):1545-9.
Misawa S, Kuwabara S, Ogawara K, Kitano Y, Hattori T. Strength-duration properties and glycemic control in human diabetic motor nerves. Clinical Neurophysiology 2005b;116(2):254-8.
Misawa S, Kuwabara S, Ogawara K, Kitano Y, Yagui K, Hattori T. Hyperglycemia alters refractory periods in human diabetic neuropathy. Clinical Neurophysiology 2004;115(11):2525-9.
Misawa S, Sakurai K, Shibuya K, Isose S, Kanai K, Ogino J, . . . Kuwabara S. Neuropathic pain is associated with increased nodal persistent Na+ currents in human diabetic neuropathy. Journal of the Peripheral Nervous System 2009;14(4):279-284.
Mishra R, Chesi A, Cousminer DL, Hawa MI, Bradfield JP, Hodge KM, . . . Zemel BS. Relative contribution of type 1 and type 2 diabetes loci to the genetic etiology of adult-onset, non-insulin-requiring autoimmune diabetes. BMC Medicine 2017;15(1):1.
Mogyoros I, Kiernan MC, Burke D. Strength-duration properties of human peripheral nerve. Brain 1996;119(2):439-447.
Mogyoros I, Kiernan MC, Burke D, Bostock H. Excitability changes in human sensory and motor axons during hyperventilation and ischaemia. Brain 1997;120 (Pt 2):317-25.
Moldovan M, Lange KHW, Aachmann-Andersen NJ, Kjær TW, Olsen NV, Krarup C. Transient impairment of the axolemma following regional anaesthesia by lidocaine in humans. The Journal of Physiology 2014;592(13):2735-2750.
Monami M, Dicembrini I, Nreu B, Andreozzi F, Sesti G, Mannucci E. Predictors of response to glucagon-like peptide-1 receptor agonists: a meta-analysis and systematic review of randomized controlled trials. Acta Diabetologica 2017;54(12):1101-1114.
Morano S, Tiberti C, Cristina G, Sensi M, Cipriani R, Guidobaldi L, . . . Di Mario U. Autoimmune markers and neurological complications in non-insulin-dependent diabetes mellitus. Human immunology 1999;60(9):848-854.
Moustafa PE, Abdelkader NF, El Awdan SA, El-Shabrawy OA, Zaki HF. Liraglutide ameliorated peripheral neuropathy in diabetic rats: Involvement of oxidative stress, inflammation and extracellular matrix remodeling. Journal of neurochemistry 2018;146(2):173-185.
265 References
Müller LJ, Marfurt CF, Kruse F, Tervo TM. Corneal nerves: structure, contents and function. Experimental eye research 2003;76(5):521-42.
Murray JE, Jankelowitz SK. A comparison of the excitability of motor axons innervating the APB and ADM muscles. Clinical Neurophysiology 2011;122(11):2290-2293.
Muscogiuri G, DeFronzo RA, Gastaldelli A, Holst JJ. Glucagon-like Peptide-1 and the Central/Peripheral Nervous System: Crosstalk in Diabetes. Trends in Endocrinology and Metabolism 2017;28(2):88-103.
Namadurai S, Yereddi NR, Cusdin FS, Huang CLH, Chirgadze DY, Jackson AP. A new look at sodium channel β subunits. Open biology 2015;5(1):140192-140192.
Namazi G, Asa P, Sarrafzadegan N, Pourfarzam M. Decreased Na+/K+-ATPase Activity and Altered Susceptibility to Peroxidation and Lipid Composition in the Erythrocytes of Metabolic Syndrome Patients with Coronary Artery Disease. Annals of nutrition & metabolism 2019;74(2):140-148.
National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. American Journal of Kidney Diseases 2002;39(2 Suppl 1):S1-266.
National Kidney Foundation. KDOQI Clinical Practice Guideline for Diabetes and CKD: 2012 Update. American Journal of Kidney Diseases 2012;60(5):850-886.
Neher E, Sakmann B. Single-channel currents recorded from membrane of denervated frog muscle fibres. Nature 1976;260(5554):799-802.
Ng LL, Hockaday TD. Non-esterified fatty acids may regulate human leucocyte sodium pump activity. Clinical science 1986;71(6):737-42.
Nielsen VK. The peripheral nerve function in chronic renal failure. V. Sensory and motor conduction velocity. Acta medica Scandinavica 1973;194(5):445-54.
Nilsson PM, Tuomilehto J, Rydén L. The metabolic syndrome – What is it and how should it be managed? European Journal of Preventive Cardiology 2019;26(2_suppl):33-46.
Ninčević V, Omanović Kolarić T, Roguljić H, Kizivat T, Smolić M, Bilić Ćurčić I. Renal Benefits of SGLT 2 Inhibitors and GLP-1 Receptor Agonists: Evidence Supporting a Paradigm Shift in the Medical Management of Type 2 Diabetes. International Journal of Molecular Sciences 2019;20(23):5831.
Noda M, Shimizu S, Tanabe T, Takai T, Kayano T, Ikeda T, . . . Numa S. Primary structure of Electrophorus electricus sodium channel deduced from cDNA sequence. Nature 1984;312(5990):121-127.
O'Brien PD, Hinder LM, Callaghan BC, Feldman EL. Neurological consequences of obesity. The Lancet Neurology 2017;16(6):465-477.
266 References
Oates PJ. Aldose reductase, still a compelling target for diabetic neuropathy. Current Drug Targets 2008;9(1):14-36.
Ogawa Y, Schafer DP, Horresh I, Bar V, Hales K, Yang Y, . . . Rasband MN. Spectrins and ankyrinB constitute a specialized paranodal cytoskeleton. The Journal of Neuroscience 2006;26(19):5230-9.
Oliveira-Soto L, Efron N. Morphology of corneal nerves using confocal microscopy. Cornea 2001;20(4):374-84.
Opie EL. On the Relation of Chronic Interstitial Pancreatitis to the Islands of Langerhans and to Diabetes Mellitus. The Journal of experimental medicine 1901;5(4):397-428.
Pape HC. Queer current and pacemaker: The hyperpolarization-activated cation current in neurons. Annual Review of Physiology1996. p. 299-327.
Park SB, Lin CS, Krishnan AV, Goldstein D, Friedlander ML, Kiernan MC. Oxaliplatin-induced neurotoxicity: changes in axonal excitability precede development of neuropathy. Brain 2009;132(Pt 10):2712-23.
Parvanova AI, Trevisan R, Iliev IP, Dimitrov BD, Vedovato M, Tiengo A, . . . Ruggenenti P. Insulin Resistance and Microalbuminuria: A Cross-Sectional, Case- Control Study of 158 Patients With Type 2 Diabetes and Different Degrees of Urinary Albumin Excretion. Diabetes 2006;55(5):1456-1462.
Peles E, Nativ M, Lustig M, Grumet M, Schilling J, Martinez R, . . . Schlessinger J. Identification of a novel contactin-associated transmembrane receptor with multiple domains implicated in protein–protein interactions. The EMBO Journal 1997;16(5):978-988.
Peters KE, Davis WA, Ito J, Winfield K, Stoll T, Bringans SD, . . . Davis TME. Identification of Novel Circulating Biomarkers Predicting Rapid Decline in Renal Function in Type 2 Diabetes: The Fremantle Diabetes Study Phase II. Diabetes Care 2017.
Petropoulos IN, Alam U, Fadavi H, Marshall A, Asghar O, Dabbah MA, . . . Malik RA. Rapid automated diagnosis of diabetic peripheral neuropathy with in vivo corneal confocal microscopy. Investigative Ophthalmology and Visual Science 2014;55(4):2071-8.
Petropoulos IN, Ferdousi M, Marshall A, Alam U, Ponirakis G, Azmi S, . . . Malik RA. The Inferior Whorl For Detecting Diabetic Peripheral Neuropathy Using Corneal Confocal Microscopy. Investigative Ophthalmology and Visual Science 2015;56(4):2498-504.
Petropoulos IN, Manzoor T, Morgan P, Fadavi H, Asghar O, Alam U, . . . Malik RA. Repeatability of in vivo corneal confocal microscopy to quantify corneal nerve morphology. Cornea 2013;32(5):e83-9.
267 References
Petropoulos IN, Ponirakis G, Khan A, Gad H, Almuhannadi H, Brines M, . . . Malik RA. Corneal confocal microscopy: ready for prime time. Clinical and Experimental Optometry 2020;103(3):265-277.
Pop-Busui R, Boulton AJ, Feldman EL, Bril V, Freeman R, Malik RA, . . . Ziegler D. Diabetic Neuropathy: A Position Statement by the American Diabetes Association. Diabetes Care 2017;40(1):136-154.
Pop-Busui R, Herman WH, Feldman EL, Low PA, Martin CL, Cleary PA, . . . Albers JW. DCCT and EDIC studies in type 1 diabetes: lessons for diabetic neuropathy regarding metabolic memory and natural history. Current diabetes reports 2010a;10(4):276-82.
Pop-Busui R, Lu J, Brooks MM, Albert S, Althouse AD, Escobedo J, . . . BARI 2D Study Group. Impact of glycemic control strategies on the progression of diabetic peripheral neuropathy in the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI 2D) Cohort. Diabetes Care 2013;36(10):3208-15.
Pop-Busui R, Roberts L, Pennathur S, Kretzler M, Brosius FC, Feldman EL. The management of diabetic neuropathy in CKD. American Journal of Kidney Diseases 2010b;55(2):365-85.
Pozzilli P, Battelino T, Danne T, Hovorka R, Jarosz-Chobot P, Renard E. Continuous subcutaneous insulin infusion in diabetes: patient populations, safety, efficacy, and pharmacoeconomics. Diabetes/metabolism research and reviews 2016;32(1):21-39.
Pozzilli P, Pieralice S. Latent Autoimmune Diabetes in Adults: Current Status and New Horizons. Endocrinology and Metabolism (Seoul) 2018;33(2):147-159.
Qiao X, Zheng H, Zhang S, Liu S, Xiong Q, Mao F, . . . Lu B. C-peptide is independent associated with diabetic peripheral neuropathy: a community-based study. Diabetology & Metabolic Syndrome 2017;9(1):12.
Rachana KS, Manu MS, Advirao GM. Insulin-induced upregulation of lipoprotein lipase in Schwann cells during diabetic peripheral neuropathy. Diabetes & metabolic syndrome 2018;12(4):525-530.
Rakowski RF, Gadsby DC, De Weer P. Stoichiometry and voltage dependence of the sodium pump in voltage-clamped, internally dialyzed squid giant axon. Journal of General Physiology 1989;93(5):903-941.
Ramahi AA, Ruff RL. Action Potential, Regeneration of. In: Aminoff MJ, Daroff RB, editors. Encyclopedia of the Neurological Sciences (Second Edition). Oxford: Academic Press; 2014. p. 44-46.
Rang HP, Ritchie JM. On the electrogenic sodium pump in mammalian non‐ myelinated nerve fibres and its activation by various external cations. The Journal of Physiology 1968;196(1):183-221.
Rasband MN. It's "juxta" potassium channel. Journal of Neuroscience Research 2004;76(6):749-757.
268 References
Rasband MN. The axon initial segment and the maintenance of neuronal polarity. Nature reviews Neuroscience 2010a;11(8):552-62.
Rasband MN. Clustered K+ channel complexes in axons. Neuroscience letters 2010b;486(2):101-6.
Rasband MN, Trimmer JS, Schwarz TL, Levinson SR, Ellisman MH, Schachner M, Shrager P. Potassium channel distribution, clustering, and function in remyelinating rat axons. Journal of Neuroscience 1998;18(1):36-47.
Rash JE, Vanderpool KG, Yasumura T, Hickman J, Beatty JT, Nagy JI. KV1 channels identified in rodent myelinated axons, linked to Cx29 in innermost myelin: support for electrically active myelin in mammalian saltatory conduction. Journal of neurophysiology 2016;115(4):1836-59.
Redondo MJ. LADA: time for a new definition. Diabetes 2013;62(2):339-40.
Reid G, Scholz A, Bostock H, Vogel W. Human axons contain at least five types of voltage-dependent potassium channel. Journal of Physiology 1999;518(3):681-696.
Reina MA, Sala-Blanch X, Arriazu R, Machés F. Chapter 7 - Microscopic Morphology and Ultrastructure of Human Peripheral Nerves. In: Tubbs RS, Rizk E, Shoja MM, Loukas M, Barbaro N, Spinner RJ, editors. Nerves and Nerve Injuries. San Diego: Academic Press; 2015. p. 91-106.
Riazi S, Bril V, Perkins BA, Abbas S, Chan VWS, Ngo M, . . . Brull R. Can Ultrasound of the Tibial Nerve Detect Diabetic Peripheral Neuropathy? A cross- sectional study 2012;35(12):2575-2579.
Rios JC, Melendez-Vasquez CV, Einheber S, Lustig M, Grumet M, Hemperly J, . . . Salzer JL. Contactin-Associated Protein (Caspr) and Contactin Form a Complex That Is Targeted to the Paranodal Junctions during Myelination. The Journal of Neuroscience 2000;20(22):8354-8364.
Ritchie JM. Physiology of axons. Waxman SG, Kocsis JD, Stys PK, editors: Oxford: Oxford University Press, 1995.
Ritchie JM, Rogart RB. Density of sodium channels in mammalian myelinated nerve fibers and nature of the axonal membrane under the myelin sheath. Proceedings of the National Academy of Sciences of the United States of America 1977;74(1):211-215.
Röper J, Schwarz JR. Heterogeneous distribution of fast and slow potassium channels in myelinated rat nerve fibres. The Journal of Physiology 1989;416(1):93-110.
Rumora AE, Lentz SI, Hinder LM, Jackson SW, Valesano A, Levinson GE, Feldman EL. Dyslipidemia impairs mitochondrial trafficking and function in sensory neurons. FASEB journal 2018;32(1):195-207.
Rumora AE, LoGrasso G, Haidar JA, Dolkowski JJ, Lentz SI, Feldman EL. Chain length of saturated fatty acids regulates mitochondrial trafficking and function in sensory neurons. Journal of lipid research 2019;60(1):58-70.
269 References
Sacchetti M, Lambiase A. Neurotrophic factors and corneal nerve regeneration. Neural Regeneration Research 2017;12(8):1220-1224.
Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, . . . Williams R. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9(th) edition. Diabetes Research and Clinical Practice 2019;157:107843.
Safronov BV, Kampe K, Vogel W. Single voltage‐dependent potassium channels in rat peripheral nerve membrane. The Journal of Physiology 1993;460(1):675-691.
Said G. Uremic neuropathy. Handbook of Clinical Neurology 2013;115:607-12.
Salzer JL, Brophy PJ, Peles E. Molecular domains of myelinated axons in the peripheral nervous system. Glia 2008;56(14):1532-40.
Savelieff MG, Callaghan BC, Feldman EL. The emerging role of dyslipidemia in diabetic microvascular complications. Current Opinion in Endocrinology, Diabetes and Obesity 2020;27(2):115-123.
Scarpini E, Bianchi R, Moggio M, Sciacco M, Fiori MG, Scarlato G. Decrease of nerve Na+/K+-ATPase activity in the pathogenesis of human diabetic neuropathy. Journal of the Neurological Sciences 1993;120(2):159-167.
Schaefer K, Offermann G, von Herrath D, Schröter R, Stölzel R, Arntz HR. Failure to show a correlation between serum parathyroid hormone, nerve conduction velocity and serum lipids in hemodialysis patients. Clinical nephrology 1980;14(2):81-8.
Schofield D, Shrestha RN, Cunich MM, Passey ME, Veerman L, Tanton R, Kelly SJ. The costs of diabetes among Australians aged 45-64 years from 2015 to 2030: projections of lost productive life years (PLYs), lost personal income, lost taxation revenue, extra welfare payments and lost gross domestic product from Health&WealthMOD2030. BMJ Open 2017;7(1):e013158-e013158.
Schwarz JR, Glassmeier G, Cooper EC, Kao TC, Nodera H, Tabuena D, . . . Bostock H. KCNQ channels mediate IKs, a slow K+ current regulating excitability in the rat node of Ranvier. The Journal of Physiology 2006;573(1):17-34.
Schwarz JR, Reid G, Bostock H. Action potentials and membrane currents in the human node of Ranvier. Pflügers Archiv European Journal of Physiology 1995;430(2):283-292.
Sekiguchi K, Kohara N, Baba M, Komori T, Naito Y, Imai T, . . . Hamatani T. Aldose reductase inhibitor ranirestat significantly improves nerve conduction velocity in diabetic polyneuropathy: a randomized double-blind placebo-controlled study in Japan. Journal of diabetes investigation 2019;10(2):466-474.
Selvarajah D, Kar D, Khunti K, Davies MJ, Scott AR, Walker J, Tesfaye S. Diabetic peripheral neuropathy: advances in diagnosis and strategies for screening and early intervention. The Lancet Diabetes & Endocrinology 2019;7(12):938-948.
270 References
Shabeeb D, Najafi M, Hasanzadeh G, Hadian MR, Musa AE, Shirazi A. Electrophysiological measurements of diabetic peripheral neuropathy: A systematic review. Diabetes & metabolic syndrome 2018;12(4):591-600.
Shaheen BS, Bakir M, Jain S. Corneal nerves in health and disease. Survey of ophthalmology 2014;59(3):263-285.
Shekunova EV, Kashkin VA, Muzhikyan Acapital A C, Makarova MN, Makarov VG, Balabanyan VY. Therapeutic efficacy of arginine-rich exenatide on diabetic neuropathy in rats. European journal of pharmacology 2019:172835.
Sherratt RM, Bostock H, Sears TA. Effects of 4-aminopyridine on normal and demyelinated mammalian nerve fibres. Nature 1980;283(5747):570-572.
Shibuta Y, Nodera H, Mori A, Okita T, Kaji R. Peripheral nerve excitability measures at different target levels: the effects of aging and diabetic neuropathy. Journal of clinical neurophysiology 2010;27(5):350-7.
Shibuta Y, Shimatani Y, Nodera H, Izumi Y, Kaji R. Increased variability of axonal excitability in amyotrophic lateral sclerosis. Clinical Neurophysiology 2013;124(10):2046-53.
Silvestri E, Martinoli C, Derchi LE, Bertolotto M, Chiaramondia M, Rosenberg I. Echotexture of peripheral nerves: Correlation between US and histologic findings and criteria to differentiate tendons. Radiology 1995;197(1):291-296.
Sima AA, Brismar T. Reversible diabetic nerve dysfunction: structural correlates to electrophysiological abnormalities. Annals of neurology 1985;18(1):21-9.
Sima AA, Lattimer SA, Yagihashi S, Greene DA. Axo-glial dysjunction. A novel structural lesion that accounts for poorly reversible slowing of nerve conduction in the spontaneously diabetic bio-breeding rat. The Journal of Clinical Investigation 1986;77(2):474-484.
Sima AAF. Metabolic alterations of peripheral nerve in diabetes. Seminars in Neurology 1996;16(2):129-137.
Sima AAF. New insights into the metabolic and molecular basis for diabetic neuropathy. Cellular and Molecular Life Sciences 2003;60(11):2445-2464.
Simioni N, Berra C, Boemi M, Bossi AC, Candido R, Di Cianni G, . . . Rea LSG. Predictors of treatment response to liraglutide in type 2 diabetes in a real-world setting. Acta Diabetologica 2018;55(6):557-568.
Singleton JR, Marcus RL, Lessard MK, Jackson JE, Smith AG. Supervised exercise improves cutaneous reinnervation capacity in metabolic syndrome patients. Annals of neurology 2015;77(1):146-53.
Smith AG, Singleton JR. Obesity and hyperlipidemia are risk factors for early diabetic neuropathy. Journal of Diabetes and its Complications 2013;27(5):436-442.
271 References
Solbiati L, de Pra L, Ierace T, Bellotti E, Derchi LE. High-resolution sonography of the recurrent laryngeal nerve: Anatomic and pathologic considerations. American Journal of Roentgenology 1985;145(5):989-993.
Stämpfli R. Saltatory Conduction in Nerve. Physiological Reviews 1954;34(1):101- 112.
Standl E, Schnell O, Ceriello A. Postprandial hyperglycemia and glycemic variability: should we care? Diabetes Care 2011;34 Suppl 2:S120-7.
Stanton RC. Clinical challenges in diagnosis and management of diabetic kidney disease. American Journal of Kidney Diseases 2014;63(2 Suppl 2):S3-21.
Stevens MJ, Lattimer SA, Kamijo M, Van Huysen C, Sima AAF, Greene DA. Osmotically-induced nerve taurine depletion and the compatible osmolyte hypothesis in experimental diabetic neuropathy in the rat. Diabetologia 1993;36(7):608-614.
Stino AM, Rumora AE, Kim B, Feldman EL. Evolving concepts on the role of dyslipidemia, bioenergetics, and inflammation in the pathogenesis and treatment of diabetic peripheral neuropathy. Journal of the Peripheral Nervous System 2020;25(2):76-84.
Stino AM, Smith AG. Peripheral neuropathy in prediabetes and the metabolic syndrome. Journal of diabetes investigation 2017;8(5):646-655.
Stühmer W, Conti F, Suzuki H, Wang X, Noda M, Yahagi N, . . . Numa S. Structural parts involved in activation and inactivation of the sodium channel. Nature 1989;339(6226):597-603.
Suk JI, Walker FO, Cartwright MS. Ultrasonography of peripheral nerves. Current Neurology and Neuroscience Reports 2013;13(2):328.
Sun H-s, Feng Z-p. Neuroprotective role of ATP-sensitive potassium channels in cerebral ischemia. Acta pharmacologica Sinica 2013;34(1):24-32.
Sung JY, Park SB, Liu YT, Kwai N, Arnold R, Krishnan AV, Lin CS. Progressive axonal dysfunction precedes development of neuropathy in type 2 diabetes. Diabetes 2012;61(6):1592-8.
Sung JY, Tani J, Chang TS, Lin CS. Uncovering sensory axonal dysfunction in asymptomatic type 2 diabetic neuropathy. PLoS One 2017;12(2):e0171223.
Taddese A, Bean BP. Subthreshold sodium current from rapidly inactivating sodium channels drives spontaneous firing of tuberomammillary neurons. Neuron 2002;33(4):587-600.
Tagliafico A, Martinoli C. Reliability of side-to-side sonographic cross-sectional area measurements of upper extremity nerves in healthy volunteers. Journal of Ultrasound in Medicine 2013;32(3):457-462.
272 References
Tahrani AA, Dubb K, Raymond NT, Begum S, Altaf QA, Sadiqi H, . . . Stevens MJ. Cardiac autonomic neuropathy predicts renal function decline in patients with type 2 diabetes: a cohort study. Diabetologia 2014;57(6):1249-56.
Takakura S, Toyoshi T, Hayashizaki Y, Takasu T. Effect of ipragliflozin, an SGLT2 inhibitor, on progression of diabetic microvascular complications in spontaneously diabetic Torii fatty rats. Life sciences 2016;147:125-131.
Tamura N, Kuwabara S, Misawa S, Kanai K, Nakata M, Sawai S, Hattori T. Increased nodal persistent Na+ currents in human neuropathy and motor neuron disease estimated by latent addition. Clinical Neurophysiology 2006;117(11):2451-2458.
Tasaki I. Nervous transmission: Thomas, 1953.
Tavakoli M, Ferdousi M, Petropoulos IN, Morris J, Pritchard N, Zhivov A, . . . Malik RA. Normative values for corneal nerve morphology assessed using corneal confocal microscopy: A multinational normative data set. Diabetes Care 2015;38(5):838-843.
Tavakoli M, Mitu-Pretorian M, Petropoulos IN, Fadavi H, Asghar O, Alam U, . . . Malik RA. Corneal confocal microscopy detects early nerve regeneration in diabetic neuropathy after simultaneous pancreas and kidney transplantation. Diabetes 2013;62(1):254-260.
Tavee J. Chapter 14 - Nerve conduction studies: Basic concepts. In: Levin KH, Chauvel P, editors. Handbook of Clinical Neurology. 160: Elsevier; 2019. p. 217-224.
Taylor JL, Burke D, Heywood J. Physiological evidence for a slow K+ conductance in human cutaneous afferents. The Journal of Physiology 1992;453(1):575-589.
Tesfaye S, Boulton AJM, Dyck PJ, Freeman R, Horowitz M, Kempler P, . . . Toronto Diabetic Neuropathy Expert G. Diabetic neuropathies: update on definitions, diagnostic criteria, estimation of severity, and treatments. Diabetes Care 2010;33(10):2285-2293.
Tesfaye S, Chaturvedi N, Eaton SEM, Ward JD, Manes C, Ionescu-Tirgoviste C, . . . Fuller JH. Vascular risk factors and diabetic neuropathy. The New England journal of medicine 2005;352(4):341-350.
Themistocleous AC, Crombez G, Baskozos G, Bennett DLH. Using stratified medicine to understand, diagnose, and treat neuropathic pain. Pain 2018;159 Suppl 1:S31-S42.
Themistocleous AC, Ramirez JD, Serra J, Bennett DLH. The clinical approach to small fibre neuropathy and painful channelopathy. Practical Neurology 2014;14(6):368-379.
Thomas R, Kanso A, Sedor JR. Chronic kidney disease and its complications. Primary care 2008;35(2):329-44.
Thomas RC. Electrogenic sodium pump in nerve and muscle cells. Physiological Reviews 1972;52(3):563-594.
273 References
Thulé PM. Mechanisms of current therapies for diabetes mellitus type 2. Advances in physiology education 2012;36(4):275-283.
Tilki HE, Akpolat T, Coşkun M, Stålberg E. Clinical and electrophysiologic findings in dialysis patients. Journal of Electromyography and Kinesiology 2009;19(3):500-8.
Tomlinson SE, Burke D, Hanna M, Koltzenburg M, Bostock H. In vivo assessment of HCN channel current (Ih) in human motor axons. Muscle and Nerve 2010;41(2):247- 256.
Tomlinson SE, Howells J, Burke D. In vivo assessment of neurological channelopathies: Application of peripheral nerve excitability studies. Neuropharmacology 2018;132:98-107.
Tsantoulas C, McMahon SB. Opening paths to novel analgesics: the role of potassium channels in chronic pain. Trends in Neuroscience 2014;37(3):146-58.
Tuomi T, Groop LC, Zimmet PZ, Rowley MJ, Knowles W, Mackay IR. Antibodies to glutamic acid decarboxylase reveal latent autoimmune diabetes mellitus in adults with a non-insulin-dependent onset of disease. Diabetes 1993;42(2):359-62.
Tuttle KR, Bakris GL, Bilous RW, Chiang JL, de Boer IH, Goldstein-Fuchs J, . . . Molitch ME. Diabetic kidney disease: a report from an ADA Consensus Conference. American Journal of Kidney Diseases 2014;64(4):510-33.
Uncini A, Kuwabara S. Nodopathies of the peripheral nerve: An emerging concept. Journal of Neurology, Neurosurgery and Psychiatry 2015;86(11):1186-1195.
Vagenas D, Pritchard N, Edwards K, Shahidi AM, Sampson GP, Russell AW, . . . Efron N. Optimal image sample size for corneal nerve morphometry. Optometry and vision science 2012;89(5):812-7.
Vanholder R, De Smet R, Hsu C, Vogeleere P, Ringoir S. Uremic toxicity: the middle molecule hypothesis revisited. Seminars in Nephrology 1994;14(3):205-18.
Vanmolkot KRJ, Kors EE, Hottenga JJ, Terwindt GM, Haan J, Hoefnagels WAJ, . . . Van den Maagdenberg AMJM. Novel mutations in the Na+,K+-ATPase pump gene ATP1A2 associated with familial hemiplegic migraine and benign familial infantile convulsions. Annals of Neurology 2003;54(3):360-366.
Vellanki K, Bansal VK. Neurologic Complications of Chronic Kidney Disease. Current Neurology and Neuroscience Reports 2015;15(8):50.
Vincent AM, Callaghan BC, Smith AL, Feldman EL. Diabetic neuropathy: Cellular mechanisms as therapeutic targets. Nature Reviews Neurology 2011;7(10):573-583.
Virk SA, Donaghue KC, Cho YH, Benitez-Aguirre P, Hing S, Pryke A, . . . Craig ME. Association Between HbA1c Variability and Risk of Microvascular Complications in Adolescents With Type 1 Diabetes. Journal of Clinical Endocrinology and Metabolism 2016;101(9):3257-63.
274 References
Vlassakov KV, Sala-Blanch X. Chapter 16 - Ultrasound of the Peripheral Nerves. Nerves and Nerve Injuries. San Diego: Academic Press; 2015. p. 227-250.
Vogel W, Schwarz JR. Voltage-clamp studies in axons: Macroscopic and single- channel currents. The Axon 1995:257-280.
Vos T, Allen C, Arora M, Barber RM, Bhutta ZA, Brown A, . . . Murray CJL. Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. The Lancet 2016;388(10053):1545-1602.
Vucic S, Lin CSY, Cheah BC, Murray J, Menon P, Krishnan AV, Kiernan MC. Riluzole exerts central and peripheral modulating effects in amyotrophic lateral sclerosis. Brain 2013;136(5):1361-1370.
Vujkovic M, Keaton JM, Lynch JA, Miller DR, Zhou J, Tcheandjieu C, . . . Saleheen D. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nature Genetics 2020;52(7):680-691.
Wahren J, Foyt H, Daniels M, Arezzo JC. Long-Acting C-Peptide and Neuropathy in Type 1 Diabetes: A 12-Month Clinical Trial. Diabetes Care 2016;39(4):596-602.
Wahren J, Larsson C. C-peptide: new findings and therapeutic possibilities. Diabetes Research and Clinical Practice 2015;107(3):309-19.
Waldbillig RJ, LeRoith D. Insulin receptors in the peripheral nervous system: a structural and functional analysis. Brain Research 1987;409(2):215-220.
Wan W, Skandari MR, Minc A, Nathan AG, Zarei P, Winn AN, . . . Huang ES. Cost- effectiveness of Initiating an Insulin Pump in T1D Adults Using Continuous Glucose Monitoring Compared with Multiple Daily Insulin Injections: The DIAMOND Randomized Trial. Medical decision making 2018;38(8):942-953.
Wang C, Lu J, Lu W, Yu H, Jiang L, Li M, . . . Jia W. Evaluating peripheral nerve function in asymptomatic patients with type 2 diabetes or latent autoimmune diabetes of adults (LADA): results from nerve conduction studies. Journal of diabetes and its complications 2015;29(2):265-9.
Wang F, He Z. Chapter 9 - Axon Maintenance and Degeneration. In: Rubenstein JLR, Rakic P, editors. Cellular Migration and Formation of Neuronal Connections. Oxford: Academic Press; 2013. p. 177-189.
Wang H, Kunkel DD, Martin TM, Schwartzkroin PA, Tempel BL. Heteromultimeric K+ channels in terminal and juxtaparanodal regions of neurons. Nature 1993;365(6441):75-79.
Wang JT, Medress ZA, Barres BA. Axon degeneration: Molecular mechanisms of a self-destruction pathway. Journal of Cell Biology 2012;196(1):7-18.
275 References
Watanabe T, Ito H, Morita A, Uno Y, Nishimura T, Kawase H, . . . Seishima M. Sonographic Evaluation of the Median Nerve in Diabetic Patients. Journal of Ultrasound in Medicine 2009;28(6):727-734.
Watanabe T, Ito H, Sekine A, Katano Y, Nishimura T, Kato Y, . . . Matsuoka T. Sonographic Evaluation of the Peripheral Nerve in Diabetic Patients. Journal of Ultrasound in Medicine 2010;29(5):697-708.
Waxman SG, Ritchie JM. Molecular dissection of the myelinated axon. Annals of Neurology 1993;33(2):121-136.
Weisman A, Bril V, Ngo M, Lovblom LE, Halpern EM, Orszag A, Perkins BA. Identification and Prediction of Diabetic Sensorimotor Polyneuropathy Using Individual and Simple Combinations of Nerve Conduction Study Parameters. PLoS One 2013;8(3):e58783.
Weiss G. Sur la possibilité de rendre comparables entre eux les appareils servant à l'excitation électrique. Archives Italiennes de Biologie 1901;35:413-446.
Welt LG, Sachs JR, McManus TJ. An ion transport defect in erythrocytes from uremic patients. Transactions of the Association of American Physicians 1964;77:169-81.
Won SJ, Kim BJ, Park KS, Yoon JS, Choi H. Reference values for nerve ultrasonography in the upper extremity. Muscle and Nerve 2013;47(6):864-871.
Wood JG, Jean DH, Whitaker JN, McLaughlin BJ, Albers RW. Immunocytochemical localization of the sodium, potassium activated ATPase in knifefish brain. Journal of Neurocytology 1977;6(5):571-581.
Yamada K, Inagaki N. Neuroprotection by KATP channels. Journal of molecular and cellular cardiology 2005;38(6):945-9.
Yan A, Issar T, Tummanapalli SS, Markoulli M, Kwai NCG, Poynten AM, Krishnan AV. Relationship between corneal confocal microscopy and markers of peripheral nerve structure and function in Type 2 diabetes. Diabetic Medicine 2020;37(2):326- 334.
Yang J, Zhao Z, Yuan H, Ma X, Li Y, Wang H, . . . Qin G. The mechanisms of glycemic variability accelerate diabetic central neuropathy and diabetic peripheral neuropathy in diabetic rats. Biochemical and Biophysical Research Communications 2019;510(1):35-41.
Ylitalo KR, Sowers M, Heeringa S. Peripheral vascular disease and peripheral neuropathy in individuals with cardiometabolic clustering and obesity: National health and nutrition examination survey 2001-2004. Diabetes Care 2011;34(7):1642- 1647.
Yorek MS, Obrosov A, Shevalye H, Lupachyk S, Harper MM, Kardon RH, Yorek MA. Effect of glycemic control on corneal nerves and peripheral neuropathy in streptozotocin-induced diabetic C57Bl/6J mice. Journal of the Peripheral Nervous System 2014;19(3):205-17.
276 References
Young EA, Fowler CD, Kidd GJ, Chang A, Rudick R, Fisher E, Trapp BD. Imaging correlates of decreased axonal Na+/K+ ATPase in chronic multiple sclerosis lesions. Annals of neurology 2008;63(4):428-35.
Z'Graggen WJ, Lin CSY, Howard RS, Beale RJ, Bostock H. Nerve excitability changes in critical illness polyneuropathy. Brain 2006;129(9):2461-2470.
Zaharia OP, Strassburger K, Strom A, Bönhof GJ, Karusheva Y, Antoniou S, . . . Ziegler D. Risk of diabetes-associated diseases in subgroups of patients with recent- onset diabetes: a 5-year follow-up study. The Lancet Diabetes & Endocrinology 2019;7(9):684-694.
Zaidman CM, Al-Lozi M, Pestronk A. Peripheral nerve size in normals and patients with polyneuropathy: an ultrasound study. Muscle Nerve 2009;40(6):960-6.
Zampetti S, Campagna G, Tiberti C, Songini M, Arpi ML, De Simone G, . . . Buzzetti R. High GADA titer increases the risk of insulin requirement in LADA patients: a 7- year follow-up (NIRAD study 7). European journal of endocrinology 2014;171(6):697-704.
Zenker J, Poirot O, de Preux Charles AS, Arnaud E, Medard JJ, Lacroix C, . . . Chrast R. Altered distribution of juxtaparanodal kv1.2 subunits mediates peripheral nerve hyperexcitability in type 2 diabetes mellitus. The Journal of Neuroscience 2012;32(22):7493-8.
Zhang C, Ward J, Dauch JR, Tanzi RE, Cheng HT. Cytokine-mediated inflammation mediates painful neuropathy from metabolic syndrome. PLoS One 2018;13(2):e0192333.
Zhang CL, Ho PL, Kintner DB, Sun D, Chiu SY. Activity-Dependent Regulation of Mitochondrial Motility by Calcium and Na/K-ATPase at Nodes of Ranvier of Myelinated Nerves. The Journal of Neuroscience 2010;30(10):3555-3566.
Zhang Z, David G. Stimulation-induced Ca(2+) influx at nodes of Ranvier in mouse peripheral motor axons. The Journal of Physiology 2016;594(1):39-57.
Zheng Y, Ley SH, Hu FB. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications. Nature Reviews Endocrinology 2018;14(2):88-98.
Zhou R, Li F, Chen G, Fu Q, Gu S, Wu X. Associations Between General and Abdominal Obesity and Incident Diabetic Neuropathy in Participants with Type 2 Diabetes Mellitus. Journal of Diabetes 2020.
Ziegler D, Papanas N, Vinik AI, Shaw JE. Chapter 1 - Epidemiology of polyneuropathy in diabetes and prediabetes. In: Zochodne DW, Malik RA, editors. Handbook of Clinical Neurology. 126: Elsevier; 2014. p. 3-22.
Zimmet PZ, Tuomi T, Mackay IR, Rowley MJ, Knowles W, Cohen M, Lang DA. Latent autoimmune diabetes mellitus in adults (LADA): the role of antibodies to glutamic acid decarboxylase in diagnosis and prediction of insulin dependency. Diabetic Medicine 1994;11(3):299-303.
277 References
Zochodne DW. Chapter 2 - Clinical features of diabetic polyneuropathy. In: Zochodne DW, Malik RA, editors. Handbook of Clinical Neurology. 126: Elsevier; 2014. p. 23-30.
278