If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Musculoskeletal Biology, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, United KingdomDivision of Diabetes, Endocrinology and Gastroenterology, School of Medical Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United Kingdom
Division of Diabetes, Endocrinology and Gastroenterology, School of Medical Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United KingdomCardiovascular and Metabolic Medicine, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, United Kingdom
Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United KingdomBrain Function Research Group, School of Physiology, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
Address correspondence to: Andrew Marshall, University Hospital Liverpool, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, United Kingdom.
Musculoskeletal Biology, Institute of Life Course and Medical Sciences, University of Liverpool, Liverpool, United KingdomDepartment of Clinical Neurophysiology, The Walton Centre, Liverpool, United KingdomDivision of Neuroscience and Experimental Psychology, School of Biological Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United Kingdom
Diabetic peripheral neuropathy (DPN) is the most common complication of diabetes. Small and large peripheral nerve fibers can be involved in DPN. Large nerve fiber damage causes paresthesia, sensory loss, and muscle weakness, and small nerve fiber damage is associated with pain, anesthesia, foot ulcer, and autonomic symptoms. Treatments for DPN and painful DPN (pDPN) pose considerable challenges due to the lack of effective therapies. To meet these challenges, there is a major need to develop biomarkers that can reliably diagnose and monitor progression of nerve damage and, for pDPN, facilitate personalized treatment based on underlying pain mechanisms.
Methods
This study involved a comprehensive literature review, incorporating article searches in electronic databases (Google Scholar, PubMed, and OVID) and reference lists of relevant articles with the authors’ substantial expertise in DPN. This review considered seminal and novel research and summarizes emerging biomarkers of DPN and pDPN that are based on neurophysiological methods.
Findings
From the evidence gathered from 145 papers, this submission describes emerging clinical neurophysiological methods with potential to act as biomarkers for the diagnosis and monitoring of DPN as well as putative future roles as predictors of response to antineuropathic pain medication in pDPN. Nerve conduction studies only detect large fiber damage and do not capture pathology or dysfunction of small fibers. Because small nerve fiber damage is prominent in DPN, additional biomarkers of small nerve fiber function are needed. Activation of peripheral nociceptor fibers using laser, heat, or targeted electrical stimuli can generate pain-related evoked potentials, which are an objective neurophysiological measure of damage along the small fiber pathways. Assessment of nerve excitability, which provides a surrogate of axonal properties, may detect alterations in function before abnormalities are detected by nerve conduction studies. Microneurography and rate-dependent depression of the Hoffmann-reflex can be used to dissect underlying pain-generating mechanisms arising from the periphery and spinal cord, respectively. Their role in informing mechanistic-based treatment of pDPN as well as facilitating clinical trials design is discussed.
Implications
The neurophysiological methods discussed, although currently not practical for use in busy outpatient settings, detect small fiber and early large fiber damage in DPN as well as disclosing dominant pain mechanisms in pDPN. They are suited as diagnostic and predictive biomarkers as well as end points in mechanistic clinical trials of DPN and pDPN.
Introduction
Peripheral neuropathy is the most common complication of both type 1 (T1DM) and type 2 diabetes (T2DM), with more than one half of all patients developing nerve dysfunction in their lifetime.
Both chronic and acute diabetic neuropathies are seen, but distal length-dependent symmetrical polyneuropathy is the most common and generally referred to as diabetic peripheral neuropathy (DPN). DPN is the primary cause of diabetic foot disease, including ulceration and nontraumatic amputations.
Currently, there are no US Food and Drug Administration–approved disease-modifying therapies for diabetic neuropathy. Many clinical trials of drugs targeting presumptive mechanisms of DPN, including aldose reductase inhibitors, protein kinase C inhibitors, and benfotiamine, have failed at Phase III of development. Potential reasons for these failures include the use of subjective or insensitive end points such as physician-based clinical scores and use of biomarkers that only report function of large myelinated nerve fibers.
Consequently, there is a lack of disease-modifying medication. Beyond glycemic control and modulation of cardiometabolic risk factors, the only option is pain management. Unfortunately, pharmacologic treatments for pDPN are inconsistently effective. In patients with pDPN treated with monotherapy, 50% pain relief in 50% of cases is considered a favorable outcome.
Treatment strategies typically involve a trial-and-error approach of prescribing anti-neuropathic pain medications, which have inconsistent therapeutic benefit.
There is a major unmet need for the development of reliable biomarkers that: (1) capture the onset and progression of DPN; (2) inform mechanistic-based drug discovery; and (3) facilitate individualized treatment of neuropathic pain in pDPN.
The purpose of the present narrative review was to discuss emerging neurophysiological biomarkers of human DPN and pDPN. Clinical neurophysiology is a medical specialty in which the recording of spontaneous or evoked bioelectrical activity is used to investigate function of the brain, spinal cord, spinal nerve roots, peripheral nerves, and muscle. Clinical neurophysiology departments may also perform quantitative sensory testing (QST), such as thermal threshold analysis, but this review is restricted to electrophysiological investigations according to the definition noted (details about QST are given elsewhere
Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): somatosensory abnormalities in 1236 patients with different neuropathic pain syndromes.
). Investigations of the autonomic nervous system, including cardiac autonomic testing and sweat responses, are also not discussed (reviews are presented elsewhere
A comprehensive literature review was undertaken, including article searches in multiple electronic databases (Google Scholar, PubMed, and OVID) using key words (eg, diabetic neuropathy, neuropathic pain, pain biomarkers) and reference lists of relevant articles with the authors’ substantial expertise in DPN. This review considered seminal and novel research and summarizes emerging biomarkers of DPN and pDPN that are based on neurophysiological methods. Articles published from inception of databases to December 2020 were identified. Data from articles that were not relevant to the aim of the study were excluded from the review.
Results
In total, 145 papers were cited in the final manuscript. An evaluation of selected articles was undertaken, and relevant data were included in the current review. Authors excluded studies that were not relevant to the aims and ethos of the manuscript under the guidance of the senior author (A.M.).
A biomarker is defined by the National Institutes of Health as “a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacological responses to a therapeutic intervention.”
According to this definition, biomarkers can have a wide range of complexity ranging from very simple clinical tests (eg, deep tendon reflex) or noninvasive diagnostic investigations (eg, heart rate variability to detect cardiac autonomic neuropathy) through to complex laboratory or invasive evaluation (eg, skin biopsy, microneurography).
define several categories of biomarkers, including susceptibility/risk, diagnostic, prognostic, predictive, enrichment, monitoring, and surrogate end points. The precise characteristics of an ideal biomarker depend on its context of use
but in general terms they include the following: (1) being scalable and easy to use (eg, be simple, cost-effective, minimally invasive); (2) have high sensitivity (ie, high true-positive rate) and specificity (ie, high true-negative rate); and (3) have a high predictability (ie, a strong indicator of the condition or of a response to a particular treatment).
For DPN, the major purposes of biomarkers are to: (1) act as a diagnostic biomarker to identify individuals with the biologically defined disorder (eg, large and/or small fiber DPN); (2) monitor the change in degree of the disorder over time; and (3) act as an indicator of drug efficacy in clinical trials. Although detection of potentially fully reversible changes in nerve function may theoretically be possible (as discussed in the Nerve Excitability Testing section), in general, this requires the detection of nerve pathology or surrogates thereof. It is therefore important that a diagnostic biomarker can reliably and reproducibly detect the earliest signs or the earliest changes of the severity/extent of DPN.
The earliest clinical manifestations of DPN are typically sensory symptoms or signs in the extremities. This sensory predominant phenotype is associated with pathologic alterations in all major classes of sensory fibers: large myelinated fibers, as well as small thinly myelinated A-delta and unmyelinated C-fibers.
Sural nerve fibre pathology in diabetic patients with mild neuropathy: relationship to pain, quantitative sensory testing and peripheral nerve electrophysiology.
Therefore, DPN is a mixed large fiber neuropathy and small fiber neuropathy (SFN). Because small and large fiber deficits are detected by using different methodologies, there are distinct biomarkers of small and of large fiber neuropathy. Biomarkers of somatic SFN include skin biopsy for quantification of intra-epidermal nerve fiber density (IENFD) and corneal confocal microscopy (CCM) with quantification of corneal innervation, both of which provide information on small fiber structure, whereas QST of thermal or pain perception and laser evoked potentials are tests of small fiber function. The current preferred biomarkers of large fiber neuropathy are nerve conduction studies (NCS) and quantitative assessment of vibration perception.
The current prevailing view is that small fiber deficits can be detected before objective large fiber involvement. This is primarily based on cross-sectional studies of patients with diabetes or prediabetes that show small fiber abnormalities in a proportion of patients with normal NCS and in the large majority of those with abnormal NCS parameters.
Furthermore, normalization of hyperglycemia (diabetes remission) or intensive treatment of hyperglycemia with continuous subcutaneous insulin are associated with early improvements in measures of small, but not large, fiber neuropathy in severe DPN.
Corneal confocal microscopy shows an improvement in small-fiber neuropathy in subjects with type 1 diabetes on continuous subcutaneous insulin infusion compared with multiple daily injection.
For example, in severe DPN, although motor nerve conduction velocity did eventually increase 36 months after simultaneous pancreas and kidney transplant, this lagged behind improvements in SFN detected by using CCM.
These findings indicate that biomarkers of SFN are, in general, more sensitive at detecting early neuropathy or the earliest signs of repair in DPN. However, in mild to moderate DPN, short-term modest improvements in glycemic control and serum triglyceride levels had an independent, additive, and durable effect on restoration of nerve function with improvement in lower limb electrophysiological parameters.
in a study of recently diagnosed T2DM found that although CCM and IENFD detected small nerve fiber loss, there was a reduction in NCS parameters in a subpopulation suggestive of large fiber disease. Furthermore, a recent study of 133 patients with T2DM and DPN participants were subclassified into large, small, or mixed neuropathy on the basis of NCS and IENFD findings.
The majority (74%) had evidence of mixed large and small fiber involvement, although a significant minority had evidence of isolated large or small fiber DPN. Other studies that stratify patients based on clinical scales show that NCS and measures of SFN such as QST, CCM, and IENFD all progressively decline with increasing severity of clinical neuropathy.
Corneal confocal microscopy compared with quantitative sensory testing and nerve conduction for diagnosing and stratifying the severity of diabetic peripheral neuropathy.
Although these studies imply that small and large fiber pathology develop in parallel, longitudinal investigations are needed before inferences can be made regarding the natural history of peripheral nerve pathology. A recent multicenter longitudinal study of DPN which assessed large and small fibers showed that more rapid deterioration in SFN (loss of ≥6% of nerve fiber length/year detected by using CCM) is associated with a greater deterioration of NCS parameters (peroneal motor conduction velocity and amplitude).
However, the converse (ie, whether greater deterioration in NCS was associated with more rapid small fiber abnormalities) was not addressed. Therefore, the concomitant use of biomarkers of both large fiber neuropathy and SFN for DPN represents an appropriate strategy.
Recent data suggest a greater degree of SFN (detected by using IENFD and CCM) in patients with pDPN compared with those with equivalent large fiber neuropathy but no pain.
However, the findings in the literature are mixed. For example, other recent large-scale cross-sectional studies report no difference in structural or functional small fiber deficits between painless DPN and pDPN
The Pain in Neuropathy Study (PiNS): a cross-sectional observational study determining the somatosensory phenotype of painful and painless diabetic neuropathy.
Whether interventions that prevent or reverse small fiber loss in diabetes can affect the pain phenotype remains to be determined.
The role of biomarkers of pDPN is somewhat different from that for DPN. A biomarker of pain per se is not particularly useful because this information can be detected, and indeed quantified, by asking the patient directly. A more useful biomarker would detect a particular pain phenotype or underlying pathophysiological mechanism that can be used to predict an individual's benefit from particular treatments, to enrich patient populations for clinical trials or, potentially, to predict the risk of developing pain.
Nerve Conduction Studies: a Gold Standard Biomarker With Limitations
NCS have long been used as diagnostic, staging, and prognostic biomarkers for DPN. The investigation involves supramaximal transcutaneous stimulation of defined upper and lower limb peripheral nerves. Recordings are also taken noninvasively with surface electrodes. Motor and sensory fiber function can be assessed separately. Measurement of response amplitude serves as a correlate of axonal loss. Conduction velocity, which is measured by using single and 2-site stimulation for sensory and motor nerves, respectively, is affected by several factors such as axonal atrophy, internodal distance, and nodal membrane functions, although it is typically regarded as a biomarker of demyelination. NCS are routine procedures and, in skilled hands, provide objective data that can be compared with local normative ranges.
Although NCS are widely available, they are most often used for the assessment of atypical clinical presentations. A number of point-of-care NCS devices that may facilitate screening of large fiber neuropathy in the clinic setting are available. One such device, DPNCheck (NeuroMetrix, Waltham, Massachusetts), which measures sural nerve sensory nerve action potential amplitude and velocity, has been validated for use in DPN.
In controlled settings performed by trained nontechnical individuals, it can be rapidly performed (<5 minutes) and shows 95% sensitivity and 71% specificity for DPN detection compared with conventional NCS
NCS have been historically considered a gold standard for the diagnosis of DPN. Confirmed DPN is defined as clinical symptom(s) and/or sign(s) of neuropathy and an abnormal nerve conduction or adequate small fiber measure if NCS are normal, as classified by using the updated Toronto consensus guidelines.
measures of severity of neuropathy. Furthermore, they have prognostic significance: slowing of motor nerve conduction velocity is predictive of foot ulceration and death in DPN.
However, NCS have significant limitations. Neurophysiology departments often have normative reference ranges based on healthy individuals, but performance standards for NCS are not uniform. Reference ranges ideally should be standardized for methodology, age, and for sensory conduction especially (because of lower signal-to-noise ratio) and potentially other technical factors such as body mass index.
Establishing high-quality reference values for nerve conduction studies: a report from the normative data task force of the American Association Of Neuromuscular & Electrodiagnostic Medicine.
Furthermore, reference ranges based on populations of healthy individuals can be broad. For example, a sural sensory amplitude of ≥4 microvolts might be considered normal by some reference ranges
but would be highly abnormal for an individual with a not uncommonly encountered premorbid value of >20 microvolts. Therefore, detection of early neuropathy may require serial studies to investigate declining parameters originally within the “normal” range, which will limit segregation of patients for clinical trials based on inclusion criteria requiring “abnormal” NCS.
The major shortcoming of NCS is that they assess large myelinated fibers only. It is not possible to transcutaneously record from thinly myelinated (Aδ) fibers, which mediate nociception and cold sensation, or unmyelinated (C) fibers, which mediate warm and nociceptive sensation. Therefore, an insensitivity to detecting changes in small nerve fibers, potentially the most sensitive method for detecting DPN onset and progression, combined with questionable reproducibility and lack of measurement standardization across centers,
limit the use of NCS as a standalone biomarker of DPN. There is a remarkable paucity of studies that directly compare the ability of small and large fiber biomarkers to detect DPN. One reason for this is that NCS are often used as a criterion for diagnosis of neuropathy and thus form the gold standard against which the sensitivity and specificity of other methods are compared. Studies that have compared the diagnostic capability of NCS and SFN biomarkers in DPN used relatively crude and subjective clinical criteria as a benchmark for neuropathy diagnosis and stratification.
Corneal confocal microscopy compared with quantitative sensory testing and nerve conduction for diagnosing and stratifying the severity of diabetic peripheral neuropathy.
Nevertheless, these studies indicate that NCS parameters such as sural nerve conduction velocity perform with more-or-less equivalent sensitivity and specificity as structural and functional measures of SFN. Although NCS are established as a biomarker of large fiber neuropathy in DPN, large-scale studies involving detailed neuropathy phenotyping indicate that they do not differentiate between pDPN and painless DPN.
The Pain in Neuropathy Study (PiNS): a cross-sectional observational study determining the somatosensory phenotype of painful and painless diabetic neuropathy.
Pain-related evoked potentials involve the recording of cortical activity through scalp electrodes in response to selective stimulation of nociceptive skin afferents (Figure 1). Distinct waveforms can be obtained for A-delta and C-fiber stimulation, but for the purposes of clinical evaluation only assessment of A-delta waveforms is currently believed to be reliable.
Nociceptor stimulation is most commonly delivered by using a laser or a brief thermal stimulus, although stimulus paradigms using pinpricks or specialized surface electrodes have also been used.
Figure 1Schematic showing established and emerging biomarkers of diabetic peripheral neuropathy (DPN) and painful diabetic peripheral neuropathy (pDPN). H-Reflex = Hoffmann reflex; RDD = rate-dependent depression.
The cortical waveforms generated are termed the N1 and N2/P2 complex. Attenuated, delayed, or absent LEPs to stimulation of a pathologically involved region such as the foot in distal symmetrical polyneuropathy can provide objective evidence of a loss-of-function lesion affecting the A-delta nociceptive pathways. In a group of 45 patients with diabetes and varying degrees of DPN, the most frequent LEP abnormalities were attenuated or absent cortical waveforms. The degree of LEP abnormality correlated with the degree of NCS abnormality, suggesting that large and small fiber dysfunction occurs in parallel.
Using carefully collected age- and site-related normative data, LEPS showed 78% sensitivity and 81% specificity for the diagnosis of SFN in patients with DPN and normal NCS compared with the current gold standard diagnostic, IENFD.
Similar sensitivity and specificity were documented in a large study of 164 patients with clinical evidence of SFN and normal NCS; ~20% of the patients had pDPN. LEP abnormalities had sensitivity and specificity of 66% and 89% for detecting SFN compared with diagnostic criteria of ≥2 abnormal results from a battery of 6 investigations (LEP, QST, cardiac autonomic tests, electrodermal skin conductance, quantitative sweat testing, and IENFD).
LEP amplitudes have also been shown to be attenuated in patients with diabetes of >10 years’ duration and asymptomatic “pure” SFN diagnosed by using IENFD,
Asymptomatic small fiber neuropathy in diabetes mellitus: investigations with intraepidermal nerve fiber density, quantitative sensory testing and laser-evoked potentials.
Although this suggests a greater small fiber burden in painful neuropathy, it is important to emphasize that LEPs primarily demonstrate a lesion affecting small (A-delta) fiber pathways rather than providing evidence of hyperalgesia or pain mechanisms. However, there are some data suggesting that differential involvement of LEPs is associated with that particular symptom of pDPN. For example, in a recent study of 133 patients with T2DM and DPN, subclassified into myelinated versus SFN versus mixed on the basis of NCS and abnormal IENFD, the presence of burning pain was associated a greater loss of LEP amplitude, whereas allodynia showed a tendency for preservation of LEP amplitudes.
LEPs are noninvasive, relatively rapid to perform, and are an objective neurophysiological biomarker for SFN. Indeed, in patients with painful sensory neuropathy, LEP amplitudes positively correlate with IENFD.
However, LEPs have several drawbacks as a biomarker. The technique is expensive, not widely available, and certain safety measures (eg, wearing of eye protection) are required during the investigation. LEPs also habituate, and the waveform amplitudes are dependent on the participants’ attention.
Because LEPs assesses the whole neuraxis through the spinothalamic tracts, the method does not distinguish between central and peripheral nervous system pathology.
The principles of contact heat evoked potentials (CHEPs) are similar to those of LEPs. Stimuli are delivered with a Peltier thermode, with a rapid rate of rise in temperature to ensure rapid, synchronous activation of thermal sensitive nociceptors. N1/P1 cortical waveforms are generated that reflect activation of A-delta nociceptors. Although CHEPs are a potential biomarker of SFN in DM, this has been addressed in few studies. Two small-scale investigations involving ~30 patients with T2DM have shown that CHEP amplitudes are reduced compared with healthy control subjects and correlate with other biomarkers of SFN (QST, autonomic function tests, and IENFD) and large fiber neuropathy.
CHEPs are noninvasive and do not have safety issues as with LEPs. However, they require specialized equipment and are subject to significant habituation effects.
Both CHEPs and LEPs are unlikely to be used in a busy clinic setting.
Intra-epidermal electrical evoked potentials (IEEPs) are performed by using specialized electrodes that deliver a high current density to epidermal fibers. Longer stimulus durations preferentially stimulate nociceptors. IEEPs offer the advantage of not requiring a specialized stimulator device such as a laser or thermode. They are noninvasive and easy to perform. However, this method is currently experimental, with few studies that address the use of IEEPs in DM and none that compare their use with other SFN biomarkers. Using a concentric planar electrode in patients with DPN and normal NCS, IEEP cortical waveforms were shown to be delayed and of lower amplitude relative to those of healthy volunteers. In patients with neuropathic symptoms, 95% showed at least one abnormal finding (greater than the maximum latency and less than the minimal amplitude of recordings in healthy volunteers).
NET is performed in a similar manner to NCS but uses submaximal as well as supra-maximal stimulation according to standardized excitability protocols (stimulus–response curve, strength–duration properties, threshold electrotonus and the current/threshold relationship, and the recovery cycle) to obtain a reflection of firing threshold (Figure 1).
The results provide a surrogate of ion channel and axonal membrane potential properties at the stimulation site. Although motor axons are more commonly investigated, because recordings are more stable and less affected by artifact, sensory nerves can also be assessed. NET has the advantage in that it assesses functional measures which may occur before pathologic axonal or demyelinating features and their NCS corollary become evident.
In patients with T1DM or T2DM, alterations in excitability, including features compatible with sodium channel dysfunction, have been reported in association with subclinical neuropathy.
These alterations progress with increasing neuropathy severity and are seen in patients with normal NCS, raising the possibility that NET could provide a window of opportunity to detect abnormalities before they become irreversible.
Evidence in individuals with diabetes and no neuropathy as well as those with “mild to moderate” neuropathy indicate that glycemic control significantly affects nerve excitability. For example, the alterations in NET seen in individuals with diabetes and no neuropathy can be reversed by improvements in glycemic control,
Underlining the potential of NET to detect changes after intervention, parallel improvements in nodal sodium currents and conduction velocity were shown in a small open clinical trial with the aldose reductase inhibitor epalrestat in mild to moderate DPN.
NET is not widely available and requires specialist equipment and software. There are no clinically relevant normative ranges and, although alterations in NET can be shown despite normal NCS, it has yet to be validated as a biomarker for either DPN or pDPN. As with NCS, it does not provide information about the status of small nerve fibers. However, it is a useful, noninvasive method to explore disease pathophysiology and may prove a useful biomarker when assessing response to treatment for DPN or pDPN such as in trials of therapeutic agents that act on voltage-gated ion channels.
Neuropathic Pain in Diabetes
A major unmet need for patients with pDPN is the ability to predict whether a particular drug is likely to be efficacious. A more “personalized” and mechanistically based approach to identify the pain generator or modulatory site(s) would enable greater selectivity and targeting of drug therapy, which would limit side effects and improve overall efficacy.
Neuropathic pain drugs work in specific locations, as defined by the receptors that they target. For instance, gabapentin and pregabalin exert their analgesic effect through high-affinity binding and modulation of the calcium channel α2-δ proteins in the dorsal root ganglion.
In contrast, duloxetine relieves neuropathic pain through inhibition of serotonin and norepinephrine reuptake, which enhances descending inhibition of pain.
Consequently, prescribing drugs that target peripheral nerves when the pain is being generated or modulated in the central nervous system (CNS) is likely to fail. The following sections address neurophysiological methods that are potential biomarkers of peripheral and centrally generated/modulated pain in pDPN.
Peripherally Driven Neuropathic Pain and Nociceptor Hyperexcitability
has been proposed to cause hyperexcitability of nociceptor afferents leading to abnormal spontaneous activity and sensitization of peripheral endings. These features could contribute to spontaneous pain and hyperalgesia or could drive central sensitization.
In support of the nociceptor hyperexcitability hypothesis, it has been suggested that gain-of-function variants of genes that encode voltage-gated sodium channels which cause very rare Mendelian pain disorders may play a role in common acquired neuropathic pain disorders such as pDPN. For example, rare SCN9A variants, defined as minor allele frequency of <1% in the general population, are associated with pDPN
However, in a cohort of participants with painful and painless polyneuropathy, analysis of low-frequency variants in sodium channels, defined as minor allele frequency of <5% in the general population, did not associate with neuropathic pain.
Therefore, in certain well-defined clinical contexts, such as pDPN or SFN, rare sodium channel variants may act as risk factors for the development of neuropathic pain.
The identification of clinically relevant sodium channel variants offers the potential for personalized and more effective treatment.
clinically relevant sodium channel variants in pDPN are currently identified in only a small minority of patients. With more specific sodium channel blockers on the horizon, the identification of pDPN with nociceptor hyperexcitability pDPN is a priority.
NET has shown evidence of Na+ channel dysfunction that associates with pDPN.
However, this method assesses large myelinated fibers and cannot detect dysfunction in A-delta or C-fiber nociceptors. Indirect evidence of nociceptor hyperexcitability might be obtained by using neuropathic pain phenotyping, either based on detailed symptom-based questionnaires
Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): somatosensory abnormalities in 1236 patients with different neuropathic pain syndromes.
The Pain in Neuropathy Study (PiNS): a cross-sectional observational study determining the somatosensory phenotype of painful and painless diabetic neuropathy.
However, most patients with pDPN show a loss of function on QST rather than evidence of nociceptor hyperexcitability, and there is no clear correlation between positive symptoms for hyperexcitability such as allodynia on questionnaires and QST/clinical findings.
The Pain in Neuropathy Study (PiNS): a cross-sectional observational study determining the somatosensory phenotype of painful and painless diabetic neuropathy.
The only method that can directly detect nociceptor hyperexcitability is microneurography.
Microneurography
Microneurography involves insertion of a fine tungsten electrode into a peripheral nerve. For evaluation of peripheral neuropathy, this typically involves recording from the peroneal or superficial peroneal nerves. The electrode is micromanipulated to enable recording of unmyelinated fibers (Figure 1). A typical paradigm involves stimulation of the cutaneous receptive field(s) of the unmyelinated fibers by using electrical stimuli. Using this method, several C-fibers can be recorded at one time. A raster plot is generated showing action potentials time-locked to the electrical stimulus with latencies appropriate for the slow conduction velocity of C fibers. This enables differentiation of nociceptor subtypes (eg, polymodal nociceptors vs silent nociceptors).
Normally, there is a stable baseline latency to low-frequency stimulation of the receptive field, but when a fiber is spontaneously active an irregular “saw-tooth” baseline is seen.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
The method can therefore provide evidence of both spontaneous activity and sensitization. Patients with painful peripheral neuropathy have been shown to have abnormal hyperexcitability of nociceptor fibers.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
Microneurographic studies in DPN have shown variable findings. In patients with documented large fiber neuropathy, one study found an altered distribution of nociceptor subtypes with a higher proportion of mechanically insensitive to mechanically sensitive C-nociceptors, as well as evidence of loss of mechanical sensitivity in normally sensitive fibers.
Although a small proportion of nociceptor fibers were spontaneously active, this did not significantly differ between pDPN and DPN. Other studies have shown a higher proportion of hyperexcitable nociceptor fibers in patients with pDPN.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
reported spontaneous activity in 57% of 109 recorded nociceptor fibers in patients with pDPN but of undefined neuropathy severity. Although this study did not use a DPN group without pain, the rates of spontaneous activity (~5%) were higher than in age-matched healthy individuals.
Although microneurography is an invaluable tool for investigating the presence of nociceptor hyperexcitability, its use as a biomarker is limited. The technique is invasive, albeit minimally so, and time-consuming for both the patient and investigator. Furthermore, it requires specialist equipment and highly trained operators, and at present is performed only in a limited number of centers worldwide. These factors significantly affect the availability and cost-effectiveness of microneurography. Also, age and potentially disease-specific definitions of normative and abnormal data are needed so that its sensitivity and predictive value can be determined, and it can be established as a clinical tool. A further drawback is that each session may yield only small numbers of nociceptor recordings that are suitable for analysis. For example, in one study, only 1 to 3 suitable nociceptor fibers were recorded per session.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
This limits diagnostic applicability and capacity for monitoring interventions on an individual level. It is not known how microneurography compares with the indirect, but considerably more widely available, QST techniques in identifying patients with hyperexcitable nociceptors. Furthermore, the relationship between abnormalities detected by microneurography and structural changes of small fibers in the skin is unknown. The method is potentially of use in clinical trials, especially those with a focus on underlying pain mechanisms, in which patients with and without evidence of hyperexcitability can be segregated.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
may generate and modulate neuropathic pain in patients with diabetes. Multiple pathophysiological changes can occur in the CNS as a result of peripheral nervous system pathology.
These secondary CNS changes can inappropriately amplify or fail to suppress incoming signals from the periphery, a mechanism of pain called disinhibition.
The inhibitory neurotransmitters γ aminobutyric acid (GABA) and glycine, active through local inhibitory interneurons, play critical roles in regulating spinal nociceptive processing. In animal models, pharmacologic or genetic interventions that enhance or diminish spinal inhibition result in decreased and increased behavioral indices of pain, respectively.
Muscimol prevents long-lasting potentiation of dorsal horn field potentials in rats with chronic constriction injury exhibiting decreased levels of the GABA transporter GAT-1.
In diabetic rodents, spinal inhibitory processes are dysfunctional. Allodynia in the streptozotocin (STZ)-rat model of T1DM is driven by spinal disinhibition in which GABA, acting via spinal GABA-A receptors, is no longer inhibitory and becomes pro-nociceptive.
The mechanism involves downregulation of the postsynaptic chloride pump KCC2, which alters the chloride equilibrium potential that dictates the direction of ion flow through the GABA-A receptor. Accordingly, spinal GABA-A blockade, which under normal circumstances is pro-nociceptive, reverses allodynia in STZ-rats.
The H-reflex arc comprises Ia afferent fibers, derived from muscle spindles, forming strong monosynaptic connections with alpha motor neurons within the spinal cord. Although originally considered a monosynaptic trans-spinal reflex,
The H-reflex is measured in humans using a simple modification of traditional NCS. The stimulation protocol evokes 2 waveforms: a direct nerve to muscle M wave and a longer latency trans-spinally mediated H wave (Figures 1 and 2). Rate-dependent depression (RDD) is the measure of the change in amplitude of the H-wave component over consecutive stimulations. In normal rats, RDD is driven via activation of inhibitory spinal GABA-A receptors.
Loss of RDD occurs in both humans and animals after disinhibition of sensory processing caused by spinal cord injury, and in rats this is linked to reduced spinal KCC2 expression and subsequent loss of GABAergic inhibition.
Development of GABA-sensitive spasticity and rigidity in rats after transient spinal cord ischemia: a qualitative and quantitative electrophysiological and histopathological study.
Loss of RDD can be used as a biomarker to separate rats with behavioral markers of neuropathic pain caused by spinal disinhibition from rats with similar behavioral manifestations that are of peripheral origin.
Intriguingly, the selective serotonin and norepinephrine reuptake inhibitor duloxetine, acting via spinal 5HT2A receptors, not only alleviates behavioral indices of pain in STZ-rats
Figure 2Representative electromyogram traces showing M and H waves in response to 3 consecutive stimulations at 1 Hz frequency in a patient with painless (upper triplicate) and painful (lower triplicate) diabetic neuropathy. The H-wave amplitude declines in the traces from the patient with painless neuropathy and illustrates rate-dependent depression. DPN = diabetic peripheral neuropathy.
There is preclinical evidence that loss of RDD and indices of neuropathic pain share a common pathogenic mechanism involving spinal KCC2 depletion and disinhibition caused by inversion of GABA-A receptor function and exhibit common responses to spinally acting analgesics outside the root pathogenic mechanism. This evidence supports the concept that RDD status may be a viable biomarker for both identifying the dominant generator site in individual patients with pDPN and for predicting efficacy of therapeutic strategies that alleviate spinal disinhibition.
Our research has previously measured the magnitude of RDD in patients with T1DM and painful or painless neuropathy and age-matched healthy control subjects.
There was loss of RDD in patients with pDPN compared with both healthy control subjects and patients with painless DPN (Figure 2). These findings were independent of glycemic control and severity of large fiber neuropathy and SFN. Loss of RDD correlated with pain severity, indicating that more severe pain is associated with the loss of RDD and, by inference, spinal disinhibition.
Expanding these data from patients with T1DM, we have recently shown that, similar to findings in the Zucker Diabetic Fatty rat model of T2DM,
impaired RDD is also seen in patients with T2DM and pDPN (unpublished observations, Marshall 2020). Importantly, not all subjects with pDPN had impaired RDD (defined as a H3:H1 ratio >2 × SD of control group mean). This supports the hypothesis that impaired RDD may serve as a clinical biomarker in a subset of patients in whom pain arises primarily from spinal disinhibition. Approximately 60% of patients with diabetes will develop neuropathy, approximately one third of those will develop pDPN, and, from our exploratory studies, 40% of those will exhibit RDD deficits.
This heterogeneity could plausibly be used to enable definition of abnormal values for predicting response to spinally acting drugs. It is important to acknowledge that a proportion of patients in these studies were taking prescribed antineuropathic pain medication and that therapeutic responses were not studied in a systematic manner. Similarly, whether antineuropathic pain medication affects or normalizes RDD either in a general or drug-specific manner is unknown.
The measurement of RDD is noninvasive and potentially as widely available as current NCS. One drawback is that the H-reflex can be absent or difficult to elicit with increasing severity of DPN
or in patients with co-existing S1 radiculopathy. This may limit the use of RDD as a biomarker in patients with severe neuropathy. Further larger scale studies controlling for the use of antineuropathic pain medication which systematically assess therapeutic response are also required to determine the sensitivity and predictive value of RDD, both as a biomarker of spinal disinhibition and as a predictor of neuropathic pain medication efficacy in patients with pDPN. It is not yet known if spinal disinhibition and impaired RDD in pDPN
are associated with a distinctive (and measurable) pain phenotype or whether they relate to other putative mechanisms that facilitate ascending nociceptive drive from the spinal cord such as wind-up or an abnormal descending pain modulatory system
; all of these are potentially easier to measure in an outpatient setting using targeted QST.
Conclusions
The early diagnosis of DPN and management of pDPN continue to pose considerable challenges. NCS are a long-established biomarker for DPN that predict clinically relevant outcomes. However, biomarkers that capture SFN are also required to facilitate the earliest detection of small, as well as large, fiber DPN. Noninvasive neurophysiological methods such as pain-related evoked potentials, which assess the function of small fiber pathways, and NET, which has the potential to detect nerve dysfunction before it is irreversible, are potential biomarkers capable of detecting early DPN. Although these approaches are currently not practical for use in busy outpatient settings, they may be suited as diagnostic biomarkers and end points in mechanistic clinical trials of DPN.
For pDPN, emerging methods such as H-reflex RDD and microneurography have the potential to segregate patients on the basis of pain mechanisms. They may also allow assessment of the balance between peripheral versus spinal drive in pDPN. Thus, they have the potential to function as biomarkers for trial enrichment and for segregating patients for mechanistic informed clinical trials in pDPN. Unlike microneurography, assessment of spinal disinhibition with RDD is a noninvasive and potentially widely available method that could be used to inform physicians about the optimal choice of drug for individual patients on the basis of pain mechanisms.
Conflicts of Interest
The authors have indicated that they have no conflicts of interest regarding the content of this article.
Acknowledgements
This study was supported by American Diabetes Association Award 1-17-ICTS-062 (Andrew G Marshall PhD and Nigel Calcutt PhD).
References
Abbott C.A.
et al.
Prevalence and characteristics of painful diabetic neuropathy in a large community-based diabetic population in the U.K.
Quantitative sensory testing in the German Research Network on Neuropathic Pain (DFNS): somatosensory abnormalities in 1236 patients with different neuropathic pain syndromes.
Sural nerve fibre pathology in diabetic patients with mild neuropathy: relationship to pain, quantitative sensory testing and peripheral nerve electrophysiology.
Corneal confocal microscopy shows an improvement in small-fiber neuropathy in subjects with type 1 diabetes on continuous subcutaneous insulin infusion compared with multiple daily injection.
Corneal confocal microscopy compared with quantitative sensory testing and nerve conduction for diagnosing and stratifying the severity of diabetic peripheral neuropathy.
The Pain in Neuropathy Study (PiNS): a cross-sectional observational study determining the somatosensory phenotype of painful and painless diabetic neuropathy.
Establishing high-quality reference values for nerve conduction studies: a report from the normative data task force of the American Association Of Neuromuscular & Electrodiagnostic Medicine.
Asymptomatic small fiber neuropathy in diabetes mellitus: investigations with intraepidermal nerve fiber density, quantitative sensory testing and laser-evoked potentials.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial.
Muscimol prevents long-lasting potentiation of dorsal horn field potentials in rats with chronic constriction injury exhibiting decreased levels of the GABA transporter GAT-1.
Development of GABA-sensitive spasticity and rigidity in rats after transient spinal cord ischemia: a qualitative and quantitative electrophysiological and histopathological study.
00155-7&id=gr1.jpg){kind=link}
00155-7&id=gr2.jpg){kind=link}