Authors: Sanjeev Sharma, Gerry Rayman
Categories: Endocrinology, diabetes, early diagnosis diabetes neuropathy, large fibre neuropathy, polyneuropathy, small fibre neuropathy
Source: Frontiers in Endocrinology
Diabetes sensory polyneuropathy (DSPN) is a significant complication of diabetes affecting up to 50% of patients in their lifetime and approximately 20% of patients suffer from painful diabetes neuropathic pain. DSPN – both painless and painful - leads to considerable morbidity including reduction of quality of life, increased lower limb amputations and is associated with worsening mortality. Significant progress has been made in the understanding of pathogenesis of DSPN and the last decade has seen newer techniques aimed at its earlier diagnosis. The management of painful DSPN remains a challenge despite advances made in the unravelling the pathogenesis of pain and its transmission. This article discusses the heterogenous clinical presentation of DSPN and the need to exclude key differential diagnoses. Furthermore, it reviews in detail the current diagnostic techniques involving both large and small neural fibres, their limitations and advantages and current place in the diagnosis of DSPN. Finally, the management of DSPN including newer pharmacotherapies are also discussed.
Keywords: diabetes, polyneuropathy, small fibre neuropathy, large fibre neuropathy, early diagnosis diabetes neuropathy
The term diabetes sensorimotor polyneuropathy (DSPN) refers to a heterogeneous group of neurological disorders – either clinically evident or subclinical – that occur in the setting of diabetes mellitus and cannot be attributed to other aetiologies of peripheral neuropathy (1). It is by far the he commonest form of diabetes polyneuropathy (DPN) ( Table 1 ) affecting up to 50% of people with diabetes, while its yearly incidence amounts to approximately 2% (3, 4). DSPN is defined as a symmetrical, length-dependent polyneuropathy attributable to not only chronic hyperglycaemia-mediated microvascular alterations but also contributed by other cardiovascular factors including dyslipidaemia, hypertension and smoking (1).
The classification of DSPN varies and is based on a number of factors including presence or absence of painful symptoms, pattern of neural involvement and the setting of its evaluation – clinical or research. Chronic painful DSPN is found in up to 25% of subjects with diabetes and is defined as persistent or recurrent pain lasting 3 months and caused by a lesion or disease of the somatosensory system due to diabetes and after exclusion of other causes (5, 6). ( Table 2 ) The position statement by American Diabetes association (ADA) proposed a clinical classification of DSPN into either primarily large and small fibre or mixed based on the type of neural fibres involved (2). This differentiation will be more elaborately discussed later.
A practical definition of DSPN in the clinical setting is the presence of neuropathic symptoms as reported by the patient and/or signs of peripheral nerve dysfunction in diabetes subjects after exclusion of other causes. Historically – in busy diabetes clinics – the presence of abnormal ankle reflex, abnormal 128 Hz tuning fork sensation or insensitivity to 10-gm Semmes-Weinstein monofilament (SWMF) was considered to be valid evidences to diagnose DPN. More accurate assessments like nerve conduction studies were time consuming and resource-specific and hence reserved for atypical presentation including truncal radiculopathy, entrapment neuropathy and amyotrophy (7, 8). Whist these methods were undoubtedly useful in assessing risk of patients of foot ulceration, they are limited in their capabilities to assess mostly Aα and Aβ nerve myelinated fibres ( Figure 1 ) which serve touch, proprioception, position sense, vibration and muscle control. Consequently, they are affected later in the natural history of diabetes (10). There is now a large body of evolving evidence that small neural fibres (thinly myelinated Aδ and unmyelinated C fibres) are affected early in the natural history of DSPN and precedes large fibre involvement (11). Hence the specific assessment of this group offers hope for earlier diagnosis of DSPN which can lead to quicker adoption of disease modifying strategies such as life style modification, glycaemic, blood pressure and lipid control or novel therapeutics to prevent onset and progression of worsening neuropathy (9).
Figure 1 Types of neural fibres [adapted from Sharma S et al. (9)].
The commonest clinical presentation of DSPN is a distal symmetrical type of progressive sensorimotor polyneuropathy which is length dependent i.e., the lower limbs are affected earlier than upper extremities and with progression of the condition, there is distal to proximal axonal degeneration; hence many patients present with the typical “stocking and glove” distribution of symptoms (2). Up to 50% of subjects remain asymptomatic and hence the need for at least annual review of their feet sensations to prevent ulceration. Symptoms at presentation can be either “negative” (loss of sensations including touch, vibration, position or deficits including ataxia) or “positive” (pain, paraesthesia, allodynia). Not uncommonly patients report the subjective symptoms of “numbness” which could indicate either of the above group of symptoms and it is up to the clinicians to make this differentiation to enable treatment (12). It is important to note that autonomic dysfunction is common in patients with DSPN but is largely undetected unless heralded by symptoms as mentioned in Table 1 (2).
Common determinants for progression of DSPN include patient characteristics (advancing age, height and obesity), clinical parameters (presence of hypertension and uncontrolled dyslipidaemia) and adverse glycaemic matrices (poor glycaemic control, longer duration of diabetes, beta cell insufficiency) (1, 3, 13). Whilst it is beyond the scope of this article to examine each of these determinants individually, research into early diabetes neuropathy using methods of small fibre function have shown that the impact of glycaemic control in the progression of microangiopathy is much more important in type 1 diabetes whilst in type 2 diabetes, the impact is more multifactorial (14, 15).
The traditional and still mostly prevalent screening tool assessment of diabetes feet in busy clinical environments remains the use of a 10gm-SWMF or 128-Hz tuning form (detects large fibres) and pin-prick sensation (detects small fibres) (2). It must be highlighted that these bedside techniques are useful for assessing diabetes feet at risk of future ulceration and are not sensitive for early detection of DSPN (16). More recently, the Ipswich Touch test has been shown to be an equally sensitive method for evaluating at-risk feet with a strong concordance with the 10-gm monofilament (17, 18).
The diagnosis of painful DSPN is based on presence of painful neuropathic symptoms and a clinical diagnosis of DSPN after exclusion of other neuropathies as shown in Table 2 .
Over the years, many scored clinical assessments have been proposed as standardized objective and quantitative measures for screening and grading of severity of DSPN. Their utility is particularly useful for epidemiological studies looking at prevalence of DSPN in larger populations. Of these, the Michigan Neuropathy Screening Instrument (MNSI) (2, 19), Toronto Clinical Neuropathy Score (20), modified Toronto Neuropathy Score (21), Neuropathy Disability Score (16), Neurological Disability Score (22), Neuropathy Symptom Score (22) and Utah Early Neuropathy (23) are commonly used. It is beyond the scope of this article to examine each score individually but their usage is limited by the time required in busy clinical settings. Recently, machine-learning severity prediction tools based on the MNSI have been proposed to save time (24).
For painful DSPN, useful scoring tools include the Neuropathy Total Symptom Score-6 (25), DN4 (26), PainDETECT (27) and Neuropathic Pain Symptom Inventory (28).
In order to overcome the shortcomings of screening tools and to reduce the time taken by scoring systems in clinical practice, a number of point-of-care devices have evolved in the past few years which have the ability to reduce bias and provide more accurate results. The Vibratip™ for objective assessment of vibration sense is one such device (29). Another device is DPNCheck™ which a hand-held device providing sural nerve conduction velocity and amplitude in under five minutes correlates well with Neuropathy Disability score (30, 31). Similarly the SUDOSCAN™ is another well-established device which by assessing electrochemical skin conductance as a measure sudomotor function has been shown to show early small fibre changes (autonomic system) in patients at risk of DSPN (32, 33).
Small fibre neuropathy (SFN) is a sub-type of peripheral neuropathy affecting the thinly myelinated Aδ or the non-myelinated C fibres ( Figure 1 ). These fibres constitute 79.6% to 91.4% (39, 40), of peripheral nerve fibres and mediate several key functions including temperature and pain perception, sweating, and tissue blood flow - all of which, when impaired are pathophysiologically related to adverse outcomes associated with foot ulcerations in people with diabetes
Small fibre methods can be broadly classified into methods for assessing small fibre function and small fibre structure. Table 3 provides a summary of methods currently available.
Figure 2 The laser doppler imager (LDI
flare) is a non-invasive technique of measuring C fibre-mediated cutaneous vasodilatation in the foot skin in response to thermal heating. The heating is done by a 1cm2 probe and the area of induced hyperaemia measured by a 610nm laser probe. The scan image on left is a 11cm^2^ flare area in a healthy control while on the right is 1.1cm^2^ flare in a subject with small fibre neuropathy.
Autonomic dysfunction is common in up to 50% of subjects with DSPN but remains either asymptomatic or undiagnosed. The presenting features include impairment of cardiovascular, gastrointestinal, urogenital, thermoregulatory, sudomotor, and pupillomotor function. The accurate diagnosis of the autonomic neuropathies has been enhanced by the availability of physiological tests that measure autonomic function, and more recently, structural studies of the autonomic cutaneous innervation (57). Presence of cardiovascular autonomic neuropathy (CAN) in the presence of DSPN is an independent risk factor increased mortality and morbidity (58). The Composite Autonomic Symptom Score (COMPASS) 31 score is a validated, easy-to-use, quantitative assessment tool for autonomic symptoms in diabetic neuropathy, with a fair diagnostic accuracy for both cardiovascular autonomic neuropathy and diabetic polyneuropathy (59). A detailed description for autonomic neuropathy tests is beyond the scope of this article where the focus in mainly on DSPN.
Both the brain and the spinal cord have been implicated in the genesis of painful DSPN and broadly involves a shift towards excitation and reduced inhibition of transmission of pain signals (60). The brain network involved in such chronic pain states is partially distinct from those involved in acute pain. Magnetic resonance spectroscopy has been used to determine the neuronal function in the thalamus and primary somatosensory cortex in patients with and without painful DSPN. Early results have shown thalamic neuronal dysfunction was found in only advanced painless DSPN but preserved in both subclinical and painful DSPN. These methods show promise in improving our understanding of painful DSPN and might have a role in future targeted pharmacotherapy (61).
The pillars of diabetes management include early diagnosis of complications and steps to prevent their worsening. For other microangiopathic complications, advances have been made for their early detection e.g., retinal screening for diabetes retinopathy and urinary microalbuminuria for diabetes nephropathy (62). Yet in contrast, The current ADA position statement on DPN endorses the “Assessment for distal symmetric polyneuropathy should include a careful history and assessment of either temperature or pinprick sensation (small-fibre function) and vibration sensation using a 128-Hz tuning fork (for large-fibre function). All patients should have annual 10-g monofilament testing to identify feet at risk for ulceration and amputation” (2, 4). It should be emphasized that the ADA’s position is to “identify feet at risk for ulceration and amputation” rather than early diagnosis of neuropathy; in essence the major focus of this guidance is the prevention of disabling diabetes lower limb complications.
A comprehensive review of various methodologies for the assessment of DSPN has been done but it is to be noted that none of them, except the screening bedside tools have found their way to any established guidance for detection of DPN. There are many reasons for firstly, the evidences available are mostly cross-sectional and only 6 studies have been performed longitudinally – the longest being at 5 years (63). Secondly, the lack of globally agreed-upon normative data for these tests make them less practical for routine use. So far, only confocal microscopy (CCM) and intra-epidermal nerve fibre density (IEFND) have acceptable normative values (64, 65). Thirdly, age-related decline in neural function and structure is well recognised. Hence normative data should also be age-linked and only the LDIFLARE technique has shown such age-linked reduction of small fibre function (43).
Furthermore, it well known that with duration of diabetes, there is progression of other microangiopathic complications including diabetes kidney disease and it is well known that progressive uraemia affects neural health (66). Hence normative values should also be linked to renal health i.e. linked to measures of creatinine and eGFR. Finally, the methodologies of various techniques are still not uniform with various centers using their own adaptions to improve performance. Hence in view of considerable heterogeneity involved with various techniques, it is unlikely that they will appear in guidelines soon unless a much-needed consensus is agreed upon by investigators and equally importantly accepted by legislators.
The pillars of management of DSPN can be categorized as follows;
Adoption of healthy lifestyle measures and weight loss remain the cornerstone of non-pharmacological management of DSPN. Whilst there are epidemiological studies which have shown the prevalence of less neuropathic symptoms in type 2 diabetes subjects with lower body weight (67). Interestingly, in a recent prospective cohort study of 131 obese participants attending a medical weight-management program, improvements were observed on the MNSI questionnaire, two Quality of Life in Neurological Disorders subdomains, and quantitative sensory testing cold thresholds whilst IENFD remained stable (68).
The impact of intensive glycaemic control in retarding the progression of DSPN has been well demonstrated in type 1 diabetes in the DCCT/EPIC study (69) but specific evidence to suggest the same is lacking in type 2 diabetes (70, 71). The reason for the latter could be due to the difference of pathogenesis of neuropathy in both types of diabetes and the effect of metabolic syndrome in type 2 diabetes. Nevertheless, it is well accepted that improved glycaemic control reduces the onset and progression of DSPN.
The EURODIAB IDDM study (13) of 3250 type 1 diabetes subjects showed significant correlations between the presence of diabetic peripheral neuropathy with age, duration of diabetes, quality of metabolic control, height, the presence of background or proliferative diabetic retinopathy, smoking, dyslipidaemia and the presence of cardiovascular disease. However prospective studies are lacking to indicate that control of any of the above factors are helpful in retarding the progression of DSPN although it is widely believed that hypertension and dyslipidaemia need to be aggressively managed to limit microangiopathic complications (72, 73).
Several compounds are available which target major pathways implicated in the pathogenesis of DSPN including polyol pathway, hexosamine pathway, protein kinase C (PKC) activity, and advanced glycation end products (AGEs) pathway (29). The AGE inhibitor Benfotiamine and antioxidant A-lipoic acid are licensed as drugs for treatment of DSPN in many countries (74, 75). In the NATHAN 1 trial, neuropathic deficits improved after 4 years in patients with mild to moderate largely asymptomatic DSPN (76). Actovegin, a poly(ADP-ribose) polymerase (PARP) inhibitor, is authorized in mainly eastern European countries, while the aldose reductase inhibitor Epalrestat is being used in India and Japan for DPN (77, 78).
Similarly for painful DSPN, several pharmacotherapies have been shown to reduce painful symptoms by their mechanisms of action on the pathways mentioned above. Treatment with A-lipoic acid 600mg twice daily for 6 months has been shown to reduce symptoms of pain, paresthesia and numbness in symptomatic DSPN (79). A-lipoic acid infusions over 3 weeks have also been shown to significantly reduce neuropathic symptoms (75). Larger randomized controlled trials are needed to study their efficacy in multicentered population subsets to strengthen the rationale of use in DSPN.
Painful DSPN can often be a crippling symptom leading to significant reduction of quality of life for patients, many of whom also have neuropathic deficits leading to complications. The management of painful DSPN is often challenging due to difference in efficacy between individual agents, balance between side-effects and working patterns and drug interactions. Monotherapy is usually not tolerated in maximal dosages and recent evidences from the OPTION-DM study showed that combination treatment was well tolerated and led to improved pain relief in patients with suboptimal pain control with a monotherapy (80, 81). The broad group of mediations in this class are shown in Table 4 .
In diabetes subjects with refractory pain, spinal cord stimulation has been shown to significant pain relief and improved quality of life. It is an expensive invasive procedure needing insertion of percutaneous leads placed epidurally and connected to an implantable pulse generator typically placed in the low back. Whilst the mechanism of action is yet to fully understand, early results are promising but long-term results are awaited (87, 88). Other methods such as cognitive behavioral therapy, transcutaneous electrical nerve stimulation and acupuncture have a low level of evidence but can be successful in individual subjects (89).
DSPN is a major complication of diabetes. By the time a diabetes subject develops foot ulceration, the cardiovascular risk for death is increased by 50% and their 5-year mortality is worse than most cancers. With the worldwide prevalence of diabetes increasing to pandemic standards, it is anticipated that the consequences of DSPN in diabetes will continue to be on rise leading to serious challenges in both personal wellbeing and global health expenditure. Whilst on one hand, considerable advances in the understanding of the mechanisms of DSPN have been made, as discussed in this article, its early diagnosis and effective management remains uncertain and lacks uniformity. Whilst medications like the glucagon-like-peptide 1 receptor agonists and sodium-glucose co-transporters 2 have shown significant benefit in reducing cardiovascular mortality, heart failure and chronic kidney disease associated with diabetes, there is still a substantial unmet need in the therapeutics governing DSPN and other aspects of diabetes polyneuropathy. It is our combined responsibility to pool our resources, invest in uniform management algorithms and convince legislators to embrace newer diagnostic techniques so that burden and sequalae of DSPN can be lessened.
SS and GR researched, wrote and reviewed all parts of this manuscript.
The authors would like to thank the colleagues of the Diabetes Trials unit and Diabetes & Endocrine centre at Ipswich hospital for their continued support.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.