Assessing decreased sensation and increased sensory phenomena in diabetic polyneuropathies

Abstract

Loss of sensation and increased sensory phenomena are major expressions of varieties of diabetic polyneuropathies needing improved assessments for clinical and research purposes. We provide a neurobiological explanation for the apparent paradox between decreased sensation and increased sensory phenomena. Strongly endorsed is the use of the 10-g monofilaments for screening of feet to detect sensation loss, with the goal of improving diabetic management and prevention of foot ulcers and neurogenic arthropathy. We describe improved methods to assess for the kind, severity, and distribution of both large- and small-fiber sensory loss and which approaches and techniques may be useful for conducting therapeutic trials. The abnormality of attributes of nerve conduction may be used to validate the dysfunction of large sensory fibers. The abnormality of epidermal nerve fibers/1 mm may be used as a surrogate measure of small-fiber sensory loss but appear not to correlate closely with severity of pain. Increased sensory phenomena are recognized by the characteristic words patients use to describe them and by the severity and persistence of these symptoms. Tests of tactile and thermal hyperalgesia are additional markers of neural hyperactivity that are useful for diagnosis and disease management.

Figures

FIG. 1.

Density of Meissner corpuscles visualized by the pseudo-cholinesterase reaction in punch biopsy specimens of the skin of the toe of healthy humans to illustrate the influence of age on their density and distribution. Upper insets show Meissner corpuscles in the terminal phalanx of the first toe of a 4-year-old boy (A), 43-year-old man (B), and 76-year-old woman (C). D: Meissner corpuscle density of healthy subjects plotted on the age in years. There is a rapid decrease in density with development (as the surface area increases) and a further decline with aging. (The figure is redrawn from data in Bolton et al. A quantitative study of Meissner’s corpuscles in man. Neurology 1966;16:1–9.)

FIG. 2.

Illustrated are graded Dyck monofilaments, a modification of Semmes Weinstein monofilaments (North Coast Medical, Inc., Morgan Hill, CA) used in quantitative testing of touch-pressure sensation, altered to provide exponential increases of force suitable for neurosensory testing. Monofilaments A, B, C - - - I shown in A and B provide static loads that increase exponentially from −3 ln g to 5 ln g. In monofilament testing, to avoid a variable degree of impact, the monofilaments should first be brought to within 1 or 2 mm of the skin (C), gently lowered to make contact with the skin and bent to five-sixths of its extended length (D), and then slowly released, with the entire stimulus event taking 1.5–2 s. If null stimuli are used (e.g., in 2:1 alternative forced-choice testing), the observer should go through all the motions of stimuli testing but without making contact with the skin. E: The CASE IVc thermode is shown for evaluation of cooling and heat as pain threshold testing. The thermoelectric technology used allows giving of pyramidal- and trapezoid-shaped thermal stimuli (F). The CASE IVc system is manufactured by WR Medical Electronics. Typical patterns of hyperalgesia, normal response, and hypoalgesia using the CASE IVc system are shown in G. (The panels are reformatted from P.J.D.’s previous publications.) HP, heat as pain; JND, just noticeable difference.

FIG. 3.

This figure outlines the methodology and steps used to estimate sensation loss of predetermined standard cutaneous fields of the body’s surface area to estimate sensation loss of large and small sensory fibers in symmetrical length-dependent sensorimotor polyneuropathy (e.g., DSPN) and other sensorimotor polyneuropathies (e.g., familial amyloid polyneuropathy). The approaches were developed by P.J.D. and then were programmed for CASE IVc by WR Medical Electronics. The 95th and 99th percentile values for touch pressure and HP5 for each of the 10 sites were provided by P.J.D. and colleagues. As described in the text, the algorithm is designed to test only one side of the body (because the pathological process is assumed to be symmetrical); not continuing testing when thresholds are found to be 99th percentile; and testing only lateral leg and forearm sites and only a few additional sites depending on length dependence of sensation loss. Most aspects of testing (finding the threshold, comparing the results to reference values, selecting a subsequent site to be tested, and printout of results) are automated (smart QSTing). TP, touch pressure.

FIG. 4.

The top figure provides composite NC normal deviate score (nds) values (from percentiles) of healthy subjects (open circles) and five patients with diabetes (DM) and borderline and abnormal values (solid circles). In the bottom figure, the combined normal deviate values of the composite NC score normal deviate values of ENFs/1 mm of the same healthy subjects and diabetic patients are shown. A 50th and 2.5th percentile line is shown for both top and bottom figures. Estimating abnormality using both NC (large-fiber function) and ENFs/1 mm as compared with use of only NC appears not to have altered the pattern of abnormality a great deal, but patients D and E, who were low normal by NC criteria, are just abnormal when assessed by both NC and ENFs/1 mm. *Abnormality is in the lower tail; †based on n = 330 Rochester Diabetic Neuropathy Study of Healthy Subjects. (Data are from Engelstad et al. ENFs – confidence intervals and continuous measures with NC. Neurology 2012;79:2187–2193.)

FIG. 5.

Plotted are the HP thresholds (HP5 and HP0.5 using the CASE IVb system) at various times (days) before and after intracutaneous injections of minute amounts of NGF (open circles) into the skin of the volar forearm of healthy subjects as compared with injection of the contralateral forearm with saline. Subjects and observers were masked as to which side was injected with NGF. Sixteen healthy subjects were assessed over a period of 28 days. A significant decrease in HP0.5 and HP5 followed NGF injection, and this was significant for HP5 for all periods between 3 h and 21 days after NGF injection; for HP0.5, significant thermal hyperalgesia was found for 1, 3, and 7 days after injection. The data provide convincing evidence that the methodology used in CASE IVc is a useful methodology for the demonstration of thermal hyperalgesia. In this study, mechanical hyperalgesia was also demonstrated. (Data are from Dyck et al. Intradermal recombinant human NGF induces pressure allodynia and lowered heat-pain threshold in humans. Neurology 1997;48:501–505.) †, ††Statistically significant differences in the heat pain responses at the time interval shown.