Cold-evoked potentials in Fabry disease and polyneuropathy

Abstract

Background: Fabry disease (FD) causes cold-evoked pain and impaired cold perception through small fiber damage, which also occurs in polyneuropathies (PNP) of other origins. The integrity of thinly myelinated fibers and the spinothalamic tract is assessable by cold-evoked potentials (CEPs). In this study, we aimed to assess the clinical value of CEP by investigating its associations with pain, autonomic measures, sensory loss, and neuropathic signs.

Methods: CEPs were examined at the hand and foot dorsum of patients with FD (n = 16) and PNP (n = 21) and healthy controls (n = 23). Sensory phenotyping was performed using quantitative sensory testing (QST). The painDETECT questionnaire (PDQ), FabryScan, and measures for the autonomic nervous system were applied. Group comparisons and correlation analyses were performed.

Results: CEPs of 87.5% of the FD and 85.7% of the PNP patients were eligible for statistical analysis. In all patients combined, CEP data correlated significantly with cold detection loss, PDQ items, pain, and autonomic measures. Abnormal CEP latency in FD patients was associated with an abnormal heart frequency variability item (r = -0.684; adjusted p = 0.04). In PNP patients, CEP latency correlated significantly with PDQ items, and CEP amplitude correlated with autonomic measures (r = 0.688, adjusted p = 0.008; r = 0.619, adjusted p = 0.024). Furthermore, mechanical pain thresholds differed significantly between FD (gain range) and PNP patients (loss range) (p = 0.01).

Conclusions: Abnormal CEPs were associated with current pain, neuropathic signs and symptoms, and an abnormal function of the autonomic nervous system. The latter has not been mirrored by QST parameters. Therefore, CEPs appear to deliver a wider spectrum of information on the sensory nervous system than QST alone.

Keywords: Fabry disease; cold-evoked potentials; diagnostic workup; neuropathic pain; neurophysiology; polyneuropathy.

Figure 1

Overview of the study population (controls, PNP and FD patients) and the performed statistical analyses with a summary of significant results after correction for multiple testing. For comprehensive data, see the Results section of the manuscript. CDT, cold detection threshold; CEP, cold-evoked potential; CPT, cold pain threshold; HPT, heat pain threshold; FD, Fabry disease; MDT, mechanical detection threshold; MWU, Mann–Whitney U; PDQ, painDETECT questionnaire; PNP, polyneuropathy; TDM, time-domain measurement; QST, quantitative sensory testing; WDT, warm detection threshold; VDT, vibration detection threshold.

Figure 2

Quantitative sensory testing (QST) profiles of controls and PNP and FD patients. The asterisks indicate significant group differences; *** ≙ p < 0.001; ** ≙ p < 0.01; * ≙ p < 0.05; group symbols are indicated on the right side of the figure. FD patients exhibit a loss of Aδ-function, while the PNP group exhibits a wide range of dysfunction with functional loss of small and large fibers. The controls are within the normative range of Z-values. CDT, cold detection threshold; WDT, warm detection threshold; CPT, cold pain threshold; HPT, heat pain threshold; MPT, mechanical pain threshold; MDT, mechanical detection threshold; VDT, vibration detection threshold; FD, Fabry disease; PNP, polyneuropathy.

Figure 3

Grand averages of CEPs with standard deviations. Grand averages of CEPs derived from the hand and foot are displayed for each group (controls, FD and PNP patients). The black line equals the averaged EP of each cohort, the dashed line equals one standard deviation of the averaged EP data. N2 and P2 markers indicate the CEP potential. Within the PNP group, at the foot, there was no identifiable grand-average potential due to heterogenous EP latencies caused by heterogenous loss of function in different individuals. CEP, cold-evoked potential; EP, evoked potential; FD, Fabry disease; PNP, polyneuropathy.

Figure 4

Overlay of grand-averaged CEPs. After performing the grand average of the CEP data there were no visible averaged CEPs within the PNP group at the hand and foot, due to the heterogeneity of normal and abnormal values. N2 and P2 markers indicate the CEP potential where applicable. CEP, cold-evoked potential; EP, evoked potential; PNP, polyneuropathy.

Figure 5

Dot blot and box blot of CEP data. N2 and P2 latencies as well as N2P2 amplitudes of controls and FD and PNP patients are shown here. The dots indicate individual patient data. The box blots indicate minimum and maximum values (T-bars) first quartile, median, and third quartile. There is a strong latency difference between PNP patients and controls, but not between FD patients and controls, supporting that PNP patients exhibit a more severe small fiber dysfunction than FD patients (see also Figure 2). FD, Fabry disease; PNP, polyneuropathy.

Figure 6

Quantitative sensory testing (QST) correlations. Scatter plots for the significant QST correlations (which withstood correction for multiple testing) within both patient groups combined (A) and FD patients (B). In the case of the PNP patients, the calculations did not withstand correction for multiple testing. (A) The correlation between CDT and the PDQ score indicates that a loss of cold fiber function is associated with a neuropathic pain (NP) component and the sensory symptom of pain attacks. (B) Cold fiber function loss is strongly correlated with NP in FD patients. The loss of C-fiber function was associated with higher ratings in the FabryScan and a stronger sensation of numbness. CDT, cold detection threshold; WDT, warm detection threshold; QST, quantitative sensory testing; FD, Fabry disease; PNP, polyneuropathy; PDQ, painDETECT questionnaire.

Figure 7

EP correlations within all patients combined. Scatter plots for the significant CEP correlations (which withstood correction for multiple testing) within both patient groups combined. A loss of function indicated by negative CDT Z-scores was associated with a CEP amplitude reduction. A prolonged N2 latency was indicative of higher average and maximum pain as well as stronger tingling sensation and numbness. Interestingly, a CEP amplitude decrease was associated with a functional loss of the autonomic nervous system. As a proof of concept, N2 correlated significantly with height (bottom right). CEP, cold-evoked potential; CDT, cold detection threshold; NRS, numeric rating scale; PDQ, painDETECT questionnaire; RMSSD, root mean square of successive square differences; TDM, time-domain measurement.

Figure 8

EP correlations for PNP and FD patients each. Scatter plots for the significant CEP correlations (which withstood correction for multiple testing) within PNP patients (A) and FD patients (B) each. (A) A prolonged N2 latency was associated with neuropathic pain (NP) and sensory symptoms (PDQ score and PDQ items). A CEP amplitude reduction was associated with a functional loss of the autonomic nervous system. (B) In FD patients, a prolonged N2 latency was associated with a functional loss of the autonomic nervous system. There were no other significant correlations. CEP, cold-evoked potential; FD, Fabry disease; NRS, numeric rating scale; PDQ, painDETECT questionnaire; PNP, polyneuropathy; RMSSD, root mean square of successive square differences; TDM, time-domain measurement.