Sensory profile in primary restless legs syndrome and restless legs syndrome associated with small fibre neuropathy.

Sir, The association of restless legs syndrome with polyneuropathy, especially with involvement of small sensory fibres, is suggested by several observations (Iannaccone et al., 1995; Polydefkis et al., 2000; Gemignani et al., 2006); however, it is still controversial (Hattan et al., 2009). Bachmann et al. (2010) investigated 34 consecutive outpatients with restless legs syndrome from their Movement Disorders Clinic and identified 13 patients (38%) with small fibre neuropathy. The main scope of the study was to compare the sensory profile—according to a validated, comprehensive quantitative sensory test protocol—in primary restless legs syndrome versus secondary restless legs syndrome associated with small fibre neuropathy, and the conclusion was that thermal hypoaesthesia characterizes the latter. We have several concerns regarding the methodology of this study. Differential diagnosis of secondary restless legs syndrome associated with small fibre neuropathy was based on clinical symptoms, i.e. patients with small fibre neuropathy were identified as having persistent plus symptoms, mainly painful, whereas patients with primary restless legs syndrome had tingling and dragging paraesthesias; however, even these latter symptoms may be features of small fibre neuropathy (Lacomis, 2002; Hoitsma et al., 2004). The diagnosis of small fibre neuropathy was confirmed by skin biopsy, which was, however, not performed in any of the patients with putative primary restless legs syndrome. In addition, not all patients with small fibre neuropathy had skin biopsy findings consistent with definite abnormalities (55 intraepidermal nerve fibres per mm), but values were between 5 and 7 in some patients (‘possible’ small fibre neuropathy). Some features of the patient subgroups were perplexing. Mean age of restless legs syndrome onset was not significantly different between the two groups (41 years in primary restless legs syndrome and 46.7 in secondary restless legs syndrome), in contrast with the observation that primary and secondary restless legs syndrome segregate in different ranges of onset age (Allen and Earley, 2000; Whittom et al., 2007). On the other hand, the age at onset in patients with small fibre neuropathy is usually older, in the 6–7th decade (Devigili et al., 2008; Gemignani et al., 2010); thus, it may be that patients with restless legs syndrome–small fibre neuropathy in this series represented a peculiar subpopulation of small fibre neuropathy. A positive family history was present in 38% of patients with secondary restless legs syndrome, and this again is in contrast with the notion that genetic factors play a specific role in primary restless legs syndrome (Paulus et al., 2007; Whittom et al., 2007). The argument that Hattan et al. (2009) identified 37% of patients with restless legs syndrome and neuropathy reporting a positive family history is not pertinent as their series included many patients with genetic neuropathy. It is also puzzling that in the primary restless legs syndrome group there were patients with an elevated glycosylated haemoglobin (7/19, if we consider normal range 56.1%, as indicated in the article), implicating a possible diabetic aetiology, rather than primary restless legs syndrome. In summary, some patients classified with secondary restless legs syndrome displayed features usually associated with primary restless legs syndrome, whereas in the primary group there were patients who could be suspected to have small fibre neuropathy, and this could even suggest that in some patients, restless legs syndrome was generated by interacting central and peripheral factors (Polydefkis et al., 2000; Gemignani, 2010). In the comparison of the sensory profiles, there were similarities between the two groups that were almost as impressive as differences. Pinprick and pressure hyperalgaesia were of identical degree, whereas dynamic mechanic allodynia was absent in both groups; this latter finding is in contrast with the current view that doi:10.1093/brain/awq291 Brain 2011: 134; 1–2 | e167