Sensory symptoms are not uncommon in Parkinson disease (PD).1 2 It is unclear whether these symptoms are of primary3 or of secondary origin attributable to motor symptoms such as rigidity and bradykinesia.4 Mutations in the PINK1 (PARK6) gene have been identified as one cause of recessively inherited parkinsonism.5 Asymptomatic carriers of PINK1 mutations represent individuals at risk who may develop PD later in life. Therefore, they may serve as a model to detect sensory abnormalities before subjects show typical clinical signs of parkinsonism.

The aim of this study was to elucidate whether sensory abnormalities are present and may precede motor symptoms in familial parkinsonism by characterising sensory function in symptomatic and asymptomatic PINK1 mutation carriers.

METHODS

Fourteen members of one German family with PINK1 mutations6 and 14 healthy age-matched controls without PINK1 mutations (mean age 46.4 years, range 31–65 years, nine women, five men) were included in this study. Subjects underwent quantitative sensory testing (QST) bilaterally of the hands and feet according to a shortened form of the QST protocol of the German Research Network on Neuropathic Pain,7 clinical examination and neurography of the sural and peroneal nerves. Family members who were aware of their parkinsonian signs were classified as “symptomatic,” and individuals unaware of PD signs as “asymptomatic” (table 1). No subject was on regular pain medication. The examinations were performed in accordance with the declaration of Helsinki. All participants gave written informed consent.

A three-way analysis of variance (ANOVA) and Kendalls-Tau-b test were used for statistical analysis, and p<0.05 was considered significant. Values are given as mean (SEM).

RESULTS

Three subjects with homozygous PINK1 mutation had definite PD and received antiparkinsonian treatment (table 1). All heterozygous PINK1 mutation carriers were subjectively asymptomatic, that is they had not noticed any signs of parkinsonian. However, six of them showed mild clinical signs of PD, such as mild bradykinesia, rigidity and signs of postural instability (table 1).6 Two family members described delayed detection of heat stimuli (table 1).

The occurrence of pain did not differ between mutation carriers and control subjects in our study group (table 1) as five control subjects also reported occasional and intermittent pain due to lower-back pain, headache, neck pain or arthrosis (mean intensity 4.8 on the NAS, range 3 to 7). No subject was symptomatic on the day of testing.

Neurography revealed no differences between PINK1 mutation carriers and controls for peroneal motor action potential (10.7 (1.0) vs 9.6 (0.7) mV, p NS) and nerve conduction velocity (NCV, 46.6 (0.9) vs 46.9 (0.8) m/s, p NS) as well as sural nerve action potential (8.1 (0.5) vs 11.2 (1.5) μV, p NS) and NCV (46.3 (0.8) vs 48.6 (2.2) m/s, p NS).

Since QST results did not differ between the right and left sides, data were pooled for further analyses. The results are shown in fig 1. The thresholds for mechanical detection (MDT), mechanical pain (MPT) and pressure pain (PPT) were higher in PINK1 mutation carriers compared with controls (fig 1). To exclude the possibility that increased thresholds would be largely due to the three PINK1 mutation carriers with PD, data for asymptomatic (n = 11) PINK1 mutation carriers were compared with controls. Interestingly, this comparison also revealed higher thresholds for mechanical detection, mechanical pain and pressure pain in asymptomatic PINK1 mutation carriers (fig 1). Among the PINK1 mutation carriers, only A-beta fibre mediated stimuli (MDT, vibration detection = VDT) increased with Unified Parkinson Disease Rating Scale (UPDRS) III scores (MDT hands R = 0.64, p<0.01, MDT feet R = 0.6, p<0.01, VDT hands R = 0.67, p<0.01, VDT feet R = 0.64,p<0.01).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Thresholds (SD) of hands (black symbols) and feet (open symbols) for cold detection (A), warm detection (B), cold pain (C), heat pain (D), mechanical detection (E), vibration detection (F), mechanical pain (G) and pressure pain (H) in all (n = 14) and asymptomatic (n = 11) PINK1 mutation carriers compared with controls (n = 14). *p<0.05, **p<0.01 for PINK1 mutation carriers versus controls; #p<0.05, ##p<0.01 for asymptomatic PINK1 mutation carriers versus controls. VDT was significantly different between asymptomatic PINK1 mutation and controls only on the feet. Note that statistical calculations were performed with z scores according to Rolke et al7 and not with raw data shown in this figure.

DISCUSSION

These data suggest that PINK1-associated parkinsonism is associated with a primary hypofunction of nociceptive and non-nociceptive afferent systems, as we found (1) increased thresholds of mechanical detection as well as mechanical pain and pressure pain, and (2) increased thresholds even in asymptomatic, clinically not or only mildly affected PINK1 mutation carriers. These findings propose that pronounced somatosensory abnormalities can already be found at the time when motor signs of PD are only subtle. This points towards a primary origin.

These data are in line with several studies that found an impairment of different sensory qualities in PD patients—for example of tactile stimulus location8 or two-point-discrimination.9 One study found significantly higher thresholds for touch pressure on the sole of the foot in PD10 and suggested that reduced sensitivity of the plantar side of the foot may contribute to impaired postural control in PD.

However, somatosensory profiles of PINK1 mutation carriers differ from those reported in studies on idiopathic PD patients who showed lower heat pain, but normal warm and mechanical detection thresholds compared with controls.11 Since the methods of investigation in both studies were similar to those that we used, the differing results might be due to (1) differences in subjects regarding the severity of their PD, as the mean UPDRS III score of our PINK1 mutation carriers was lower compared with the idiopathic PD study sample11; or (2) different aetiologies of PD that might lead to different somatosensory abnormalities—for example idiopathic versus familial PINK1-associated PD. In contrast to patients with idiopathic or non-classified PD, in which abnormally low heat pain thresholds suggest a predisposition for the development of pain,11 pain was not more frequent in PINK1 mutation carriers than in controls, and some of the PINK1 mutation carriers in our study even reported elevated heat pain thresholds. This might suggest PINK1-related PD to be a distinctly different clinical entity from idiopathic or non-classified PD.

It is unlikely that elevated thresholds are due to psychomotor slowing, that is a prolonged reaction time during the examination, because (1) higher thresholds were also found in asymptomatic heterozygous PINK1 mutation carriers who had only moderate or subtle signs of PD, and (2) only some, but not all, QST parameters might depend on reaction time. The strongly abnormal MDT in PINK1 mutation carriers and the MPT do not depend on reaction time.

Recently, a reduction of epidermal sensory nerve fibres and a loss of Meissner corpuscles have been shown in patients with idiopathic PD. As the latter correlated with disease severity not only central but also peripheral mechanisms of sensory abnormalities in PD have been suggested.12 However, for several reasons, we propose that the somatosensory impairment in PINK1 mutation carriers is related to a disturbed central sensory processing of mechanical, noxious mechanical and pressure pain stimuli. First, MDT was higher in PINK1 mutation carriers compared with controls, whereas neurography was normal and did not differ between the two groups. Second, MDT was significantly impaired, while VDT remained almost unchanged in PINK1 mutation carriers. Both QST parameters, MDT and VDT, examine the function of A-beta fibres. Accordingly, in case of impaired peripheral processing, one should expect an increase in both VDT and MDT thresholds. In contrast, in patients with lesions of the central nervous system, changes in thresholds in only one of these parameters have been observed.13 From a clinical point of view, this is reasonable, as humans do not perceive mechanical stimuli as vibratory, and vice versa. Furthermore, usually VDT is the parameter to be affected first in patients with a polyneuropathy (peripheral origin). Third, hypofunction was present in nociceptive as well as non-nociceptive mechanical afferent systems, whereas, fourth, noxious and innocuous temperature detection thresholds (warm, cold, heat pain, cold pain) were not affected. Fifth, mean values of some of the tested parameters showed higher thresholds on the hands than on the feet. In peripheral disorders, for example polyneuropathy, the longest nerves are often lesioned first, that is sensorimotor disturbances are usually more pronounced on the feet than in the hands. Furthermore, dysfunction of MDT and VDT increases with a higher severity of parkinsonian motor signs. This association between abnormal sensory thresholds and central motor signs may suggest a central cause of the somatosensory dysfunction. Accordingly, a PET study previously performed on three asymptomatic PINK1 mutation carriers showed a reduction in 18F-dopa uptake in nigrostriatal neurons compared with healthy controls.14 These subjects were also included in a study sample of asymptomatic PINK1 mutation carriers that showed higher mean temporal discrimination thresholds compared with controls, which was interpreted as a disturbance in central temporal processing of sensory inputs.15

Unfortunately, we were not able to enlarge the study sample size of PINK1 mutation carriers due to the very low prevalence of PINK1 mutations in the PD population. It has to be kept in mind that both the development and progression of PD in PINK1 mutation carriers are usually slow, and only a minority of heterozygous PINK1 mutation carriers will develop parkinsonian signs during their lifespan.16 However, we and others have recently shown in clinical,6 behavioural and imaging studies that heterozygous mutations may play a role in the pathophysiology of PD, as documented, for example, in asymptomatic heterozygous carriers of Parkin mutations,17 18 the most frequent cause of recessive early-onset PD. Accordingly, follow-up examinations of the asymptomatic PINK1 mutation carriers including skin biopsies are necessary to determine whether sensory abnormalities, which can be detected prior to the development of pronounced clinical motor symptoms of PD, might possibly help to identify PINK1 mutation carriers at risk of developing PD. Furthermore, it will be necessary to recruit a larger sample size of PINK1 mutation carriers in order to prove whether the abnormal sensory profile of our family can be generalised to other families of PINK1 mutation carriers and to elucidate the role of homo- versus heterozygote mutation carriers in different cohorts.