Large-Fiber Neuropathy in Parkinson’s Disease: Clinical, Biological, and Electroneurographic Assessment of a Romanian Cohort

(1) Background: Increased attention has lately been given to polyneuropathy in Parkinson’s Disease (PD). Several papers postulated that large-fiber neuropathy (PNP) in PD is related to vitamin B12 deficiency and L-Dopa exposure. (2) Methods: Using a cross-sectional, observational study, we evaluated 73 PD patients without a previously known cause of PNP using clinical scores (UPDRS II and III and Toronto Clinical Scoring System), biological evaluation of vitamin B12 and folic acid, and nerve conduction studies to assess the prevalence and features of PNP. (3) Results: The prevalence of PNP was 49.3% in the study group. In the L-Dopa group, the frequency of PNP was 67.3% as compared to PNP in the non-L-Dopa group, where one subject had PNP (χ2 = 23.41, p < 0.01). PNP was predominantly sensory with mild to moderate axonal loss. Cyanocobalamin correlated with L-Dopa daily dose (r = −0.287, p < 0.05) and L-Dopa duration of administration (r = −0.316, p < 0.05). L-Dopa daily dose correlated with the amplitudes of sensory nerve action potentials of the superficial peroneal and radial nerves (r = −0.312, p < 0.05) (r = −0.336, p < 0.05), respectively. (4) Conclusions: PNP is more frequent in L-Dopa-treated patients than in L-Dopa-naïve patients. The results imply that longer exposure to high doses of L-Dopa may cause vitamin B12 and folate imbalance and PNP, secondarily.


Materials and Method
In this cross-sectional, observational study, we included patients that suffer from PD according to the UK Parkinson's Disease Society Brain Bank [45] and newly reviewed criteria by the Movement Disorders Society [46] with and without L-dopa treatment. The subjects were admitted to the Neurology I Department of the County Emergency Clinical Hospital in Cluj-Napoca, a general neurology department, from January 2017 to June 2019.
Patients with previously known hereditary, metabolic (including diabetes mellitus), toxic, inflammatory, or autoimmune causes for PNP were excluded. In addition, patients with entrapment neuropathies, moderate and severe radiculopathies, plexopathies, or having undergone spinal surgery or gastric resection were not included. Patients known to have Biermer anaemia undergoing treatment or those over the age of 80 were also excluded.

Clinical Features
All patients underwent a complete neurological examination using The Unified Parkinson's Disease Rating Scale (UPDRS ) part II and III for evaluation of motor symptoms and the Toronto Clinical Scoring System (TCSS) [47] for the clinical assessment of peripheral nerve involvement. TCSS is a validated tool for the clinical diagnosis of polyneuropathy in diabetes mellitus, with a cut-off value above 6 points. The TCSS also asseses the severity of polyneuropathy and has been used in other similar studies [5,7,12,13,28].
The progression of PD was evaluated using the Hoehn and Yahr scale. Patients' charts were reviewed for PD using the following criteria: disease duration, L-Dopa daily dose (LDD) over the last three months, and duration of administration of L-Dopa (LDDA). Other dopaminergic drugs were not considered and the L-Dopa equivalent dose was not calculated.

Biological Features
Prior to enrollment in the study, all potential subjects were screened for possible causes of polyneuropathy-a jeun glycemia, red and white blood cells, renal function, liver enzymes, electrolytes, protein electrophoresis, and thyroid function. For all the enrolled subjects, seric folic acid and cyanocobalamin levels were dosed, and a jeun seric glucose was repeated. Tests for homocysteine, vitamin B6, and methylmalonic acid were not available in our laboratory and, consequently, were not conducted. Subjects having two repeated fasting seric glucose levels above 125 mg/dL were not included. The hospital laboratory worked-up blood samples on a UniCel DxI 600 analyzer (Beckman Coulter Inc., Brea, California, CA, USA). The process is automatic chemiluminescence. Reference values, according to the laboratory, were 180-914 pg/mL for vitamin B12 and 5.9-23.2 ng/mL for folic acid, respectively.

Nerve Conduction Studies
Based on the fact that large-fiber polyneuropathy in PD is considered to be distal and symmetrical, predominantly axonal, and sensory-motor, affecting primarily, the sensory fibers [37] resembling the polyneuropathy in diabetes mellitus, all patients underwent electroneurographic assessment (ENoG) on a Keypoint Medtronic machine using surface electrodes. All patients had unilateral assessment of the sural, superficial peroneal, tibial, and common peroneal nerves. The following parameters were obtained for each sensory nerve: amplitude of sensory action potential (aSNAP) and sensory nerve conduction velocity (sNCV) of the sural and superficial peroneal nerve. For motor nerves, tibial, and common peroneal nerves, the distal motor latency (DL), the amplitude of compound muscle action potential (aCMAP), and the motor nerve velocity (mNCV) were obtained. The sural and tibial nerves were examined bilaterally to assess the symmetry of the parameters. For most of the subjects, the sensory response and sensory nerve velocity of the radial nerve and motor and sensory response of the median nerve (unilaterally) were also evaluated to assess the extent of a possible axonal loss, if impaired values were found in the lower limbs. The examination was made in optimal environmental conditions-a warm room with a skin temperature over 33 • C. All obtained values were adjusted for height and weight. All sensory nerve evaluations were carried out in antidromic conditions, with the stimulation point being situated at a distance of 12 to 14 cm proximal to the recording site, as stated in the handbook [48], and the average aSNAP and sNCV was recorded and taken into account.
Axonal loss is considered to be present when the aSNAP for sensory fibers or aCMAP for motor nerves is reduced, but with normal velocities (above 70% of the lower normal limit (LNL)) and mild or no prolongation of motor or sensory distal latency (less than 125% of the upper normal limit (UPL)) [48]. Demyelinating features were considered if reduced velocities or prolonged distal latencies were observed.
The amount of axonal loss was quantified as follows: • Mild sensory axonal loss: a reduction of the SNAP amplitude of more than 15% but no more than 50% of LNL for the sural and superficial peroneal nerves. LNL values for the sural and superficial peroneal nerves were considered to be 6 µV and 20 µV for the radial nerve, respectively ( Figure 1).

•
Moderate sensory axonal loss: a reduction of the SNAP amplitude of more than 50% of LNL for the sural and superficial peroneal nerves and a decrease of the SNAP amplitude of more than 15% but no more than 50% of LNL for the radial nerve ( Figure 2).

•
Severe sensory axonal loss: a reduction of the SNAP amplitude of more than 50% of LNL for sural, superficial peroneal nerve, and radial nerves ( Figure 3). 12 to 14 cm proximal to the recording site, as stated in the handbook [48], and the average aSNAP and sNCV was recorded and taken into account. Axonal loss is considered to be present when the aSNAP for sensory fibers or aCMAP for motor nerves is reduced, but with normal velocities (above 70% of the lower normal limit (LNL)) and mild or no prolongation of motor or sensory distal latency (less than 125% of the upper normal limit (UPL)) [48]. Demyelinating features were considered if reduced velocities or prolonged distal latencies were observed.
The amount of axonal loss was quantified as follows: • Mild sensory axonal loss: a reduction of the SNAP amplitude of more than 15% but no more than 50% of LNL for the sural and superficial peroneal nerves. LNL values for the sural and superficial peroneal nerves were considered to be 6 µV and 20 µV for the radial nerve, respectively ( Figure 1). • Moderate sensory axonal loss: a reduction of the SNAP amplitude of more than 50% of LNL for the sural and superficial peroneal nerves and a decrease of the SNAP amplitude of more than 15% but no more than 50% of LNL for the radial nerve ( Figure 2).  • Severe sensory axonal loss: a reduction of the SNAP amplitude of more than 50% of LNL for sural, superficial peroneal nerve, and radial nerves ( Figure 3). • Motor axonal loss: a reduction of the CMAP amplitude of more than 25% of LNL for the tibial nerve (LNL = 4 mV) or/and the peroneal nerve (LNL = 2 mV), or/and the median nerve (LNL = 4 mV). • Severe sensory axonal loss: a reduction of the SNAP amplitude of more than 50% of LNL for sural, superficial peroneal nerve, and radial nerves ( Figure 3). • Motor axonal loss: a reduction of the CMAP amplitude of more than 25% of LNL for the tibial nerve (LNL = 4 mV) or/and the peroneal nerve (LNL = 2 mV), or/and the median nerve (LNL = 4 mV). One impaired value of the SNAP or CMAP amplitude or one unexcitable sensory nerve were not criteria for axonal loss.
According to these ENoG changes, the following subtypes of neuropathy that can be found in the group, from an electrodiagnostic point of view, were established:

1.
Mild axonal sensory polyneuropathy-the presence of mild sensory axonal loss, absence of motor axonal loss, and demyelinating features; 2.
Moderate axonal sensory PNP-the presence of moderate sensory axonal loss, absence of motor axonal loss, and demyelinating features; 3.
Severe axonal sensory with motor features-the presence of severe sensory axonal loss and motor axonal loss with or without demyelinating features.
A PD patient was considered to have PNP if they met the criteria of likelihood, as stated by J.D England et al. in the special article for AAN on "Distal symmetric polyneuropathy: a definition for clinical research" [49].

Statistical Analyses
Descriptive statistics for normally distributed continuous variables (e.g., age) were presented as a mean ± standard deviation. The statistical significance of differences between two independent groups was tested using the independent samples t-test, while comparisons between three groups were performed using the Analysis of Variances (ANOVA) followed by Scheffe post-hoc tests to identify the pairs of groups with significant differences. Correlations between variables were assessed using the Pearson correlation coefficient.
Descriptive statistics for categorical variables (e.g., gender) were presented as counts and proportions, and for statistical comparisons between groups, the Chi-square test was used. For all inferential analysis, a two-sided p-value <0.05 was considered statistically significant.
Statistical analysis was performed using the SPSS 20.
Signed informed consent was obtained for each patient. This research was carried out under the Helsinki Declaration.

L-Dopa Group Versus Non-L-Dopa Group
In an attempt to analyze the differences between L-Dopa-treated patients and those without L-Dopa, the group was separated into two subgroups-the L-Dopa group, which consisted of 52 subjects receiving L-Dopa on a daily basis, and the 21 L-Dopa-naïve subjects (the non-L-Dopa group) on other dopaminergic therapies. Demographic, clinical, biological features and differences between the L-Dopa group and the Non-L-Dopa group are summarized in Table 1. Relevant nerve conduction studies results are represented in Table 2. The frequency of PNP in the L-Dopa group was 67.3% (35 subjects) as compared to PNP in the non-L-Dopa group, where one subject fulfilled the criteria for PNP (χ 2 = 23.41, p < 0.01). Nerve conduction studies revealed a predominant sensory axonal loss.

Differences Between the L-Dopa-PNP Group, the L-Dopa-Non-PNP Group Versus the Non-L-Dopa-Non-PNP Group
Further, due to the heterogeneity of the two groups regarding age, PD duration, motor impairment, and ENoG parameters, the following three subgroups were formed: More characteristics of the subgroups and differences are reported in Table 3.

Discussion
The prevalence of PNP in the studied PD group was 49.3%, and it resembled that described in some recently published studies [3,4,13,14] but was higher than in others [9,10]. The prevalence of PNP in the L-Dopa group was 67.3%, similar to two other studies that compared neuropathy in PD patients undergoing LCIG and oral administration [5,6].
The mean age of patients with associated PNP was 68 years old, similar to the age of the subjects with PNP in research conducted by Jugel et al. [5]. However, the subjects featured in the present work were younger than those included in other studies [8,[12][13][14][15]. Age was found to correlate to certain sensory ENoG parameters, as expected because the amplitude of sensory nerve action potentials decreases with age [48].
Limits of this study include the small sample size, lack of a control group, and the single-center basis of the study. Other restrictions are the lack of fasting homocysteine, methylmalonic acid, and pyridoxine that other studies evaluated and determined to be a positive correlation between hyperhomocysteinemia and PNP [3,[5][6][7][12][13][14]24,43]. Also, the occurrence of small-fiber neuropathy was not assessed as this was not our primary objective.
One recent study found that low B6 plasma values correlated to PNP and L-Dopa doses [6]. Recently, Cossu et Melis postulated that prolonged L-Dopa administration, due to its metabolic conversion to dopamine, modifies the peripheral nerve homeostasis by determining homocysteine accumulation and depletion of pyridoxine, folate, and vitamin B12, secondarily [25].
To our knowledge, this is the first study to assess axonal sensory loss at the level of the radial nerve in PD patients, also correlating the ENoG parameters to clinical and biological features. From what we know, this is the first study to assess neuropathy in Romanian PD patients.
Muller et al., in 2004, found a significant correlation between the SNAP amplitude of the sural nerve and increased levels of homocysteine in PD patients undergoing long-term L-Dopa therapy [35]. In the present study, significant correlations between the aSNAP of the sural nerve and the superficial peroneal nerve to vitamin B12 and folate were found. Furthermore, the aSNAP of the radial nerve described the same trend of association with cobalamin, but B12 and folic acid deficits are a known treatable cause of neuropathy [9]. This positive correlation suggests that higher plasma values of vitamin B12 and folic acid may even have a protective role.
The aSNAP of the superficial peroneal and radial nerves negatively correlated to daily intake and exposure to L-Dopa, suggesting that L-Dopa may affect axons, hence the axonal sensory loss. This possible relation was undermined by the correlation of vitamin B12 to the LDD and LDDA. These results imply that prolonged exposure to high doses of L-Dopa can be associated with vitamin B12 and folate imbalance and, secondarily, with PNP.
The severity of axonal sensory and motor polyneuropathy is usually assessed by ENoG at the level of lower limbs, but the pattern of this neuropathy is "length-dependent" [37]. Impaired values are found initially at the level of nerves in the lower limbs and, as axonal loss progresses, nerves in the arms are affected [48,49]. The Sural Radial SNAP amplitude ratio can be used to evaluate mild sensory axonal neuropathy in PD [37,51]. Although the radial nerve in almost all subjects was assessed, this ratio was not calculated. However, correlations that imply that the severity of PNP in PD is related to L-Dopa daily intake and prolonged exposure were found. This implication was also found by Toth et al. in 2008 and2010 [12,14] when they revealed a higher L-Dopa equivalent daily dose in the PNP group compared to the non-PNP group, that positively correlated to PNP severity, as assessed by the TCSS [12]. Age and PD duration contribute indirectly to the severity of PNP in PD. Nonetheless, a clear link between PD duration, age, and PNP in PD, relations that were albeit revealed in other published studies [6][7][8][9]15,52], could not be established in the present work. On the other hand, the risk for PNP was not stratified, in contrast to the study of Ceravolo et al. [15].
In a recently published trial, the authors did not report any cases in which PNP developed as an adverse effect of L-Dopa therapy, when the investigators followed up patients with early PD who took under 300/75 mg L-Dopa/Carbidopa a day for 80 weeks [53].
Bearing in mind that PNP in PD can affect motor functions and that L-Dopa therapy remains the standard therapy in PD, improving motor functions and quality of life, screening for vitamin B12 and folic acid deficiencies should be taken into consideration by any clinician when treating a PD patient. Furthermore, correcting those deficiencies is mandatory, and prevention may be taken into account even if it is still debatable, at least regarding vitamin B6 due to its interactions with decarboxylase inhibitors [25].

Conclusions
The reported results strongly suggest that large-fiber neuropathy in PD is common in L-Dopa-treated patients. The axonal injury is predominantly sensory and mainly mild to moderate. PNP in PD seems to be primarily related to L-Dopa treatment, correlating to the L-Dopa daily dose and longer exposure, and secondarily to cobalamin and folate deficiency that derives from L-Dopa metabolism.
Further long-term prospective studies on larger cohorts are needed for a possible task force to be able to recommend evidence-based strategies for the prevention or appropriate treatment of large-fiber neuropathy in PD.

Conflicts of Interest:
The authors declare no conflict of interest.