The methods used in the four postmortem studies were comparable in regard to tissue preparation, with the autopsy samples being fixed in formalin or formaldehyde solution (in 3.75% or 10% solution for more than 24 hours), embedded in paraffin, sectioned (at 5–6 µm), and stained.13143940 One major difference was the antibodies used: two studies selected antibodies against p-αSN, while the other two studies used antibodies against total αSN. Another significant methodological difference was the method of image analysis, with two studies evaluating αSN deposition using light microscopy1439 and the others using fluorescence microscopy.1340 The two postmortem studies that used light microscopy identified multiorgan involvement of αSN pathology, but no cutaneous αSN deposition in either the patient or control groups.1439 The two postmortem studies using fluorescence microscopy found αSN deposition in the dermal fibers of patients, each using a different antibody; the study of Ikemura et al.13 showed a sensitivity of p-αSN immunoreactivity of 70% in PD and PD with dementia and 40% in Lewy-body disease, and a specificity of 100%. This suggests that Lewy-body-related pathology also involves cutaneous nerves in Lewy-body disease, and that skin biopsies may be of value in PD or Lewy-body disease presenting with advanced autonomic failure. The more-recent study of Gibbons et al.40 assessed total αSN with immunostaining and confocal microscopy in autopsy samples obtained from abdominal skin and the scalp of 11 individuals with PD, and compared this to 5 controls with nonsynucleinopathy. That study found that the amount of αSN deposited in cutaneous autonomic nerves was significantly greater in subjects with a postmortem confirmation of a diagnosis of PD than in controls (sensitivity of 100%).40 Comparing the results across studies according to the antibodies used reveals that those using p-αSN antibodies showed higher specificity than those using antibodies against total or nonphosphorylated αSN by providing an ordinal yes/no result. In contrast, total αSN provided a deposition gradient that was correlated with the disease severity but was also present in control subjects. The other main reason underlying the different results appears to be the imaging method used: fluorescence microscopy may provide αSN data that are superior to those attainable using light microscopy because it can visualize specific features of small samples such as skin sections and also enhance three-dimensional characteristics on small scales by attaching fluorescent tags to antibodies attached to targeted structures and using a high-intensity light source.

Twelve studies have included biopsies in patients with a clinically confirmed diagnosis of synucleinopathy, and they had various designs.202122232425262728293031 Most of the studies focused on PD patients only. Nine case–control studies compared the differences among PD patients, controls, and/or patients with other synucleinopathies or neurodegenerative diseases assumed not to have αSN deposits.202122232425262728 The early study of Miki et al.,20 which was a case series of 20 PD patients, found abnormal αSN accumulation in paraffin-embedded chest skin samples in 2 (10%) of the patients, suggesting that αSN in a skin biopsy sample assessed with conventional immunohistochemistry is not a sensitive diagnostic marker. However, more-recent studies support the role of cutaneous αSN deposition as a potential biomarker for disease. The different results may be attributable to technical and methodological differences in the fixation method, the use of different antibodies (polyclonal versus monoclonal), and the analysis of cryosections with different thicknesses [thin sections (5–6 µm) versus thick sections (50–60 µm)]. These methodological differences are addressed in detail later in this review.

The sensitivity of αSN deposition in skin biopsy samples may vary depending on the underlying disease. Studies of PD patients have shown high specificities but a wide range of sensitivities: two studies found αSN deposits in only 5.2–10% of PD patients using skin samples of the chest wall and distal leg,2025 one study found αSN deposition in 52% of patients (16/31) using skin samples of the finger, proximal leg, distal leg, and back,21 and one study found that 100% of PD patients had αSN deposits in biopsy samples of the volar forearm.24 Looking at a possible length-dependent distribution, a few studies showed a proximal-to-distal gradient of cutaneous αSN deposition in PD patients: 100% of the patients had αSN depositions at the cervical site, 52–75% at the thigh, and 24–31% at the leg.2229 Comparing with healthy controls and patients with essential tremor, tauopathy, or parkinsonisms assumed not to have αSN deposits, the specificity of αSN was 90% when using antibodies against total αSN and 100% when using antibodies against p-αSN.212223242728 When looking at the class of nerve fibers affected, studies involving PD, DLB, and PAF patients found αSN accumulations in autonomic fibers of the skin,202122232425262728293031 while in MSA patients they were observed mainly in unmyelinated somatosensory fibers of the subepidermal plexus but not in dermal autonomic fibers.232425 These findings suggest that the involvement of postganglionic autonomic nerve fibers contributes to autonomic symptoms in PD, DLB, and PAF, which contrasts with cutaneous autonomic fibers appearing to be preserved in MSA. Two comparative studies found that 100% of PAF patients presented with αSN depositions in autonomic nerve fibers.2930 The detection of cutaneous αSN deposition could therefore be a sensitive tool for PAF. Moreover, one of these studies found p-αSN deposits with an even distribution pattern in PAF, whereas all of the PD and DLB patients had a proximal-to-distal pattern.3031 This suggests that there are variations in the pathomechanisms underlying synucleinopathies.222931

The various results obtained in the previous studies should be interpreted with caution, since they had quite heterogeneous designs and methodologies, as well as numerous limitations: the sample was small in many studies, most studies had a case–control design, and technical aspects of the study methodologies differed according to the biopsy site, antibodies used, the type of microscopy, tissue thickness, tissue preparation and fixation methods, and how positive results were defined. The recruited patients were diagnosed based mainly on clinical criteria, which limits the comparability of cases across studies from different referral centers. In addition, a study with a case–control design might be affected by selection bias.

Several studies have evaluated the correlations of cutaneous synuclein deposition with disease severity and autonomic dysfunction. One study yielded correlations between the αSN ratio and the score on the Hoehn and Yahr scale or measures of autonomic function tests, which suggested that greater αSN deposition was associated with greater autonomic dysfunction and more-advanced stages of PD.26 The same group recently described that αSN deposition is present also in early stages of the disease and when PD has been diagnosed recently (within 0.5–3 years), even among PD patients without autonomic failure, which indicates that cutaneous synuclein deposition may be present also during the premotor stages.27 Those authors also found that αSN ratios were higher in individuals with autonomic failure and were correlated with the severity of PD, suggesting that the αSN ratio can provide a sensitive and specific biomarker for PD patients.27 Other studies have investigated if αSN deposition is correlated with disease severity and autonomic dysfunction, quantifying p-αSN deposition as the percentage of autonomic structures or nerve bundles showing positive staining, or a more-sophisticated index of total p-αSN staining.222429 However, those studies did not demonstrate a correlation between deposits of p-αSN and clinical or nuclear imaging measures including the disease duration, motor involvement, autonomic dysfunction graded on the Composite Autonomic Scoring Scale, or the heart-to-mediastinum denervation ratio obtained from nuclear imaging.222429 Therefore, while a correlation of p-αSN with a clinical measure remains to be identified, the total αSN ratio may be appropriate for evaluating the disease severity or autonomic dysfunction in PD patients based on cutaneous synuclein depositions. Although questions remain, cutaneous synuclein deposition might reflect the disease stage or degree of autonomic failure in synucleinopathies. Future studies are needed to evaluate the associations of disease severity and autonomic dysfunction with cutaneous αSN deposition.

Autopsy samples are routinely fixed in formalin or formaldehyde solution for more than 24 hours and then embedded in paraffin, whereas in vivo samples are typically fixed in a 2% paraformaldehyde or Zamboni's solution (paraformaldehyde mixed with picric acid and NaOH) for 18–24 hours for the optimal detection of peripheral nerve fibers. Overfixation with formalin can impair the ability to quantify the peripheral nerve density and also to detect αSN.40

There are also differences in the methods used to section fixed skin biopsy tissue samples. Autopsy samples are generally cut into 5- to 6-µm-thick tissue sections from paraffin blocks, while in vivo samples are cut into 50- to 60-µm-thick sections. αSN is deposited within nerves in the dermal layer of the skin (i.e., in vasomotor nerve fibers of the blood vessels, sudomotor nerve fibers of the sweat glands, and pilomotor nerve fibers of the pilomotor muscles), and is therefore spread out across this structure with dimensions of approximately 2 mm. Since a section that is 5–6 µm thick has only one-tenth the sampling region of a section that is 50–60 µm thick, it is far less likely to contain dermal nerve fibers in which potential αSN depositions could be detected.40

The thickness of tissue sections also has a profound effect on the detection rate of studies involving in vivo samples. The overall sensitivity of immunostaining has been higher in studies using 50-µm-thick sections than in those involving sections that are 10–20 µm thick. Important data for clarifying this problem has come from the recent study of Wang et al.,41 who compared the detection of p-αSN in cutaneous autonomic nerve fibers across in vivo samples with section thicknesses of 10, 20, and 50 µm. This study, which has only been reported on abstract form, found that immunostaining with 50-µm-thick tissue sections was superior to using sections that were 20 or 10 µm thick for detecting of p-αSN in PD.41 This difference is probably due to the quantity of intact nerve fibers in each section.

As mentioned above, immunohistochemical studies measuring total αSN use polyclonal antibodies, whereas those measuring p-αSN use monoclonal antibodies. The selection of antibodies significantly affects the results obtained: 4 of 12 studies involving clinically diagnosed subjects used antibodies against total αSN, with the remaining 8 using antibodies against p-αSN (Table 2). Although there have been no direct comparisons, performing immunohistochemical staining with antibodies against total αSN may provide higher sensitivity compared to using anti-p-αSN antibodies. Previous studies observed p-αSN deposits intermittently within cutaneous nerve fibers, and detected them inside small dermal nerve bundles or in autonomic small fibers innervating blood vessels, usually in the deeper layers of the dermis.2122 These p-αSN deposits could be missed if deeper layers are not analyzed and if the p-αSN deposits are not colocalized with a panaxonal marker such as protein gene product (PGP) 9.5. In contrast, total αSN deposits can be identified more easily than p-αSN deposits. However, αSN deposits are also present in healthy controls, albeit in smaller amounts, and therefore require quantification to distinguish between healthy and diseased states. Thus, p-αSN is more specific to synucleinopathy and the choice of detection antibody against p-αSN or total αSN can have a large effect on study outcomes.

Nonspecific background noise and staining artifacts are inevitable in most immunohistochemical studies, and will depend on the choice of polyclonal or monoclonal antibodies, type of staining methods used, fixation technique, tissue thickness, and use of light or immunofluorescence microscopy. A recent unpublished study observed a low-intensity signal of staining artifacts and background noise in certain areas around autonomic structures in sections in which p-αSN-positive nerve fibers were not colocalized with PGP-9.5-positive fibers, irrespective of the tissue thickness.41 These artifacts could be misinterpreted as p-αSN-immunoreactive fibers. Therefore, applying a double-staining method to p-αSN with the panaxonal marker PGP 9.5 could be very helpful for decreasing the rates of false-positive and false-negative results.