Introduction

The clinical diagnosis of Parkinson’s disease (PD) currently relies on clinician assessments and the application of established diagnostic criteria. However, a definitive diagnosis of PD can only be confirmed through post-mortem autopsy1. As a result, there is a longstanding need to develop reliable biomarkers that can aid in confirming the diagnosis and monitoring disease progression. Moreover, the early or prodromal diagnosis of PD is critical for the development and effective implementation of disease-modifying therapies, underscoring the growing importance of biological biomarkers2,3.

The pathognomonic pathological hallmark of PD is the presence of Lewy bodies and neurites, whose primary component is the aggregation of alpha-synuclein (AS) proteins. AS accumulation has also been identified in peripheral tissues, including the gastrointestinal (GI) tract, skin, and submandibular glands4, highlighting its potential as an in vivo pathological biomarker. However, the diagnostic utility of pathological confirmation via biopsy is limited due to its low sensitivity.

The real-time quaking-induced conversion (RT-QuIC) assay is a highly sensitive technique for detecting pathological proteins by amplifying their seeding activity5. Therefore, the RT-QuIC assay could enhance the sensitivity of detecting pathological AS-seeding activity in peripheral tissues. To date, RT-QuIC studies have been conducted on various tissue types, including the skin6,7,8,9,10,11, olfactory mucosa12,13, submandibular gland14, and GI tract15,16,17. Based on these results, AS RT-QuIC assays of the skin have been proposed as one of the potential tools for the biological definition of PD2,3.

The GI tract is not only a site where peripheral AS accumulation is detected but also plays a substantial role in the pathophysiology of PD via intensive neural connections and communications, a relationship commonly referred to as the gut–brain axis18. Alterations in the enteric microbiome can induce toxic environmental changes, promote inflammatory conditions in the gut wall and subsequent AS accumulation18. These pathological changes have been observed to occur even before the manifestation of overt PD symptoms19. Furthermore, disease progression in neuropathology correlates with the frequency of AS accumulation in the GI tract20.

Given this background, the measurement of AS accumulation in the GI tract using RT-QuIC assays has potential as a biomarker of disease progression and pathophysiology. However, RT-QuIC studies of the GI tract have reported variable positivity rates for pathological AS-seeding activity, ranging from 10% to 95.7%15,16,17,21. Moreover, although a few studies have assessed the clinical characteristics of participants15,21, none have evaluated cognitive function.

Therefore, this study aimed to evaluate the diagnostic accuracy of the RT-QuIC assay using stomach biopsies from patients with early PD and to investigate its associations with various clinical characteristics.

Results

A total of 22 patients with PD and 17 healthy controls were recruited. The clinical characteristics of all the participants are summarized in Table 1. The mean duration from motor symptom onset to biopsy in patients with PD was 2.1 ± 1.1 years (mean ± SD). Motor symptoms were right-dominant in 8 patients (36.4%), left-dominant in 10 patients (45.5%), and bilaterally dominant in 3 patients (13.6%). Regarding PD subtypes, 14 patients (63.6%) were classified as akinetic-rigid, six (15.4%) as tremor-dominant, and one (4.5%) as mixed-type.

Table 1 Clinical characteristics of patients with PD and controls
Full size table

The mean MDS-UPDRS Part III motor score of patients with PD was 15.6 ± 7.8, and all patients were at HY stage 2 or below (stage 1: 7 [31.8%]; stage 2: 15 [68.2%]). The scores of assessments for non-motor symptoms were as follows: MoCA, 25.4 ± 3.8; FAB, 15.9 ± 2.1; GDS, 10.2 ± 6.1; NMSS, 15.5 ± 14.4; Brief Smell Identification Test, 5.6 ± 2.3; and RBDSQ, 3.0 ± 2.9. We identified RBD in 4 (18.2%) patients and orthostatic hypotension in 6 (27.3%) patients with PD. The LEDD in patients with PD was 352.3 ± 193.2 mg. Based on these assessments, two patients fulfilled the level I diagnostic criteria for PD with mild cognitive impairment (PD-MCI)22.

Pathological AS-seeding activity in stomach biopsies from patients with PD and controls

The pathological AS-seeding activity evaluated using the RT-QuIC assay is summarized in Figs. 1 and 2 and Supplementary Table 1. The AS-seeding activity was detected in 10 patients with PD (45.5%) but none in the healthy controls (0%; p = 0.001). One patient in the PD group showed positivity in only a single well and was classified as negative for pathological AS-seeding activity. Therefore, the RT-QuIC assay demonstrated sensitivity and specificity of 45.5% and 100%, respectively. In patients with PD + AS, the mean lag time was 36.0 ± 5.4 h, and the mean PAR value was 0.0287 ± 0.0042 (Fig. 2A, B).

Fig. 1: Pathological AS-seeding activity in stomach biopsies from patients with PD and controls.
figure 1

Pathological AS-seeding activity was positive in 10 (45.5%) patients with PD (A), but none in controls (B). The figure displays the mean fluorescence values of four wells per participant, which may appear lower than the actual values of wells with pathological AS-seeding activity in confirmed positive cases. Due to differences in the positivity thresholds used for each analysis plates ([Plate 1: 26,168], [Plate 2: 25,680], [Plate 3: 28,837]), the positivity thresholds are not marked in the figure. AS alpha-synuclein, RT-QuIC real-time quaking-induced conversion, PD Parkinson’s disease.

Full size image
Fig. 2: Characteristics of the kinematic parameters of the RT-QuIC assay in patients with PD.
figure 2

Both mean (A) lag time and (B) PAR of the four wells of the AS-seeding assay are significantly different between patients with PD and controls. Detailed representative values are presented in Supplementary Table 1. In the PD + AS group, MoCA score was positively correlated with both (C) Lagmed1 and (D) Lagmed2, which were significantly after adjustment of age at biopsy. ** p-value < 0.01. AS alpha-synuclein, RT-QuIC real-time quaking-induced conversion, PAR protein aggregation rate, PD Parkinson’s disease, MoCA Montreal Cognitive Assessment, Lagmed1, median of the lag time of each positive replicate, Lagmed2 median of the first two replicates reaching the threshold in each positive test.

Full size image

Clinical differences between PD + AS and PD–AS groups

The differences in clinical characteristics between the 10 patients in the PD + AS group and the 12 patients in the PD–AS group are summarized in Table 2. There were no significant differences between the two groups in terms of demographics, dominant side of motor symptoms, PD subtype, or motor and non-motor symptom assessment scores. Detailed comparisons of NMSS domain scores (Supplementary Table 2) showed that the NMSS Domain 6: GI tract score was slightly higher in the PD–AS group (0.6 ± 0.8) than that in the PD + AS group (0.1 ± 0.3), but this difference did not reach statistical significance (p = 0.072). Among the two patients with PD-MCI, one was classified to the PD + AS group and the other to the PD–AS group.

Table 2 Comparison of clinical characteristics between PD + AS and PD–AS groups
Full size table

Correlations between kinematic parameters and clinical characteristics in the PD + AS group

Correlation analyses within the PD + AS group (Table 3) revealed that the MoCA score were positively correlated with Lagmed1 (Spearman’s ρ = 0.742; p = 0.014) and Lagmed2 (Spearman’s ρ = 0.762; p = 0.010) (Table 3, Fig. 2C, D). After adjusting for age at biopsy, a partial Spearman correlation analysis showed a statistically significant positive correlation between the MoCA score and Lagmed1 (Partial Spearman’s ρ = 0.738; p = 0.023) and Lagmed2 (Partial Spearman’s ρ = 0.742; p = 0.022). The NMSS domain 6: Attention/memory was also correlated Lagmed1 (Spearman’s ρ = -0.647; p = 0.043), which was not significant after age adjustement (Partial Spearman’s ρ = -0.616; p = 0.078).

Table 3 Correlation matrix of kinematic parameters with clinical characteristics in the PD + AS group (n = 10)
Full size table

Discussion

In this study, pathological AS-seeding activity was observed in 45.5% (10/22) of stomach biopsies obtained from patients with PD. The positivity rates observed in previous RT-QuIC studies investigating the GI tract were as follows: Fenyi et al. found a positivity rate of 55.5% (10/18) using biopsy tissues from the antrum, sigmoid colon, or rectum17; Vascellari et al. reported a positivity rate of 95.7% (22/23) using duodenal biopsies15; Emmi et al. observed positivity rates of 40.1% (9/22) using duodenal biopsies and 59.1% (13/22) using stomach biopsies21; and Shin et al. documented a positivity rate of 10% (2/20) using stomach surgical specimens16.

These conflicting results could be attributed to multiple factors. First, the clinical characteristics of the patient cohort may have influenced the outcomes. The mean disease duration in patients with PD in the study by Vascellari et al. was 14 ± 5 years. In contrast, the disease durations in the study by Shin et al.16 and the present study were 2.7 ± 5.1 years and 2.1 ± 1.1 years, respectively, suggesting earlier stages of PD. Notably, in the study by Emmi et al., patients were stratified into early PD (median disease duration, 5.5 years) and advanced PD groups which included patients experiencing motor fluctuations and undergoing percutaneous levodopa intrajejunal infusion therapy (disease duration unspecified; duration since the diagnosis of PD, 11.5 years)21. In this study, the positivity rates for duodenal and stomach biopsies in early PD were 20% and 50%, respectively, whereas the rates increased to 58.3% and 66.7%, respectively, in patients with advanced PD. Moreover, pathological studies have demonstrated that the degree of AS accumulation in the GI tract correlates with the progression of AS pathology in the brain23,24 and that the extent of AS accumulation increases with longer disease duration in patients with PD20. These patient cohort characteristics may have contributed to the varying positivity rates among studies.

Another possible explanation is the rostrocaudal gradient, which is the distribution pattern of AS accumulation in the GI tract20,25,26. The rostrocaudal gradient refers to the higher frequency of AS accumulation in the proximal parts of the GI tract (including the esophagus and stomach) than in the distal parts (including the colon and rectum). This feature was the key reason for focusing on stomach biopsies in the present study. Fenyi et al. also identified this gradient by analyzing multiple regions of the GI tract17. In their study, AS amplification was detected in both patients whose tissues were sampled from the antrum, whereas the detection rates for the sigmoid colon and rectal tissues were 58% (7/12) and 25% (1/4), respectively. However, it is important to consider the limited sample size in each area and the lack of data on the disease duration in this study. Additionally, further research is needed to confirm whether the rostrocaudal gradient persists because the high sensitivity of the RT-QuIC assay may obscure this characteristic.

Finally, the tissue processing and preservation methods may have also contributed to these differences. In a study by Shin et al., formalin-fixed, paraffin-embedded GI tissues were used16. The processes involved in formalin fixation, cutting, and deparaffinization may have influenced the loss of AS seeds. This is supported by a pathological study in which AS immunofluorescence assays using paraffin-embedded skin tissue from patients with PD showed a lower positivity rate than those using frozen tissue27.

The diagnostic accuracy of RT-QuIC assays at other anatomical sites showed different results. Studies of RT-QuIC assays using the skin and cerebrospinal fluid (CSF) have demonstrated pooled sensitivities of 0.92 and 0.90, respectively28. Moreover, RT-QuIC assays using serum have recently shown high sensitivities ranging from 80.5% to 98.8%29,30,31. Therefore, if we consider only diagnostic accuracy, the RT-QuIC assay of the GI tract has less potential as a biomarker in patients with PD than assays using other anatomical sites.

This study employed a prospective design to clinically evaluate various motor and non-motor symptoms in patients with PD at the time of biopsy. These evaluations enabled the assessment of their association with pathological AS-seeding activity. Interestingly, in the PD + AS group, lag time parameters (Lagmed1 and Lagmed2) showed a robust correlation with the MoCA scores, even after adjusting for age. These kinematic parameters were suggested to reflect pathological AS-seeding activity more accurately than the simple arithmetic mean lag time32. Recently, Mastrangelo et al. conducted a longitudinal study involving repeated CSF sampling and demonstrated a progressive decline in Lagmed1 and Lagmed2 over time in asymptomatic LBD participants, defined as individuals exhibiting pathological AS-seeding activity in the absence of clinical symptoms33. Furthermore, a shorter baseline lag time (Lagmed1) was associated with an increased risk of dementia in both the whole cohort (hazard ratio [HR] = 0.91) and the PD subgroup (HR = 0.87). Similarly, a greater reduction in lag time over time (Lagmed1) was also associated with a higher risk of dementia in the whole cohort (HR = 0.76) and in the PD subgroup (HR = 0.69)33. Another CSF RT-QuIC study also showed that lag time was positively correlated with the MoCA scores in patients with PD and dementia with Lewy bodies34. Consistent with these CSF-based studies, the present results provide the first evidence of an association between pathological AS-seeding activity of stomach biopsy and cognitive function in patients with PD.

From the perspective of the gut–brain axis, the observed association between cognitive decline and the degree of pathological AS-seeding activity in the stomach may reflect a pattern consistent with the “body-first” progression model in PD35. However, other clinical features typically associated with the body-first model, such as motor asymmetry, PD subtypes, RBD, GI symptoms, orthostatic hypotension, and anosmia, did not differ between the PD + AS and PD–AS groups, nor were they correlated with seeding activity. Therefore, it is difficult to conclude whether the present findings fully support the body-first progression model. Given the small sample size, the results should be interpreted with caution. Nevertheless, the strong correlation between the MoCA scores and pathological AS-seeding activity suggests the potential of pathological AS accumulation in the stomach as a biomarker for cognitive function in patients with PD.

Previous studies on the correlation between pathological AS-seeding activity in the GI tract and clinical manifestations in patients with PD have reported varying results. Vascellari et al. reported no correlation between pathological AS-seeding activity and UPDRS Part III motor score, constipation score, or disease duration15. Conversely, Emmi et al. identified certain clinical associations with pathological AS-seeding activity21. Specifically, in cases stratified according to positive RT-QuIC results from duodenal biopsies, patients with positive results exhibited a longer disease duration. Additionally, when stratified by positive gastric biopsy results, the positive group demonstrated higher HY stages and elevated total scores on the Movement Disorder Society Non-Motor Rating Scale. However, direct comparison with the present study is limited because these studies did not include cognitive assessments.

The following limitations should be considered when interpreting the results of this study. First, the sample size was small. Therefore, the interpretation of the results, particularly regarding their association with clinical presentation, should be approached with caution. Second, other Parkinsonian syndromes, such as multiple system atrophy or progressive supranuclear palsy, were not assessed. These limitations restricted the evaluation of the diagnostic potential of the AS RT-QuIC assay as a biomarker. Additionally, brain imaging studies, such as dopamine transporter imaging, were not performed. Therefore, the associations between brain lesions and the pathological AS-seeding activity in the stomach could not be evaluated. Finally, biomarkers associated with Alzheimer’s disease, such as amyloid imaging or fluid biomarkers, were not assessed in this study. Therefore, it was not possible to determine whether the observed cognitive decline in patients was related to underlying Alzheimer’s pathology. Further studies specifically designed to investigate the association between cognitive function and pathological AS-seeding activity will be necessary.

In conclusion, the RT-QuIC assay using stomach biopsy tissues demonstrated a moderate positivity rate in patients with early PD. The lag time of AS-seeding activity was positively correlated with a cognitive function in patients with PD. These findings suggest that the AS seed amplification assay using the stomach tissue may serve as a biomarker reflecting disease pathophysiology of the gut–brain axis in PD. Further research using large-scale longitudinal cohorts is necessary to validate its potential as a pathological biomarker using the stomach biopsy in patients with PD.

Methods

Study design and participants

Patients with PD and healthy controls were prospectively recruited from the Seoul National University Hospital (SNUH) between November 2022 and August 2023. The inclusion criteria for participants were: (1) age ≥ 45 years and (2) a diagnosis of PD based on the Movement Disorder Society clinical diagnostic criteria for PD1. The exclusion criteria were: (1) familial PD with confirmed genetic abnormalities; (2) significant cognitive decline or impaired capacity to provide informed consent; (3) currently taking oral anticoagulants; (4) inability to undergo gastric tissue biopsy because of GI conditions; and (5) individuals deemed unsuitable for participation at the discretion of the investigator.

Healthy controls were recruited based on the following inclusion criteria: (1) age ≥ 45 years and (2) absence of neurological disorders. The exclusion criteria for the controls were: (1) current use of oral anticoagulants; (2) inability to undergo gastric tissue biopsy because of GI conditions; and (3) individuals deemed unsuitable for participation at the discretion of the investigator.

All recruited participants underwent gastric endoscopic biopsies. Comprehensive clinical assessments were conducted in patients with PD. The study protocol was approved by the Institutional Review Board of the SNUH (no. 2109-096-1255). Written informed consent was obtained from all participants at the time of their recruitment.

Clinical evaluation

Patients with PD enrolled in this study were assessed for the severity of Parkinsonian symptoms using the Movement Disorder Society-sponsored revision of the Unified Parkinson’s Disease Rating Scale (MDS-UPDRS)36 and the modified Hoehn and Yahr (HY) stage37. Based on the MDS-UPDRS scores, the dominant side of motor symptoms in patients (bilateral dominant, left dominant, or right dominant) was identified. Based on previous studies, PD subtypes were categorized as akinetic-rigid, tremor-dominant, or mixed38. The doses of medications taken by the patients were calculated using the levodopa equivalent daily dose (LEDD)39. Cognitive function was evaluated using the Montreal Cognitive Assessment (MoCA)40 and the Frontal Assessment Battery (FAB)41. The MoCA is a widely used cognitive screening tool that evaluates various cognitive domains, including visuospatial/executive function (5 points), naming (3 points), attention (6 points), language (3 points), abstraction (2 points), delayed recall (5 points), and orientation (6 points), for a total score of 30 points. It has been shown to be a reliable and appropriate tool for assessing cognitive function in patients with PD22. To assess non-motor symptoms, the Non-Motor Symptoms Scale (NMSS)42, Geriatric Depression Scale (GDS)43, and Rapid Eye Movement Sleep Behavior Disorder Screening Questionnaire (RBDSQ)44 were used. The presence of RBD in patients with PD was determined by an RBDSQ score of six or higher44. The Brief Smell Identification Test (Sensonics International, NJ, USA) was used to assess the olfactory dysfunction. Orthostatic hypotension was defined as a decrease in blood pressure of at least 20/10 mm Hg after 5 min of standing. Orthostatic hypotension was not measured in two (9.1%) patients.

Specimen collection

Following the clinical evaluation, the participants underwent endoscopic examination and biopsy at the Division of Gastroenterology, Department of Internal Medicine, SNUH. Four gastric biopsy specimens were obtained (two from the fundus and two from the antrum) from each participant. We selected the biopsy sites based on the regional distribution pattern of AS accumulation in previous studies that demonstrated a rostrocaudal gradient20,25,26. The collected tissue samples were placed in Eppendorf tubes containing phosphate-buffered saline and stored at −80 °C. The stored biopsy tissues were sent to the Korea Brain Research Institute and processed for the RT-QuIC assay, as described later. To analyze a more proximal region of the stomach, the RT-QuIC assay was performed using a combined sample of the two biopsy specimens from the fundus. The antrum samples were preserved for potential future studies. All tissues were anonymized at this stage to ensure a blinded evaluation during the RT-QuIC assay.

Stomach homogenate preparation

Frozen stomach tissues obtained from patients with clinically suspected PD or non-PD controls patients were processed as previously described5. Briefly, approximately 10–20 mg of stomach tissue was homogenized in 19 volumes of phosphate-buffered saline (pH 7.4) containing complete EDTA-free protease and phosphatase inhibitors (Roche Applied Science, Penzberg, Germany) using a Precellys 24 tissue homogenizer (Bertin Instrument, Montigny-le-Bretonneux, France). The resultant 5% (w/v) homogenates were kept in aliquots at −80°C following clarification at 2000 × g for 2 min.

RT-QuIC assay

The RT-QuIC assay for detecting pathological AS-seeding activity was performed as previously described45 with minor modifications. This method was developed to detect AS pathological amplification using post-mortem brain tissues from patients with dementia with Lewy bodies and PD, and has shown consistent and reliable results in previous studies5,16,46. On the day of the experiment, 5% stomach homogenates were sonicated5 and serially diluted to 0.1% in the homogenization buffer (10⁻³ dilution). Recombinant full-length human wild-type AS (rPeptide) was used as the substrate. The RT-QuIC reactions were conducted in quadruplicate using 96-well black plates with clear bottoms (Nalgene Nunc, NY, USA). Each well of the plate was preloaded with four glass beads (1.0–1.25 mm in diameter), followed by the addition of 100 μL of reaction mixture containing 2 μL of stomach homogenate at 10⁻³ dilution, 0.1 mg/mL of recombinant AS, 10 μM thioflavin T, 170 mM NaCl, 0.00125% sodium dodecyl sulfate (SDS), and 40 mM phosphate (pH 8.2). The plates were incubated at 37 °C in a FLUOstar Omega plate reader (BMG Labtech, Ortenberg, Germany) with alternating 1-min shakes (400 rpm, double orbital) and 1-min rest cycles for approximately 50 h. Thioflavin T fluorescence in relative fluorescence units was measured at 1-h intervals from the bottom of the wells (440 nm excitation and 480 nm emission, gain of 1,900; 20 flashes per well). Samples were considered positive for AS-seeding activity when at least two of the four replicates exceeded the fluorescence threshold. The threshold was defined as the mean fluorescence of the first five measurements across all samples plus seven standard deviations (SDs). Samples from patients and controls were randomly distributed across three plates, and RT-QuIC assays were conducted using a single batch of recombinant AS protein. The fluorescence threshold values were set differently for each plate (Plate 1—26,168; Plate 2—25,680; and Plate 3—28,837).

Statistical analysis

The clinical characteristics and RT-QuIC assay results were compared between patients with PD and healthy controls. To evaluate the kinematic features of pathological AS-seeding activity, the lag time (the time needed to reach the fluorescence threshold in each run) and protein aggregation rate (PAR, the inverse of the lag time) were calculated47,48. A subgroup analysis was performed to compare the clinical characteristics between patients with AS positivity (PD + AS) and those without (PD–AS). Student’s t-test was used for continuous variables, and Pearson’s chi-square test was used for binomial variables. When the assumptions for parametric tests were not satisfied, compatible nonparametric tests were used. We further calculated Lagmed1 (the median of the lag time of each positive replicate) and Lagmed2 (the median of the first two replicates reaching the threshold in each positive test). Recent studies have suggested that these kinematic parameters were associated with the Lewy body disease (LBD) stage and the development of dementia32,33. The correlations between kinematic parameters and clinical characteristics in patients with PD + AS were assessed using Spearman’s rank correlation coefficient. Partial Spearman’s rank correlation was used for age-adjustment for significant variables in the correlation analysis. All statistical analyses were performed using SPSS version 30.0.0.0 (IBM Corp., Armonk, NY, USA) and R version 4.4.3 (R Core Team, 2024). Statistical significance was set at p < 0.05.