Introduction

Growing evidence supports the involvement of immunity in Parkinson’s disease (PD), a neurodegenerative disorder characterised by loss of dopaminergic neurons in the substantia nigra. Recent studies suggest that both peripheral and central immune responses contribute to neuronal death in PD, potentially preceding clinical symptoms by several years1. In the past two decades, in vitro, ex vivo and in vivo studies have consistently shown the involvement of inflammatory responses mediated by microglia and astrocytes, which are linked to neurodegeneration and disease progression. Apart from that, more recent studies have demonstrated important changes in the peripheral immune profile within both the innate and adaptive compartments, particularly involving T lymphocytes and monocytes. However, changes in peripheral immune cell subsets are somewhat inconsistent across human studies, possibly due to differences in clinical phenotype and disease stage of patients with PD.

A better understanding of the mechanisms of immunity in PD could lead to the identification of targets for therapies to slow disease progression, as well as to the detection of new biomarkers to aid earlier diagnosis, allow immune-based stratification for clinical trials and monitoring of treatment response. Therefore, there is a need to improve the quality of observational studies devoted to elucidating the contribution of immunity in the development and progression of PD.

The clinical studies developed to date generally include small sample sizes and are widely heterogeneous in terms of inclusion and exclusion criteria, as well as in the clinical and genetic characteristics of patient cohorts. In addition, there is considerable heterogeneity in methods of sample collection and storage, which can have a substantial impact on biomarker analysis and results. The absence of clear guidelines for designing clinical studies specifically aimed at studying immunity may be delaying our understanding of its role in PD.

In this context, to enhance the quality and consistency of observational studies exploring the role of peripheral immunity in PD, a panel of IMMUPARKNET experts, comprising leading researchers in PD and neuroimmunology, was convened. IMMUPARKNET (The role of IMMUnity in tackling PARKinson’s disease through a Translational NETwork) is an Action supported by the European Cooperation in Science and Technology (COST), that supports a multidisciplinary research network to foster collaboration among scientists and clinicians studying immunity in PD2. The panel aimed to develop a guidance document for the clinical and biomedical research communities. This document provides recommendations for designing and conducting robust observational studies to investigate immunity’s role in PD, to ensure reproducible and clinically translatable findings. It discusses criteria for participant selection and clinical characteristics tailored to study objectives. It also includes guidance on neuroimaging biomarkers of neuroinflammation. Additionally, it provides fundamental principles to support clinicians and researchers in developing effective sample collection and processing strategies for reliable immune biomarker analysis.

Given the scarcity of high-level evidence in the rapidly evolving field of immune system involvement in PD, a systematic review was not conducted, and recommendations were not graded. Instead, the expert panel performed a narrative review of the literature, which facilitated the development of informed and consensus-driven recommendations. The panel’s collective expertise enabled a nuanced synthesis of fragmented evidence. Detailed narrative justifications, citing key studies and acknowledging evidence limitations, have been incorporated to ensure transparency and provide a robust and practical framework for researchers in the field.

Criteria for inclusion and exclusion of participants

The first consideration is which participants, both PD patients and control subjects, should be included. Participant selection should be guided by the specific objectives of each study, while considering the relevance of clinically meaningful comparator groups. In large epidemiological studies, it is important to be as inclusive as possible in terms of participant recruitment to enable the exploration of diverse immune-related mechanisms and their associations with the presence and progression of PD. However, in studies focusing on detailed characterisation of inflammatory profiles in PD, it is important to carefully consider criteria for participant selection to minimise confounding factors.

To address both perspectives, we propose a two-step approach: (1) to characterise system dysregulation in PD versus controls and (2) to investigate interactions and common pathways with immune-related comorbidities.

For Step 1, we propose that initially core studies should be conducted, consisting of patients with PD and controls without immune-related comorbidities, such as cancer, inflammatory or infectious diseases. Such studies will enable detailed immune profiling to identify PD-specific immunological signatures, free from confounding factors. We recommend the following pragmatic inclusion and exclusion criteria for defining core study cohorts, as outlined below and summarised in Table 1.

Table 1 Recommended pragmatic inclusion and exclusion criteria for recruitment to the core studies for detailed characterisation of immune and inflammatory profiles in Parkinson’s disease (PD) and/or prodromal PD compared to controls
Full size table

For Step 2, we propose extended studies including subjects with immune-related comorbidities. These will investigate whether these comorbidities share immune dysregulation pathways with PD or influence its pathogenesis and progression.

Inclusion criteria

There must be diagnostic confidence in the diagnosis of PD. Currently, the most widely accepted criteria are the Movement Disorder Society (MDS) Diagnostic Criteria for PD3, which requires the presence of Parkinsonism, together with supportive features balanced against ‘red flags’. However, these criteria are not well suited to recently diagnosed cases in whom supportive criteria and some ‘red flags’ may not have had time to evolve, and so for early/de novo cohorts, the revised MDS criteria for clinically established early PD may be more suitable4. There is an ongoing debate about the need for additional biomarker testing, such as the alpha synuclein seeding assay, dopamine transporter imaging, and genetic testing to provide biological confirmation of the diagnosis5,6. Whilst these newly proposed classifications of PD may strengthen confidence in the diagnosis, they still require further validation and may limit recruitment in centres where such testing is not readily available.

Prodromal PD cases may be interesting to include as they can help us to understand early disease mechanisms and to identify predictive biomarkers. In this line, accumulating evidence has shown an immune activation in patients with rapid eye movement (REM) sleep behaviour disorder (RBD)7,8 known to be at high risk of developing PD. A diagnosis of prodromal PD can be made by employing the MDS criteria9. Furthermore, including carriers of variants in PD-associated genes who are considered at risk of developing PD, and comparing their immune profiles between those who eventually develop the disease and those who do not, may provide valuable insights into both causative and potentially protective immune responses10,11,12.

In terms of the selection of control participants, matching to the age range and sex of the PD population is critical, given the well-established impact of ageing and sex on the immune system13,14. Furthermore, sex-dependent changes in the immune system have been reported in PD15,16.

Exclusion criteria

For studies investigating the relationship between the immune system and PD, it is important to consider the impact of other clinical factors which may confound any alteration in the immune profile. A wide range of comorbidities and medications may influence inflammatory status. As mentioned at the beginning of this section, a balance between designing a robust study which explores PD-relevant immune changes and obtaining a representative sample of patients should be considered. Consistent with the two-step approach, the following exclusion criteria are recommended for core studies in the first step.

Firstly, we would recommend excluding any participants with inflammatory or autoimmune disease17. For conditions with an immune component but which are characterised by intermittent flares (for example gout18), an alternative approach would be to allow recruitment of these participants, ensuring this comorbidity is inactive at the time of assessment. Participants with either acute or chronic infection should be excluded19. Both should ideally be assessed with a clinical review. However, a blood test for C-reactive protein can also be a useful screen for an acute condition, with a cut-off of >10 mg/L being most adopted as an exclusion criterion20. Exclusion of patients with active solid organ malignancy is warranted21. In the case of haematological malignancy, not only active but also previous malignancy may be considered, given the possibility of residual immune perturbation22,23. Recent major surgery or trauma should also be an exclusion criterion24,25, with the timeframe being a matter of clinical judgement, but our consensus opinion being at least 6 weeks. Diabetes mellitus is associated with changes in both innate and adaptive immunity26. Given that this is most pronounced in hyperglycaemic states, well-controlled diabetes need not necessarily be a reason for exclusion.

Participants currently taking immunosuppressant medication should be excluded. The majority of immunosuppressants have long-lasting effects27. Therefore, either a sufficient wash-out period after their discontinuation must be an exclusion criterion (for example 1 year in the case of broad-spectrum agents such as azathioprine or methotrexate), or it may be appropriate to exclude participants with any prior use, particularly where high potency agents such as cyclophosphamide, rituximab or alemtuzumab have been administered, which have long-lasting effects on the immune system. Steroids could potentially be treated differently, given their shorter immunosuppressive effect, with one study showing a change in immunoglobulin levels lasting 4 weeks following a short course28. Therefore, a reasonable criterion would be no oral steroid use in the past 3 months. There is some evidence of a systemic effect of inhaled, topical or nasal steroids29,30, and hence it may be appropriate to exclude current users of these therapies.

Anti-inflammatory medication is commonly used in this age group, and therefore, there are challenges in adopting this as an exclusion criterion for studies of immunity in PD, given the large number of patients who would be rendered ineligible. We suggest not excluding patients based on the use of low-dose aspirin (75 mg). However, we would suggest that participants should be excluded if they have used aspirin at a dose of over 75 mg or non-steroidal inflammatory agents within the previous 2 weeks31,32.

Vaccinations are known to lead to an altered immune profile and are frequently relevant in this age group. Considering that the primary response to a vaccination typically occurs over 2 weeks33, it would seem reasonable to wait at least four weeks post-vaccination until recruitment into the study.

For control participants, clinical screening to exclude neurological disease, in addition to the factors discussed above, is essential. This recommendation is intended to minimise false positives and false negatives among control participants. However, future investigations could also incorporate other neurodegenerative conditions (e.g. other synucleinopathies, tauopathies, and amyloid-beta pathologies) as controls in comparator groups to determine whether the observed immune profiles are unique to PD or shared across neurodegenerative conditions34,35,36. This strategy will enhance specificity and help to distinguish PD-related immune mechanisms from those in related pathologies.

Exclusion of people with a first-degree relative with PD as controls is also recommended, given the possibility of carriage of a PD risk gene, which may be associated with an altered immunophenotype11,12.

In the second step of our recommended approach, the focus shifts to exploring interactions and shared pathways with immune-related comorbidities. Accordingly, both PD patients and control participants with such conditions may be included under more flexible exclusion criteria. Comparing PD and control groups with inflammatory or autoimmune disorders may provide valuable insights into common and divergent immunological mechanisms between PD and other chronic inflammatory conditions37,38. This strategy ultimately enhances the specificity and interpretability of the findings.

Study design

Clinical studies in PD are usually hampered by a long duration and slow progressive course of the disease, the clinical heterogeneity and the variability of symptoms and signs over the day related to the time of medication and polytherapy. In this regard, the approach to a successful clinical study design is challenging and requires consideration of several aspects, including whether to adopt a cross-sectional or longitudinal timeline, selection of the most appropriate clinical measures, and the consideration of stratification into subgroups for analysis.

Cross-sectional versus longitudinal design

Cross-sectional studies are easier to perform, faster and less expensive than longitudinal studies. In a field that is largely underexplored, cross-sectional studies are conducted with the main aim of generating research hypotheses. For example, in a cross-sectional study, peripheral blood T cell phenotypes differed significantly between cognitively intact and impaired PD patients39. Such findings may suggest a relationship between the peripheral immune profile and cognitive evolution of PD, but a cause-and-effect relationship can only be demonstrated through a longitudinal study, in which both the immunophenotype and cognitive status are repeatedly recorded over time, starting from disease onset, when patients were still drug-naive. Such a study design would obviously require a long observation, due to the relatively low rate of PD patients converting to dementia in a short timeframe (about 10% per year)40, but this issue could be mitigated by including neuroimaging data along with fluid biomarkers associated with neuronal damage as surrogate markers. Longitudinal studies are also essential to address the question of when the immune response is most prominent in PD and how it changes with disease course, which is important to guide the timing of future immune-based therapeutic interventions. Studies which follow patients from the prodromal stage through their disease course are needed to further investigate this.

However, researchers often face several challenges when performing longitudinal studies41. Patient retention is one of the most significant issues, particularly in neurodegenerative diseases like PD, where the progression of symptoms can affect patient participation42,43. To address this, flexible study designs that offer adaptable scheduling for visits, as well as early and clear communication about the long-term nature of the study and its potential benefits, can help encourage continued participation. Additionally, providing incentives such as regular updates on the study’s progress or personalised health reports can encourage continued participation and a sense of involvement. By maintaining regular follow-ups and personalised communication, participants are more likely to feel like active contributors rather than passive subjects. Another key challenge in longitudinal studies is the cost associated with repeated immune monitoring, which requires multiple sample collections, clinical visits, sample storage and other logistical components over time. Lastly, ensuring the reliability of biomarker assays over time is crucial. Variability between timepoints can be enhanced by running batch analyses of samples collected across multiple visits, in order to minimise longitudinal drift in assay performance. To facilitate this, it is essential to implement standardised and validated protocols for sample collection, processing, and storage to minimise sample degradation. Relevant recommendations on this are included in the following section of this manuscript.

Clinical data and patient stratification

Given that conducting clinical studies can be limited by the clinical setting (lack of time and clinicians and/or resource facilities), two levels of clinical data collection are suggested. The first covers essential and minimal data that should be collected to allow basic clinical characterisation, and the second provides a more comprehensive assessment for specialised centres with more research resources. A schematic summary is displayed in Table 2.

Table 2 Two-level clinical data collection is recommended for clinical studies investigating the contribution of immunity to the development and progression of Parkinson’s disease (PD)
Full size table

Demographic data

The age and sex of the participants are essential to allow careful matching of PD and control cohorts and may need to be considered as stratification variables. Age is well known to influence the immune system13, and age of PD onset may also be important, as some studies have indicated that immune factors have a more prominent role in late-onset PD, compared to early-onset44. Subgrouping of participants by sex within the study design should be considered based on previous findings of sex-dependent immune changes in PD15,16,45. In addition, ethnic differences in immune response have been reported in several studies46,47, therefore, patient ethnicity should be considered when studying the role of immunity in PD.

Genetic status

Genetic factors must be considered in the complex aetiopathogenesis of PD and may warrant consideration in the selection of participants for studies of immunity. Several genes have been associated with PD, including variants in genes responsible for Mendelian forms of the disease (such as leucine-rich repeat kinase 2, LRRK2) and genetic risk factors (such as heterozygous variants in glucocerebrosidase, GBA). These genes encode proteins involved in specific cellular pathways related to immunity and inflammation48,49 and patients with variants in PD-associated genes have shown different alterations in the peripheral immune profile50,51. Interestingly, different inflammatory pathogenic mechanisms involved among the genetic and sporadic forms could also be responsible for inconsistent results on immunity in PD reported to date. In addition, patients with variants in PD-associated genes have distinct clinical features and disease progression, which at the same time have been linked to different inflammatory profiles49,50. Therefore, it is highly recommended to consider the genetic background, at least of the most frequent genes associated with PD, when designing studies investigating immunity in PD.

Comorbidities

Common age-related comorbidities such as metabolic syndrome, cardiovascular disease, hypertension, diabetes and dyslipidemia are recommended to be recorded. These conditions can indirectly affect immune system activation and may interact with immune dysregulation in neurodegenerative diseases, including PD52,53,54. Integrating these factors into study design and data analysis will enable researchers to determine whether immune dysregulation in PD is influenced by or independent of these prevalent comorbidities in the second-step approach.

Environmental and lifestyle factors

To gain a more comprehensive understanding of environmental contributors to the neuroinflammatory processes implicated in PD, it is recommended to capture detailed lifelong exposure histories along with relevant lifestyle information55. First, detailed records of geographical and occupational histories, as well as incidents of trauma, should be collected. Exposure to pesticides, insecticides or heavy metals warrants consideration during patient selection and stratification, as there is accumulating evidence from experimental and epidemiological studies suggesting their contribution to neuroinflammation in PD56. Additionally, vaccination records, including those not administered, as well as histories of past infections, should be documented57, and serological studies may be conducted to provide a more detailed history of exposure of common viruses.

On the other side, lifestyle factors such as diet58, exercise59 and sleep60 have been shown to modulate immune function and could significantly impact the immune profiles of PD patients. Stratifying participants based on their habits can help identify how lifestyle influences immune responses in PD.

Dopaminergic treatment

Concurrent dopaminergic treatment should also be considered. Emerging evidence has shown dopamine as a key transmitter between the nervous and the immune systems as well as a mediator produced and released by immune cells themselves61. Dopamine receptors are widely expressed on immune cells, which can mediate stimulatory or inhibitory responses, and exogenous dopamine has been shown to have a range of effects on different immune populations in vitro62,63.

Given this evidence, treatment status must be carefully considered in study design. Distinguishing—and where possible, comparing—treatment-naïve and treated PD patients is essential when studying immune profiles. While some studies take the approach of only recruiting drug-naïve patients, this approach may not always be feasible depending on local clinical practice and timing of patient referrals into research programmes. Alternatively, dopaminergic medication should be recorded at the time of sampling using a standardised method, such as calculation of a levodopa equivalent daily dose (LEDD)64, so that this can be controlled for within subsequent analyses.

Disease symptoms and signs

PD clinical heterogeneity may be mediated by different underlying pathogenic mechanisms, including immunity. For example, a peripheral proinflammatory profile has been associated with a more severe motor and cognitive phenotype and faster progression39,65,66,67,68,69,70. Thus, a more detailed stratification based on clinical phenotype should be considered.

  • Disease duration

    Disease duration/stage is an important consideration because there is evidence to suggest that immune activation may play a more prominent role in the early stages of the disease71,72. Failure to stratify based on disease duration is likely to introduce heterogeneity into immune profiling and lead to difficulty interpreting findings. It should be noted that disease duration is typically measured from the date that the diagnosis was given, but with the caveat that the underlying PD pathology is well established by this point. Duration from symptom onset is an alternative approach, although it can be affected by recollection bias and is problematic in studies recruiting later-stage patients.

  • Motor function

    A wide variety of rating instruments are available for assessing motor function, including Hoehn and Yahr (HY) staging and the MDS-Unified Parkinson’s disease Rating Scale (UPDRS), and specific scales to report motor complications, such as the freezing of gait questionnaire73. The need for a comprehensive motor evaluation needs to be balanced against time constraints in clinical settings. We consider that the HY scale is essential for motor severity stratification. However, it is widely acknowledged that each stage on the scale can encompass considerable heterogeneity74. Therefore, the MDS-UPDRS75 is recommended as a comprehensive tool for motor evaluation in research settings.

  • Non-motor symptoms

    Immune changes and neuroinflammation have been strongly implicated in cognitive impairment and dementia in PD39,69,70,76. For classifying patients according to their cognitive status, the current accepted diagnostic criteria for mild cognitive impairment (PD-MCI)77 and dementia (PD-D)78 should be followed. Cognitive scales that measure global cognitive performance can be useful tools for screening and can provide reasonable coverage of multiple cognitive domains. For this purpose, the scales most recommended by the Movement Disorders Society are the Mattis Dementia Rating Scale, the Montreal Cognitive Assessment (MoCA), and the Parkinson’s Disease-Cognitive Rating Scale (PD-CRS)79. A more comprehensive neuropsychological assessment, consisting of tests within each cognitive domain, would allow a more detailed exploration of the association between immune changes and cognition, although these tests are time-consuming, require specialised training, and may not be feasible in a busy clinical practice.

To date, little is known about the link between the immune system and other non-motor symptoms in PD. However, there is strong evidence that depression and anxiety, which are common features of PD, are linked to immune dysfunction80, and it is therefore important to consider depression status when evaluating immune changes in PD, which can be done using a short patient-completed questionnaire (see Table 2). Assessment of gut symptoms should also be considered, given the accumulating evidence that gut dysfunction and associated microbial changes cause a local inflammatory response, which may influence PD progression via a gut-immune-brain axis81,82. Gut symptoms can be quantified using the Gastrointestinal Dysfunction Scale for Parkinson’s disease (GIDS-PD), which also captures information on diet83. Olfactory dysfunction is a clinically relevant and early non-motor feature of PD, which has been associated with both immune mechanisms and α-synuclein pathology within the olfactory bulb84. Standardised and validated approaches for olfactory testing include the University of Pennsylvania Smell Identification Test (UPSIT)85 and the Sniffin’ Sticks test86. A scale encompassing a broader range of non-motor symptoms, such as the NMSS, should also be considered in research settings87.

Brain imaging of biomarkers of neuroinflammation

Neuroinflammation can be assessed in vivo using imaging techniques, which have the advantage of providing spatially informative and potentially quantitative measurements in the brain.

Positron emission tomography (PET) imaging

PET imaging is regarded as the gold standard to depict the neuroimmune endophenotype of many brain conditions. Several PET radiotracers of candidate biomarkers of neuroinflammation have been developed. Initially identified as the peripheral benzodiazepine receptor, the translocator protein (TSPO) remains the most extensively studied molecular target to assess neuroinflammation using PET in humans88,89,90. The central nervous system (CNS) normally expresses TSPO at low levels, making it notable when there is significant upregulation associated with a response to injury and disease.

To date, several dozen TSPO radioligands have undergone preclinical or clinical evaluation. The first PET radiotracer used in PD is the isoquinoline derivative (R)-[11C]PK11195. However, its intrinsic imaging limitations led to the development of 'second-generation' TSPO radioligands aimed at providing a better signal-to-noise ratio. An important limitation regarding the clinical application of these second-generation radiotracers is their universal sensitivity to the rs6971 single-nucleotide polymorphism of the TSPO gene on chromosome 22q13.2, so that study participants may have a high-affinity, mixed-affinity or low-affinity binder phenotype, depending on their particular TSPO gene polymorphism and TSPO binding features. Individuals with two copies of the rare allele (i.e. low-affinity binders) bind these radioligands with a lower affinity than people with two copies of the major allele (i.e. high-affinity binders), and people who are heterozygous for this allele (i.e. mixed-affinity binders) express both high-affinity and low-affinity binding sites in similar proportions. The allelic abundance in humans is ~5:2 in favour of the high-affinity binding allele (HAB), such that 10% of humans will show little or no TSPO-specific binding, and approximately 40% will show intermediate binding. Thus, individuals with similar TSPO expression but different genotypes will produce different PET signals using second-generation TSPO tracers, while the prototypic tracer [11 C]-(R)-PK11195 is indifferent to this allelic status. This obstacle could be partly addressed by performing TSPO genotyping prior to PET imaging to exclude low-affinity binders89.

In PD patients, cross-sectional studies using first- and second-generation PET radioligands have mainly shown increased TSPO binding, predominantly in the midbrain, compared with matched control participants91. Longitudinal assessment of TSPO binding is lacking in PD. Nevertheless, PET studies consistently showed that regional increases in TSPO binding can be detected early in the disease course76,92,93, including in de novo patients94, individuals in the premotor phase of the disease95 and some asymptomatic gene mutation carriers at risk for PD96. These results provide evidence suggesting that it is unlikely to be confounded by dopaminergic treatment, which can interfere with both the inflammation response97 and TSPO binding quantification98.

Interestingly, TSPO is just one marker of the inflammatory response, and new molecular targets to assess neuroinflammation using molecular PET imaging are currently being developed, targeting both microglia and astrocytes at different activity states99. None have been extensively investigated in PD.

The drawbacks of molecular PET imaging include the low availability of radiochemistry facilities, ionising radiation exposure, which limits its use in vulnerable populations and longitudinal studies, and low spatial resolution, making it unsuitable to image small structures satisfactorily.

Magnetic resonance imaging (MRI)

In the future, these drawbacks could be overcome by molecular MRI, but the latter is still in infancy. In the meantime, an emerging application of MRI in the field of neuroinflammation is to investigate biomarkers of brain—border immune niches, which are believed to contribute to the development of neurodegenerative disorders100. However, the exact pathophysiological mechanisms underlying many of these new biomarkers still need to be cross-validated in humans. Currently, the most well-established application of MRI is to allow anatomical co-localisation of molecular PET findings of neuroinflammation. A common approach consists of the acquisition of high-resolution T1-weighted MR images of the entire brain using a 3D magnetisation-prepared rapid gradient echo (MPRAGE) or an MP2RAGE sequence at ultra-high field strengths101.

Biological samples and immune biomarkers

Sample collection, handling and storage are crucial steps in the study of immune-related biomarkers, where a standardised approach is essential to ensure reliability. In this section, we will discuss basic principles that can be useful to clinicians and researchers to help determine appropriate sample collection and processing strategies to enable reliable immune biomarker analysis.

Immune biomarkers can include cell-based measures of phenotype and function, as well as measurement of inflammatory mediators (cytokines and chemokines). Alterations in the frequency of blood immune cell subsets, as well as functional changes, are well described in PD1. There is evidence for lower lymphocyte numbers in PD patients compared to controls, but there are conflicting results on specific T-cell and B-cell subpopulation dysregulations102. Changes in T cell and monocyte subsets have also been reported in the cerebrospinal fluid (CSF) in PD, although studies have been limited to date103,104. Notably, research on the role of neutrophils in PD is lacking, due to the standard approach of isolating peripheral blood mononuclear cells (PBMCs; comprising lymphoid and myeloid cells), which discards the polymorphonuclear cells (PMNCs, i.e. granulocytes). The observed discrepancies in immunophenotyping results across studies may be related not only to differences in the clinical phenotype of PD patients, but also to the methodology used in sample collection and analyses. Several key pre-analytical variables can affect cell quality, as discussed below in relation to specific sample types.

Cytokines and chemokines have been reported to be increased in serum, plasma and CSF in PD, and some studies have shown links with disease stage and progression68,102,105,106. However, a wide disparity exists between studies in terms of which markers are altered. As for cell-based studies, discrepancies may relate to differences in methodology employed in sample collection, processing and storage. There is an intricate relationship between the clock gene system, the circadian rhythm, and the pathology underlying PD107,108,109. Circadian control of innate immune cell movement and pathogen response has been shown, together with day/night differences in the adaptive immune cell response110. Therefore, it is very important to keep the timing for biosample collection constant within studies. Correlation between blood and CSF cytokine levels has been shown to be poor, likely reflecting local cytokine production within the CNS, and suggesting that compartment-specific analyses are needed111. Other biological samples of interest to study inflammatory mediators include faecal samples and saliva.

Below, we address key considerations for the collection and processing of individual sample types and optimal biomarker analyses. Table 3 provides a summary of samples and biomarkers recommended for immune profiling studies in PD. While biospecimen selection should ultimately be guided by the specific research objectives, a two-level collection framework is proposed, considering both scientific value and feasibility. For example, CSF and functional assays offer particularly critical insights, but they have practical feasibility constraints that need to be considered. Table 4 and Fig. 1 further outline recommendations for sample collection, processing and storage.

Fig. 1
figure 1

Standardized workflow recommendations for blood sample handling.

Full size image
Table 3 Two-level sample and biomarkers collection recommended for clinical studies investigating the contribution of immunity to the development and progression of Parkinson’s disease (PD), considering both scientific value and feasibility
Full size table
Table 4 Summary of recommendations for sample collection, processing, and storage
Full size table

PBMCs

The analytical endpoint use of PBMCs can feed into two assay categories, with implications for the collection and processing method: (1) assays not requiring viable cells, such as nucleic acid (NA)-based genetic analysis, for which immediate stabilisation after isolation is critical to preserve NAs. Cells should be stabilised in an appropriate NA stabiliser and then stored frozen until NA extraction112. (2) assays requiring intact/viable cells, such as immunophenotyping and functional assays. These assays are sensitive to the conditions under which PBMCs are collected, processed, and stored113.

  • PBMC isolation methodology

    The choice of blood collection tubes is important when collecting blood for PBMC extraction. PBMCs are typically isolated from peripheral blood samples collected in tubes containing anticoagulants such as lithium heparin. Although ethylenediaminetetraacetic acid (EDTA) tubes provide higher PBMC isolation efficiency compared to heparin114, citrate-based anticoagulants like acid citrate dextrose (ACD) and citrate phosphate dextrose (CPD) are beneficial in preserving cell functionality over longer storage periods115.

    Common methods for PBMC isolation include density gradient centrifugation with Ficoll-Paque, and isolation with ready-to-use cell preparation tubes, such as CPT or Leucosep/Lymphoprep, which contain a density gradient medium that enables PBMC isolation while preserving cell integrity and viability116. CPT or Leucosep have been shown to be simple and fit-for-purpose for the isolation of high-quality immune cell subpopulations117. They have the advantages of ease of handling, timeliness, safety and reduced variability. However, manual isolation with EDTA tubes renders a higher cell yield than other anticoagulants114. It is highly recommended to test and document the type of anticoagulant and isolation method used for each study and maintain consistency within the same study.

    Delayed processing can adversely affect PBMC viability and function, so minimising time between collection and processing is crucial for ensuring quality118,119. Samples should ideally be processed within two hours at room temperature, followed by immediate cryopreservation. If this is not feasible, as a consensus, blood can be left for up to 8 h (never >24 h) at room temperature on a shaker to prevent coagulation before processing. However, it should be noted that delayed processing or prior cold storage can selectively bias the survival of specific immune cell populations as well as compromise RNA integrity120,121,122.

  • Volume of blood

    The number of PBMCs required depends on the specific downstream application, i.e. flow cytometry (FACS), magnetic-activated cell sorting (MACS) or stimulation assays. Researchers should refer to specific assay protocols to determine the number of cells needed. Standard PBMC isolation using a Ficoll-Paque gradient technique typically yields 0.5–2 million PBMCs per millilitre, and hence the appropriate blood volume needed can range from 10 to 50 mL per assay117.

  • Fresh versus cryopreserved PBMCs

    The choice between fresh and frozen PBMCs will depend on several factors, such as cell type, cell functionality, the downstream assay and logistical and timing flexibility. In general terms, fresh PBMCs ensure maximum cell functionality and accurate immunophenotyping results123,124,125. Frozen storage is limited for genetic analysis, considering the instability of DNA after long-term storage of PBMCs126,127,128, but also for the design of immune cell profiling studies129,130.

    Despite these limitations, freezing PBMCs allows for more efficient bulk analyses, which is particularly important in longitudinal studies and multicentre projects. To minimise damage during freezing, PBMCs should be cryopreserved using a controlled-rate freezing process and stored in cryoprotective medium (10% DMSO in foetal calf serum) in liquid nitrogen131,132,133. Storage of PBMCs at −80 °C must be avoided since it increases their apoptosis rate and induces substantial genomic modifications134,135. Of note, to minimise inter-centre variability, a cell concentration greater than 6 × 106 PBMC/mL has been associated with improved viability136.

  • Quality control of PBMCs samples

    Finally, a systematic assessment of cell counting, cell viability and cell contamination (by dead cells, platelets and/or red blood cells (RBC)) is essential to improve sample purity, downstream data accuracy, as well as reliability and interpretability of immunological data. Cell counting can be done by using different standardised counting methods137,138. Systematic verification of cell viability can be carried out by using specific stains or fluorescent dyes to provide accurate discrimination of live and dead cells during flow cytometry analysis137. Platelet contamination can be minimised by washing and centrifugation, and identified in flow cytometry either by their smaller size during gating or by specific staining or markers139,140,141. Regarding RBC contamination, the main approaches include chemical lysis or the use of nucleic acid–binding dyes such as acridine orange and propidium iodide, as well as the application of specific markers for flow cytometry identification138,140.

PMNCs

There is a knowledge gap on the role of neutrophils on PD, mainly due to methodological issues1.

Percoll gradient is the most commonly adopted strategy for PMNC isolation142,143,144 and has been reported to achieve a slightly higher degree of purity compared to Ficoll145. Furthermore, isolation of PMNCs using a Percoll gradient or spontaneous sedimentation technique reduces PMNC priming, which is crucial when planning functional assays rather than immunophenotyping143,146. However, when the amount of blood is limited, PMNC and PBMC isolation can be done from the same sample using LeucosepTM tubes, with density gradient centrifugation using Ficoll, resulting in the formation of a layer of PBMCs and a layer of PMNCs with red blood cells. Isolation of PMNCs and PBMCs can then proceed in parallel147. Both cell types can be combined following isolation for the purposes of immunophenotyping.

Serum and plasma

Serum and plasma are separated from whole blood through centrifugation at 1000–2000G for 10–15 min. For plasma, blood is collected into a tube containing anticoagulants, whereas serum is obtained from blood collected in a plain tube, which is left to clot for at least 15 min prior to centrifugation. Factors such as the use of anticoagulants, storage tube contamination with endotoxins, and delays in blood processing (centrifugation) can have a major impact on measured cytokine concentrations in both blood components. In terms of the choice of sample for analysis, plasma has been suggested to be a more sensitive matrix for detecting changes in low-abundance cytokines compared to serum148. Serum has been found to have higher levels of some cytokines than plasma, which may be due to the release of cytokines by activated platelets during the coagulation process149. On the other hand, prior to plasma separation from whole blood, leucocytes can secrete cytokines in vitro, thus altering cytokine levels; thus, it is important to minimise sample processing delays. The choice of anticoagulant in the blood collection tube may also influence cytokine levels in plasma150,151,152. Plasma collection with EDTA has been shown to produce the most consistent results for cytokine analyses and more closely resembles data obtained in serum148.

Importantly, delays in blood processing can lead to changes in cytokine levels due to cellular release or degradation; therefore, it is advisable to process samples within 60 min of collection. Keeping samples on ice or at 4 °C during processing can help maintain stability. At room temperature, serum cytokines are more susceptible to delayed processing compared to plasma153. Since repeated freeze-thaw cycles can lead to cytokine degradation, it is recommended to aliquot samples into small volumes (e.g. 200 µL) to avoid multiple freeze-thaw cycles, and this point should be considered when planning sample collection. In addition, it is recommended to use low-binding tubes (polypropylene rather than polystyrene tubes) to minimise cytokine adsorption to the tube walls. Long-term storage should be at −80 °C to minimise protein degradation.

CSF

CSF should be stored on ice as soon as possible post collection. It should be ideally processed by centrifugation (for 500–1500×g for 10–15 min at 4 °C) within 60 min to separate cells from supernatant154,155. The cell pellet should be used immediately for immunophenotyping using flow cytometry, as due to the low number of cells, the cell yield post freezing is generally insufficient for meaningful analysis156. The supernatant is stored at −80 oC for subsequent protein measurement, extracellular vesicle extraction, or cell-free RNA analysis154,157. Single-cell analyses require specialised processing and freezing conditions to maintain cellular integrity and transcriptomic profiles, as detailed in specific manuscripts158,159.

Faecal samples

Cytokine analysis in stool can potentially provide valuable insights into gastrointestinal inflammation and immune responses in the gut160. Moreover, it has been shown that there are correlations between cytokine levels and microbiota subpopulations, shedding light on how microbial changes may contribute to inflammatory processes and disease severity in PD161. However, stool analysis presents unique challenges due to the complex and heterogeneous nature of the human faecal matter and methodologies for assessing cytokine alterations remain poorly standardised162.

Sample collection can be hampered by difficulties in collection (e.g. avoiding cross-contamination upon sample collection, or patient compliance with home sampling) as well as gut symptoms such as constipation in PD. For adequate sample collection, sterile non-reactive containers designed for the purpose should be used, ensuring that a representative sample of stool (typically 1–2 g) is collected, without being contaminated with urine or water. The faecal samples should be aliquoted to prevent degradation and unnecessary repeated freeze-thaw cycles. They should be cryopreserved (−80 °C) for long-term storage within a 2-h window after collection. As an alternative, they can be stored temporarily at 4 °C, but not longer than overnight162.

Saliva

There is increasing interest in using saliva to measure inflammatory biomarkers, as it is non-invasive, requires a lower biosafety classification compared to blood, and requires less specialised personnel to collect. Salivary inflammatory markers have scarcely been explored in PD163. The concentration of levels of inflammatory markers in saliva is relatively high, and some studies show concentrations higher than detected in blood164. There are some unique considerations when collecting saliva, such as oral health status (e.g. participants should be excluded if there is evidence of ulceration, dental abscess or recent dental work) and restrictions on behaviours prior to saliva collection (e.g. avoiding food and drink for 1 h prior to collection). Specific particularities of saliva collection, handling and processing are described elsewhere164.

Conclusions

Immune processes are strongly implicated in the pathogenesis of PD, though there is a lack of data from large-scale human studies on how immune markers map onto clinical subtypes and disease stages. Furthermore, there are considerable discrepancies in the results of clinical studies, and methodological heterogeneity in study design and sample collection may be limiting our progress and our ability to obtain reliable and robust data. This review aims to guide the scientific community in optimal design of observational studies to advance our understanding of the role of immunity in the development and progression of PD, with the ultimate goal of developing new immune-targeting disease-modifying therapies. This manuscript was developed by a panel of experts from the IMMUPARKNET consortium, integrating current evidence with consensus-based recommendations. As the field advances, future systematic reviews and formal evidence grading will be essential to further strengthen these guidelines as additional data becomes available.

Data Sharing

Data sharing is not applicable to this article, as no datasets were generated or analysed during the current study