Effects of Montelukast on Neuroinflammation in Parkinson's Disease: An Open Label Safety and Tolerability Trial with CSF Markers and [ ¹¹C]PBR28 PET

Abstract

Background

Dysregulated leukotriene signaling is proposed to be involved in pathogenesis of Parkinson's disease (PD).

Objective

The objective was to examine the safety and tolerability of montelukast, a cysteinyl‐leukotriene receptor1 and GPR17 antagonist, in patients with PD. Secondary outcomes were target engagement, effects on PD signs/symptoms, and central neuroinflammation.

Methods

Fifteen PD patients were recruited to a 12‐week open‐label trial of 20 mg bi‐daily montelukast treatment. Patients underwent ratings with the Movement Disorder Society Unified Parkinson Disease Rating Scale (MDS‐UPDRS), the Montreal Cognitive Assessment (MoCA), Beck's Depression Inventory (BDI), Parkinson's Disease Questionnaire‐39 (PDQ‐39), [¹¹C]PBR28‐PET, and lumbar punctures before and during montelukast treatment.

Results

All patients completed the study. Three patients reported loose stool. No serious adverse events related to treatment were reported. MDS‐UPDRS‐Total scores improved by 6.9 points. Very low levels of montelukast were detected in all cerebrospinal fluid (CSF) samples and resulted in a reduction in inflammation/metabolism markers. [¹¹C]PBR28 binding was lowered in high, but not mixed, affinity binders after montelukast.

Conclusions

Montelukast crosses the blood–brain barrier at very low levels and is well tolerated and safe in PD patients. Preliminary effects on neuroinflammation and clinical scores motivate a future randomized controlled trial (RCT) in PD. © 2025 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Increasing evidence suggests complex interactions between lipids and the immune system in the pathogenesis of Parkinson's disease (PD). ¹ Of particular interest is the involvement of leukotrienes, bioactive lipid mediators derived from arachidonic acid through the 5‐lipoxygenase pathway. ² , ³ Recent studies have identified dysregulation in leukotriene signaling and synthesis in alpha‐synucleinucleopathies including PD. ⁴

Leukotrienes, traditionally associated with asthma and allergic reactions, have also gained recognition for their function in neuroinflammation and neurodegeneration. ⁵ Cysteinyl leukotrienes (leukotrienes C4, D4, and E4) mainly bind cysteinyl leukotriene receptors (CysLTR) 1 and 2. CysLTR1 is a Gq/11‐family G‐protein‐coupled receptor (GPCR). ² CysLTR1 was elevated by MPTP in mice ⁶ and by rotenone in BV2 microglial‐like cells. ⁷

GPR17, which is activated by leukotrienes and purines, is enriched in oligodendrocyte precursor cells in the CNS, but also in other organs that are susceptible to ischemia. ⁸ Accordingly, GPR17 antagonism counteracts both ischemic damage after stroke and demyelination in multiple sclerosis. ⁹ , ¹⁰

Montelukast is a CysLTR1 and GPR17 antagonist that was introduced in the late 1990s as an adjuvant therapy for asthma. ¹¹ , ¹² In a retrospective follow‐up of asthma patients, high‐dose montelukast was associated with a lower incidence of PD. ¹³ Likewise, montelukast has shown to protect dopaminergic neurons and inhibit microglial activation in 6‐OHDA‐treated animals. ¹⁴ In a mouse model of dementia with Lewy bodies, montelukast improved memory and reduced α‐synuclein load. ¹⁵ Furthermore, montelukast boosts neurogenesis by blocking GPR17 in neuronal stem cells in vitro. ¹⁶

There is a concern of neuropsychiatric side effects of montelukast, especially in children and adolescents. The symptoms include sleep disturbances, depression, and suicidal ideation. ¹⁷ Because depression is common in PD, ¹⁸ any novel potential anti‐parkinsonian therapy that may worsen depression must be evaluated thoroughly at an early stage.

Here we present data from an open‐label study, investigating the safety and tolerability of montelukast in PD patients with particular emphasis on psychiatric side effects. We also report data on target engagement, effects on PD signs/symptoms, and central neuroinflammation.

Patients and Methods

This open‐label trial of 12‐week 20 mg BID montelukast therapy in early PD patients was conducted at Academic Specialist Center, Center for Neurology, Stockholm (EudraCT:2020‐000148‐76). The study was approved by the Swedish Ethical Review Authority (Dnr:2020‐06035) and the Swedish Medical Products Agency (Dnr:5.1–2020‐95,486). The study was conducted according to good clinical practice along with the Declaration of Helsinki. Patients were recruited between February 9, 2021, and March 16, 2022.

A higher dosage than used for asthma was chosen due to the expected lower bioavailability in the CNS. Neuroprotective effects of montelukast on an Alzheimer's disease (AD) mouse model were shown to be dose‐dependent ¹⁹ and justified a high dose of montelukast in an AD trial (NCT03402503).

Eligible study participants were PD patients between 35 and 80 years of age with Hoehn & Yahr ≤2. All participants were high or mixed‐affinity binders on translocator protein (TSPO) rs6971 polymorphism gene sequencing. ²⁰ A full list of inclusion and exclusion criteria can be found in Data S1.

All patients were on dopaminergic therapy, and levodopa equivalent daily dosage (LEDD) is included in Table 1. Patients attended the trial visits in ON medication state. The patients were asked to refrain from taking anti‐inflammatory medications during the study.

TABLE 1.

Baseline characteristics of study participants

Variable	Overall
No.	15
Sex, male	7 (46.7)
Age, years, mean (SD)	63.1 (9.4)
Time since diagnosis, years, mean (SD)	1.7 (1.5)
Levodopa equivalent daily dose, mean (SD)	240.9 (167.7)
MDS‐UPDRS 3, mean (SD)	17.3 (9.4)
MDS‐UPDRS Total, mean (SD)	33.1 (18.2)
Hoehn & Yahr, mean (SD)	1.5 (0.5)
BDI, mean (SD)	5.9 (4.9)

Note: Numbers are No. (%) unless otherwise noted.

Abbreviations: SD, standard deviation; MDS‐UPDRS, Movement Disorder Society‐Unified Parkinson's Disease Rating Scale; BDI, Beck's Depression Inventory.

Assessments were performed by a single assessor (P.S.). Measurement instruments included the Movement Disorder Society Unified Parkinson Disease Rating Scale (MDS‐UPDRS), the Montreal Cognitive Assessment (MoCA), Non‐Motor Symptoms Questionnaire (NMSQ), Parkinson's Disease Questionnaire‐39 (PDQ‐39), and Beck's Depression Inventory (BDI).

Plasma and cerebrospinal fluid (CSF) were collected at baseline and after 4 weeks of treatment to measure montelukast concentrations and for proximity extension assay (PEA) to assess global changes in inflammatory and metabolic biomarkers. Montelukast concentrations were measured at the Swedish Metabolomics Centre. Samples for PEA were prepared as previously described. ²¹ Cysteinyl leukotriene E4 (LTE4) in CSF was analyzed by Karolinska Institute Small Molecule Mass Spectrometry Core Facility (KI‐SMMS).

[¹¹C]PBR28‐PET was performed as previously described ²² at baseline and after 12 weeks of montelukast treatment to assess microglia activation. TSPO levels detected by [¹¹C]PBR28 were quantified using the two‐tissue compartment model providing volume of distribution (V _T), which is described in more detail in Data S1.

A linear mixed model was chosen for analysis of PET measurements. Paired t tests were chosen for PEA immunoassay and clinical rating scales. Pearson correlation coefficient was calculated for drug concentration in plasma and CSF. Two‐tailed P‐values below 0.05 were regarded as statistically significant. All data analyses were performed using R version 4.3.1.

A full description of TSPO genotyping, [¹¹C]PBR28‐PET, mass spectrometry (MS), and PEA can be found in Data S1.

Results

Fifteen early‐stage PD patients (7 men and 8 women) were included in the study (Table 1). Patient survey and collection of remaining study drug suggested that compliance with the study drug was very high. All patients completed the study. A consort diagram can be found in Figure S1.

The study drug was well tolerated with few reported adverse events. Three patients experienced loose stool but not severe enough to affect quality of life or motivate a reduction in the medication. One serious adverse event was reported. The patient had a pulmonary embolism that was treated successfully with anticoagulants. The patient had undergone an orthopedic procedure and was immobilized. No relation with the study drug was suspected, and the patient completed the trial. One patient experienced worsening of tremor during the study that persisted after study completion. No psychiatric adverse events were reported during the study. A full list of adverse events can be found in Table S1.

Twelve weeks after the treatment began, the mean MDS‐UPDRS‐Total score had improved with 6.9 points (95 CI [confidence interval]: −10.99 to −2.88, P = 0.0025) and MDS‐UPDRS‐Part 3 with 2.8 points (95 CI: −5.18 to −0.42, P = 0.024) (Fig. 1A,B). No significant changes were observed in NMSQ or PDQ‐39 (Fig. 1C,D). Statistically significant improvements were observed in MoCA with 0.73 points (95 CI: 0.09 to 1.38, P = 0.028) and BDI with −1.27 points (95 CI: −2.48 to −0.06, P = 0.042) (Fig. 1E,F).

FIG. 1.

(A–F) Changes in clinical rating scales between baseline and 12 weeks of montelukast. Height of bars represents group means. Error bars represent mean + 1 SD. (G) Volcano plot of proximity extension assay of cerebrospinal fluid (CSF) proteins. The estimated difference between baseline and 4 weeks is plotted on the x‐axis and the negative 10‐log P‐value on the y‐axis. The horizontal dotted line indicates a P value less than 0.05. (H) Representation of [¹¹C]PBR28 PET volume of distribution (V _T) in the three TSPO high‐affinity binders (subjects M103, M110, and M111) before and after 12 weeks of treatment. The blue end of spectrum represents lower binding and red higher binding. [Color figure can be viewed at wileyonlinelibrary.com]

Plasma and CSF were collected before and after 4 weeks of treatment from 14 and 13 patients, respectively. Average plasma concentration of montelukast after 4 weeks of treatment was 1621.26 ± 361.11 ng/mL with two outliers around 500 ng/mL. At the same time point, CSF concentration of montelukast was 3.66 ± 1.31 ng/mL. Montelukast was detected in the two outliers at an average concentration of 0.1 ng/mL, indicating low compliance or high pharmacokinetic metabolism. Montelukast was undetectable in baseline CSF and plasma samples. There was a significant correlation between plasma and CSF drug concentrations (r = 0.686, P = 0.010), including the two outliers. Plasma and CSF concentrations are visualized in Figure S2.

Of 184 analytes, 125 were detectable by PEA in CSF. Assays that had more than 30% of samples under the limit of detection were excluded for further analysis. Normalized Protein Expression (NPX) values of six proteins were significantly changed in CSF after treatment (Fig. 1G; Table S2). All six proteins were shifted to the left indicating a general lowering of inflammatory and metabolic pathway proteins after montelukast. There was no significant change in plasma detectable by PEA (Table S3).

LTE4 was detected in CSF using MS in 9 out of 13 provided samples. The mean concentration of LTE4 was 1.07 ± 0.45 pg/mL before treatment and 0.89 ± 0.43 pg/mL after treatment. This difference was not statistically significant.

[¹¹C]PBR28 PET was performed twice on 13 patients. Three patients were high‐affinity binders to TSPO, and 10 were mixed‐affinity binders. ²³ A linear mixed‐effects model was done with brain region V _T as a dependent variable, visit (baseline/12 weeks of treatment) and TSPO binding affinity as interaction terms considering subject variability. Model results show a significant decrease in binding in the high‐affinity group after 12 weeks in many brain regions (Fig. 1H; Table S4). V _T values showed increased binding in the high‐affinity group in all brain regions, which confirms differences in results due to TSPO polymorphism. No statistically significant change was seen in the group as a whole in any brain region after 12 weeks of treatment (Tables S4 and S5; Fig. S3). There was no significant correlation between change in MDS‐UPDRS‐Total score and change in PET V _T (Table S6).

Discussion

This study represents the first clinical trial studying effects of montelukast on PD patients. The primary objective, to ascertain the safety and tolerability of high‐dose montelukast, was met. No psychiatric adverse events were reported on a group level. However, the treatment period of 12 weeks is short and might not be enough to conclusively rule out psychiatric side effects of montelukast. The most common side effect was loose stool, but it was not severe enough to affect quality of life.

Clinically, montelukast was well tolerated. Only one patient reported a worsening of resting tremor during the study that also persisted after treatment stopped. On a group level, MDS‐UPDRS‐III was improved by 2.9 points, and MDS‐UPDRS‐Total was improved by 6.9 points on average, which is deemed clinically significant according to a suggested minimal clinically important difference (MCID) of 3.5 points. ²⁴ , ²⁵ Improvements in MoCA and BDI were not clinically meaningful. MoCA improvement is probably explained by repeating the test during a short time frame.

Montelukast was detected in CSF in all patients, even in the two outliers that had a much lower concentration of montelukast in plasma. This confirms passage over the blood–brain barrier albeit at very low levels. The main CNS targets are believed to be microglia through CysLTR1 antagonism and oligodendrocytes through GPR17 antagonism. Very low and similar amounts of LTE4 was detected in CSF before and during montelukast treatment.

Results of CSF proteomics indicate a lowering of proteins, including CCL25 and CASP8, after treatment with montelukast. CCL25 is a pro‐inflammatory chemokine that is elevated in PD patients and correlated with disease severity. ²⁶ CASP8 is involved in apoptosis and is linked to death of dopaminergic neurons and shown to be elevated in CSF of AD patients. ²⁷ , ²⁸ , ²⁹ No significant changes were seen in plasma proteins upon montelukast treatment.

A lower microglia activation could be seen in the high‐affinity group in the PET examination. There were only 3 patients in this group, and no significant change was seen in the group as a whole in any brain region. A follow‐up study including only high‐affinity binders and a placebo arm would be desirable to adequately assess effects of montelukast on [¹¹C]PBR28 binding. It is unclear if TSPO genotype has any impact on glial function, which would imply a different possible treatment effect in high‐affinity binders. In a study of AD patients, TSPO genotype was not associated with a change in glial activation, synaptic and axonal damage in early AD. ³⁰

There are limitations to this study. As with all open‐label studies, there is a risk of both subject and observer expectancy effects, which most likely affected the clinical assessment. The subjects were examined only in the ON medication state, which complicates interpretations of montelukast effects. Furthermore, participants had relatively low baseline BDI and a higher female‐to‐male ratio compared to the PD population in general.

In conclusion, high‐dose montelukast passes the blood–brain barrier at very low levels and appears safe in PD patients. CSF proteomics suggest a reduction in neuroinflammation, which, in combination with an improvement of clinical scores, motivates future randomized controlled trials with montelukast in PD.

Author Roles

(1) Research project: A. Conceptualization, B. Methodology, C. Investigation, D. Data curation, E. Formal analysis; (2) Statistical analysis: A. Design, B. Execution, C. Review and critique; (3) Manuscript preparation: A. Writing of the first draft, B. Review and critique; (4) Other: A. Funding acquisition, B. Supervision.

J.W.: 1A, 1B, 1C, 1D, 1E 2A, 2B, 3A

A.F.: 1B, 1C, 1D, 1E, 2C, 3B

P.S.: 1A, 1B, 1C, 2C, 3B, 4A, 4B

Full financial disclosures of all authors for the preceding 12 months

Johan Wallin: Research grant from Stockholm City Council. Anton Forsberg: None. Per Svenningsson: Honoraria for advisory boards and speaker engagements from AbbVie, Astra Zeneca, Lundbeck and Bial. Research grants from Stockholm City Council, Swedish Parkinson Fund, Swedish Brain Fund, Swedish Research Council, ASAP, Micheal J Fox Foundation, Knut and Alice Wallenberg Foundation.

Supporting information

Data S1. Supporting Information.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.