Participants in the individual case of off-label use study
Two patients with video-polysomnography (PSG) confirmed diagnosis of iRBD26 gave written informed consent to receive therapy with ADLL under “individual case of off-label use rules”— according to CARE guidelines and in compliance with the Declaration of Helsinki principles.
Our research complies with all relevant ethical regulations. Ethical approval for individual case of off-label use was given by the Ethics Committee, Faculty of Medicine, Philipps-University Marburg (Chairperson of the committee: Prof. Carola Seifart, MD). As this article is a case report, no statistical method was used to predetermine sample size. No data were excluded from the analyses. The experiments were not randomized. Apart from the unblinded treating physician (first author), all investigators were blinded to allocation during experiments and outcome assessment.
Patients were treated with oral ADLL (5 g per day)6,7,9,10 for a period of >18 months. Both patients were otherwise healthy. They were also participants in a parallel long-term observational natural history study called REMPET: Rapid-Eye-Movement (REM) Sleep Behavior Disorder and Fluorodeoxyglucose Positron Emission Tomography (FDG-PET). The study REMPET investigates the progression of the metabolic “Parkinson-Disease-related-Pattern (PDRP)” in FDG-PET of iRBD-patients over 10 years18. As part of the REMPET study, both patients received serial molecular imaging with DAT-SPECT and FDG-PET, annual clinical monitoring, and a cardiac 123I-metaiodobenzylguanidine scintigraphy (123I-MIBG) scan4,19,27 (patient 1 received the MIBG scan at baseline in 2014, 9 years before treatment with ADLL; patient 2 received the MIBG scan in 2022, under ADLL treatment) showing in both patients a reduction in cardiac sympathetic innervation. In addition, their olfactory function was tested with the Sniffin Sticks method (Threshold (T), Discrimination (D), Identification (I)) which provides a TDI-sum-score on olfactory function (see below)4,19,28,29.
Clinical evaluation
Both patients receiving ADLL therapy were evaluated with the Unified Parkinson Disease Rating Scale (UPDRS—original version; motor part: UPDRS III—range from 0 to 104, higher scores indicate greater impairment) or the Movement Disorder Society (MDS)-UPDRS (revised version; motor part: MDS-UPDRS III—range from 0 to 137, higher scores indicate greater impairment)30,31. Both scores include the item “action tremor”, which is not a cardinal motor sign of PD. We therefore calculated the UPDRS III and the MDS-UDPDRS III with and without the item “action tremor”. The clinical evaluation also included the screening test Montreal Cognitive Assessment (MoCA—range from 0 to 30, higher score indicating better cognitive performance; the threshold for mild cognitive impairment was defined below 26)32 and the SCOPA-AUT (Non-motor Autonomic Symptoms Questionnaire—range from 0 to 69, higher scores signify greater impairment of autonomic functions)33 at baseline and in about 6–12 month intervals for follow-up—all performed by blinded personnel.
Patient 1
In 2013, a female subject reported symptoms of RBD starting in 2006. PSG26 confirmed the diagnosis of iRBD in 2011. For the symptomatic treatment of iRBD, she had been prescribed a daily dose of 1 mg clonazepam at night since 2018. She underwent four pretreatment DAT-SPECT scans (123I-FP-CIT SPECT, marketed as DaTSCAN) in 02/2013, 03/2014, 04/2016, and 01/2019. In the REMPET study, two pretreatment FDG-PETs were performed in 2014 and 2018. In November 2021, she began taking 5 g/d ADLL, i.e., about three years after the last “baseline” DAT-SPECT in 2019 and after the second FDG-PET in 2018. After 3 months of ADLL treatment, she received the next DAT-SPECT scan in February 2022 and a further DAT-SPECT scan after 22 months of continuous ADLL therapy in September 2023. Likewise, after taking 13 months of continuous ADLL therapy, she underwent a third FDG-PET in December 2022.
Patient 2
A male subject reported RBD symptoms in 2018. PSG26 confirmed the diagnosis in June 2020. He was not taking any symptomatic therapy for iRBD nor any other concomitant medication. He received a baseline DAT-SPECT scan in August 2020 and an FDG-PET scan in June 2021, 17 and 7 months before starting the continuous ADLL therapy (5 g/d) in January 2022. After 12 months of continuous ADLL therapy, he underwent a second FDG-PET scan in January 2023 and after 18 months of continuous ADLL therapy a second DAT-SPECT scan in July 2023.
Treatment regime with ADLL
ADLL is commercially available under the trade name Tanganil® in France. It has been registered for the indication “vertigo” since 1960. The drug contains the racemate of acetyl-leucine, i.e., the inactive D-form and the bioactive enantiomer, the L-form of acetyl-leucine in equal parts. The daily dosage of ADLL in previous clinical studies for other indications was 5 g/day (2 g in the morning, 1.5 g at noon and 1.5 g at night)6,9,10. This dosage is similar to the equivalent dosage in animal studies (0.1 g per kg and day)7. For this case report, we administered the already investigated total oral daily dosage of 5 g/day. However, the distribution throughout the day was changed and the highest dose was taken in the evening with a recommended dosage of 1 g in the morning, 1–1.5 g at noon and 2.5–3 g in the late evening. If considered necessary, the dosage was slightly changed during the treatment period of 18–22 months. The highest ADLL dose taken was 5.5 g with 1 g in the morning and at noon and 3.5 gram in the late evening (patient 1, see Section C in Supplementary Information).
“Disease control” untreated iRBD patients—18F-Fluorodeoxyglucose-PET
For comparison, the FDG-PET scans generated in 2014, 2018, and 2022 from 12 other iRBD participants in the REMPET study, who did not receive ADLL, were utilized and analyzed in parallel to the three FDG-PET scans of patient 1 (Fig.3c, gray circles; see section FDG-PET below). All these “control” untreated iRBD patients had a marked hyposmia or anosmia at baseline in 2014 and in 9 of these 12 iRBD patients an MIBG scan was performed in 2014 and showed a reduced cardiac sympathetic innervation, an indicator that the respective iRBD patients presented a prodromal stage of PD4.
Phenoconversion of these untreated iRBD patients to PD, DLB or (rarely) multiple system atrophy (MSA) was searched for and either excluded or confirmed by a neurologist (AJ) according to the published diagnostic criteria34,35.
Inclusion criteria for all participating iRBD patients
The diagnosis made by video-assisted polysomnography (vPSG) was mandatory. Patients were informed that they would be allowed to take symptomatic therapy for RBD—such as clonazepam or melatonin in normal or slow-release preparations—if requested or necessary during the study at the discretion of the principal investigator.
Exclusion criteria for all participating iRBD patients
The presence of symptoms and signs for a manifest PD, DLB or MSA or any other neurological or psychiatric disorder were exclusionary. Cognitive impairment, as defined by a MoCA score below 26 at baseline, was exclusionary. The iRBD patients did not suffer from the following diseases: heart/kidney failure, myocardial infarction in the last five years, diabetes, amyloid or other neuropathy, pheochromocytoma. They were not taking any medications (reserpine, opioids, labetalol, phenylpropanolamine, phenylephrine), which might affect 123I-MIBG results.
Sniffin’ Sticks olfactory function test
Olfactory function was assessed at baseline and at most dates of the imaging procedures (Fig.2) with the full range Sniffin’ Sticks method consisting of a test for threshold of odor recognition (T), a test for discrimination between odors (D), and a test for identification of odors (I)4,19,28. Each subtest results in a subscore T, D, and I, respectively, and the sum of all three subscores is presented as the TDI sum-score (range 0–48, TDI sum-score of ≥32 was defined as normal, TDI sum-score between 16 and 31 was defined as hyposmia and anosmia was defined as a TDI sum-score ≤15 (see Section D in Supplementary Information for test details and Fig.2a, b as well as Supplementary Table1 for results). During the olfactory testing procedure, the patients were blindfolded. The olfactory function test was performed by a specially trained study nurse who was blinded for the purpose of the investigation4,28,29. The TDI sum-score is not agreed as a prodromal progression marker in RBD4 and has not been shown so far to be therapy-responsive. Therefore, the TDI sum-score is considered as an exploratory measure.
Clinicaloutcome Measure
As a clinical outcome parameter for measuring the severity of the subjective RBD phenotype, we used an RBD severity sum-score. After the start of ADLL treatment, the severity and frequency of RBD symptoms (dream enactment) were recorded daily in the morning by the patient and spouse in a standardized form according to a modified version of a published RBD diary36. The modified RBD diary consists of two separate parts: (1) severity - daily ranking scale ranging from 0 to 4 (0—no RBD symptoms; 1—speaking or mild jerks; 2—shouting or complex, non-aggressive movements; 3—aggressive movements with the risk of injuring her-/himself or the partner; 4—movements, which are so violent that the person falls out of bed); (2) frequency - weekly ranking scale ranging from 0 to 3 (0—no RBD symptoms; 1—RBD symptoms 1 to 2 nights per week; 2—RBD symptoms 3 to 5 nights per week; 3—RBD symptoms 6 to 7 days per week). Filled-out forms covering 7 daily ratings of the week were sent to the principal investigator every Sunday evening.
For this study, daily scores were analyzed for a period of 561 days. The daily scores of the first 18 weeks and the last 18 weeks are presented in Supplementary Fig.1 (see Section C in Supplementary Information). To present the data in Fig.1 of the main manuscript, we added up the daily RBD severity scores of 21 consecutive days to create a 3-week RBD-severity sum-score (RBD-SS-3—range from 0 to 84; the higher the score, the greater the severity).
Imaging outcome measures
Dopamine-transporter (DAT) ligand-binding imaging (DAT-SPECT)
As key outcome parameter we employed serial presynaptic dopamine-transporter (DAT) ligand-binding SPECT-imaging by ¹²³I-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl)-nortropane (123I-FP-CIT) (DAT-SPECT) to monitor the dopaminergic nigrostriatal synaptic density in the two patients who received ADLL therapy.
Patient 1 underwent pretreatment DAT-SPECT scans in 02/2013, 03/2014, 04/2016, and 01/2019. The DAT-SPECT scan in 01/2019 was selected as the pretreatment “baseline image” for further comparisons. The absolute time interval between the acquisition of the “baseline” DAT-SPECT scan in January 2019 and initiation of ADLL therapy in November 2021 was 34 months. Patient 1 received the first DAT-SPECT scan “under ADLL therapy” after 3 months of ADLL therapy in February 2022. The second DAT-SPECT scan under continuous ADLL therapy was performed in September 2023, that is 22 months after the initiation of ADLL therapy. In patient 1, ADLL was withdrawn for 10 days before the DAT-SPECT “under ADLL therapy” (plasma terminal half-life (t1/2) of the active enantiomer Acetyl-L-Leucine is 0.96 (±0.18) hours) in order to avoid any interference of the drug per se with the DAT-SPECT procedure.
Patient 2 received a “baseline” DAT-SPECT scan in August 2020. The absolute time interval between acquisition of the pretreatment “baseline” DAT-SPECT scan in 2020 and initiation of ADLL therapy in January 2022 was 17 months. He received his DAT-SPECT scan “under ADLL therapy” in July 2023, i.e., 18 months after the start of the continuous ADLL therapy. Due to miscommunication, patient 2 did not withdraw ADLL for the DAT-SPECT scan in July 2023.
The DAT-SPECT procedure including the acquisition and reconstructions of the DAT-SPECT scans has been published37,38,39 (see also Section D in Supplementary Information). Apart from the last DAT-SPECT of patient 1 performed in 09/2023, all DAT-SPECT investigations were performed in the Department of Nuclear Medicine, University of Marburg. Thus, the DAT-SPECT scans of patient 1 in 2013, 2014, 2016, 2019, and 2022 and the DAT SPECT scans of patient 2 in 2020 and 2023 were carried out on the same SPECT camera. After receiving the results of the second DAT-SPECT scan of patient 2 (showing an increase in the specific striatal DAT binding ratios in 07/2023 (1.72, right putamen) in comparison to the DAT-SPECT in 08/2020 (1.42, right putamen) (Fig.3b and Table1), we decided to perform the next and last DAT-SPECT scan in patient 1 in a different Department of Nuclear Medicine under blinded conditions. Patient 1 therefore agreed to travel to the Department of Radiology and Nuclear Medicine, University of Amsterdam, Amsterdam UMC (Prof. J. Booij). No one on the team at the Department of Nuclear Medicine in Amsterdam had any information about the purpose of the DAT-SPECT appointment. In fact, the patient was thought to be undergoing a routine DAT-SPECT investigation as part of her diagnostic work-up37.
After the last DAT-SPECT scan of patient 1 had been performed in Amsterdam, the group of Prof. J. Booij received all reconstructed data files of all previous DAT-SPECT investigations—generated in Marburg – without any further information about the purpose of the study.
All DAT-SPECT scans were subsequently analyzed by blinded personnel at the Department of Radiology and Nuclear Medicine, University of Amsterdam, Amsterdam UMC (Prof. J. Booij), The Netherlands, using the Brain Registration & Analysis Software Suite (BRASS; HERMES Medical, Sweden). Specific striatal binding ratios were determined and corrected for age as described earlier38,39. Specific to non-specific binding ratios were calculated in the striatum, the caudate nucleus and the putamen for each site, using the occipital cortex as a reference to assess non-specific binding. The numeric values of the striatal binding ratios are listed in Table1. Transverse images for Fig.3 were reconstructed as previously reported18.
18F-Fluorodeoxyglucose PET imaging (FDG-PET)
We employed serial 18F-Fluorodeoxyglucose PET imaging (FDG-PET) to identify the pathologic metabolic “Parkinson-Disease-related -Pattern (PDRP)” over time. Expression-z-scores for the PDRP were quantified in all scans (higher z-score value indicating stronger expression of PDRP). In order to minimize variation in the FDG-PET procedure, patient 1 (but not patient 2 (see below)), and all other “disease control” untreated iRBD patients underwent FDG-PET imaging at the Department of Nuclear Medicine and Molecular Imaging, University Medical Center Groningen (UMCG), The Netherlands.
FDG-PET analysis of 12 untreated iRBD patients for reference
FDG-PET scanning was performed for 12 untreated iRBD patients at baseline, first follow-up (approximately 4 years after baseline) and second follow-up (approximately 8 years after baseline). During the 8-year course of the study, 4 iRBD patients converted to manifest Parkinson’s disease (PD), 1 to manifest Dementia with Lewy bodies (DLB) and 7 remained in the prodromal stage iRBD and were thus defined as “non-converters”.
All baseline and follow-up scans were performed on a Siemens Biograph mCT64 or mCT40 PET/CT camera (Siemens, Munich, Germany) at the Department of Nuclear Medicine and Molecular Imaging, University Medical Center of Groningen (UMCG), The Netherlands, using a static imaging protocol. Images were reconstructed with OSEM3D (3 iterations, 21 subsets), time-of-flight, point-spread-function, and smoothed with a Gaussian 8-mm full-width-at-half-maximum spatial filter according to the EANM guidelines18,40. The matrix size was 256 (corresponding to a voxel size of 2 mm × 3.18 mm × 3.18 mm).
Central nervous system depressants and any RBD-related medications (i.e., melatonin or clonazepam) were discontinued in all subjects for at least 24 hours before scanning. Likewise, in the 5 iRBD patients, who had converted to manifest PD or DLB during the 8-year period of the REMPET study, levodopa or dopamine agonists were discontinued at least 24 h before scanning.
All scans were spatially normalized to an FDG-PET template in Montreal Neurological Institute brain space18,41 using SPM12 software (Welcome Centre for Human Neuroimaging, London, UK) implemented in MATLAB (version R2019a; MathWorks, Natick, MA, USA)40.
The Parkinson’s disease-related pattern (PDRP) that was used in this study was previously identified in FDG-PET scans of a cohort of 19 PD patients (13 M/6 F, age 63.9 ± 7.8), in the off-levodopa state, in comparison to a cohort of 17 healthy controls (12 M/5 F, age 61.5 ± 7.5 years)41. PDRP subject scores were calculated for each scan18. For this study, PDRP subject scores were z-scored to a cohort of 12 age- and gender-matched healthy controls (10 male/2 female, age 65.96 ± 6.21 years). These controls only underwent baseline FDG-PET imaging. By definition, healthy controls had an average z-score of 0, with a standard deviation of 1.
Average PDRP-z-scores (mean ± SD) of the iRBD patients for baseline, follow-up 1 and follow-up 2 visits were plotted for converters (n = 5) and non-converters (n = 7) separately (Fig.3c).
FDG-PET in patient 1
Patient 1 was part of the REMPET3 study. FDG-PET scans were performed as described above in Groningen, with an identical protocol.
FDG-PET in patient 2
Patient 2 also participated in the REMPET3 study. FDG-PET imaging was, however, performed at the Department of Nuclear Medicine, University of Marburg on a Siemens Biograph 6—and not at the UMCG in Groningen—using a static imaging protocol. Images were reconstructed with OSEM3D (3 iterations, 21 subsets), point-spread-function, and smoothed with a Gaussian 4-mm full-width-at-half-maximum spatial filter. The matrix size was 336 and Zoom in position 2. Reconstructed FDG-PET data were sent to the UMCG, Groningen, The Netherlands for further analysis (see above)37,40.
It is well known that variations in scanners and image reconstruction algorithms can impact PDRP expression scores. One way to resolve this is to apply a z-transformation to healthy control data from the same camera with an identical reconstruction protocol18,37.
However, FDG-PET scans of healthy controls were not available from the Marburg site. Therefore, the FDG-PET scans of patient 2 were preprocessed in a similar way as the UMCG data, and the PDRP was calculated in the same manner, using the same UMCG healthy control cohort (n = 12) for z-scoring. As a consequence, the z-scores for patient 2 cannot be compared to the scores of the reference iRBD cohort (n = 12) or of patient 1. Nevertheless, because any noise from a difference in acquisition protocols is systematically present in both scans, the difference in PDRP-z-scores between the two scans of patient 2 can still be appreciated. After initiation of ADLL therapy in 01/2022, the PDRP-z-score decreased from 1.02 (06/2021) to 0.30 (01/2023).
In both patients, ADLL was withdrawn for 10 days before the FDG-PET “under ADLL therapy” in order to avoid any interference of the drug per se with the FDG-PET procedure.
[123I]-Metaiodobenzylguanidine cardiac scintigraphy (MIBG)
Patient 1 receiving ADLL therapy and 9 out of 12 disease control iRBD participants underwent cardiac [123I]MIBG scintigraphy at baseline performed according to the standard operating procedures (see Section D in Supplementary Information) of the Department of Nuclear Medicine, Marburg, Germany. Patient 2 received the MIBG investigation under ADLL therapy in May 2022. Regions of interest (ROI) were manually placed on planar anterior images. A rectangular ROI was used for the mediastinum and a circular ROI for the left ventricle of the heart. According to the in-house code, a heart-to-mediastinum ratio of [123I]MIBG-binding of <1.5 was considered pathological4.
Statistical reporting section
This article reports two cases. Therefore, data of the RBD severity scores RBD-SS and striatal binding ratios for the serial DAT-SPECT images are presented as individual values. No further statistical analysis was performed. z-scores of the Parkinson-Disease-related-Pattern (PDRP) of the FDG-PET images were calculated according to published algorithms18,37. Mean ± standard deviation were calculated for the z-scores of the converters and the non-converters, respectively, in the untreated “disease control” group of iRBD patients. Data analysis was performed using GraphPad Prism (version 8.3.1 GraphPad Software, USA). All figures were created with Adobe Illustrator version 25.1 (Adobe Systems).
Reporting summary
Further information on research design is available in theNature Portfolio Reporting Summary linked to this article.
