Journal of Parkinson’s Disease 11 (2021) 187–198 187 DOI 10.3233/JPD-202224 IOS Press Research Report A European Observational Study to Evaluate the Safety and the Effectiveness of Safinamide in Routine Clinical Practice: The SYNAPSES Trial Giovanni Abbruzzesea,∗ , Jaime Kulisevskyb , Bruno Bergmansc , Juan C. Gomez-Esteband , Georg Kägie , Jason Rawf , Alessandro Stefanig , Tobias Warneckeh and Wolfgang H. Josti , the SYNAPSES Study Investigators Group a DINOGMI, University of Genoa, Genova, Italy b Sant Pau Hospital, Universitat Autònoma de Barcelona, CIBERNED, Barcelona, Spain c Department of Neurology, AZ St-Jan Brugge-Oostende AV, Campus Brugge, Brugge & Ghent University Hos- pital, Ghent, Belgium d Biocruces Research Institute, Barakaldo, Spain e Department of Neurology, Cantonal Hospital St.Gallen, Switzerland f Fairfield General Hospital, Bury, Greater Manchester, UK g Parkinson Center, University Policlinico Tor Vergata, Roma, Italy h Department of Neurology, University of Muenster, Muenster, Germany i Parkinson-Klinik Ortenau, University of Freiburg, Wolfach, Germany Accepted 1 October 2020 Pre-press 22 October 2020 Abstract. Background: Safinamide modulates both dopaminergic and glutamatergic systems with positive effects on motor and non- motor symptoms of Parkinson’s disease (PD). The drug utilization study SYNAPSES was designed to investigate the use of safinamide in routine clinical practice, as recommended by the European Medicines Agency. Objective: To describe the occurrence of adverse events in PD patients treated with safinamide in real-life conditions. Methods: The SYNAPSES trial is an observational, European, multicenter, retrospective-prospective cohort study. Patients were followed up to 12 months with analyses performed in the overall population and in patients aged >75 years, with relevant comorbidities and with psychiatric conditions. Results: Of the 1610 patients included, 82.4% were evaluable after 12 months with 25.1% of patients >75 years, 70.8% with relevant comorbidities and 42.4% with psychiatric conditions. During observation 45.8% patients experienced adverse events, 27.7% patients had adverse drug reactions and 9.2% patients had serious adverse events. The adverse events were those already described in the patients’ information leaflet. The majority were mild or moderate and completely resolved and no differences were detected between the subgroup of patients. Clinically significant improvements were seen in the UPDRS motor score and in the UPDRS total score in ≥40% of patients, according to the criteria developed by Shulman et al. ∗ Correspondence to: Giovanni Abbruzzese, MD, DINOGMI, University of Genoa, Largo Paolo Daneo 3, 16132 Genova, Italy. Tel.: +39 010 3537039; E-mail: giabbr@csita.unige.it. ISSN 1877-7171/21/$35.00 © 2021 – The authors. Published by IOS Press. This is an Open Access article distributed under under the terms of the Creative Commons Attribution License (CC BY 4.0). 188 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice Conclusion: The SYNAPSES study confirms the good safety profile of safinamide even in special groups of patients. Motor complications and motor scores improved with clinically significant results in the UPDRS scale maintained in the long-term. Keywords: Parkinson’s disease, safinamide, MAO-B inhibitor, real-life evaluation INTRODUCTION medicines in everyday clinical use, complementing with evidence from randomized control trials [7]. Parkinson’s disease (PD) is the second most com- During the initial marketing authorization pro- mon neurodegenerative illness following Alzhei- cedure, the European Medicines Agency (EMA) mer’s disease, characterized by the loss of neuro- recommended to provide additional real-world data melanin—containing neurons in the substantia nigra. on safinamide treatment in some categories of Approximately 1.2 million people live with PD patients not well represented in clinical trials, namely throughout Europe and the number is expected to those aged >75, with relevant comorbidities and double in the next 10 years [1]. with concomitant psychiatric conditions such as The main motor symptoms of PD are resting psychosis, bipolar disorder and severe depression. tremor, bradykinesia and rigidity. The disease is Following this request, a Drug Utilization Study also associated with non–motor symptoms such as (DUS) called “SYNAPSES” (“european multicen- depression, apathy, sleep disorders, pain and gastroin- ter retrospective-prospective cohort StudY to observe testinal disturbances, with a considerable reduction safiNAmide safety profile and pattern of use in clini- of patients’ quality of life [2]. The current pharmaco- cal Practice during the firSt post-commErcialization logical management is largely based on symptomatic phaSe”) was designed to investigate how safinamide drugs. Traditional pharmacotherapies for PD aim is prescribed and used in routine clinical practice to restore depleted dopamine levels in the brain and to collect safety data. The aim of the study was but are limited by long-term complications, such merely descriptive and there were no pre-specified as motor fluctuations and dyskinesia. Moreover, the hypotheses. existing medications usually do not alleviate non- motor symptoms [3]. Other neurotransmitters beyond MATERIALS AND METHODS dopamine, in particular glutamate, are believed to play important roles in the pathogenesis of primary Study design symptoms, motor fluctuations, dyskinesia and possi- bly neuronal cell loss [4]. SYNAPSES (EU PAS Register Number EUPAS- Safinamide is a multimodal drug with a dual mech- 13745) is a multinational, multicenter, retrospective- anism of action, dopaminergic (reversible mono- prospective cohort observational study. The study amine oxidase-B inhibition) and non-dopaminergic design was conceived to include potentially all (modulation of the abnormal glutamate release), that patients treated with safinamide as per clinical prac- offers an innovative approach to the management of tice. The prospective observation was chosen because motor and non-motor symptoms and motor compli- in most countries the study onset was expected to cations [5]. None of the drugs for PD already on coincide with the drug commercialization, while the the market have this peculiar double mechanism of retrospective part was performed to include also action, therefore the Movement Disorder Society has patients starting treatment before the study onset. The included safinamide in a class of drugs different from countries involved were Belgium, Germany, Italy, selegiline and rasagiline [6]. Spain, Switzerland and United Kingdom. The study Despite that the international guidelines recom- was conducted in 128 neurology and geriatric centers, mend double-blind, placebo or active-controlled specialized in PD treatment. Physicians participating studies to assess the efficacy and safety of a new ther- in the study got an appropriate compensation. apeutic treatment, they do not provide information Both the protocol and patient materials were on the “real-word” clinical use of these drugs. Many approved by Independent Ethics Committees and regulatory authorities are thus encouraging the gener- Health Authorities in all the participating countries ation of “real-world data” that offer the possibility to and the study was conducted according to the eth- derive novel insights on the use and performance of ical standards of the institutional and/or national G. Abbruzzese et al. / Safinamide in Routine Clinical Practice 189 research committee and according to the Declaration treated with safinamide in real-life conditions for up of Helsinki. to one year. The analysis was conducted overall and A total of 1,610 patients were enrolled. This study in the targeted subgroups, namely in patients aged avoided any selection of patients by means of broad >75, those with relevant comorbidities and those with inclusion/exclusion criteria in order to observe the psychiatric conditions. largest population of patients treated with safinamide. The secondary objectives were to describe the Patients were eligible provided they met the follow- characteristics of patients treated with safinamide ing inclusion criteria: male and female patients aged according to clinical practice and to describe safi- ≥18 years who started treatment with safinamide at namide treatment patterns in real-life settings. the enrolment visit or in the previous four months according to clinical practice, with signed informed Data source and measurements and privacy consent forms. Patients were excluded if they were participating in any clinical trial with Investigators were asked to record the data in elec- safinamide at study inclusion. The criteria were more tronic standardized case report forms (eCRFs). Given compelling in Germany due to local Health Author- the observational nature of the study all activities ity requirements: in compliance with § 67 section 6 concerning patient management were conducted in AMG (German Drug Law) only patients with con- compliance with the clinical practice of each center. firmed diagnosis of PD for whom safinamide was For all patients enrolled in the study after the sig- prescribed in accordance with its summary of product nature of the Informed Consent, data on treatments characteristics (SmPC) were included, while patients and on safety outcomes were prospectively recorded cannot be enrolled in case of contraindications to safi- by the Investigator both by retrieving those already namide as listed in safinamide SmPC. According to available from the medical charts and by interview- the non-interventional study type, the treating physi- ing the patient. For patients starting treatment with cians were not given any guidelines regarding patient safinamide before the enrolment visit, data were col- selection and treatment administration and no random lected retrospectively from each patient’s medical procedure was applied. Investigators were asked only charts at the moment of the enrolment in the study; to enroll patients in the study consecutively, accord- they were later integrated/updated with information ing to the inclusion/exclusion criteria, provided they routinely collected interviewing the patient at each were using safinamide in their clinical practice. This subsequent study visit. was an observational study, hence the physician’s The assessments were made in compliance with decision of starting treatment with safinamide must Good Pharmacoepidemiology Practices and with be taken before the patient’s inclusion in the study regulatory and institutional requirements for the pro- and was completely independent from the study tection of confidentiality of patients. Personal data protocol. were collected, stored and processed exclusively in As for inclusion criteria, adult patients giving their pseudonymized form. The eCRF used for the study consent to participate in the study were eligible if they was validated according to Good Automated Manu- start treatment at the enrolment visit or in the previous facturing Practice version 5 (GAMP5) [8]. four months according to clinical practice. This limit Patients were also evaluated with the Unified was set in order to allow a reasonable time point for Parkinson’s Disease Rating Scale (UPDRS) during which data are retrospectively collected, leading to ON time, measuring the changes from baseline to the exclusion of a negligible number of patients. All each follow-up visit in the UPDRS total score and patients were followed for 12 months after the start in the UPDRS part III score. The UPDRS [9] was of treatment. If a patient discontinued treatment with chosen because it is widely known and is the rat- safinamide during the study, the observation contin- ing scale most commonly used in clinical studies and ued. Data at baseline (treatment start) and 4, 8, and 12 in routine clinical practice to follow the longitudinal months after baseline were recorded. Data of patients clinical course of PD. enrolled at the start of treatment were prospectively collected. Otherwise, data of patients enrolled after Statistical methods the start of treatment were partially retrospective and updated in continuum during the course of the study. The statistical analysis was done on all “evalu- The primary objective of the study was to describe able patients for the Full Analysis Set” defined as the occurrence of adverse events (AEs) in patients the patients satisfying all inclusion criteria and not 190 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice violating any exclusion criteria. All study endpoints • Description of safinamide treatment duration, were provided using descriptive statistics. In fact, including safinamide dose adjustments, inter- the aim of the study was merely descriptive and ruptions, discontinuation and episodes of over- there were no pre-specified hypotheses. Categorical dose. variables were described by means of absolute and • Description of changes in other concomitant PD relative frequencies, while continuous variables by therapies. means of mean, standard deviation, quartiles, min and • Motor evaluation, as measured by UPDRS III max. Data collected on all patients were pooled for (UPDRS scores was summarized at each time statistical analyses. Stratifications by country were point). not foreseen because no differences among countries were detected. Patients with missing values were not A total of 1,600 patients were expected to be excluded from the analysis, but their data were not enrolled in the study. This sample size was defined replaced; frequency of missing data was given for based on safety data available from pivotal clinical all analyzed variables. However, in order to evaluate trials, where about 66% of patients experienced treat- the potential impact of recall bias on the primary end- ment emergent AEs after 6 months of treatment. With point, the proportion of patients experiencing any AE a 20% drop-out rate, 1280 evaluable patients allow to was provided as sensitivity analysis. observe the same proportion rate of AEs with a 95% confidence interval between 63% and 68%. Study results were clinically reviewed and valued Primary objective of the study by the Study Outcome Review Board. SAS for Win- The number of patients with AEs (portion of dows Version 9.4 and SAS Enterprise Guide 7.1 were patients experiencing at least one AE from the start used for statistical analyses. of treatment with safinamide until the end of the observation period), serious adverse events (SAEs) RESULTS and adverse drug reactions (ADRs) related to safi- namide, serious or not, were described. The analyses Descriptive data were provided overall and for subgroups of interest: patients aged >75, those with relevant comorbidities Demography and those concomitantly suffering from psychiatric The patients’ overview is shown in Table 1. Out of conditions. The concomitant relevant comorbidities the 1610 patients enrolled in the SYNAPSES study, (including psychiatric conditions) were those con- 1,558 (96.8%) were evaluable for the analysis, with sidered by the study Investigators, based only on more than 80% followed up prospectively for one their own clinical judgement, as clinically significant year. and/or causing a significant deterioration of patients’ Overall, 37.1% of patients had Hoehn and Yahr conditions or interfering with PD treatment. Serious- (H&Y) stage >2 [11]. Lower H&Y stages were ness, severity, relation with safinamide according to observed for younger patients (vs older ones), for Investigator judgment, action taken and outcome of patients without (vs patients with) relevant comor- the event were also summarized. Adverse Event terms bidities and for patients without (vs patients with) (AEs, SAEs, and ADRs) were coded with the Medi- psychiatric conditions. cal Dictionary for Regulatory Activities (MedDRA) Motor symptoms were reported for 1,151 (99.6%) version 21.1 [10]. Coding was performed at the end patients (this information was missing for the remain- of the study and periodically during data collection. ing 7 patients) and about 88% of subjects had also Any AE typed in a language different from the one non-motor symptoms. Tremor and postural instabil- used for encoding (English) was translated and then ity were more frequent in older patients and in patient encoded. At the end of data collection 100% of AEs with comorbidities, while postural instability and were reconciled with the Sponsor’s safety database. rigidity were preeminent in patients with psychiatric conditions. Secondary objectives of the study Regarding non-motor symptoms, older patients The following endpoints were evaluated: showed a higher frequency of cognitive, cardiovas- cular, gastrointestinal and urinary symptoms than • Description of demographic and clinical base- younger patients. Sleep disorders, psychiatric and line characteristics. cognitive symptoms, fatigue and pain were more G. Abbruzzese et al. / Safinamide in Routine Clinical Practice 191 Table 1 Patients’ overview FAS Patients Patients with Patients with (N = 1,558) aged > 75 y relevant psychiatric (N = 391) comorbidities conditions (N = 1,103) (N = 661) Gender (N, %) Male 961 (61.7%) 221 (56.5%) 665 (60.3%) 356 (53.9%) Female 597 (38.3%) 170 (43.5%) 438 (39.7%) 305 (46.1%) Age at enrolment (y) Mean (SD) 68.4 (9.7) 79.7 (3.1) 70.0 (8.7) 68.3 (9.4) Race (N, %) Caucasian 1,543 (99.0%) 389 (99.5%) 1094 (99.2%) 658 (99.5%) Other 15 (1.0%) 2 (0.5%) 9 (0.8%) 3 (0.5%) Diagnosis (N, %) Idiopathic PD 1,542 (99.0%) 389 (99.5%) 1094 (99.2%) 653 (98.8%) Atypical Parkinsonisms 12 (0.8%) 2 (0.5%) 7 (0.6%) 6 (0.9%) Other∗ 4 (0.2%) 0 (0.0%) 2 (0.2%) 2 (0.3%) Time from PD diagnosis (y): mean (SD) 7.9 (5.3) 7.9 (5.3) 7.8 (5.3) 8.4 (5.5) Time from PD onset of symptoms (y): 8.8 (5.5) 8.9 (5.5) 8.7 (5.4) 9.3 (5.5) mean (SD) Age at onset of symptoms (y): mean (SD) 59.3 (11.0) 70.9 (6.5) 61.0 (10.1) 58.8 (10.7) Hoehn & Yahr stage 1 92 (5.9%) 6 (1.5%) 50 (4.5%) 22 (3.4%) 2 888 (57.0%) 167 (42.5%) 610 (55.3%) 332 (50,4%) 3 473 (30.3%) 167 (42.9%) 361 (32.7%) 234 (35.2%) 4 99 (6.4%) 46 (11.9%) 77 (7.0%) 68 (10.3%) 5 6 (0,4%) 5 (1.2%) 5 (0.5%) 5 (0.7%) Percentages (%) were computed by column. FAS, Full Analysis Set; N, number of patients; SD, standard deviation; PD, Parkinson’s disease. ∗ Patients with other diagnosis had juvenile Parkinson’s disease. frequent in patients with comorbidities and those with Table 2 psychiatric conditions. Concomitant PD treatments at the start of safinamide At the start of treatment, the most frequent comor- Categories Total number of evaluable patients (FAS) bidities (>10%) were hypertension and heart dis- (N = 1,558) eases, metabolic disorders and joint, bone and pain Levodopa 1,537 (98.7%) disorders. The most frequently reported psychiatric Dopamine agonist 912 (58.5%) conditions (>10%) were depression (mild or moder- COMT inhibitors 424 (27.2%) ate) and anxiety (severe depression, bipolar disorders, Amantadine 178 (11.4%) Anticholinergics 34 (2.2%) psychosis and apathy were also present at a lower MAO-B inhibitors 11 (0.7%) frequency), with about 28% of patients assuming Percentages are computed out of the total number of patients antidepressants [mainly selective serotonin reuptake evaluable for FAS. A patient could have more than one PD inhibitors (SSRIs), serotonin–norepinephrine reup- treatment. PD, Parkinson’s disease; FAS, Full Analysis Set; N, take inhibitors (SNRIs) and tricyclics]. number of patients; COMT, catechol-o-methyltransferase; MAO- B, monoamine-oxidase B. Concomitant PD medications Motor fluctuations Almost all patients (N = 1,556, 99.9%) had ≥ 1 As shown in Table 3, at start of safinamide treat- ongoing treatment for PD at the start of safi- ment 1437 (92.2%) patients had motor fluctuations, namide therapy. As shown in Table 2, 1,537 patients the most frequent one being wearing off (74.3%). (98.7%) were treated with levodopa (L-dopa), 912 The percentage of patients with motor fluctuations (58.5%) dopamine agonists, 424 (27.2%) catechol-O- decreased during the study due to safinamide treat- methyltransferase (COMT) inhibitors, 178 (11.4%) ment: at 4, 8, and 12 months after start of treatment, amantadine, 34 (2.2%) anticholinergics and 11 the percentage of patients with motor fluctuations (0.7%) monoamine oxidase-B (MAO-B) inhibitors was respectively 73.5%, 70.6%, and 67.4%, with the (rasagiline). Over the study period, the overall mean highest reduction for wearing-off and early morning dose of levodopa (alone or plus COMT inhibitors) fluctuations. The improvements seen after 4 months and the overall mean doses of the dopamine-agonist indicate a rapid onset of efficacy of the drug. The treatments did not change significantly, confirming percentage of patients with fluctuations has been cal- what has been previously observed in the pivotal trials culated at each follow-up visits based on the available [12, 13]. data and on the total number of the evaluable patients. 192 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice Table 3 Fluctuations at the start of treatment with safinamide and during the follow-up Motor complications FAS Total evaluable Total evaluable Total evaluable (N = 1,558) patients at patients at patients at 4 months 8 months 12 months (N = 1,373) (N = 1,323) (N = 1,326) Any 1,437 (92.2%) 1,009 (73.5%) 934 (70.6%) 894 (67.4%) Wearing-off 1,158 (74.3%) 752 (54.8%) 704 (53.2%) 701 (52.9%) Early morning fluctuations 363 (23.3%) 198 (14.4%) 196 (14.8%) 182 (13.7%) Unpredictable fluctuations 264 (16.9%) 152 (11.1%) 149 (11.3%) 133 (10.0%) Delayed on 177 (11.4%) 108 (7.9%) 109 (8.2%) 109 (8.2%) Dyskinesia 610 (39.2%) 469 (34.2%) 409 (30.9%) 369 (27.8%) Other 87 (5.6%) 62 (4.5%) 52 (3.9%) 32 (2.4%) Percentages were computed out of the number of patients evaluable for the FAS and for each follow up visit. FAS, Full Analysis Set; N, number of patients. Other AEs with a frequency ≤ 3% of the total num- ber of events were hallucinations (N = 41, 2.9%), fall (N = 25, 2.3%), nausea (N = 18, 1.7%), consti- pation (N = 14, 1.3%) and headache (N = 14, 1.3%). The AEs observed were those already described in the patients’ leaflet and no differences were detected in term of nature, frequency, causal relationship or severity. Forty-four eye disorders (3.1% of the total number Fig. 1. Adverse events and adverse reactions during observation of AEs) were observed, all non serious: the most fre- (overall). AE, Adverse Event; SAE, Serious Adverse Event; ADR, quent ones (<1%) were blurred vision, reduced visual Adverse Drug Reaction; SADR, Serious Adverse Drug Reaction. acuity and diplopia. It must be noted that data on the concomitant use of safinamide in patients with Primary endpoint retinopathy of any type are not available because these subjects were not enrolled. Nevertheless, no Adverse events and serious adverse events relevant frequency of cataract was reported thus no As reported in Fig. 1, during observation 714 signals of ophthalmological relevance were detected. (46%) patients experienced at least one AE and 432 As previously reported, 194 serious adverse events (28%) patients experienced at least one ADR; 143 (SAEs) occurred in about 9% of patients. SAEs were patients (9%) had at least one SAE and 36 patients about 14% of all adverse events. The most frequently (2%) had at least one SADR. The percentage of reported SAEs by system organ class were infec- patients experiencing AEs during one year of treat- tions and infestations (N = 38, 3.0% of all occurred ment with safinamide in real-life conditions was 30% AEs), injury, poisoning and procedural complications lower compared to the percentage observed in six- (N = 31, 2.0%) and nervous system disorders (N = 29, months pivotal trials [12, 13]. 2.0%). Adverse events were mainly mild (62.0%) or mod- Among the above, the most frequent were pneumo- erate (28.0%): only 2% were considered to have nia (N = 9), urinary tract infections (N = 6), dyskinesia a definite relationship with safinamide and in the (N = 5) and femur fractures (N = 5). majority of the cases no action was taken regard- The majority of SAEs was completed resolved. No ing the treatment and the outcome was resolved. The SAEs had a definite relationship with safinamide and monthly incidence rate of AEs was very low, 0.07 only eight (4.0% of all SAEs) and four (2.0% of all AEs per patient per month. In Table 4 the AEs are SAEs) had a possible or probable relationship with classified by system organ class (SOC). safinamide, respectively. Dyskinesia was the most frequently reported Patients with a partially retrospective observa- AE among nervous system disorders, although it tion period (having started treatment with safinamide occurred in a lower frequency in the SYNAPSES before study inclusion) were 779 (50% of evaluable study compared to previous pivotal trials (13.7% vs patients). No differences were detected in term of AE 18%). or SAE frequency, based on the sensitivity analysis, G. Abbruzzese et al. / Safinamide in Routine Clinical Practice 193 Table 4 Distribution of adverse events severity by system organ class System Organ Class Any Mild Moderate Severe N = 1,435 (100%) N = 888 (62%) N = 404 (28%) N = 143 (10%) Nervous system disorders 480 (33.4%) 345 (24.0%) 110 (7.7%) 25 (1.7%) Psychiatric disorders 205 (14.3%) 121 (8.4%) 68 (4.7%) 16 (1.2%) Gastrointestinal disorders 107 (7.4%) 73 (5.1%) 29 (2.0%) 5 (0.3%) Injury, poisoning and procedural complications 99 (6.9%) 36 (2.5%) 41 (2.9%) 22 (1.5%) Musculoskeletal and connective tissue disorders 94 (6.5%) 55 (3.8%) 37 (2.6%) 2 (0.1%) General disorders 93 (6.5%) 65 (4.5%) 21 (1.5%) 7 (0.5%) Infections and infestations 75 (5.2%) 32 (2.2%) 18 (1.3%) 25 (1.7%) Eye disorders 44 (3.1%) 34 (2.4%) 10 (0.7%) 0 (0%) Vascular disorders 37 (2.5%) 23 (1.6%) 12 (0.8%) 2 (0.1%) Respiratory, thoracic and mediastinal disorders 32 (2.3%) 16 (1.1%) 8 (0.6%) 8 (0.6%) Skin and subcutaneous tissue disorders 27 (1.9%) 22 (1.5%) 4 (0.3%) 1 (0.1%) Renal and urinary disorders 26 (1.8%) 16 (1.1%) 7 (0.5%) 3 (0.2%) Cardiac disorders 25 (1.7%) 5 (0.3%) 9 (0.6%) 11 (0.8%) Metabolism and nutrition disorders 18 (1.2%) 12 (0.8%) 4 (0.3%) 2 (0.1%) Investigations 12 (0.8%) 9 (0.6%) 3 (0.2%) 0 (0%) Neoplasms benign, malignant and unspecified 12 (0.9%) 1 (0.1%) 3 (0.2%) 8 (0.6%) Surgical and medical procedures 12 (0.8%) 3 (0.2%) 7 (0.5%) 2 (0.1%) Ear and labyrinth disorders 11 (0.8%) 11 (0.8%) 0 (0%) 0 (0%) Hepatobiliary disorders 7 (0.5%) 0 (0.0%) 6 (0.4%) 1 (0.1%) Blood and lymphatic system disorders 6 (0.4%) 3 (0.2%) 2 (0.1%) 1 (0.1%) Reproductive system and breast disorders 5 (0.4%) 3 (0.2%) 1 (0.1%) 1 (0.1%) Endocrine disorders 2 (0.2%) 1 (0.1%) 1 (0.1%) 0 (0%) Percentages were computed out of the total number of AEs and data sorted by descending frequency of SOC. N, number of patients. between patients who started safinamide before or (a similar percentage was reported in pivotal trials). after study inclusion. Main reasons for discontinuation were adverse events A total of 233 (16.9%) evaluable patients had (N = 161, 10.3%), patient choice (N = 81, 5.1%) and non-appropriate use of safinamide (e.g., absence disease progression (N = 20, 1.3%). No relevant of fluctuations, safinamide starting dosage differ- differences regarding the treatment patterns with safi- ent from 50 mg/day, no concomitant levodopa use, namide emerged between the subgroup of patients diagnosis different from idiopathic PD or concomi- (elderly, with relevant comorbidities and with psy- tant treatment with other MAO-B inhibitors). These chiatric conditions). data were captured in the SYNAPSES study because they represent the real-practice of the PD treatment. Although not initially foreseen by the study proto- Subgroup of patients col, a descriptive analysis on the frequency of safety The summary of the safety results of the three sub- events was provided in patients undergoing appro- group of interest (patients aged >75 years, patients priate use of safinamide vs inappropriate use: no with relevant comorbidities and patients with psychi- differences were detected between the two groups. atric conditions) is shown in Fig. 2. Elderly patients. During observation 47.3% patients older than 75 years had at least one AE, Secondary endpoints 26.1% at least one ADR, 13.6% at least one SAE and 2.3% at least SADR. Safinamide treatment patterns The proportions of patients experiencing at least Safinamide was administered at an initial dose of one AE, ADR or SADR were similar in the popula- 50 mg/day to 94% of patients; during the observa- tion of patients aged >75 compared to patients aged tion period, 58.0% of patients had a dose increase ≤75, while was higher for SAE in patients aged >75 from 50 to 100 mg due to “clinician decision” and (13.6%) compared to patients aged ≤75 (7.7%). No 6% a dose decrease from 100 to 50 mg due to relevant differences emerged in severity, causal rela- “patient choice”. Safinamide temporary discontinu- tionship or action taken between the two age groups. ation was observed for 68 patients (4.4%), while 336 Dyskinesia was still the most frequent AE in elderly patients (21.6%) permanently discontinued the drug patients, but with a lower frequency than the overall 194 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice Fig. 2. Safety summary in subgroups of patients. AE, Adverse Event; SAE, Serious Adverse Event. sample and the younger patients (9.5% vs 13.7%) UPDRS scores while hallucinations (5.1% vs 2.9%) were slightly The mean (SD) of UPDRS (part I, II, III, IV, total) more frequent in patients aged >75. scores at the beginning of treatment with safinamide Patients with comorbidities. During observation and at 4-, 8-, 12-month follow up are shown in Fig. 3. 49.1% patients with relevant comorbidities had at A substantial stability of disease severity over time least one AE, 28.6% had at least one ADR, 11.1% emerged with improvements since the first 4-month had at least one SAE and 2.5% at least one SADR. follow-up in the UPDRS part II (Activities of Daily The proportion of patients with AEs and SAEs was Living) part III (Motor Examination), part IV (Com- higher in patients with relevant comorbidities than in plication of Therapy) and UPDRS Total Score. patients without (49.1% vs 37.8% and 11.1% vs 4.6% According to the criteria developed by Shulman respectively). No other differences were detected. [14], a difference >4.3 points for the UPDRS Total Dyskinesia was the most frequent AE in the sub- score (subscales I, II, III and IV) and >2.5 points for group of patients with comorbidities, with a slightly the UPDRS Part III (Motor Examination score) is lower frequency than the overall sample and the considered clinically significant. The percentage of patients without comorbidities (12.3% vs 13.7%). patients with clinically relevant differences between The distribution of the occurrence of the other AEs baseline and 12-month follow up visit is reported was similar in the two subgroup of patients (with and in Fig. 4. After one-year treatment with safinamide, without comorbidities) and the overall sample. 39% and 45% patients showed a clinically significant Patients with psychiatric conditions. During obser- improvement in UPDRS Total and Motor Examina- vation, 47.8% patients with psychiatric conditions tion scores, respectively. had at least one AE, 31.0% at least one ADR, 10.3% at least one SAE and 3.5% at least one SADR. DISCUSSION No relevant differences were observed in patients with or without psychiatric conditions regarding the The SYNAPSES study documented for the first severity, causal relationship and action taken for time the real-world use of safinamide in six European adverse events. No serotoninergic syndromes were countries, particularly in patients’ populations such reported despite 28% of patients were assuming as patients older than 75, with psychiatric illnesses antidepressant drugs including fluoxetine, thus con- or with relevant comorbidities, where limited infor- firming that the concomitant use of safinamide and mation was available in the pre-authorization clinical SSRIs/SNRIs/tricyclics is safe and well tolerated. trials. The results have been analyzed and presented Dyskinesia was the most frequent AE in the sub- as pooled data of the two safinamide doses, anyway group of patients with psychiatric conditions, with a during the study 58% of patients increased the dose slightly higher frequency than in the overall sample of safinamide from 50 to 100 mg with no differences and in patients without psychiatric conditions (15.6% in term of adverse events, including dyskinesia, con- vs 13.7%). firming what has been reported in the literature. G. Abbruzzese et al. / Safinamide in Routine Clinical Practice 195 Fig. 3. UPDRS (part I, II, III, IV, total scores) at start of treatment with safinamide and at 4-, 8-, and 12-month follow-up (overall). UPDRS, Unified Parkinson’s Disease Rating Scale. However, it is important to note that most PD patients who complained of dyskinesia have presented these abnormal movements since the beginning of the study with no further aggravation. The treatment with safinamide did not require any change in the concomitant dopaminergic ther- apies, allowing a better motor control with no safety problems. Despite its dopaminergic activity, Fig. 4. Percentage of patients with clinically important differ- impulse control disorder or sleep deterioration were ence (improvement) in the UPDRS scores (difference between not reported after safinamide treatment. Safinamide 12-months follow-up and baseline). UPDRS, Unified Parkinson’s increases dopamine levels in the putamen, the region Disease Rating Scale. devoted to motor control, but does not affect regions involved in reward circuitry and impulse control Safinamide was safe and well tolerated and no disorders (ICDs) such as the nucleus accumbens major or unexpected safety concerns were identi- [18]. Moreover, glutamate levels seem to mediate fied. The percentage of patients experiencing AEs reward-seeking behavior within this nucleus [19]. was 30% lower compared to the percentage observed Safinamide is a MAO-B inhibitor and a glutamate in previous pivotal trials. This is remarkable because modulator and this second mechanism of action, pivotal trials have strict inclusion/exclusion criteria in addition to the non-influence on the mesolimbic and enroll generally patients with less comorbidities structures, may contribute to avoid any ICD deterio- than subjects in a practice setting. ration. The rate of AEs observed with safinamide was also At the start of the study about 60% of the patients 50% lower than the percentage reported in the same were receiving dopamine agonists (DA) as concomi- patients’ population for rasagiline in the PRESTO tant treatment with L-dopa. It is reported in the study (45.7% vs 93% respectively) [15] and 40% literature that 15% to 20% of patients on DA exhibit lower than the percentage reported for opicapone in impulse control disorder behavior (e.g., gambling the OPIPARK study (45.7% vs 74.9% respectively) and hypersexuality), imposing a significant economic [16]. burden and a quality of life deterioration of patients The most frequent reported AE was dyskinesia, as and their caregivers [20]. The adjunct of safinamide expected with drugs that increase levodopa bioavail- improved the symptoms without tolerability prob- ability [17], although at very lower frequency than lems and with no need to increase the DA dose or those observed in previous pivotal clinical trials. switch to another drug. 196 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice Regarding sleep disorders, it is known that dopa- 4 months. This significant and rapid-onset effect of mine agonists alone or combined with levodopa can safinamide may be explained by its dual mechanism cause sleep disturbances and episodes of sudden sleep of action, dopaminergic and glutamatergic [27]. The attacks [21]. Glutamate is an important wake-active pathogenesis of motor fluctuations, in fact, suggests neurotransmitter producing a sustained behavioral that glutamate and other neurotransmitters, in addi- and electro-encephalogram arousal [22]. The absence tion to dopamine, contribute to the appearance of the of sleep disturbances in PD patients taking safinamide symptoms [28]. may be related to the effects of the drug on the glu- Consistent with the benefits observed in motor fluc- tamatergic system. tuations, UPDRS scores improved with safinamide The safety profile of safinamide in patients over 75 as adjunct therapy after only 4 months of treatment years, with comorbidities and with psychiatric con- with a clinically important effect in ≥40% of patients ditions appeared to be broadly similar to the overall according to the criteria developed by Shulman et al. population. The slightly higher percentage of SAEs [14]. In term of absolute values, the magnitude of the observed in elderly patients and in patients with effect of safinamide on motor symptoms is similar comorbidities is an expected phenomenon since they to that reported for other MAO-B inhibitors and for are generally affected by several relevant concomi- COMT inhibitors by Stowe et al. [29]. As expected, tant pathologies treated with polytherapies. This is the improvements are higher at 4 and 8 months due confirmed by the fact that ADRs did not occur more to the progression of the pathology, nevertheless the frequently in these patients, suggesting that there results obtained for the motor score are noteworthy were no contraindications to the use of safinamide because the patients were already receiving an opti- in these subjects. Safinamide in fact, unlikely rasag- mized dopaminergic therapy (based on clinician’s iline, is not metabolized by CYP1A2 and thus does judgement) and further improvements were unex- not have any important drug-drug interaction. The pected. The annual decline in early untreated PD slightly higher frequency of hallucinations in elderly patients is 5–6 points for UPDRS III motor score [30]. patients reflects, as known, age as risk factor for hal- Dopaminergic therapies reduce the decline in motor lucinations [23]. function to an annual downfall of 3.3 points [31]. The The absence of any safety issue in psychiatric add-on of safinamide to a standard levodopa therapy patients, despite their concomitant treatment with compensates this progressive deterioration in 45% of antidepressant drugs, suggests that safinamide does patients allowing for a substantial stability of their not require special safety precautions for these sub- motor symptoms. jects. Safinamide, in fact, increases the synaptic The main strengths of this study are the large num- availability of dopamine without a direct interaction ber of fluctuating patients included, with a broad with dopamine D2 receptors that are known to be range of disease severity (Hoehn and Yahr stages involved in psychotic behaviors [18]. I–IV), the real-life clinical practice setting and the Motor fluctuations, and in particular wearing-off relevant sample sizes for the subgroups of interest. and early morning fluctuations, affect the majority of Limitations include the open-label design without a patients with PD: as observed in the DEEP study, placebo or active control, typical of the observational 63–75.6% of patients treated with L-dopa experi- studies, where the expectations of the patients are enced wearing-off with a significant deterioration of generally higher than in double-blind trials. Patients their quality of life [24]. The pharmacological treat- diaries for ON and OFF time were not adopted, as ment of motor fluctuations is difficult and remain a suggested by EMA, because they are not used in rou- real unmet need [25]. There is a significant associa- tine clinical practice. There was also a potential risk of tion between motor fluctuations and annual costs of under-estimation of AEs from both the patients and PD: the mean costs of patients with motor fluctuations the clinicians: however, no relevant difference was is generally two-three times greater than the costs observed when considering the frequency of AEs or of patients without [26]. Treatment strategies capa- SAEs in retrospective or prospective patients. ble of delaying the onset or attenuating the severity of motor fluctuations could be expected to improve Conclusions QoL and reduce some of the economic burden of PD. Safinamide was shown to reduce of about 40–50% The management of motor complications in PD motor fluctuations, in particular wearing-off and early remains a significant challenge in which all available morning fluctuations, with visible efficacy already at pharmacologic options involve a risk of exacerbating G. Abbruzzese et al. / Safinamide in Routine Clinical Practice 197 adverse events. The SYNAPSES study, conducted in E, De Fábregues O, Escalante Arroyo S, Espinosa a real word setting in six European countries, con- Rosso R, Esquivel López A, Freire E, Garcı́a Cobos firmed the safety and tolerability of safinamide, as E, Garcı́a Moreno JM, Gomez-Esteban JC, González- adjunct therapy, in fluctuating patients and in spe- Ardura J, Grandas Perez F, Kulisevsky J, Kurtis M, cial groups of subjects. Neither age, comorbidities, Juni J, Legarda I, Leiva C, López Aristegui N, López nor psychiatric conditions seem to have any relevant Manzanares L, Lozano JJ, Luquı́n MR, Martinez Cas- effect on its safety profile. trillo JC, Martı́ Domenech MJ, Martı́nez I, Mata M, Motor complications and motor scores improved Mir Rivera P, Pascual Sedano B, Rodrı́guez Oroz MC, with clinically significant results in the UPDRS scale Rodrı́guez Uranga JJ, Sanchez S, Santos Garcı́a D, maintained in the long-term. These results suggest Solano B, Vaamonde Gamo J. that safinamide can be an effective and safe option for Switzerland: Accolla E, Bohlhalter S, Kälin A, the management of motor fluctuations in levodopa- Kägi G, Michelis J. treated patients. U.K.: Carrol C, Henderson E, Raha S, Raw J, Silva N, Silverdale M. ACKNOWLEDGMENTS REFERENCES The authors thank the investigators and the patients involved in the trial. [1] de Lau LM, Giesbergen PC, de Rijk MC, Hofman A, Koud- The study was funded by Zambon S.p.A. staal PJ, Breteler MM (2004) Incidence of parkinsonism and Parkinson’s Disease in a general population: The Rotterdam Study. Neurology 63, 1240-1244. CONFLICT OF INTEREST [2] Chaudhuri KR, Shapira AH (2009) Non-motor symptoms of Parkinson’s disease: Dopaminergic pathopshysiology and treatment. Lancet Neurol 8, 464-474. G. Abbruzzese, J. Kulisevsky, W. H. Jost, B. [3] Hauser RA (2009) Levodopa: Past, present and future. Eur Bergmans, J.C. Gomez-Esteban, G. Kägi, J. Raw, A. Neurol 62, 1-8. Stefani, and T. Warnecke are members of the Scien- [4] Chase TN, Bibbiani F, Oh JD (2003) Striatal glutamater- tific Advisory Board of Zambon SpA. gic mechanisms and extrapyramidal movement disorders. Neurotox Res 5, 139-146. [5] Müller T, Foley P (2017) Clinical pharmacokinetics and SYNAPSES Study Investigators Group pharmacodynamics of safinamide. Clin Pharmacokinet 56, 251-261. Belgium: Bergmans B, Bourgeois P, Cras P, De [6] Fox SH, Katzenschlager R, Lim SY, Barton B, de Bie RMA, Seppi K, Coelho M, Sampaio C; Movement Disor- Klippel N, Dethy S, Franco G, Garraux G, Geens K, der Society Evidence-Based Medicine Committee (2018) Jacquerye P, Jeanjean A, Santens P, Supiot F, Van der International Parkinson and Movement Disorder Society Linden C. Evidence-Based Medicine review: Update on treatments for Germany: Blersch WK, Delf M, Hellwig B, Herbst the motor symptoms of Parkinson’s disease. Mov Disord 33, 1248-1266. HP, Kupsch A, Jost WH, Lang M, Muhlack S, Nastos [7] Cave A, Kurz X, Arlett P (2019) Real-word data for regu- I, Oehlwein C, Schlegel E, Schwarz J, Warnecke T, latory decision making: Challenges and possible solutions Woitalla D. for Europe. Clin Pharmacol Ther 106, 36-39. Italy: Abbruzzese G, Aguggia M, Avarello T, [8] ISPE: The Good Automated Manufacturing Practice (GAMP) guide for validation of automated systems in phar- Barone P, Baruffaldi R, Belgrado E, Bentivoglio AR, maceutical manufacture. GAMP 5: A risk-based approach to Bosco D, Calabresi P, Callegarini C, Cannas A, Cen- compliant GxP computerized systems. International society tonze D, Ceravolo R, Colosimo C, Comi C, Contardi for pharmaceutical engineering (February 2008). S, Cortelli P, Cossu G, D’Amelio M, De Pandis MF, [9] Fahn S, Elton R, Members of the UPDRS Development Committee (1987) The Unified Parkinson’s Disease Rat- Denaro A, Di Lazzaro V, Fabbrini G, Gasparoli E, ing Scale. In Recent Developments in Parkinson’s Disease, Guidi M, Iliceto G, Lopiano L, Manganotti P, Mar- Vol. 2, Fahn S, Marsden CD, Calne DB, Goldstein M, eds. coni R, Marini C, Marsala SZ, Mauri M, Moleri M, McMellam Health Care Information, Florham Park, pp. Monge A, Morgante F, Negrotti A, Nordera G, Onofrj 153-163. [10] Medical Dictionary for Regulatory Activities (MedDRA) M, Pacchetti C, Padovani A, Pontieri FE, Priori A, version 21.1 (September 2018). Quatrale R, Sensi M, Stefani A, Tamma F, Tessi- [11] Hoehn M, Yahr M (1967) Parkinsonism: Onset, progression tore A, Tinazzi M, Vitale C, Volontè MA, Zappia M, and mortality. Neurology 17, 427-442. [12] Borgohain R, Szasz J, Stanzione P, Meshram C, Bhatt Zecchinelli AL. M, Chirilineau D, Stocchi F, Lucini V, Giuliani R, For- Spain: Arbelo Gonzalez JM, Bayés A, Blazquez M, rest E, Rice P, Anand R; Study 016 Investigators (2014) Calopa Garriga M, Callen A, Campos Arillo V, Cubo Randomized trial of safinamide add-on to levodopa in 198 G. Abbruzzese et al. / Safinamide in Routine Clinical Practice Parkinson’s disease with motor fluctuations. Mov Disord 29, [22] Pedersen NP, Ferrari L, Venner A, Wang JL, Abbott SBG, 229-237. Vujovic N, Arrigoni E, Saper CB, Fuller PM (2017) Supra- [13] Schapira AH, Fox SH, Hauser RA, Jankovic J, Jost WH, mammillary glutamate neurons are a key node of the arousal Kenney C, Kulisevsky J, Pahwa R, Poewe W, Anand R system. Nat Commun 8, 1405. (2017) Assessment of safety and efficacy of safinamide [23] Sanchez-Ramos JR, Ortoll R, Paulson GW (1996) Visual as a levodopa adjunct in patients with Parkinson’s disease hallucinations associated with Parkinson’s disease. Arch and motor fluctuations. A randomized clinical trial. JAMA Neurol 53, 1265-1268. Neurol 74, 216-224. [24] Stocchi F, Antonini A, Barone P, Tinazzi M, Zappia M, [14] Shulman LM, Gruber-Baldini AL, Anderson KE, Fishman Onofrj M, Ruggieri S, Morgante L, Bonuccelli U, Lopiano PS, Reich SG, Weiner WJ (2010) The clinically important L, Pramstaller P, Albanese A, Attar M, Posocco V, Colombo difference on the Unified Parkinson’s Disease rating scale. D, Abbruzzese G; DEEP study group (2014) Early DEtec- Arch Neurol 67, 64-70. tion of wEaring-off in Parkinson disease: The DEEP study. [15] Parkinson Study Group (2005) A randomized placebo- Parkinsonism Relat Disord 20, 204-211. controlled trial of rasagiline in levodopa-treated patients [25] Olanow CW, Watts RL, Koller WC (2001) An algorithm with Parkinson’s disease and motor fluctuations: The (decision tree) for the management of Parkinson’s disease: PRESTO study. Arch Neurol 62, 241-248. Treatment guidelines. Neurology 56, S1–S86. [16] Reichmann H, Lees A, Rocha JF, Magalhes D, Soares- [26] Keränen T, Kaakkola S, Sotaniemi K, Laulumaa V, Haa- da-Silva P, OPTIPARK investigators (2020) Effectiveness paniemi T, Jolma T, Kola H, Ylikoski A, Satomaa O, and safety of opicapone in Parkinson’s disease patients Kovanen J, Taimela E, Haapaniemi H, Turunen H, Takala with motor fluctuations: The OPTIPARK open-label study. A (2003) Economic burden and quality of life impairment Transl Neurodegener 9, 9. increase with severity of PD. Parkinsonism Rel Disord 9, [17] Freitas ME, Hess CW, Fox SH (2017) Motor complications 163-168. of dopaminergic medications in Parkinson’s disease. Semin [27] Cattaneo C, Sardina M, Bonizzoni E (2016) Safinamide as Neurol 37, 147-157. add-on therapy to levodopa in mid- to late-stage Parkinson’s [18] Caccia C, Maj R, Calabresi M, Maestroni S, Faravelli L, disease fluctuating patients: post-hoc analyses of studies 016 Curatolo L, Salvati P, Fariello RG (2006) Safinamide: From and SETTLE. J Parkinsons Dis 6, 165-173. molecular targets to a new anti-Parkinson drug. Neurology [28] Blandini F, Porter RH, Greenamyre JT (1996) Glutamate 67, S18-S23. and Parkinson’s disease. Mol Neurobiol 12, 73-94. [19] Roussakis AA, Lao-Kaim N, Piccini P (2019) Brain imaging [29] Stowe R, Ives N, Clarke CE, Handley K, Furmston A, Deane and impulse control disorders in Parkinson’s disease. Curr K, van Hilten JJ, Wheatley K, Gray R (2011) Meta-analysis Neurol Neurosci Rep 19, 67. of the comparative efficacy and safety of adjuvant treatment [20] Bostwick JM, Hecksel KA, Stevens SR, Bower JH, Ahlskog to levodopa in later Parkinson’s disease. Mov Disord 26, JE (2009) Frequency of new-onset pathologic compulsive 587-598. gambling or hypersexuality after drug treatment of idio- [30] Poewe W, Mahlknecht P (2009) The clinical progression of pathic Parkinson disease. Mayo Clin Proc 84, 310-316. Parkinson’s disease Parkinsonism Relat Disord 15S, S28- [21] Frucht S, Rogers JD, Greene PE, Gordon MF, Fahn S (1999) S32. Falling asleep at the wheel: Motor vehicle mishaps in person [31] Schrag A, Dodel R, Spottke A, Bornschein B, Siebert U, taking pramipexole and ropinirole. Neurology 52, 1908- Quinn NP (2007) Rate of clinical progression in Parkinson’s 1910. disease. A prospective study. Mov Disord 22, 938–945.