Neuropsychiatric Adverse Events Associated With Foslevodopa/Foscarbidopa Continuous Subcutaneous Infusion in Clinical Practice: A Multicenter Study

ABSTRACT

Background

Foslevodopa/foscarbidopa continuous subcutaneous infusion (LDp/CDp CSI) has emerged as an effective and well‐tolerated therapy for reducing OFF and increasing non‐troublesome ON in advanced Parkinson's disease (PD). Neuropsychiatric adverse events (AEs) have been reported in both clinical trials and real‐world studies, with some real‐world cohorts suggesting higher rates among patients with prior hallucinations or cognitive impairment. The present study aimed to determine the incidence and risk factors of neuropsychiatric AEs in a large prospective real‐world cohort.

Methods

We analyzed data from the DATs‐PD GETM Spanish Registry, an observational, prospective, multicenter, open‐label study.

Results

214 patients treated with LDp/CDp CSI were included. Median age was 69 years, and median disease duration was 12 years. At baseline, 35% had cognitive impairment, 25.7% hallucinations/psychosis, and 26.2% impulse control disorders (ICDs). During follow‐up after initiation (median 163 days), 19.2% developed at least one neuropsychiatric AE, mostly mild–moderate, and only 2.3% required device removal. Most events occurred more than 1 month after treatment initiation. In adjusted Cox, none of the evaluated variables were associated with the development of hallucinations/psychosis/confusion. The presence of ICD at baseline was associated with an increased risk of ICD‐related AEs.

Conclusion

Neuropsychiatric AEs, mainly hallucinations/psychosis, occurred in a clinically relevant proportion of patients treated with LDp/CDp CSI. However, they were generally mild‐to‐moderate and rarely led to treatment discontinuation. Except for ICD, baseline cognitive and psychotic features were not associated with higher incidence. These findings support its use in appropriately selected patients while highlighting the importance of individualized careful clinical monitoring.

In 214 patients with advanced Parkinson's disease receiving foslevodopa/foscarbidopa continuous subcutaneous infusion, neuropsychiatric adverse events occurred in 19.2%. Baseline cognitive impairment and hallucinations did not increase risk, whereas preexisting impulse control disorders predicted ICD‐related adverse events. Neuropsychiatric complications rarely led to treatment discontinuation.

1. Introduction

Parkinson's Disease (PD) is characterized by the degeneration of dopaminergic and non‐dopaminergic neurons, leading to progressive motor and non‐motor symptoms [1]. As the disease advances, pulsatile oral therapies often become insufficient to provide adequate symptom control, resulting in fluctuations and dyskinesias that substantially impair patients' quality of life [2]. Several advanced therapeutic approaches, including deep brain stimulation and infusion‐based treatments, have demonstrated efficacy in these stages of the disease [3]. Among these, continuous subcutaneous infusion (CSI) of levodopa has recently emerged as a promising therapy that enables continuous dopaminergic delivery without the need for invasive intestinal procedures [4, 5].

CSI of levodopa has been tested in two different formulations: foslevodopa/foscarbidopa (LDp/CDp) (ABBV‐951), and levodopa/carbidopa (ND0612 and DIZ102); however, only ABBV‐951 has received regulatory approval to date [5, 6, 7]. In pivotal phase III clinical trials, both ABBV‐951 and ND0612 have demonstrated significant increases in ON time without troublesome dyskinesia, as well as reductions in OFF time [5, 7]. Moreover, these trials have shown a favorable benefit–risk profile, with adverse events (AEs) predominantly being local and of mild‐to‐moderate severity [5, 7, 8, 9, 10]. Among nonlocal AEs, neuropsychiatric symptoms have raised the greatest concern. Neuropsychiatric AEs, including hallucinations, psychosis, confusion, and impulse control disorders (ICD), were reported in a relevant proportion of patients in clinical trials (up to 18.5%), with higher rates than those observed in oral levodopa control groups [5, 7, 8, 9]. Notably, these events were most often mild hallucinations and treatment discontinuation only occurred in < 5% of patients [5, 7, 8]. However, these trials included few patients with prior psychosis or cognitive impairment, potentially limiting the generalizability of their findings [11, 12]. Neuropsychiatric complications have also been reported with other infusion‐based therapies for advanced PD, including continuous subcutaneous apomorphine infusion (CSAI) and levodopa‐carbidopa intestinal gel (LCIG), although reported frequencies vary considerably across studies and direct comparisons remain challenging due to differences in patient populations and outcome definitions.

Available real‐world evidence regarding LDp/CDp CSI has reported rates of neuropsychiatric AEs ranging from approximately 8% to over 30%, broadly overlapping with the incidence reported in clinical trials, but differences in patient characteristics and baseline neuropsychiatric burden complicate direct comparisons [11, 13, 14, 15]. These events typically manifest as visual hallucinations or delirium, confusion, or ICD have also been reported rarely [14, 15, 16]. Whether these events arise as a consequence of preexisting neuropsychiatric comorbidities or are influenced by therapeutic management strategies remains a matter of debate [12, 15, 16].

Although existing real‐world evidence has provided valuable clinical insights, the available data remain limited with only a few studies published to date, each based on small cohorts and lacking formal statistical analyses. Therefore, this study aims to investigate the incidence of neuropsychiatric AEs and to identify associated risk factors in a large, prospective, multicenter cohort of PD patients treated with LDp/CDp CSI.

2. Methods

2.1. Study Design and Participants

Data for the analyses in this study were obtained on October 31, 2025, from the Device‐Aided Therapies in Parkinson's Disease GETM Spanish Registry (DATs‐PD GETM Spanish Registry) [17]. This is an observational, prospective, multicenter, open‐label study including PD patients treated with DATs under routine clinical practice at over 40 centers across Spain. Consecutive PD patients treated with DATs and fulfilling registry inclusion criteria are prospectively enrolled at participating centers. Eligibility criteria for the registry are reflective of routine clinical practice, including: (1) diagnosis of PD according to the Movement Disorders Society (MDS) criteria [18]; (2) initiation of a DAT from January 1, 2024 onwards; (3) voluntary willingness to participate and (4) provision of written informed consent. The study protocol was approved by the local Ethics Committee (Comité de Ética de la Investigación Clínica de Galicia, Spain; 2024/109). Written informed consent was obtained from all participants. The study was conducted in accordance with ICH‐GCP E6 (R2), the Declaration of Helsinki, the Oviedo Convention, and Spanish biomedical research regulations (14/2007). The structured methodology and data collection approach of the DATs‐PD GETM Spanish Registry have also been applied in previous multicenter studies derived from this cohort [3, 17, 19].

For the present study, we included all patients treated with LDp/CDp CSI and at least one follow‐up visit after treatment initiation. All patients were considered to have achieved optimized treatment. Optimization was determined according to the treating neurologist's clinical judgment in routine clinical practice, based on the overall assessment of therapeutic response, including motor fluctuations, motor and non‐motor symptoms, tolerability, patient satisfaction, and expected clinical benefit.

2.2. Assessments

All clinical assessments were performed by movement disorders neurologists. Baseline motor symptom assessments included OFF‐ and ON‐state Unified Parkinson's Disease Rating Scale part III (UPDRS‐III) and Hoehn and Yahr (H&Y). Non‐motor symptom variables included cognitive impairment, hallucinations or other psychotic phenomena, and ICD at the time of initiation of the CSI. The presence or absence of all non‐motor symptoms was assessed via clinical interviews with patients and caregivers conducted by the movement disorders specialist, rather than using standardized rating scales. Cognitive status was established according to clinician judgment based on routine clinical evaluation and/or validated cognitive screening tools including the Montreal Cognitive Assessment (MoCA) or Parkinson's Disease‐Cognitive Rating Scale (PD‐CRS), according to local practice and specialist criteria. Data related to LDp/CDp CSI included the daily infusion rate at the first and last visit recorded, nighttime infusion rate at the first visit, the setting in which therapy optimization was performed, and the number of daily infusion hours. Additional variables included age at treatment initiation, sex, disease duration, levodopa‐equivalent daily dose (LEDD), and use of any oral antiparkinsonian add‐on drug at treatment initiation.

The primary neuropsychiatric AEs evaluated in this study were grouped in: hallucinations, other psychosis‐related AEs or confusion, including both new‐onset and patient‐reported worsening of preexisting symptoms; and ICD‐related AEs, including both new‐onset ICD and patient‐reported worsening of preexisting ICD. As described above for baseline non‐motor symptoms, the presence or absence and severity of all AEs were determined based on the clinical interview conducted by a movement disorders specialist.

2.3. Statistical Analysis

Descriptive statistics were used to summarize demographic and clinical variables. Normality of continuous variables was tested using the Shapiro–Wilk test, and none fulfilled criteria for a normal distribution. Accordingly, continuous variables are reported as median and interquartile range (IQR), and between‐group comparisons were conducted using the Mann–Whitney U test. Categorical variables are presented as counts and percentages, and group differences were assessed using the χ ² test.

Kaplan–Meier survival curves were generated, and univariate and multivariate Cox proportional hazards regression models were fitted to estimate the risk of developing neuropsychiatric AEs. Crude and adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated. Multivariate models adhered to the recommended ratio of approximately one covariate per 10 events. The variables used for model adjustment are specified in the Results section. The proportional hazards assumption was assessed using the Schoenfeld test, confirming the appropriateness of the models.

All statistical analyses were performed using R statistical software (v4.4.2). A two‐sided p value < 0.05 was considered statistically significant for all tests.

3. Results

A total of 282 patients with PD initiated LDp/CDp CSI and were registered in the DATs‐PD GETM Spanish Registry during the study period. Of these, 214 patients from 31 centers had at least one follow‐up visit after treatment initiation and were included in the present analysis. The included patients had a median follow‐up of 163 days (IQR 32–242) after treatment initiation. The median age at therapy initiation was 69 years (IQR 63.5–75) and the median disease duration was 12 years (IQR 8–16). Median time to optimization was 19 days (IQR 7–40.5). At baseline, cognitive impairment was present in 72 patients (35%), hallucinations/psychosis in 53 (25.7%) and ICD in 54 (26.2%). The median daily infusion rate at therapy initiation was 0.34 mL/h (IQR 0.26–0.42), 83.2% of patients were treated with a 24‐h infusion regimen, and initiation was performed predominantly in the outpatient setting (97.2%). Additional baseline characteristics are summarized in Table 1.

TABLE 1.

Characteristics of the cohort (n = 214).

Age	69 (63.5–75)
Biological sex, women	91 (42.5%)
Disease duration, years	12 (8–16)
UPDRS III‐OFF	37 (29.2–45) a
UPDRS III‐ON	19 (11–26) b
H&Y‐OFF	3 (2–4) c
H&Y‐ON	2 (2–3) c
LEDD before initiation, mg	1188 (900–1561)
LEDD at optimization, mg	257 (100–971)
Dopamine agonist intake	87 (40.7%)
Amantadine intake	55 (25.7%)
ICOMT intake	89 (41.6%)
IMAO intake	122 (57%)
Cognitive impairment	72 (34.9%) d
Hallucinations/Psychosis	53 (25.7%) d
ICD	54 (26.2%) d
Daytime infusion rate (first visit), mL/h	0.34 (0.26–0.42)
Daytime infusion rate (last visit), mL/h	0.36 (0.28–0.44)
Nightime infusion rate (first visit), mL/h	0.20 (0.15–0.22)
24‐h infusion	178 (83.2%)
Outpatient initiation	208 (97.2%)
Follow‐up after initiation, days	163 (32–242)

Note: Data are presented as median [IQR] for continuous variables and n (%) for categorical variables.

Abbreviations: H&Y, Hoehn and Yahr scale; ICD, impulse‐control disorder; LEDD, levodopa equivalent daily dose; UPDRS III, Unified Parkinson's Disease Rating Scale Part III.

^a n = 162.

^b n = 166.

^c n = 207.

^d n = 206.

During follow‐up, 41 patients (19.2%) developed at least one neuropsychiatric AE. Visual hallucinations/psychosis were the most frequent manifestations, affecting 34 patients (15.8%). Of these, 22 cases (10.2%) were classified as new‐onset, while 12 patients (5.6%) experienced an exacerbation of preexisting symptoms. Most of these AEs were categorized as mild (19 patients) or moderate (21 patients), and only four patients (1.9%) required device removal. Confusion was reported in only two patients; in one case, it resolved without sequelae, while the other required device removal. A total of nine patients (4.2%) developed new‐onset ICD (three patients) or experienced a worsening of preexisting ICD (six patients). The majority of these ICD‐related AEs were mild (eight patients), and none required device removal.

Patients reporting hallucinations, psychosis, or confusion AE were older (p = 0.018), while no significant differences were observed in disease duration, motor symptom severity, baseline cognitive impairment, hallucinations/psychosis, or other disease or CSI‐related features (Table 2). For patients reporting ICD‐related AE, differences between groups were only found for the presence of ICD at baseline (p = 0.049).

TABLE 2.

Differences between patients with and without neuropsychiatric AEs.

Neuropsychiatric AE	No Hall/Psyc/Conf	Hall/Psyc/Conf	p	No ICD	ICD	p
N	178 (83.2%)	36 (16.8%)		205 (95.8%)	9 (4.2%)
Age	68.5 (62–75)	72.5 (69–76)	0.018	69 (63.2–75)	70 (65–75)	0.975
Biological sex, women	75 (42.1%)	16 (44.4%)	0.944	89 (43.4%)	2 (22.2%)	0.361
Disease duration, years	12 (8–16)	12 (9–15.8)	0.999	12 (8–16)	12 (11–16)	0.864
UPDRS III‐OFF	37 (29–47)	37 (30.8–41.5)	0.667	37 (30–45.5)	37 (31–39.5)	0.717
H&Y‐OFF	3 (2–4)	3 (2.8–4)	0.609	3 (2–4)	3 (2.8–3.2)	0.705
LEDD before initiation, mg	1203 (925–1565)	1100 (881–1447)	0.320	1188 (925–1555)	920 (700–1675)	0.394
LEDD at optimization, mg	260 (100–1013)	188 (100–758)	0.291	257 (100–988)	150 (50–660)	0.379
Dopamine agonist intake	73 (41%)	14 (38.9%)	0.960	85 (41.5%)	2 (22.2%)	0.422
Amantadine intake	47 (26.4%)	8 (22.2%)	0.669	53 (25.8%)	2 (22.2%)	1.000
ICOMT intake	76 (42.7%)	13 (36.1%)	0.465	88 (42.9%)	1 (11.1%)	0.086
IMAO intake	100 (56.2%)	22 (61.1%)	0.859	116 (56.6%)	6 (66.7%)	0.745
Cognitive impairment	59 (34.3%) a	13 (38.2%) b	0.808	70 (35.4%)	2 (25%)	0.823
Hallucinations/Psychosis	41 (23.8%) a	12 (35.3%) b	0.237	51 (25.8%)	2 (25%)	1.000
ICD	46 (26.7%) a	8 (23.5%) b	0.860	49 (24.7%)	5 (62.5%)	0.049
Daytime infusion rate (first visit), mL/h	0.34 (0.26–0.42)	0.32 (0.27–0.42)	0.735	0.34 (0.26–0.42)	0.34 (0.10–0.43)	0.709
Daytime infusion rate (last visit), mL/h	0.35 (0.28–0.44)	0.37 (0.27–0.45)	0.690	0.36 (0.28–0.44)	0.35 (0.29–0.47)	0.951
Nightime infusion rate (first visit), mL/h	0.20 (0.15–0.22)	0.20 (0.15–0.22)	0.788	0.16 (0–0.20)	0.20 (0.15–0.22)	0.338
24‐h infusion	148 (83.1%)	30 (83.3%)	0.755	170 (89.5%)	8 (100%)	1.000
Outpatient initiation	173 (97.2%)	35 (97.2%)	1.000	200 (97.6%)	8 (88.9%)	0.230

Note: Data are presented as median [IQR] for continuous variables and n (%) for categorical variables. Bold values indicate statistical significance.

Abbreviations: AE, adverse event; H&Y, Hoehn and Yahr scale; Hall/Psyc/Conf, hallucinations, psychosis or confusion; ICD, impulse‐control disorder; LEDD, levodopa equivalent daily dose; UPDRS III, Unified Parkinson's Disease Rating Scale Part III.

^a Number of patients included in the calculation of the corresponding variables n = 172.

^b Number of patients included in the calculation of the corresponding variables n = 34.

Kaplan–Meier estimates showed no differences in AE‐free survival for hallucinations/psychosis/confusion AEs between patients with and without prior hallucinations/psychosis or cognitive impairment (log‐rank p = 0.44) (Figure 1). In contrast, for ICD, Kaplan–Meier estimates revealed a significant difference in time to ICD‐related AEs between patients with and without baseline ICD (log‐rank p = 0.019). The majority of hallucinations/psychosis/confusion AEs (80.6%) and ICD‐related adverse events (88.9%) occurred beyond 1 month after treatment initiation.

FIGURE 1.

Kaplan–Meier curves for time to neuropsychiatric adverse events after LDp/CDp CSI initiation. The top graph shows the time to hallucinations, psychosis, or confusion‐related adverse event, comparing patients with and without previous cognitive impairment or hallucinations (p = 0.44). The bottom graph displays the time to the development of impulse‐control disorders (ICD)‐related adverse event, showing a significant difference between patients with and without a prior history of ICD (p = 0.019). AE, Adverse event; CI, Cognitive impairment; Hall/Psyc/Conf: Hallucinations, psychosis, or confusion; ICD, Impulse‐control disorder.

In both the univariate and adjusted Cox proportional hazards analyses, no variables were significantly associated with the development of hallucinations/psychosis/confusion following LDp/CDp CSI initiation, whereas baseline ICD was associated with an increased risk of ICD‐related AEs (Table 3).

TABLE 3.

Univariate and multivariate Cox proportional hazards regression estimating the risk of neuropsychiatric AEs.

	Crude HR (95% CI)	p	Adjusted HR (95% CI)	p
Hallucinations/Psychosis/Confusion
Age	1.03 (0.99–1.08)	0.098	1.03 (0.99–1.07)	0.182
Disease duration	1.00 (0.96–1.04)	0.882	—	—
UPDRS III‐OFF	0.99 (0.96–1.02)	0.454	—	—
LEDD before initiation	1.00 (1.00–1.00)	0.241	—	—
Dopamine agonist intake	1.13 (0.58–2.22)	0.715	—	—
Cognitive impairment	1.22 (0.61–2.44)	0.570	1.09 (0.51–2.35)	0.828
Hallucinations/Psychosis	1.81 (0.90–3.66)	0.098	1.70 (0.79–3.68)	0.176
Daytime infusion rate (first visit)	1.30 (0.11–15.17)	0.833	—	—
Daytime infusion rate (last visit)	0.65 (0.05–9.04)	0.748	—	—
24‐h infusion	1.13 (0.40–3.22)	0.820	—	—
ICD
Age	0.98 (0.91–1.05)	0.499	—	—
Disease duration	0.98 (0.90–1.06)	0.625	—	—
UPDRS III‐OFF	0.99 (0.93–1.05)	0.717	—	—
LEDD before initiation	1.00 (1.00–1.00)	0.321	—	—
Dopamine agonist intake	2.50 (0.52–11.97)	0.256	2.28 (0.46–11.36)	0.314
ICD	4.72 (1.13–19.77)	0.034	4.77 (1.14–20.01)	0.032
Daytime infusion rate (first visit)	0.39 (0.01–77.61)	0.729	—	—
Daytime infusion rate (last visit)	0.28 (0.01–59.50)	0.644	—	—

Note: Bold values indicate statistical significance.

Abbreviations: CI, confidence interval; HR, hazard ratio; ICD, impulse‐control disorder; LEDD, levodopa equivalent daily dose; UPDRS III, Unified Parkinson's Disease Rating Scale Part III.

4. Discussion

In this study, 19.2% of patients developed neuropsychiatric AEs during follow‐up after initiation of LDp/CDp CSI. Most events were mild‐to‐moderate in severity, predominantly manifesting as hallucinations and/or psychosis, and permanent treatment discontinuation was required in only five patients (2.3%). No baseline characteristics were associated with the development of hallucinations/psychosis/confusion, while only baseline ICD was associated with an increased risk of ICD‐related AEs. Overall, these findings suggest that significant neuropsychiatric AEs after LDp/CDp CSI initiation are relatively uncommon and rarely lead to treatment discontinuation.

To date, evidence from controlled trials suggests that LDp/CDp CSI is associated with a relatively low frequency of significant neuropsychiatric AEs, whereas real‐world clinical data have been more heterogeneous and raised greater concern regarding these potential complications. Soileau et al. [5] reported hallucinations/psychosis in 15% of participants, with most events classified as nonserious and only one discontinuation (1.3%). Similarly, Aldred et al., in an open‐label phase 3 study, observed hallucinations and psychosis in 19.7% of patients, predominantly mild–moderate, with treatment discontinuation in 4.1% [8]. Interestingly, the clinical trial evaluating ND0612 reported a markedly lower rate of neuropsychiatric AEs (< 2%), a finding that warrants confirmation in future real‐world studies to determine whether this formulation may have a more favorable safety profile [7]. ICDs were not reported in clinical trials of LDp/CDp CSI, with only one patient (1%) developing ICD in the ND0612 trial [5, 7, 8]. Real‐world cohorts of LDp/CDp CSI are more heterogeneous, with some including a relatively high proportion of patients with prior psychosis (62%) or cognitive impairment (> 50%) [15, 16]. In these studies, some cohorts reported rates comparable to our cohort, with hallucinations occurring in 17%–21% of patients and discontinuation rates of 7.5% due to these events [14, 20]. In contrast, others have described substantially higher rates of neuropsychiatric AEs (47%–56%), with low rates of discontinuation, whereas some have reported lower overall rates (8%) but with a high proportion (75%) requiring treatment discontinuation [13, 15, 16]. Data regarding ICD‐related AEs in clinical practice remain limited, with only one study reporting a 5.6% incidence [14].

In our cohort, hallucinations/psychosis/confusion AEs occurred in 16.8% of patients and ICD‐related AE in 4.2%. In both cases, events were predominantly mild–moderate and rarely led to device discontinuation, which was required in 2.3% of the overall cohort due to hallucinations/psychosis/confusion and in none of the patients due to ICDs. No other neuropsychiatric AEs were observed, such as low mood or increased aggression reported in isolated cases in previous studies [14]. Taken together, our findings align with both clinical trial and real‐world evidence, indicating that neuropsychiatric AEs, particularly hallucinations and psychosis, occur in a clinically relevant proportion of patients treated with LDp/CDp CSI. However, these events are typically mild‐to‐moderate, manageable in routine practice and only occasionally require treatment discontinuation.

The development of hallucinations/psychosis and ICDs in the general PD population has been associated with several demographic and disease‐related factors including age, disease duration, or exposure to dopaminergic therapies, particularly dopamine agonists [21]. However, given the recent approval of LDp/CDp CSI, data on risk factors in this specific population are still limited. Neither clinical trials nor real‐world studies have systematically evaluated the relationship between patient characteristics and the development of neuropsychiatric symptoms in this setting, and the available data are largely restricted to descriptive reports. In our cohort with 25% of baseline hallucinations/psychosis, 35% of cognitive impairment, and 26% of ICD, we did not identify any baseline characteristics significantly associated with the development of incident hallucinations/psychosis, and only the presence of ICD at baseline was associated with an increased risk of subsequent ICD worsening. These findings support the use of LDp/CDp CSI even in patients with a history of cognitive impairment, psychosis, or ICD, or in those receiving concomitant dopaminergic therapies. Nevertheless, given the well‐established neuropsychiatric risk profile of dopaminergic treatments, individualized therapeutic decisions and close monitoring remain essential. Further studies are warranted to determine the generalizability of these findings across broader populations.

When contextualizing these findings within the field of infusion‐based therapies for advanced PD, our results appear broadly comparable, while acknowledging the inherent limitations of cross‐study comparisons, including differences in patient selection, baseline vulnerability, and study design. Evidence from CSAI studies provides a relevant benchmark. In the TOLEDO trial, hallucinations, psychosis, or confusion were reported in 7.5% of patients, leading to discontinuation of CSAI in 1.8%, while ICDs occurred in 5.6% and did not require device removal [22]. In a retrospective cohort evaluating CSAI treatment, in which all participants had cognitive impairment and 71% had visual hallucinations at baseline, 20% experienced worsening hallucinations during treatment, and only 2% discontinued therapy due to neuropsychiatric complications [23]. These observations suggest that, although behavioral and cognitive symptoms warrant careful monitoring, they should not be considered absolute contraindications to CSAI when appropriately managed [24]. Comparisons with LCIG similarly place our findings within the expected range for continuous dopaminergic delivery strategies. Although results across LCIG studies have shown some heterogeneity, most reports describe relative stability of neuropsychiatric symptoms over time, with no consistent evidence of clinically meaningful worsening following therapy initiation [24, 25, 26, 27]. Although head‐to‐head comparative studies are still needed, our findings suggest that, from a neuropsychiatric perspective, LDp/CDp CSI appears comparable to LCIG while being less invasive. In addition, LDp/CDp CSI may offer certain advantages over CSAI, including a greater potential for monotherapy and a more favorable AE profile [28].

Several limitations should be acknowledged. First, non‐motor symptoms were assessed through structured clinical evaluation rather than uniformly administered standardized rating scales across centers, which may have introduced inter‐rater variability and measurement bias. This likely represents a minor limitation for hallucinations/psychosis or ICD, given that assessments were conducted by movement disorder specialists and that some official scales, such as the MDS‐UPDRS Part I, often rely on single‐item questions. However, it is a more relevant limitation for cognitive impairment, as clinician and patient impressions are far from the gold standard for diagnosing PD‐associated cognitive impairment, although previous studies have reported high correlations [29]. Second, the low event rate for ICDs restricted our statistical power. Additionally, individual ICD subtypes were not specified in our registry, limiting our ability to differentiate between them and their respective clinical trajectories. Third, the real‐world clinical nature of this study introduced inherent challenges, including staggered recruitment, potential selection bias and variable follow‐up intervals across patients. Although the registry was designed to consecutively include patients treated with DATs in participating centers, complete capture of all patients receiving LDp/CDp CSI cannot be fully guaranteed in routine clinical practice. Although Cox regression partially mitigates differences in follow‐up duration, heterogeneity in follow‐up may still have influenced the detection and timing of neuropsychiatric AEs. Furthermore, a degree of reporting bias cannot be entirely excluded, as mild or transient neuropsychiatric symptoms may have been underreported by both patients and clinicians. In addition, because medication adjustments were recorded according to predefined registry visits rather than continuously throughout follow‐up, detailed longitudinal analyses evaluating the relationship between dynamic changes in dopaminergic load and neuropsychiatric AEs were limited. Likewise, although information regarding nocturnal infusion flow at treatment initiation and the use of 24‐h infusion regimens was available, longitudinal data on nighttime flow adjustments during follow‐up or at the time neuropsychiatric AEs occurred were not systematically collected. Consequently, the potential influence of dynamic nocturnal infusion changes on neuropsychiatric outcomes could not be adequately assessed. Fourth, the limitations inherent to the methodology used, namely the extraction of data from a registry, such as incomplete recording of some variables in certain cases. Finally, the introduction of a novel therapy, alongside heightened awareness of potential neuropsychiatric AEs, may have introduced selection bias, favoring patients with fewer baseline neuropsychiatric symptoms. Conversely, this study represents one of the largest real‐world clinical cohorts evaluating PD patients treated with LDp/CDp CSI providing highly valuable longitudinal insights into the neuropsychiatric AE profile of this therapy in routine clinical practice.

In conclusion, neuropsychiatric AEs occur in a clinically relevant proportion of patients treated with LDp/CDp CSI. However, these events are generally mild‐to‐moderate, well tolerated, and only rarely require treatment discontinuation. Except for baseline ICD, cognitive and other neuropsychiatric features were not associated with an increased risk of neuropsychiatric AEs. Overall, these findings suggest that LDp/CDp CSI is generally associated with a manageable neuropsychiatric AE profile and highlight the importance of individualized clinical monitoring over overly restrictive patient selection.

Author Contributions

Ana Vinagre‐Aragón: resources, writing – review and editing. Nuria Caballol: resources, writing – review and editing. Elisabet Mondragón: writing – review and editing, resources. Guillermo González‐Ortega: writing – review and editing, resources, software. Javier Ruiz‐Martinez: funding acquisition, writing – review and editing, supervision, resources. Lara Pardina Vilella: writing – review and editing, resources. Nuria López‐Ariztegui: writing – review and editing, resources. Iria Cabo: resources, writing – review and editing. Inés Legarda: writing – review and editing, resources. Arnau Puig‐Davi: methodology, writing – review and editing, formal analysis, data curation. Jon Rodriguez‐Antiguedad: conceptualization, investigation, writing – original draft, methodology, writing – review and editing, formal analysis. Juan Carlos Romero Fábrega: writing – review and editing, resources. Rocío García Ramos: writing – review and editing, resources. Marta Blázquez‐Estrada: writing – review and editing, resources. María Álvarez Sauco: writing – review and editing, resources. Silvia Martí Martinez: writing – review and editing, resources. Diego Santos‐García: writing – review and editing, resources, supervision, project administration, validation. Pablo Mir: writing – review and editing, resources. David Campo‐Caballero: conceptualization, writing – original draft, formal analysis, writing – review and editing, methodology, investigation. Pilar Sánchez‐Alonso: writing – review and editing, resources.

Funding

No commercial funding was received specifically for this study. This project has been developed with resources obtained by the Degen Foundation in collaboration with the Spanish Society of Neurology and from Carlos III Health Institute as a result of the Río Hortega grant 2024 (CM24/00220).

Ethics Statement

The project is being conducted in accordance with the ICH Good Clinical Practice version 6 Revision 2 standard, the fundamental ethical principles established in the Declaration of Helsinki and the Oviedo Convention, as well as the Spanish legal requirements for biomedical research (Biomedical Research Law 14/2007). The Project has been approved on 02/APR/2024 by the IRB “Comité de Etica de la Investigación Clínica de Galicia from Spain” with code number 2024/109. Written informed consents from all participants in this study will be obtained. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines. The protocol, statistical analysis plan and unidentified participant data will be available on request.

Conflicts of Interest

D.C.‐C. has received honoraria for educational presentations by Stada and Zambon and grants from III Carlos Health Institute as a result of the Río Hortega grant 2024 (CM24/00220). J.R.‐A. has nothing to report. A.P.‐D. has nothing to report. J.R.‐M. has nothing to report. A.V.‐A. has nothing to report. E.M. has nothing to report. L.P.V. has nothing to report. N.L.‐A. has received honoraria for educational presentations and advice service or travel grants by Abbvie, Italfarmaco, Stada, Lundbeck, UCB, Esteve, Abbott, Zambon, and Bial. N.C. has received honoraria for advisory services and educational presentations, and support for conference attendance, from Italfármaco, Orion Pharma, Abbvie, Stada, KRKA, and Zambon. G.G.‐O. has received honoraria for educational purposes from ABBIE, Zambon, Bial, Esteve, and Italfarmaco. I.C. has received honoraria for educational presentations and advice service by Abbvie, UCB Pharma, Zambon, Bial, Italfarmaco, Esteve, Stada, and Orion Pharma. R.G.R. has received honoraria and grants for lecturing, advisory services from Abbvie, Zambon, Bial, Merk, and Stada. I.L. has received honoraria for educational presentations and advice service by Abbvie, UCB Pharma, Zambon, Bial, and Teva. M.B.‐E. has nothing to report. P.S.‐A. has received honoraria for educational presentations and advice service by Abbvie, UCB Pharma, Lundbeck, KRKA, Zambon, Bial, and Teva. P.M. has received financial support from the Spanish Ministry of Science and Innovation (RTC2019‐007150‐1), the Instituto de Salud Carlos III‐Fondo Europeo de Desarrollo Regional (ISCIII‐FEDER) (PI16/01575, PI18/01898, PI19/01576, PI20/00613, PI21/01875, PI22/01704, PI23/00512), the Consejería de Economía, Innovacion, Ciencia y Empleo de la Junta de Andalucía (CVI‐02526, CTS‐7685), the Consejería de Salud y Bienestar Social de la Junta de Andalucía (PI‐0471‐2013, PE‐0210‐2018, PI‐0459‐2018, PE‐0186‐2019), the Consejería de Transformación Económica, Industria, Conocimiento y Universidades de la Junta de Andalucía (PY20_00896), and support for attending meetings and/or travel or honorarium for lecturing from Abbott, Allergan, Abbvie, Bial, Britannia, Italfarmaco, Merz, UCB, Teva, and Zambon. J.C.R.F. has nothing to report. S.M.M. has nothing to report. M.A.S. has nothing to report. D.S.‐G. has received honoraria for educational presentations and advice service by Abbvie, UCB Pharma, Lundbeck, KRKA, Zambon, Bial, Italfarmaco, Teva, Archímedes, Esteve, Stada, Orion Pharma, financial compensation for INIBIC/Fundación Professor Novoa Santos in relation to activity as PI in phase 2 to phase 4 clinical trials for Parkinson's disease and other movement disorders, compensation for Fundación Degen in relation to different projects, and grants from the “Fundación Professor Novoa Santos” as a result of the “CONVOCATORIA DE AYUDAS PARA LA REALIZACIÓN DE PROYECTOS DE INVESTIGACIÓN PARA GRUPOS EMERGENTES Y ASOCIADOS DEL INIBIC (2023/2024)”.

Appendix A.

Nuria López Ariztegui (Hospital Universitario de Toledo), Maria Isabel Morales Casado (Hospital Universitario de Toledo), Guillermo Tabar Comellas (Hospital Universitario de Toledo), Asunción Ávila Rivera (Complex Hospitalari Moisès Broggi), Núria Caballol Pons (Complex Hospitalari Moisès Broggi), Anna Planas‐Ballvé (Complex Hospitalari Moisès Broggi), Alejandro Peral Quirós (Complex Hospitalari Moisès Broggi), Dolors Lozano (Complex Hospitalari Moisès Broggi), Diego Santos García (Complejo Hospitalario Universitario de A Coruña), Jose Manuel Paz González (Complejo Hospitalario Universitario de A Coruña), Carlos Cores Bartolomé (Complejo Hospitalario Universitario de A Coruña), Lucia Belen Samaniego Vinueza (Complejo Hospitalario Universitario de A Coruña), Ángela Solleiro Vidal (Complejo Hospitalario Universitario de A Coruña), Iria Cabo (Complejo Hospitalario Universitario de A Coruña), Alejandro Rivero de Aguilar Pensado (Complejo Hospitalario Universitario de A Coruña), Rocío García Ramos (Hospital Clínico Universitario San Carlos), Ana Fernández Revuelta (Hospital Clínico Universitario San Carlos), Eva López Valdés (Hospital Clínico Universitario San Carlos), Ana Aldaz (Hospital Clínico Universitario San Carlos), Carmen Ribacoba Díaz (Hospital Clínico Universitario San Carlos), Inés Legarda Ramírez (Hospital Universitario Son Espases), Bàrbara Vives Pastor (Hospital Universitario Son Espases), María Fuensanta Valero García (Hospital Universitario Son Espases), Marta Blázquez Estrada (Hospital Universitario Central de Asturias), Esther Suárez San Martín (Hospital Universitario Central de Asturias), Ciara García Fernández (Hospital Universitario Central de Asturias), Pilar Tartiere (Hospital Universitario Central de Asturias), Patricia Prendes Fernández (Hospital Universitario Central de Asturias), Pilar Sánchez‐Alonso (Hospital Universitario Puerta de Hierro), Sabela Novo Ponte (Hospital Universitario Puerta de Hierro), Elisa Gamo González (Hospital Universitario Puerta de Hierro), Raquel Martín García (Hospital Universitario Puerta de Hierro), Pablo Mir Rivera (Hospital Universitario Virgen del Rocío), Laura Muñoz Delgado (Hospital Universitario Virgen del Rocío), Astrid Daniela Adarmes Gómez (Hospital Universitario Virgen del Rocío), Elena Ojeda Lepe (Hospital Universitario Virgen del Rocío), Silvia Jesús Maestre (Hospital Universitario Virgen del Rocío), Daniel Macías García (Hospital Universitario Virgen del Rocío), Fátima Carrillo García (Hospital Universitario Virgen del Rocío), Ana María Castellano Guerrero (Hospital Universitario Virgen del Rocío), Manuela San Eufrasio Martínez (Hospital Universitario Virgen del Rocío), Cristina Pérez Calvo (Hospital Universitario Virgen del Rocío), María López González (Hospital Universitario Virgen del Rocío), Andrés García Herruzo (Hospital Universitario Virgen del Rocío), Cristina Gómez Rapela (Hospital Universitario Virgen del Rocío), Lorena Garrote Espina (Hospital Universitario Virgen del Rocío), Natalia González Rojas (Hospital Universitario Virgen del Rocío), Juan Carlos Romero Fábrega (Hospital Universitario Virgen de las Nieves), Francisco Escamilla Sevilla (Hospital Universitario Virgen de las Nieves), Lucía Triguero Cueva (Hospital Universitario Virgen de las Nieves), Carlos Javier Madrid Navarro (Hospital Universitario Virgen de las Nieves), Silvia Martí Martínez (Hospital General Universitario de Alicante), María Álvarez Sauco (Hospital General Universitario de Alicante), Eric Freire Álvarez (Hospital General Universitario de Alicante), Jessica González Ardura (Hospital de Cabueñes), Belén Castaño García (Hospital de Cabueñes), Antonio Sánchez Rodríguez (Hospital de Cabueñes), Déborah Alonso Modino (Hospital Universitario La Candelaria), Jesús Norelis Lorenzo Brito (Hospital Universitario La Candelaria), Maria Esther Rojas Pérez (Hospital Universitario La Candelaria), Carolina Hernández Javier (Hospital Universitario La Candelaria), Jorge Hernández Vara (Hospital Universitario Vall d'Hebron), Maria Victoria González (Hospital Universitario Vall d'Hebron), Sara Belmonte (Hospital Universitario Vall d'Hebron), Débora María Cerdán Santacruz (Hospital General de Segovia), Amelia Mendoza Rodríguez (Hospital General de Segovia), Ana Castrillo Sanz (Hospital General de Segovia), Lorena Caballero Sánchez (Hospital General de Segovia), Claudia Gómez López de San Román (Hospital General de Segovia), Álvaro Sánchez‐Ferro (Hospital Universitario 12 de Octubre), Pablo Rábano Suárez (Hospital Universitario 12 de Octubre), Antonio Méndez Guerrero (Hospital Universitario 12 de Octubre), Daniel Pérez Rangel (Hospital Universitario 12 de Octubre), Jesús Ramírez (Hospital Universitario 12 de Octubre), Cristina del Álamo Criado (Hospital Universitario 12 de Octubre), Javier Ruíz Martínez (Hospital Universitario Donostia), Elisabet Mondragón Rezola (Hospital Universitario Donostia), Ana Vinagre Aragón (Hospital Universitario Donostia), Lara Pardina Vilella (Hospital Universitario Donostia), David Campo Caballero (Hospital Universitario Donostia), Ana Belén Perona Moratalla (Complejo Hospitalario Universitario de Albacete), Gustavo Fernández‐Pajarín (Complejo Hospitalario Universitario de Santiago de Compostela), Maria Pilar Gil Villar (Hospital Universitari Arnau de Vilanova), Agustín Sancho Saldana (Hospital Universitari Arnau de Vilanova), Laura Quibus Requena (Hospital Universitari Arnau de Vilanova), Ángela Monterde Ortega (Hospital Universitario Joan XXIII), Juan Carlos Gómez Esteban (Hospital Universitario de Cruces), Tamara Fernández Valle (Hospital Universitario de Cruces), Marta Ruiz López (Hospital Universitario de Cruces), Beatriz Tijero Merino (Hospital Universitario de Cruces), Susana Cantarero Duque (Hospital Universitario de Móstoles), Guillermo González‐Ortega (Hospital Universitario de Móstoles), Esther Cubo Delgado (Hospital Universitario de Burgos), Beatriz Alegre García (Hospital Universitario de Burgos), Lydia López Manzanares (Hospital Universitario La Princesa), Inés Muro (Hospital Universitario La Princesa), Elena Casas Peña (Hospital Universitario La Princesa), Pablo Lorenzo Barreto (Hospital Universitario La Princesa), Emma Cañada Lahoz (Hospital Universitario La Princesa), Sara Lozano Veiga (Hospital Universitario La Princesa), Lydia Vela (Hospital Universitario Fundación de Alcorcón), Irene Martínez‐Torres (Hospital Universitari i Politècnic La Fe), Carlos Morata‐Martínez (Hospital Universitari i Politècnic La Fe), Raquel Baviera‐Muñoz (Hospital Universitari i Politècnic La Fe), Marina Campins Romeu (Hospital Universitari i Politècnic La Fe), Isabel Sastre Bataller (Hospital Universitari i Politècnic La Fe), Elena Navarro Mocholí (Hospital Universitari i Politècnic La Fe), Roberto López Blanco (Hospital Universitario Severo Ochoa), Sonia Escalante Arroyo (Hospital Virgen de la Cinta), Tania Delgado Ballestero (Parc Taulí Hospital Universitari).

Campo‐Caballero D., Rodriguez‐Antiguedad J., Puig‐Davi A., et al., “Neuropsychiatric Adverse Events Associated With Foslevodopa/Foscarbidopa Continuous Subcutaneous Infusion in Clinical Practice: A Multicenter Study,” European Journal of Neurology 33, no. 7 (2026): e70693, 10.1111/ene.70693.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.