Characteristics and Hospitalization Details of Patients with Advanced Parkinson’s Disease Receiving Continuous Subcutaneous Infusion of Foslevodopa/Foscarbidopa: Real-World First-Year Descriptive Study from Japan Authors: Toru Baba, Toshiki Nozaki, Shunji Toya, Hideyuki Migita, Kairi Ri, Atsushi Takeda Published: 2026 ABSTRACT IntroductionThe real-world characteristics of patients with advanced Parkinson’s disease (PD) who receive foslevodopa/foscarbidopa (LDp/CDp) continuous subcutaneous infusion therapy (CSCI) remain poorly understood. This study aimed to describe the characteristics and hospitalization details of patients with advanced PD hospitalized for LDp/CDp CSCI in Japan.MethodsThis study was a retrospective descriptive analysis conducted using a hospital-based database provided by Medical Data Vision, Co. Ltd. Patients with advanced PD who underwent LDp/CDp CSCI during hospitalization between July 2023 and June 2024 were enrolled. The index date was defined as the date of the first LDp/CDp prescription. Baseline patient characteristics, levodopa-equivalent dose (LED), number of pills/transdermal patches, medication classes for concomitant PD medications, and non-motor symptom medications assessed within 4 weeks prior to and including the index date, along with hospitalization details, were collected.ResultsWe enrolled 96 patients with a mean age of 65.6 years [standard deviation (SD), 10.2], of whom 58 (60.4%) were female. The median Barthel Index was 90.0 [interquartile range (IQR): 44.2], and the hospitalization duration was 15.0 days (IQR: 9.0). Eighteen patients (18.8%) discontinued LDp/CDp CSCI during hospitalization. The mean LED of any PD medications prior to LDp/CDp CSCI initiation, the mean number of PD medication classes, and the daily pill/transdermal patch count for PD medications were 1,111.0 mg/day (SD 575.7), 3.9 (SD 1.3), and 10.6 (SD 5.0), respectively. Except for levodopa (L-Dopa), dopamine agonists (78.1%) were the most commonly used concomitant PD medications.ConclusionsIn real-world clinical practice in Japan, patients with LDp/CDp CSCI tend to be younger and have preserved activities of daily living. Our results implied that patients are often managed with multiple concomitant PD medications rather than with L-Dopa dose titration prior to initiation. Further research is warranted to maximize the benefits of LDp/CDp CSCI and assess its real-world impact on pill burden.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s40120-026-01016-9. Key Summary Points Why carry out this study? The integration of foslevodopa/foscarbidopa (LDp/CDp) continuous subcutaneous infusion therapy (CSCI) for advanced Parkinson’s disease (PD) in routine clinical practice is warranted because it increases on-time without troublesome dyskinesia and reduces off-time and pill burden.The real-world characteristics of patients with advanced PD receiving LDp/CDp CSCI remain poorly understood. What was learned from the study? In real-world Japanese settings, patients with advanced PD who received LDp/CDp CSCI tended to be younger and have preserved activities of daily living.Although caution is warranted in interpreting these findings due to the descriptive nature of the study, patients with advanced PD in Japanese real-world clinical settings tend to be managed with concomitant PD medications rather than dose titration of levodopa, particularly dopamine agonists, compared with other countries. Why carry out this study? The integration of foslevodopa/foscarbidopa (LDp/CDp) continuous subcutaneous infusion therapy (CSCI) for advanced Parkinson’s disease (PD) in routine clinical practice is warranted because it increases on-time without troublesome dyskinesia and reduces off-time and pill burden.The real-world characteristics of patients with advanced PD receiving LDp/CDp CSCI remain poorly understood. What was learned from the study? In real-world Japanese settings, patients with advanced PD who received LDp/CDp CSCI tended to be younger and have preserved activities of daily living.Although caution is warranted in interpreting these findings due to the descriptive nature of the study, patients with advanced PD in Japanese real-world clinical settings tend to be managed with concomitant PD medications rather than dose titration of levodopa, particularly dopamine agonists, compared with other countries. Introduction Parkinson’s disease (PD), a progressive neurological disorder, is the second most common neurodegenerative disorder after Alzheimer’s disease [1, 2]. In Japan, approximately 290,000 patients have PD, with a prevalence of 100–150 per 100,000 individuals [3]. Levodopa (L-Dopa), commonly administered along with dopa decarboxylase inhibitors, is the primary treatment for PD. The therapeutic window of L-Dopa narrows with disease progression, causing off episodes when the medication wears off; thus, patients with advanced PD generally require more frequent administration of dopaminergic agents [4]. Patients usually also take adjunctive PD medications such as monoamine oxidase-B (MAO-B) and catechol-O-methyltransferase (COMT) inhibitors as well as medications for non-motor symptoms and comorbidities; thus, a pill burden is a major concern for patients with advanced PD. One prior systematic review showed that 40% of patients with advanced PD took more than five medications every day, whereas 18% took more than ten [5]. The pill burden in patients with advanced PD is associated with nonadherence [6], which increases the risk of an off-episode, deteriorating the patient’s quality of life (QoL) [7], and increasing caregiver burden [8]. Device-aided therapy (DAT) can reduce this burden [9]. Conventional DAT includes deep brain stimulation (DBS), levodopa–carbidopa intestinal gel therapy (LCIG), both of which require surgical intervention as well as continuous subcutaneous apomorphine infusion (CSAI), which does not require surgical intervention; however, CSAI is not available in Japan. Foslevodopa/foscarbidopa (LDp/CDp), which has been available in Japan since July 2023, is a soluble formulation of L-Dopa and carbidopa prodrugs delivered as a continuous subcutaneous infusion (CSCI) via an infusion set connected to a portable pump. LDp/CDp CSCI is markedly less invasive than DBS and LCIG. The efficacy and safety of LDp/CDp CSCI were first reported in a randomized controlled trial (RCT) in 2022, which showed an increase in on-time without troublesome dyskinesia and reduced off-time [10]. Another single-arm open-label study showed that this novel treatment may reduce the pill burden in patients with advanced PD, showing that the percentage of patients using two or more concomitant PD medications decreased from 42.6% in 1 week to 35.0% during the 5–52 weeks of the maintenance period [11]. This evidence from clinical trials indicates that the integration of LDp/CDp CSCI into routine clinical practice is warranted. Evidence of the real-world utilization patterns of LDp/CDp CSCI remains limited because of its recent introduction. In general, clinical trial populations tend to overrepresent healthy, young, white, and male individuals; this tendency has also been observed in PD studies [12–14]. General considerations for candidate selection for the use of LDp/CDp CSCI have been published, recommending it for patients with preserved cognitive function, who can manage and maintain the LDp/CDp delivery system, have no severe psychiatric symptoms, and are free from any dermatological and immunological decompensation that would make CSCI difficult [15]. To our knowledge, only two European observational studies have investigated the real-world characteristics of patients undergoing LDp/CDp CSCI [16, 17]. However, limited evidence is available in Japan [18]. Although these reports provide valuable insights, further large-scale studies are warranted. Patients who use and benefit from LDp/CDp CSCI in routine clinical practice may differ from the clinical trial populations; therefore, a real-world descriptive study is required. Furthermore, although the pill burden for patients with advanced PD may be particularly high in Japan, few studies have investigated this issue [19]. The introduction of LDp/CDp CSCI may alleviate pill burden in Japanese patients with advanced PD. Thus, describing baseline pill burden is necessary to quantify the potential pill burden-alleviating impact of this new treatment. Consequently, this study aimed to describe the patient characteristics, hospitalization, and concomitant medication information for patients with LDp/CDp CSCI introduced during hospitalization in Japan using claims data. Methods Study Design and Data Source This retrospective descriptive study used data from the Medical Data Vision (MDV) database, a large-scale administrative claims database in Japan. The MDV database comprises anonymized inpatient and outpatient data from approximately 515 acute hospitals, representing 29% of all acute-care hospitals in Japan, and encompasses over 48 million patient records as of June 2024. Hospitals included in the MDV dataset use a Diagnosis Procedure Combination (DPC) claims-based bundled payment system that enables patient tracking based on diagnoses and procedures [20]. Anonymized data in the database included demographic characteristics, diagnoses, treatments, procedures, prescriptions, and hospitalizations. Details of the MDV database have been described previously [21, 22]. Permission to use data from the MDV database was obtained before conducting the analyses. Study Population We identified patients diagnosed with PD (ICD-10 code G20) who initiated LDp/CDp CSCI between July 2023 and June 2024, and who were presumed to have advanced PD according to treatment indications. The initiation of LDp/CDp CSCI was identified using prescription records; details are shown in Supplementary Table S1. The index date was defined as the first recorded prescription date of LDp/CDp CSCI. Patients were excluded if they met any of the following criteria: (1) patients who did not have prescription records for any PD medication before the index date, as it is likely that these patients had previously been treated in a hospital not included in the MDV database and the absence of such data may introduce bias; (2) patients who were not continuously enrolled for ≥ 4 weeks before the index date; and (3) patients who had an LDp/CDp prescription initiated without any hospitalization record. In Japan, LDp/CDp CSCI was introduced during hospitalization in the study period (between July 2023 and June 2024), and LDp/CDp CSCI initiation without a hospitalization record lacks reliability. The patient population was stratified into subgroups depending on whether the patient had discontinued LDp/CDp CSCI treatment during hospitalization or had undergone DBS. Variables and Outcomes The patient characteristics included age, sex, height, weight, body mass index (BMI), comorbidities, activities of daily living (ADL), and hospitalization duration. ADL was assessed using the Barthel Index (BI), while comorbidities were assessed using the Charlson Comorbidity Index (CCI). The hospitalization duration related to LDp/CDp CSCI initiation was obtained from the difference between the dates of admission and discharge from the hospitalization records containing the index date. Information on concomitant PD medications and non-motor symptom treatments included the number and classes of PD medications and their L-Dopa equivalent doses (LED). Variables were measured during the baseline period, defined as 4 weeks prior to and including the index date. Because a transdermal patch could also increase the patient burden, both pill and transdermal patch medications were included. While the daily consumption of tablets could be directly identified from the database, only the total prescribed amount was available for patches; therefore, the daily patch usage was determined based on the assumption that one patch was used per day. When multiple records for the same product overlapped, each record was considered one patch per day, resulting in multiple patches per day for that period. It should also be noted that all prescription records for L-Dopa were included in the analysis, although we recognize that L-Dopa is sometimes used as a rescue medication or on an as-needed basis. Prescriptions of PD medications were identified using the EPHMRA Anatomical Classification of Pharmaceutical Products, which categorizes products based primarily on indications and usage patterns rather than substances alone. Medications included in this study as PD medications had ATC-4 code N04A (Supplementary Table S1). The LED was calculated based on the LED conversion formula proposed in a recent systematic review [23]. For example, if patients were prescribed pramipexole at 0.5 mg/day for 10 days during the 28-day baseline period, their LED was calculated as 17.9 mg/day. Statistical Analysis Descriptive statistics were used to summarize patient characteristics. Categorical variables were reported as frequencies and percentages, and continuous variables were presented as means with standard deviations (SD) or medians with interquartile ranges (IQR), as appropriate. Patient characteristics were also summarized according to whether patients discontinued LDp/CDp CSCI during hospitalization or underwent DBS. Because this was a descriptive study, no statistical hypothesis testing or formal quantitative comparisons between groups were performed. Descriptive analyses were conducted using R software version 4.3.2. Ethical Approval This study was conducted in accordance with the Declaration of Helsinki and the Ethical Guidelines for Medical and Health Research Involving Human Subjects in Japan, and all relevant Japanese legal and regulatory requirements. We used anonymized secondary data and waived the requirement for individual informed consent because an opt-out approach was applied. Data providers ensured confidentiality. This study was approved by the Institutional Review Board of Kitamachi Clinics (approval number: ABB10580). Results Baseline Characteristics Among the 118 patients who were initiated on LDp/CDp between July 2023 and June 2024, a total of 96 who met the eligibility criteria were included (Fig. 1). The baseline characteristics of the patients are summarized in Table 1. The mean age of this cohort was 65.6 (SD: 10.2), and 58 (60.4%) were female. Notably, the median baseline BI was 90 (IQR: 44.2), and the median hospitalization duration was 15.0 days (IQR: 9.0). The item-specific BI scores are presented in Supplementary Table S2.Fig. 1Flow diagram for patient selectionTable 1Basic demographics, ADL, and hospitalization duration of all patientsa with hospitalization records, and subgroups stratified by discontinuation of LDp/CDp CharacteristicOverallLDp/CDp discontinuation during hospitalizationDiscontinuedContinuedN = 96N = 18N = 78Age Mean (SD)65.6(10.2)68.1(8.5)65.0(10.6) < 40 years, n (%)0(0.0%)0(0.0%)0(0.0%) ≥ 40, < 50 years, n (%)8(8.3%)0(0.0%)8(10.3%) ≥ 50, < 60 years, n (%)15(15.6%)3(16.7%)12(15.4%) ≥ 60, < 70 years, n (%)37(38.5%)6(33.3%)31(39.7%) ≥ 70, < 80 years, n (%)31(32.3%)8(44.4%)23(29.5%) ≥ 80 years, n (%)5(5.2%)1(5.6%)4(5.1%)Sex, n (%) Male38(39.6%)5(27.8%)33(42.3%) Female58(60.4%)13(72.2%)45(57.7%)Height (cm) Mean (SD)158.8(10.5)154.4(11.2)159.8(10.1) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)Weight (kg) Mean (SD)54.3(13.6)51.7(13.5)55.0(13.7) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)BMI Mean (SD)21.4(4.3)21.7(5.3)21.3(4.0) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)Barthel index Median (IQR)90.0(44.2)62.5(40.8)95.0(40.0) Missing, n (%)5(5.2%)0(0.0%)5(6.4%) Use of DBS5(5.2%)2(11.1%)3(3.8%)CCI category, n (%) 038(39.6%)6(33.3%)32(41.0%) 129(30.2%)6(33.3%)23(29.5%) 217(17.7%)2(11.1%)15(19.2%) 37(7.3%)4(22.2%)3(3.8%) 43(3.1%)0(0.0%)3(3.8%) 51(1.0%)0(0.0%)1(1.3%) 60(0.0%)0(0.0%)0(0.0%) 7 1(1.0%)0(0.0%)1(1.3%) ≥ 80(0.0%)0(0.0%)0(0.0%)Hospitalization duration (days) Median (IQR)15.0(9.0)17.0(25.0)15.0(8.0) ≥ 1, < 4 days, n (%)3(3.1%)0(0.0%)3(3.8%) ≥ 4, < 7 days, n (%)3(3.1%)1(5.6%)2(2.6%) ≥ 7, < 14 days, n (%)31(32.3%)7(38.9%)24(30.8%) ≥ 14, < 21 days, n (%)36(37.5%)2(11.1%)34(43.6%) ≥ 21, < 28 days, n (%)9(9.4%)1(5.6%)8(10.3%) ≥ 28 days, n (%)14(14.6%)7(38.9%)7(9.0%) Missing, n (%)0(0.0%)0(0.0%)0(0.0%)LDp/CDp pre-initiation hospitalization duration (days) Median (IQR)1.0(1.6)2.0(6.1)1.0(1.0) 0 days, n (%)12(12.5%)1(5.6%)11(14.1%) ≥ 1, < 4 days, n (%)67(69.8%)11(61.1%)56(71.8%) ≥ 4, < 7 days, n (%)7(7.3%)1(5.6%)6(7.7%) ≥ 7, < 14 days, n (%)6(6.2%)3(16.7%)3(3.8%) ≥ 14, < 21 days, n (%)3(3.1%)1(5.6%)2(2.6%) ≥ 21, < 28 days, n (%)0(0.0%)0(0.0%)0(0.0%) ≥ 28 days, n (%)1(1.0%)1(5.6%)0(0.0%) Missing, n (%)0(0.0%)0(0.0%)0(0.0%)LDp/CDp post-initiation hospitalization duration (days) Median (IQR)14.0(8.2)16.5(21.3)13.5(7.1) ≥ 1, < 4 days, n (%)6(6.2%)2(11.1%)4(5.1%) ≥ 4, < 7 days, n (%)5(5.2%)0(0.0%)5(6.4%) ≥ 7, < 14 days, n (%)36(37.5%)6(33.3%)30(38.5%) ≥ 14, < 21 days, n (%)32(33.3%)2(11.1%)30(38.5%) ≥ 21, < 28 days, n (%)6(6.2%)2(11.1%)4(5.1%) ≥ 28 days, n (%)11(11.5%)6(33.3%)5(6.4%) Missing n (%)0(0.0%)0(0.0%)0(0.0%)Patients without discharge records were excluded because discontinuation of LDp/CDp could not be definedBMI body mass index, CCI Charlson Comorbidity Index, DBS deep brain stimulation, IQR interquartile range, LDp/CDp foslevodopa/foscarbidopa, SD standard deviation Flow diagram for patient selection Table 1Basic demographics, ADL, and hospitalization duration of all patientsa with hospitalization records, and subgroups stratified by discontinuation of LDp/CDp CharacteristicOverallLDp/CDp discontinuation during hospitalizationDiscontinuedContinuedN = 96N = 18N = 78Age Mean (SD)65.6(10.2)68.1(8.5)65.0(10.6) < 40 years, n (%)0(0.0%)0(0.0%)0(0.0%) ≥ 40, < 50 years, n (%)8(8.3%)0(0.0%)8(10.3%) ≥ 50, < 60 years, n (%)15(15.6%)3(16.7%)12(15.4%) ≥ 60, < 70 years, n (%)37(38.5%)6(33.3%)31(39.7%) ≥ 70, < 80 years, n (%)31(32.3%)8(44.4%)23(29.5%) ≥ 80 years, n (%)5(5.2%)1(5.6%)4(5.1%)Sex, n (%) Male38(39.6%)5(27.8%)33(42.3%) Female58(60.4%)13(72.2%)45(57.7%)Height (cm) Mean (SD)158.8(10.5)154.4(11.2)159.8(10.1) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)Weight (kg) Mean (SD)54.3(13.6)51.7(13.5)55.0(13.7) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)BMI Mean (SD)21.4(4.3)21.7(5.3)21.3(4.0) Missing, n (%)2(2.1%)0(0.0%)2(2.6%)Barthel index Median (IQR)90.0(44.2)62.5(40.8)95.0(40.0) Missing, n (%)5(5.2%)0(0.0%)5(6.4%) Use of DBS5(5.2%)2(11.1%)3(3.8%)CCI category, n (%) 038(39.6%)6(33.3%)32(41.0%) 129(30.2%)6(33.3%)23(29.5%) 217(17.7%)2(11.1%)15(19.2%) 37(7.3%)4(22.2%)3(3.8%) 43(3.1%)0(0.0%)3(3.8%) 51(1.0%)0(0.0%)1(1.3%) 60(0.0%)0(0.0%)0(0.0%) 7 1(1.0%)0(0.0%)1(1.3%) ≥ 80(0.0%)0(0.0%)0(0.0%)Hospitalization duration (days) Median (IQR)15.0(9.0)17.0(25.0)15.0(8.0) ≥ 1, < 4 days, n (%)3(3.1%)0(0.0%)3(3.8%) ≥ 4, < 7 days, n (%)3(3.1%)1(5.6%)2(2.6%) ≥ 7, < 14 days, n (%)31(32.3%)7(38.9%)24(30.8%) ≥ 14, < 21 days, n (%)36(37.5%)2(11.1%)34(43.6%) ≥ 21, < 28 days, n (%)9(9.4%)1(5.6%)8(10.3%) ≥ 28 days, n (%)14(14.6%)7(38.9%)7(9.0%) Missing, n (%)0(0.0%)0(0.0%)0(0.0%)LDp/CDp pre-initiation hospitalization duration (days) Median (IQR)1.0(1.6)2.0(6.1)1.0(1.0) 0 days, n (%)12(12.5%)1(5.6%)11(14.1%) ≥ 1, < 4 days, n (%)67(69.8%)11(61.1%)56(71.8%) ≥ 4, < 7 days, n (%)7(7.3%)1(5.6%)6(7.7%) ≥ 7, < 14 days, n (%)6(6.2%)3(16.7%)3(3.8%) ≥ 14, < 21 days, n (%)3(3.1%)1(5.6%)2(2.6%) ≥ 21, < 28 days, n (%)0(0.0%)0(0.0%)0(0.0%) ≥ 28 days, n (%)1(1.0%)1(5.6%)0(0.0%) Missing, n (%)0(0.0%)0(0.0%)0(0.0%)LDp/CDp post-initiation hospitalization duration (days) Median (IQR)14.0(8.2)16.5(21.3)13.5(7.1) ≥ 1, < 4 days, n (%)6(6.2%)2(11.1%)4(5.1%) ≥ 4, < 7 days, n (%)5(5.2%)0(0.0%)5(6.4%) ≥ 7, < 14 days, n (%)36(37.5%)6(33.3%)30(38.5%) ≥ 14, < 21 days, n (%)32(33.3%)2(11.1%)30(38.5%) ≥ 21, < 28 days, n (%)6(6.2%)2(11.1%)4(5.1%) ≥ 28 days, n (%)11(11.5%)6(33.3%)5(6.4%) Missing n (%)0(0.0%)0(0.0%)0(0.0%)Patients without discharge records were excluded because discontinuation of LDp/CDp could not be definedBMI body mass index, CCI Charlson Comorbidity Index, DBS deep brain stimulation, IQR interquartile range, LDp/CDp foslevodopa/foscarbidopa, SD standard deviation Basic demographics, ADL, and hospitalization duration of all patientsa with hospitalization records, and subgroups stratified by discontinuation of LDp/CDp Patients without discharge records were excluded because discontinuation of LDp/CDp could not be defined BMI body mass index, CCI Charlson Comorbidity Index, DBS deep brain stimulation, IQR interquartile range, LDp/CDp foslevodopa/foscarbidopa, SD standard deviation Baseline characteristics were summarized and stratified according to the presence or absence of LDp/CDp CSCI discontinuation during hospitalization. Among the 18 patients who discontinued treatment, the mean age was 68.1 (SD: 8.5) and the median BI score was 62.5 (IQR: 40.8). The median number of days of hospitalization in the discontinuation cases was 17.0 (IQR: 25.0); 13 patients discontinued treatment within 2 weeks of starting treatment and 12 were discharged within 1 week of discontinuation (Supplementary Table S3). We further identified 5 patients who underwent DBS before the cohort entry date. Their characteristics are described in Supplementary Table S4. Medication Information Medication information during the 4 weeks of the baseline period before the index date is presented in Table 2. The mean values of LED by all PD medications, L-Dopa including COMT inhibitors, and PD medications excluding L-Dopa and COMT inhibitors were 1111.0 mg/day (SD: 575.7), 750.0 mg/day (SD: 440.8), and 361.0 mg/day (SD: 258.8), respectively. The mean number of medication classes taken was 3.9 (SD: 1.3), and 86.5% (n = 83) of the patients took three or more types of PD medications, including L-Dopa. The mean number of PD medication pills taken per day was 10.6 (SD: 5.0). The most commonly prescribed medications excluding L-Dopa were dopamine agonists (78.1%, n = 75), followed by COMT inhibitors (61.5%, n = 59) and MAO-B inhibitors (50.0%, n = 48). The mean number of concomitant medication pills taken per day for non-motor symptoms was 4.0 (SD: 4.2), with the most frequently prescribed medications being laxatives (70.8%, n = 68) and analgesics (31.2%, n = 30). More than one-quarter of the participants were also prescribed hypnotics (28.1%, n = 27), medications for rapid eye movement sleep behavior disorder (anti-RBD) (25.0%, n = 24), or medications for an overactive bladder (anti-OAB) (25.0%, n = 24).Table 2Concomitant PD medications and LED for all patientsa with hospitalization records and subgroups, stratified by discontinuation of LDp/CDp CharacteristicOverallLDp/CDp discontinuation during hospitalizationDiscontinuedContinuedN = 96N = 18N = 78LED (mg/day), [2] Any PD medications Mean (SD)1111.0(575.7)1051.1(678.2)1124.5(554.1) > 1000 mg/day, n (%)57(59.4%)11(61.1%)46(59.0%) ≤ 1000 mg/day, n (%)39(40.6%)7(38.9%)32(41.0%)LED (mg/day), [2] L-Dopa, including COMT inhibitors Mean (SD)750.0(440.8)758.0(572.0)748.2(410.1)LED (mg/day), [2] Other PD medications, excluding COMT inhibitors Mean (SD)361.0(258.8)293.1(196.8)376.4(269.6)Classes of PD medications, including L-Dopa (per 4 week) Mean (SD)3.9(1.3)3.9(1.5)3.9(1.3) None, n (%)1(1.0%)1(5.6%)0(0.0%) 1 class, n (%)6(6.2%)0(0.0%)6(7.7%) 2 classes, n (%)6(6.2%)2(11.1%)4(5.1%) 3 classes, n (%)18(18.8%)1(5.6%)17(21.8%) 4 classes, n (%)36(37.5%)9(50.0%)27(34.6%) 5 classes, n (%)18(18.8%)3(16.7%)15(19.2%) 6 classes, n (%)11(11.5%)2(11.1%)9(11.5%)Number of PD medications (per day)b Mean (SD)10.6(5.0)11.0(6.3)10.5(4.7) L-Dopa, n (%)94(97.9%)16(88.9%)78(100.0%) Dopamine agonist, n (%)75(78.1%)14(77.8%)61(78.2%) Anticholinergic medications, n (%)6(6.2%)2(11.1%)4(5.1%) Amantadine, n (%)18(18.8%)6(33.3%)12(15.4%) MAO-B inhibitors, n (%)48(50.0%)6(33.3%)42(53.8%) COMT inhibitors, n (%)59(61.5%)10(55.6%)49(62.8%) Zonisamide, n (%)34(35.4%)8(44.4%)26(33.3%) Droxidopa, n (%)6(6.2%)2(11.1%)4(5.1%) Istradefylline, n (%)32(33.3%)6(33.3%)26(33.3%) LCIG, n (%)4(4.2%)1(5.6%)3(3.8%)Number of non-motor medications (per day) Mean (SD)4.0(4.2)4.8(4.2)3.8(4.1) Antidepressants, n (%)12(12.5%)2(11.1%)10(12.8%) Antipsychotics, n (%)7(7.3%)1(5.6%)6(7.7%) Anti-dementia medications (antidementia), n (%)7(7.3%)2(11.1%)5(6.4%) Medications for insomnia (hypnotics), n (%)27(28.1%)7(38.9%)20(25.6%) Medications for REM sleep behavior disorder (anti-RBD), n (%)24(25.0%)3(16.7%)21(26.9%) Medications for OAB (anti-OAB), n (%)24(25.0%)6(33.3%)18(23.1%) Medications for constipation (laxatives), n (%)68(70.8%)14(77.8%)54(69.2%) Antacids, n (%)18(18.8%)2(11.1%)16(20.5%) Analgesics, n (%)30(31.2%)8(44.4%)22(28.2%) Medications for ED (anti-ED), n (%)0(0.0%)0(0.0%)0(0.0%) Medications for RLS (anti-RLS), n (%)1(1.0%)0(0.0%)1(1.3%) Medications for OH (anti-OH), n (%)3(3.1%)0(0.0%)3(3.8%)aPatients without discharge records were excluded because discontinuation of LDp/CDp could not be definedbTransdermal patches are also includedCOMT catechol-O-methyltransferase, ED erectile dysfunction, LCIG levodopa–carbidopa intestinal gel, L-Dopa levodopa, LDp/CDp foslevodopa/foscarbidopa, LED levodopa equivalent dose, MAO-B monoamine oxidase B, OAB overactive bladder, OH orthostatic hypotension, PD Parkinson’s disease, REM rapid eye movement, RLS restless legs syndrome, SD standard deviation Table 2Concomitant PD medications and LED for all patientsa with hospitalization records and subgroups, stratified by discontinuation of LDp/CDp CharacteristicOverallLDp/CDp discontinuation during hospitalizationDiscontinuedContinuedN = 96N = 18N = 78LED (mg/day), [2] Any PD medications Mean (SD)1111.0(575.7)1051.1(678.2)1124.5(554.1) > 1000 mg/day, n (%)57(59.4%)11(61.1%)46(59.0%) ≤ 1000 mg/day, n (%)39(40.6%)7(38.9%)32(41.0%)LED (mg/day), [2] L-Dopa, including COMT inhibitors Mean (SD)750.0(440.8)758.0(572.0)748.2(410.1)LED (mg/day), [2] Other PD medications, excluding COMT inhibitors Mean (SD)361.0(258.8)293.1(196.8)376.4(269.6)Classes of PD medications, including L-Dopa (per 4 week) Mean (SD)3.9(1.3)3.9(1.5)3.9(1.3) None, n (%)1(1.0%)1(5.6%)0(0.0%) 1 class, n (%)6(6.2%)0(0.0%)6(7.7%) 2 classes, n (%)6(6.2%)2(11.1%)4(5.1%) 3 classes, n (%)18(18.8%)1(5.6%)17(21.8%) 4 classes, n (%)36(37.5%)9(50.0%)27(34.6%) 5 classes, n (%)18(18.8%)3(16.7%)15(19.2%) 6 classes, n (%)11(11.5%)2(11.1%)9(11.5%)Number of PD medications (per day)b Mean (SD)10.6(5.0)11.0(6.3)10.5(4.7) L-Dopa, n (%)94(97.9%)16(88.9%)78(100.0%) Dopamine agonist, n (%)75(78.1%)14(77.8%)61(78.2%) Anticholinergic medications, n (%)6(6.2%)2(11.1%)4(5.1%) Amantadine, n (%)18(18.8%)6(33.3%)12(15.4%) MAO-B inhibitors, n (%)48(50.0%)6(33.3%)42(53.8%) COMT inhibitors, n (%)59(61.5%)10(55.6%)49(62.8%) Zonisamide, n (%)34(35.4%)8(44.4%)26(33.3%) Droxidopa, n (%)6(6.2%)2(11.1%)4(5.1%) Istradefylline, n (%)32(33.3%)6(33.3%)26(33.3%) LCIG, n (%)4(4.2%)1(5.6%)3(3.8%)Number of non-motor medications (per day) Mean (SD)4.0(4.2)4.8(4.2)3.8(4.1) Antidepressants, n (%)12(12.5%)2(11.1%)10(12.8%) Antipsychotics, n (%)7(7.3%)1(5.6%)6(7.7%) Anti-dementia medications (antidementia), n (%)7(7.3%)2(11.1%)5(6.4%) Medications for insomnia (hypnotics), n (%)27(28.1%)7(38.9%)20(25.6%) Medications for REM sleep behavior disorder (anti-RBD), n (%)24(25.0%)3(16.7%)21(26.9%) Medications for OAB (anti-OAB), n (%)24(25.0%)6(33.3%)18(23.1%) Medications for constipation (laxatives), n (%)68(70.8%)14(77.8%)54(69.2%) Antacids, n (%)18(18.8%)2(11.1%)16(20.5%) Analgesics, n (%)30(31.2%)8(44.4%)22(28.2%) Medications for ED (anti-ED), n (%)0(0.0%)0(0.0%)0(0.0%) Medications for RLS (anti-RLS), n (%)1(1.0%)0(0.0%)1(1.3%) Medications for OH (anti-OH), n (%)3(3.1%)0(0.0%)3(3.8%)aPatients without discharge records were excluded because discontinuation of LDp/CDp could not be definedbTransdermal patches are also includedCOMT catechol-O-methyltransferase, ED erectile dysfunction, LCIG levodopa–carbidopa intestinal gel, L-Dopa levodopa, LDp/CDp foslevodopa/foscarbidopa, LED levodopa equivalent dose, MAO-B monoamine oxidase B, OAB overactive bladder, OH orthostatic hypotension, PD Parkinson’s disease, REM rapid eye movement, RLS restless legs syndrome, SD standard deviation Concomitant PD medications and LED for all patientsa with hospitalization records and subgroups, stratified by discontinuation of LDp/CDp aPatients without discharge records were excluded because discontinuation of LDp/CDp could not be defined bTransdermal patches are also included COMT catechol-O-methyltransferase, ED erectile dysfunction, LCIG levodopa–carbidopa intestinal gel, L-Dopa levodopa, LDp/CDp foslevodopa/foscarbidopa, LED levodopa equivalent dose, MAO-B monoamine oxidase B, OAB overactive bladder, OH orthostatic hypotension, PD Parkinson’s disease, REM rapid eye movement, RLS restless legs syndrome, SD standard deviation The results of subgroup analyses of LDp/CDp CSCI discontinuation are shown in Table 2. The mean LED value for any PD medication in discontinuation cases was 1051.1 (SD: 678.2). Dopamine agonists, COMT inhibitors, and MAO-B inhibitors were commonly prescribed to all eligible patients. The results of the DBS cases are presented in Supplementary Table S5. Discussion This study was conducted within 1 year of approval for LDp/CDp CSCI for advanced PD in Japan. The present study identified several similarities and discrepancies between clinical trials and real-world evidence from Japan and other countries, although differences in eligibility criteria, real-world clinical practices, healthcare system structures, drug availability, and physicians' initial experiences with novel therapies should be considered. Our study provides descriptive insights into how the characteristics of LDp/CDp users in the Japanese real-world setting align with or diverge from those in RCTs and overseas settings. Characteristics of patients who discontinued LDp/CDp CSCI appeared to include older age, a higher proportion of females, and lower baseline BI scores. However, these observations were based on descriptive analyses without statistical comparisons and should therefore be interpreted as exploratory findings. Overall, this study contributes to the understanding of the characteristics of patients who undergo LDp/CDp CSCI in real-world settings, which is crucial for identifying those who will derive the greatest benefit from treatment because of the limited generalizability of clinical trials [12–14]. Our findings indicate that the characteristics of patients with advanced PD in a Japanese real-world setting are similar to those of two prior phase 3 clinical trials in terms of age [10, 11]. However, our cohort differed from clinical trial populations in that we included a higher proportion of women and Japanese patients [10, 11]. Although the precise BI scores for the previous clinical trial population have not been reported [11], both the Hoehn and Yahr scale (H-Y) score (2.2) and our cohort’s median BI score [90.0 (IQR: 44.2)] suggest a population with preserved ADL. This speculation is supported by prior literature showing that patients with H-Y stages 3 and 4 had substantially lower average BI scores of 80.1 and 61.1, respectively [24, 25]. Given that our cohort's BI scores seemed to be higher than these values, it is reasonable to consider that both the clinical trial and our study targeted patient populations that maintained relatively high ADL. Here, the mean LED derived from L-Dopa including COMT inhibitors was 750.0 mg/day (SD: 440.8), which was lower than that in a previous clinical trial (1064.9 or 1000) [10, 11]. In addition, we observed that 83/96 eligible patients (86.5%) took three or more types of PD medications including L-Dopa, whereas the proportion of patients taking two or more types of PD medications excluding L-Dopa was reported to be 53.3% in a previous clinical trial, indicating that our study population tended to be treated with a larger number of concomitant PD medications [11]. Of note, a larger proportion of patients received concomitant dopamine agonists in our study (78.1% vs. 37%/46% in previous studies) [10, 11]. These findings indicated that patients receiving LDp/CDp CSCI, as with other patients with advanced PD, are more commonly managed with concomitant medications rather than dose titration of L-Dopa in Japanese real-world settings compared with different settings. This trend may be attributed to the differences in treatment patterns between Japan and other countries, as observed in a single-center study in Europe that showed a lower use of dopamine agonists in line with clinical trials [16, 17]. Therefore, LDp/CDp CSCI could play an important role in managing pill burden among patients with advanced PD, particularly in Japan. When initiating LDp/CDp CSCI, dose calculations and titrations should carefully consider concomitant dopamine agonists. If dopamine agonist dose reduction is required for the management of hallucinations or psychosis, it should be performed gradually over several weeks with careful symptom monitoring to minimize risks such as dopamine agonist withdrawal syndrome [27]. Moreover, to the best of our knowledge, the profiles of concomitant medications for nonmotor symptoms are yet to be well described. Therefore, an evaluation of these profiles in the present study provided meaningful insights that could serve as valuable benchmarks for understanding LDp/CDp treatment patterns. The present descriptive study further characterized LDp/CDp CSCI discontinuation and the potential heterogeneity between patients with advanced PD who discontinued or did not. The majority of discontinuations occurred within the first 2 weeks following LDp/CDp initiation (n = 13), and the proportion of patients who required discontinuation was similar to that reported in a previous RCT (22%, N = 16) [10], a study from Germany (24.5%, n = 13) [17], and a case series from Japan (19%, n = 5) [18]. Unfortunately, the reason for discontinuation was not identified in this study, although previous studies have provided valuable insights into this aspect. Clinical trials have reported infusion site adverse events such as edema and cellulitis as the main reasons for discontinuation [10], indicating that the appropriate management of adverse events during LDp/CDp CSCI induction is crucial. One real-world study that included German patients who had LDp/CDp CSCI introduced during hospitalization documented that 5 of 13 patients discontinued treatment due to skin reactions, which also supports the importance of appropriate management of adverse events [17]. Consistent with these findings, multinational real-world studies, such as the ORCHESTRA study [28] and the 6-month interim analysis of the ROSSINI study [29], also suggest that discontinuation tends to occur early after treatment initiation and that infusion site events are important factors associated with discontinuation. Furthermore, we observed that patients requiring discontinuation were more likely to be older [mean 68.1 (SD: 8.5) vs. 65.0 (SD: 10.6)] than those who did not require discontinuation, with a standardized mean difference of 0.33 between the groups; however, mean values for age remained below 70 years old, which has been mentioned as a characteristic of good candidates for LDp/CDp CSCI initiation [15]. Our findings showed that the proportion of discontinuation cases in the < 70 years old group and ≥ 70 years old group were 15.0% (9 of 60) and 25.0% (9 of 36), respectively. Since older age in patients with PD may be associated with greater impairments in ADL [26], this should also be considered an important characteristic for discontinuation. In addition to older age, a lower median BI score was observed in patients who discontinued treatment (62.5 vs. 95.0). These profiles are important for optimizing the persistence of LDp/CDp CSCI. However, it should be noted that the current evidence does not limit the use of LDp/CDp CSCI in older patients and those with impaired ADL. However, several of these aspects require further investigation. First, the median duration of hospitalization [15 days (IQR: 9 days)] should be noted. Although we could not determine whether optimization of the dose or infusion rate was complete, the observed duration of hospitalization may reflect the optimization period from a clinical standpoint. This hypothesis is supported by a previous real-world study in Europe, which identified an optimization period of 18.4 ± 6.8 days [17]. As optimization should be completed when there is no adjustment to the infusion rate for at least 15 days in clinical trials [15], the Japanese real-world setting may exhibit optimization patterns similar to those observed in clinical trials. However, the MDV database comprises hospitals operating under the DPC claims-based bundled payment system, and the reimbursement framework may have a limited hospitalization duration, potentially affecting the extent to which the recorded hospital stays reflect the actual optimization periods. Moreover, the 15-day hospitalization period may represent a barrier to treatment initiation for some patients and limit treatment options, particularly for working individuals and those who value accessibility and continuity of daily life. In this regard, a recent practical review noted that LDp/CDp could be initiated and adjusted in outpatient settings without hospitalization and that inpatient initiation may take 3–4 days or up to 10 days, depending on patient characteristics [27]. Therefore, information on shorter hospitalization strategies and outpatient initiation may be valuable in the Japanese real-world setting. Additional studies are required to evaluate the optimal hospital stay and the characteristics of patients with LDp/CDp CSCI admitted to an outpatient setting. Second, our study population was relatively young and had higher levels of functional independence than advanced PD populations in other settings [7]. While most patients complete hospitalization without discontinuation, careful management including enhanced adverse event prevention is warranted for older patients and those with lower ADL to maximize the benefits of LDp/CDp CSCI. Third, because the impact of LDp/CDp CSCI initiation on reducing pill burden was not evaluated in this study, future longitudinal investigations into utilization patterns, including changes in pill burden, LED at discharge, LDp/CDp doses during hospitalization (at admission and/or discharge), and the number of concomitant medication classes, remain highly desirable. This study has several limitations. First, several important clinical information points such as H-Y stage, duration from PD diagnosis, and reason for discontinuation of LDp/CDp CSCI were not identified. Second, due to the nature of the database, medical histories could not be accurately identified for patients in the MDV database who were treated at other medical institutions. In the present study, 13 patients were excluded owing to the lack of hospitalization records for the initiation of LDp/CDp CSCI despite meeting the other eligibility criteria, possibly because they underwent LDp/CDp CSCI in a hospital that was not included in the MDV dataset and subsequently changed hospitals. Third, we analyzed real-world first-year data from Japan. As such, physicians had limited experience with patient selection and management of the LDp/CDp treatment system during the optimization period, which may have influenced treatment practices and patient profiles and could affect the generalizability of our findings. Finally, although the MDV database included data from more than 500 acute-care hospitals, our findings are not fully applicable to other clinical settings. Nevertheless, this study provided valuable insights given that LDp/CDp CSCI is a novel treatment and it is plausible to assume that it was primarily introduced in large hospitals by practitioners specializing in neurology. Conclusion Overall, this observational study utilizing a large-scale hospital-based database in Japan revealed that patients initiating LDp/CDp CSCI within 1 year of approval tended to be younger with preserved ADL, similar to the population in clinical trial settings. Although caution is warranted in interpreting these findings due to the descriptive nature of the study, patients with advanced PD in Japanese real-world clinical settings tend to be managed with concomitant PD medications rather than dose titration of L-Dopa, particularly dopamine agonists, compared with other countries. Therefore, further research is warranted to optimize patient selection and maximize the benefits of LDp/CDp CSCI in Japanese patients with advanced PD, along with an ongoing assessment of its real-world impact on pill burden. Supplementary Information Below is the link to the electronic supplementary material.Supplementary file 1 (PDF 286 KB) Supplementary file 1 (PDF 286 KB)