Long-Term Effectiveness and Safety of LCIG in Romania. 3-year Results from a Prospective, Observational Study (DUOGLOBE)

Introduction

Parkinson’s Disease (PD) is dominated by motor and non-motor symptoms, with a high impact on the quality of life (QoL). More than 200 years after the initial description, and despite tremendous research efforts, the treatment is still symptomatic.¹

Levodopa plays a central role in the treatment strategy, and dopaminergic medications provide the initial control of the motor symptoms. Over time, the changes in the striatal dopaminergic neurons lead to fluctuations in levodopa concentrations^2,3 and require different strategies for oral medication adjustment with multiple dose fractioning of levodopa, addition, discontinuation, or replacement of anti-PD medications, according to the patient’s needs. The optimization process is challenging for clinicians and patients, however necessary throughout the years of progression.⁴ In patients with maximally optimized treatment and lack of control of PD symptoms, the “5-2-1 criteria” (oral levodopa at least five times a day, or at least 2 hours per day with OFF symptoms, or at least 1 hour of troublesome, uncontrolled dyskinesia) are useful to identify patients still uncontrolled on oral/conventional medications.^5,6 Timely initiation of a device-aided therapy (DAT) is likely to improve the overall PD status.⁵

In patients with advanced PD (APD), levodopa–carbidopa intestinal gel (LCIG) reduces the fluctuations of levodopa concentrations⁷ and stabilizes the synaptic dopamine levels.⁸ LCIG treatment was extensively studied in patients with APD in clinical trials^9,10 and large observational studies^7,11 and significant improvement of motor, non-motor symptoms and the QoL has been demonstrated. Additionally, dosing flexibility¹² and the option to combine the treatment with different anti-PD drugs contribute to the personalization of the therapeutic approach and offer an opportunity for sustained optimal outcomes for several years.

Neurologists from tertiary centers in Romania have over a decade of experience with LCIG therapy, which was the only realistically available DAT option in our country until 2021.^13,14 In previous publications, Szasz, Constantin, Szatmari et al detailed the procedure for the evaluation of motor complications severity and selection of eligible APD patients, the challenges of the testing process, all under the specific conditions of clinical practice in our country.^15–17 Analyzing the results, they evidenced both the efficacy and safety of long-term administration of LCIG therapy in patients with severe motor complications (fluctuations and complex dyskinesias).^15–18

DUOGLOBE (DUOdopa/Duopa in Patients with APD – a GLobal Observational Study Evaluating Long-Term Effectiveness; NCT02611713) was a 3-year, prospective observational, international study.¹⁹ The study was designed to assess the long-term effectiveness and tolerability of LCIG as well as the impact of LCIG treatment on the patients’ QoL and their caregivers’ burden in a large international cohort in Europe, Israel, the United States and Australia. This manuscript presents the results of Romanian patients participating in the DUOGLOBE study and the perspective of long-term LCIG treatment offered by clinical practice.

Patients and Methods

Inclusion and Exclusion Criteria

Patients were enrolled in the DUOGLOBE study if they were LCIG-naive and considered eligible to receive LCIG according to the approved local indication and reimbursement criteria, if applicable. Eligibility for LCIG treatment was decided before the preliminary discussion and irrespective of participation in DUOGLOBE study. All patients and caregivers provided written informed consent.

Patients with any contraindication for LCIG treatment or a Mini Mental State Examination (MMSE) score <24 were not included in the study. Current treatment with continuous apomorphine infusion (with <4 weeks between drug discontinuation and study inclusion) and previous surgery for PD, including deep brain stimulation procedure, were also exclusion criteria. To maintain the observational characteristic of the study, participation in a concomitant interventional trial was not permitted.

Study Design

The global, non-interventional, single-arm, post-marketing study DUOGLOBE was conducted from 2015 to 2020 and included a total of 195 patients in 55 neurology centers specialized in the treatment of APD.¹⁹

The study data were collected before and at the initiation of LCIG treatment via percutaneous endoscopic gastrostomy-with jejunal extension (PEG-J), after 3 months, 6 months, and then at 6-month intervals up to 36 months. The National Ethics Committee approved the study protocol and informed consent for the patient and caregiver. The study was conducted in accordance with the Declaration of Helsinki.

Treatment with LCIG (levodopa 20 mg/mL and carbidopa 5 mg/mL) was administered with the help of a portable pump (CADD-Legacy, Smiths Medical ASD, Minneapolis, MN)¹⁴ via PEG-J. The dose and daily duration of LCIG, as well as concomitant use of anti-PD oral drugs, were decided by the study investigator according to individual patient needs.²⁰

Study Objectives & Measures

The primary objective assessed the change in OFF time from baseline (number of hours in the day before LCIG initiation) at 36 months. The secondary objectives included changes in: dyskinesia severity (Unified Dyskinesia Rating Scale [UDysRS]), tremor severity (Unified Parkinson’s Disease Rating Scale [UPDRS] part III, item 20), motor function (UPDRS part III total score), non-motor symptoms (Non-motor Symptom Scale [NMSS], Parkinson’s Disease Sleep Scale-2 [PDSS-2], Epworth Sleepiness Scale [ESS]), complications of therapy (UPDRS part IV), the overall clinical impression of disease severity (Clinical Impression of Severity Index for Parkinson’s Disease [CISI-PD]), daily living activities (UPDRS part II), QoL (EuroQol- 5 Dimension [EQ-5D], the 8-item PD Questionnaire [PDQ-8]), as well as the caregiver’s burden (Modified Caregiver Strain Index [MCSI]) from baseline compared to 36 months. Serious adverse events (SAE), and adverse events (AE) leading to premature discontinuation were reported as part of this observational study from initiation of LCIG treatment to 30 days after the last study visit. Vital signs and information related to tube or pump replacement were collected.

Statistical Analysis

Unless noted otherwise, all analyses were performed on the Full Analysis Set (FAS) population. The FAS included all patients of the Safety population who had a baseline effectiveness evaluation and at least one assessment of any effectiveness measurement during PEG-J visits. The safety population included all patients who had nasojejunal and/or percutaneous endoscopic gastrostomy with jejunal extension placement, irrespective of whether patients had received the study drug or not.

For efficacy assessments, the mean change from baseline to each scheduled visit was estimated using a mixed-effect model repeated-measures, with a 95% confidence interval for change from baseline, and P value for the change from baseline at every visit. The reported P values are nominal and uncorrected for multiplicity. Significance for all efficacy measures was determined using a one-sample t-test compared with baseline efficacy assessments. Pearson/Spearman correlation coefficients between changes of motor symptoms (OFF time, UDysRS total score), non-motor symptoms (NMSS total score, PDSS-2 total score, ESS total score), and QoL (PDQ-8 summary index, EQ-5D summary index) were generated for each post-baseline visit.

No imputation of missing data applied. DUOGLOBE was not a confirmatory study, therefore, no hypothesis was formally tested.

Results

In Romania, a total of 51 patients were enrolled in six university centers specialized in the treatment of movement disorders. The baseline demographic and clinical results are summarized in Table 1. Most patients were male (70.6%), and more than one third (39.2%) were less than 65 years of age. Most patients had Hoehn & Yahr stage 3 or 4, measured during ON (86.0%) or OFF (92.0%) periods (Table 1).

Table 1 Baseline Demographics and Clinical Characteristics

The median disease duration at the time of LCIG initiation was 8.1 years (1.6–19.7). Before LCIG initiation, PD treatment included dopamine agonists (76.5%), levodopa (62.7%), monoamine oxidase B (MAO-B) inhibitors (56.9%), catechol-O-methyltransferase (COMT) inhibitors (45.1%), and amantadine (23.5%). The use of concomitant anti-PD medication decreased in one month from LCIG initiation and remained stable up to 36 months (Figure 1). COMT inhibitors were completely discontinued in the first 3 months and were not used in combination with LCIG thereafter.

Figure 1 Concomitant anti-PD medication by visit interval. Abbreviations: COMTinh, catechol-O-methyltransferase inhibitor; DA, dopamine agonist; M, month; MAO-Binh, monoamine oxidase B inhibitor.

The majority of patients (82.4%) maintained LCIG treatment for up to 24 months, and almost half of the initial cohort (47.1%) for up to 36 months. The mean treatment duration with LCIG was 905.4 ± 360.3 days, with a total estimated exposure of 126.4 patient-years.

LCIG was administered daily for a median of 16.0 hours from Day 1 (duration interval: 11 to 24 hours) to end of follow-up (Month 36; interval 14–24 hours). The mean ± SD LCIG dose was 1425.0 ± 524.3 mg/day at Day 1 and remained relatively stable up to Month 30 (1482.0 ± 647.3 mg/day) and was reduced to 1261.7 ± 442.8 mg/day until the end on the study (Month 36).

Motor Symptoms

The total daily duration in OFF state decreased significantly from Day 1 compared to the day before LCIG initiation (mean change at D1 −4.2 hours ± 2.27) and was maintained throughout the study period (mean change at final visit −3.4 hours ± 2.82) (p≤0.001 at all timepoints vs baseline) (Figure 2A), irrespective of sex, age and PD duration [data not shown]. Improvement of OFF time was further confirmed by a significant reduction of UPDRS Part IV item 39 score from baseline to all study visits (p≤0.001 at all timepoints vs baseline).

Figure 2 (A) Change of OFF time (hours/day) at study visits vs baseline (BL); (B) Change of time with dyskinesia (hours/day) at study visits vs BL; (C) Change of UDysRS total score at study visits vs BL. ***p≤0.001; **p≤0.01; *p≤0.05; ns, p>0.05. Abbreviations: BL, baseline; D, day; M, month; UdysRS, Unified Dyskinesia Rating Scale; SD, standard deviation.

Duration of dyskinesia was significantly reduced at Day 1 versus baseline (mean change of duration at D1 −1.8 hours ± 2.42, p≤0.001) and the improvement was maintained throughout the study period, with statistical significance at most timepoints (Figure 2B) (mean change of duration at final visit −0.8 hours ± 2.09, p≤0.05). Similar results were noted for the UDysRS total score (Figure 2C). Mean UPDRS Part IV scores related to dyskinesia (item 32, item 33, and item 34) were also numerically improved from LCIG initiation, with statistical significance observed for dyskinesia-related disability (Table 2).

Table 2 Change from Baseline in the Mean UPDRS Scores

Non-Motor Symptoms

The NMSS total score was significantly reduced by LCIG from Day 1 through the final visit (p<0.001 at all visits vs baseline, except at Month 36, p≤0.05 and at final visit, p≤0.01) (Figure 3). Most subdomain scores on the Non-motor Symptom Scale (NMSS) improved from Day 1 through the follow-up period, with statistical significance maintained up to the final visit for gastrointestinal tract and miscellaneous (Table 3). Exploring the quality of sleep in patients receiving LCIG treatment, statistically significant (p<0.001) reduction of Parkinson’s Disease Sleep Scale-2 (PDSS-2) total scores were achieved at almost all assessment points from baseline (Figure 4). The highest improvements, with changes of more than a mean of 10 points, were observed at Month 3 (−10.4 ± 14.70) and Month 6 (−11.4 ± 15.17). Daytime sleepiness during LCIG treatment was assessed with ESS. The mean score at baseline was in the higher normal range, and numeric improvements were recorded at each follow-up moment (Table 4).

Table 3 Change from Baseline in the Mean NMSS Total and Subdomains Scores

Table 4 Change from Baseline in ESS Score

Figure 3 Change of NMSS total score at study visits vs BL. ***p≤0.001; **p≤0.01; *p≤0.05. Abbreviations: BL, baseline; D, day; M, month; NMSS, Non-motor Symptom Scale; SD, standard deviation.

Figure 4 Change of PDSS-2 total score at study visits from baseline. ***p≤0.001; **p≤0.01; ns, p > 0.05. Abbreviations: BL, baseline; D, day; M, month; ns, not significant; PDSS-2, Parkinson’s Disease Sleep Scale-2; SD, standard deviation.

Quality of Life

The patients reported significant and sustained improvement in generic (EQ-5D) as well as disease-specific QoL (PDQ-8) (Figures 5A and B). Additionally, the burden of PD for the caregivers of patients receiving LCIG was reduced from Day 1 up to 36 months, with a mean change of the total Modified Caregiver Strain Index from baseline of more than 3 points after 3 years of follow-up (Figure 5C).

Figure 5 (A) Change of EQ-5D total score at study visits from baseline; (B) Change of PDQ-8 summary index at study visits from baseline; (C) Change of MCSI total score at study visits from baseline. ***p≤0.001; **p≤0.01; *p≤0.05; ns, p > 0.5. Abbreviations: BL, baseline; D, day; EQ-5D, EuroQol- 5 Dimension; M, month; MCSI, Modified Caregiver Strain Index; PDQ-8, 8-item Parkinson Disease Questionnaire; ns, not significant; SD, standard deviation.

Safety Results

In total, 25 patients (49.0%) reported at least one SAE throughout the study period (Table 5). Three events were considered as possibly related to LCIG (one case each of dyspepsia, a complication associated with the device, and delirium). Treatment-emergent SAEs leading to the discontinuation of LCIG (n=14, 27.5%) were fatal events, none of these considered related to LCIG by the study investigators. No case of chronic inflammatory radiculoneuropathy or polyneuropathy was reported in Romania.

Table 5 Overview of Safety Events

Discussion

Overall, the DUOGLOBE study provides information regarding the effectiveness and safety of LCIG treatment in a large international sample, with diverse practices, with up to 36 months of prospective follow-up,¹⁹ making it the only levodopa-based DAT with contemporary real-world evidence over such a long prospective follow-up.

The analysis of the results obtained in the Romanian cohort brings additional information on patients with APD treated with LCIG. We noted that PD diagnosis is confirmed relatively soon after the patients identify the first symptoms, in approximately one year, similarly to the international cohort.

Compared to the global study population,¹⁹ the patients enrolled in Romania were younger and had a shorter disease duration at study baseline. Moreover, DAT (LCIG in our study) was initiated earlier in Romanian patients compared to the global cohort, also taking into account the time from the onset of motor fluctuations to DAT initiation (3.8 years in the Romanian cohort and 5.6 years in the global cohort). It is likely that a more rapid DAT initiation in this cohort is the consequence of earlier referral of patients with APD to expertise centers and limited access to other DATs than LCIG at the time of study enrolment. Despite comparable baseline scores on UPDRS scales, the PDQ-8 (51.0 and 45.1) and PDSS-2 (31.4 and 26.6) mean scores indicate a higher burden of the disease for our cohort, compared to the global cohort, respectively.

Previous analysis suggests that a characteristic of APD management in our country is the more frequent use of different add-on therapeutic options prior to the initiation of LCIG treatment.^17,21 However, regardless of the time of initiation, the length of the naso-jejunal testing phase, and the management of the previous add-on dopaminergic medication were largely left to the clinician’s personal judgment and experience. There is a lack of clear recommendations from experts, although these would be particularly important for increasing long-term safety and efficacy and for lowering the discontinuation rate.^17,22 Given the fact that many effective add-on therapeutic options (safinamide, opicapone, extended-release amantadine) are not available in Romania, we consider that the “maneuvering space” for the clinicians is reduced in many cases. Under these circumstances, as a logical consequence of the significantly limited therapeutic options available to manage APD, selected patients may start available device-aided therapies at an earlier disease stage than suggested by worldwide current clinical practice.¹⁷

The continuous dopaminergic stimulation with LCIG provided a significant reduction in OFF time, dyskinesia, as well as non-motor symptoms, which are reflected in an improvement of the disease-specific QoL. In our cohort, the reduction in daily OFF time was more than 3 hours, consistent with results from other trials with LCIG^7,9,15 as well as DUOGLOBE global cohort¹⁹ exceeded the previously reported minimal clinically important difference (MCID) estimates of approximately 1 hour reported in the literature in patients with APD,²³ supporting a substantial patient-relevant treatment benefit. Similarly, the mean reduction of 9.0 points in UDysRS score at 36 months markedly exceeds previously proposed MCID thresholds for dyskinesia domains, where improvements greater than 2.1 points for ON dyskinesia and approximately 1.8 points for OFF dystonia have been shown to represent clinically perceptible benefit.²⁴

Non-motor symptoms, particularly sleep-related problems, are key contributors to overall disease burden. Findings from our study indicate that LCIG treatment significantly reduced NMSS total scores, with numerical improvements in all subdomains over long-term follow-up. In addition, reduction of PDQ-8 scores has been maintained throughout the study period, exceeding most of the time the MCID level (−5.94 to +4.91) reported by Horváth et al^25,26 In addition, the sleep quality and daytime sleepiness (PDSS-2 and ESS scores), and other benefits related to QoL (PDQ-8 and EQ-5D scores) have been maintained at most follow-up points despite the severe and progressive neurologic condition. The alleviation of non-motor symptoms and improvements in sleep quality with LCIG may contribute to the overall benefit of LCIG in both patients and their caregivers. Regarding the caregiver burden, a numeric improvement of the MCSI score has been observed at most timepoints over 3-year study period, although not statistically significant. The reduction with more than 3 points is consistent with the reports in global DUOGLOBE cohort.¹⁹

The LCIG dose remained relatively stable over time, from Day 1 up to 36 months, with a similar pattern observed in the use of concomitant anti-PD medications. It is important to note that COMT inhibitors were not necessarily used after LCIG initiation, as shown by a rapid decline of their use in the first 3 months after baseline. For our patients, LCIG has allowed therapeutic simplification, eliminating COMT inhibitors and reducing oral doses, with potential implications for adherence and economic burden.²⁷

The safety events were consistent with the established profile, with no additional safety signals over the long-term follow-up, comparable with the already established safety profile for LCIG.^9,10,16 Serious adverse events, including fatal outcomes, should be interpreted in the context of APD and high disease burden, as they may reflect the disease progression or complications rather than LCIG itself, as highlighted by a large retrospective analysis of dropout rates in one center in Romania.¹⁶ In the diverse settings of clinical practices, the experience of treating clinicians and training of patients and caregivers are key factors for maintenance of adherence and timely management of any treatment-related issues.¹⁶

A patient with APD may be a candidate for different DATs.²⁸ The choice of initial DAT, as well as their sequencing in the long-term strategy, should be based on robust evidence from clinical trials and observational, non-interventional, high-quality studies. Although the continuous subcutaneous infusion of foslevodopa/foscarbidopa has recently become available, with a more comfortable pump and administration for the patients, the LCIG may continue to be a suitable treatment for appropriately selected patients, as shown by the DUOGLOBE study and other recent real-world evidence.^29–31

While another study reporting observational data in Romania¹⁷ offers early and short-term evidence on a different DAT formulation (levodopa, carbidopa, and entacapone intestinal gel), DUOGLOBE provides the most comprehensive and long-term real-world evaluation ofLCIG in Romania to date.

As mentioned above, the Romanian study population is part of the global study cohort, and no direct comparisons were performed; the analysis of the Romanian cohort represents local descriptive data. The main limitations are associated with the non-interventional, observational design. P-values are nominal and uncorrected for multiplicity. The study did not include a comparator group, nor has it been powered to evidence the factors related to treatment continuity. No imputation of data applied, only observed cases were included.

In a population marked by motor fluctuations and dyskinesia, DUOGLOBE showed consistent and significant improvements with LCIG, with a good overall tolerability profile. A sustained beneficial effect of up to 3 years, documented by the validated scales scores, confirms the positive effects of more stable levodopa/carbidopa levels. Significant improvement in the quality of life for most of the follow-up period is remarkable considering the disease is progressive and debilitating.

Conclusion

DUOGLOBE is currently the longest observational study with LCIG treatment conducted in expert centers across Europe, Australia, and the United States.¹⁹ The cohort enrolled in Romania showed consistent benefits of LCIG on all efficacy and safety parameters included in the analysis. The baseline characteristics provide evidence for a population that was initiated on LCIG treatment at a younger age and had a shorter duration of motor fluctuations, however, with a more severe clinical picture and a higher disease burden than the global cohort. LCIG provided significant improvement of PD symptoms and quality of life, with MCIDs exceeding the average reported values in patients with APD, maintained over the long-term, up to 3 years of continuous follow-up. The safety results are consistent with the overall established safety profile of LCIG.