Non‐ergot dopamine agonist therapy for Parkinson's disease in Japan: A claims database analysis

Levodopa monotherapy and combination therapy with non‐ergot dopamine agonists (DAs; pramipexole, ropinirole, and rotigotine) are commonly used in patients with Parkinson's disease (PD). Changes in prescription preferences for each DA were reported previously; however, information regarding the selection of individual DAs is insufficient.


| INTRODUC TI ON
The number of treatment options for Parkinson's disease (PD) has increased since the early 2000s. However, a Japanese survey (including 158 physicians) reported a decrease in satisfaction with PD treatments, from 56.0% in 2010 to 43.6% in 2014, among 60 targeted diseases, including infections, neoplasms, metabolic diseases, mental diseases, and nervous diseases. 1 The dissatisfaction is probably due to the absence of an effective cure for the disease and the variety of symptoms that continue to evolve as the disease progresses, 2,3 complicating the choice of appropriate treatment.
Dopamine replacement therapy, used to ameliorate motor symptoms, remains the standard treatment for patients with PD since levodopa was introduced in the 1960s, 4,5 despite the development of new medications with various action mechanisms and effects. Recent Japanese research reported that levodopa is most commonly prescribed, followed by non-ergot dopamine agonists (DAs), both as monotherapy and in combination with levodopa. [6][7][8] In our previous study using a Japanese claims database, a preference change between levodopa monotherapy and combination therapy with DAs was observed between 2010 and 2017. 9 This change might be associated with the publication of the 2011 Japanese guidelines, which recommend DAs as the initial treatment in patients with PD. 10 An additional factor might be the availability of new DAs based on the concept of continuous dopaminergic stimulation, 4,5,11 including extended-release pramipexole (2011) and ropinirole (2012), and transdermal rotigotine patch (2013). Although these three DAs are categorized in the same class, there are some differences between them, including in efficacy and safety profiles, [12][13][14][15][16] which might influence the choice of a specific DA. Previous studies, including our study (Tsuboi et al. unpublished data, 2020), described the changes in prescription preferences for each DA. 6 However, a detailed investigation of the reasons for each DA selection (including the initial choice of DA and changes to a different one) has not been conducted.
This study examined the trends of dopamine replacement therapy, focusing on individual DAs, using a Japanese claims database.
We also investigated DA combinations, including the therapy before and after modifications and comorbidities at the time of DA treatment initiation and modification. Although the claims database did not include information about the specific reasons for selecting or changing a treatment, we believe that our study will provide crucial insights into DA treatment choices. replacement of another DA) from 2013 onwards when all target DAs were available. The patients were included in this analysis if they received a specific DA type in 2013 or later for the first time in the study period (this DA was defined as the index DA and the date of the first prescription as the index date), received a different type of DA before the index date (baseline DA); had a minimum 3-month observation period before the index date, and received levodopa at least once during the 3-month observation period before the index date and either at the index date or within 3 months afterward. The treatment change was defined as "addition" when the baseline DA was also prescribed at least once between 30 days and 3 months after the index date or as "replacement" in other cases.
The prevalence of comorbidities was calculated based on the initial DA types prescribed. Patients who received the first DA (defined as initial DA), prescribed as a single DA type in 2013 or later and had ≥3 months of observation before the first prescription of initial DA, were analyzed. Prevalence rates were calculated for all categories of diseases in the observation period, based on ICD-10 chapter categories, 18 and for specific diseases (Table S2). The specific diseases of interest were selected by medical experts among the study authors.
The prevalence was also calculated before adding or replacing a different DA type in the levodopa + DA group. Patients were included in the analysis if they were prescribed a different DA as an additional or replacement drug (this DA was defined as the index DA and the date of the first prescription as the index date), had ≥3 months of observation before the index date, and received levodopa at least once during the 3-month observation period before the index date and either at the index date or within 3 months after the index date.
We estimated 95% confidence intervals for the results. We used SAS version 9.4 (SAS Institute) and Microsoft Excel 2016 (Microsoft) for the analyses.

| Study population
Among 131,980 patients with records of a diagnosis coded as G20 (ICD-10), 50,408 patients were identified ( Figure S1). The mean (standard deviation) age at first diagnosis was 73.4 (10.3) years, and 54.2% were women.

| Trends of dopamine replacement therapy using levodopa and DA
The levodopa group had the highest number of patients among the three groups, including approximately 50% of the population over the study period ( Table 1). The levodopa + DA group included 43.7% of the population in 2014, increasing from previous years; afterward, it remained stable at approximately 42%-43%.
Regarding each DA, the percentage of patients who were prescribed pramipexole without levodopa tended to decrease over the study period; meanwhile, pramipexole in combination with levodopa peaked in 2012, then showed a decreasing trend. The percentage of patients receiving ropinirole without levodopa was stable at approximately 1%, while in combination with levodopa, it fluctuated over time, reaching its lowest level in 2012 and its peak in 2014.
The percentage of rotigotine users has been almost stable since it was launched in 2013; in combination with levodopa, it increased to approximately 10% in 2014 and surpassed 15% in 2016.
In subgroup analyses, similar prescription tendencies of each DA to that in the overall study population were observed for both the DA and levodopa + DA groups (Figure 1; Figure S2). Regarding age, in the DA group, the percentage of patients receiving pramipexole was similar in both age groups ( Figure 1A). In contrast, the percentage of ropinirole users was higher in patients younger than 80 years than in older patients, while the opposite was true for rotigotine. In the levodopa + DA group, after 2013, the decrease in the percentage of patients receiving pramipexole was larger in patients over 80 years than in younger patients ( Figure 1B). After 2015, the percentage of rotigotine users became higher in patients over 80 years of age than in younger patients.

| Addition or replacement pattern of DA treatment changes
We examined the addition or replacement pattern of DA in patients who experienced the treatment change among DAs. The number of patients who added an index DA (3756) was larger than that of those who replaced it with an index DA (384, Table 2). The most frequently observed addition pattern was pramipexole first and then rotigotine, followed by ropinirole and rotigotine, and pramipexole and ropinirole. These three combinations of baseline and index DA were also the most common for replacement patterns.

| Comorbidities before initial DA treatment or DA treatment changes
During the 3-month observation period before the initial DA prescription, the three most common comorbid disease categories were neurological, gastrointestinal, and cardiovascular diseases for all DAs ( Table 3). The difference in the prevalence between DAs was large in dermatologic diseases (highest prevalence in rotigotine, lowest in ropinirole), respiratory diseases (highest in rotigotine, lowest in ropinirole), and endocrine, nutritional, and metabolic diseases (highest in pramipexole, lowest in ropinirole). Among specific diseases, spondylopathy and insomnia had the highest prevalence rates in all DAs (Table 3). These diseases were followed by dementia for rotigotine; for the other two DAs, musculoskeletal and connective tissue disorders. The difference was large in dementia (highest prevalence in rotigotine, lowest in pramipexole), insomnia (highest in pramipexole, lowest in ropinirole), and anxiety (highest in pramipexole, lowest in ropinirole).
Before the index date, the two categories with the highest prevalence rates were the same as those before initial DA prescription: first neurological, then gastrointestinal diseases for all index DAs (Table 3). When the index DA was pramipexole, respiratory diseases had the third-highest prevalence, in contrast to musculoskeletal and connective tissue disorders for the other two DAs. The three categories with the largest prevalence difference between DAs were gastrointestinal, neurological, and musculoskeletal/connective TA B L E 1 Percentage of patients in the levodopa, levodopa + DA, and DA groups, and percentage receiving each DA in the levodopa + DA and DA groups, per calendar year.  Note: Denominator is the number of patients who were prescribed levodopa or DA at least once in the corresponding calendar year.

F I G U R E 1
Percentage of patients receiving each DA according to age (<80 years and ≥80 years) in the DA group (A) and levodopa + DA group (B). DA, dopamine agonist diseases. Rotigotine had the highest prevalence in these categories.
Among specific diseases, the two highest prevalence rates were also the same as those before initial DA prescriptions in all index DAs ( Note: Numbers in parentheses represent 95% confident intervals (lower limit, upper limit).

TA B L E 3
Prevalence of comorbidities before initial DA (A) and before DA treatment change (B).

| DISCUSS ION
Pramipexole was the most popular DA in the study period; however, its prescription has decreased since 2012, whether combined with levodopa or alone. In contrast, the prescription of ropinirole remained relatively stable, while rotigotine prescription, which was Among the three DAs, pramipexole was released first followed by ropinirole; this explains its popularity during the initial period of study and why these in prescription trends occurred after 2013.
Most commonly, pramipexole treatment was modified by either adding or replacing it with rotigotine. Ropinirole treatment was also commonly modified by either adding or replacing it with rotigotine.
We considered two possible reasons for such modifications. First, an increase in neuropsychiatric symptoms, including impulse con- Note: Numbers in parentheses represent 95% confidence intervals (lower limit %, upper limit %).

TA B L E 3 (Continued)
pramipexole who have concerns about psychiatric symptoms. The second reason is probably the similar characteristics of pramipexole and ropinirole, while rotigotine presents differences in the administration route and affinity for each dopamine receptor. 25 When modification in DA treatment is needed due to comorbidities, adverse events, or insufficient effect, a DA with different characteristics might be selected. Additionally, rotigotine may be used for switching simply because it was the newest drug since new treatments generally tend to be used to solve problems with current treatments.
Differences were observed in the prevalence of comorbidities among DAs; these diseases might have influenced the choice of a specific DA, although the actual reasons could not be extrapolated from the database. Notably, in patients using rotigotine as initial or index DA, higher prevalence of gastrointestinal and dermatological diseases was observed before starting treatment with rotigotine than with pramipexole and ropinirole. Gastrointestinal diseases probably determine the choice of rotigotine due to the administration route, a transdermal patch, which may be preferred in these patients.
Dermatologic diseases are possibly associated with prescriptions of moisturizers and/or other topical agents, which help to prevent skin disorders induced by rotigotine. 26 Previous studies reported a positive effect of rotigotine on sleep disturbances and pain 14,16 ; these effects may suggest its use in patients with insomnia and spondylopathy, explaining the higher prevalence of these diseases observed in this study among patients who received rotigotine. The higher prevalence of dementia and spondylopathy at the beginning of rotigotine treatment could be related to patients' age. Our study revealed a higher preference for rotigotine in patients over 80 years, and the dementia prevalence is reportedly higher in patients with PD with higher age. 27 Although an association between spondylopathy prevalence and age is not clear for patients with PD, this disease mainly develops in people older than 50 years, suggesting a higher prevalence in older patients. 28 Patients who were prescribed pramipexole as initial DA showed the highest prevalence of depression, insomnia, and anxiety before initiation compared with other DAs; the choice of pramipexole as baseline DA in patients with depressive symptoms could be related to previous reports of improvements of these symptoms with this drug. 12,13 Some studies investigated factors affecting PD treatment selec- Therefore, comparing the results without considering differences in patients' characteristics between treatments may be inappropriate.
Finally, we inferred the reasons for specific DA choices at the beginning of DA treatment or modifications based on the database analyses; however, the actual reasons could not be extrapolated from the database. Despite these limitations, this study illustrated the differences in patients' backgrounds, including demographics and comorbidities by type of DAs, which might be valuable when considering factors related to the choice of a DA.
In conclusion, pramipexole, the first available DA, was prescribed often at the beginning of the study period; the decrease in its prescription frequency might be related to new DA options, but other causes should also be considered. The frequent change from pramipexole to rotigotine might be associated with the different DA characteristics, such as administration route, side effects, or positive effects on comorbid diseases. The latter may be a relevant factor influencing the choice of a specific DA.

ACK N OWLED G M ENTS
This study was funded by Otsuka Pharmaceutical Co., Ltd. We would like to thank Editage (www.edita ge.com) for English language editing, which was supported by Otsuka Pharmaceutical Co., Ltd.