Abstract
Background
Dopamine receptor agonists (DAs) are widely used as first-line therapeutic agents for Parkinson’s disease. However, comparative clinical trials assessing their safety profiles are limited. This study aims to compare adverse event (AE) data across various DAs to inform personalized treatment strategies.
Methods
AE reports with DAs as the “primary suspicion (PS)” were extracted from the FDA Adverse Event Reporting System (FAERS) database, covering 67 quarters from the second quarter of 2007 to the fourth quarter of 2023. Four disproportionality analysis methods, including the reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian confidence propagation neural network (BCPNN) and multi-item gamma Poisson shrinker (MGPS), were employed to evaluate the risk of AEs.
Results
A total of 19,745,533 DA-related AEs reports were analyzed. The six DAs—pramipexole, ropinirole, cabergoline, rotigotine, bromocriptine and apomorphine—generated 269, 246, 202, 163, 146, and 135 preferred terms positive signals, respectively. Non-ergot DAs (pramipexole, ropinirole, rotigotine and apomorphine) were primarily associated with psychiatric disorders and reported more hallucinations than ergot-derived dopamine agonists (ergot-DAs), with ropinirole showing a slightly higher signal intensity than pramipexole (ROR = 15.76 vs. 11.23). Pramipexole demonstrated the most significant signal for impulse control disorders (ICDs). Compared with pramipexole and ropinirole, rotigotine generally exhibits milder signals in terms of psychiatric disorders such as hallucinations, ICDs, and sleep-related AEs. Administration site-related AEs were more prominent in rotigotine and apomorphine users. Ergot-DAs exhibited higher signal intensities for cardiac disorders, with cabergoline also showing a notable signal for amnestic symptoms (ROR = 340.54), which is not mentioned in the drug label.
Conclusion
This study elucidates the distinct safety profiles of six DAs. Non-ergot DAs are primarily associated with psychiatric AEs, while administration-related AEs are more notable for rotigotine and apomorphine. Ergot-DAs present a higher risk for cardiac valvulopathies. These findings highlight the importance of individualized treatment considerations in clinical practice, emphasizing the need to formulate appropriate treatment plans on patients’ specific conditions.
Supplementary Information
The online version contains supplementary material available at 10.1186/s40360-025-00886-3.
Keywords: Dopamine receptor agonists, Adverse events, Risk signal mining, Psychiatric disorders, FAERS
Introduction
Dopamine receptor agonists (DAs) have become the first-line monotherapy option for the initial treatment of Parkinson’s disease (PD) since the late 1990s, particularly for patients under the age of 65 to 70 years [1, 2]. These agents primarily function by stimulating the postsynaptic receptors to enhance the body’s response to dopamine and restore neurotransmitters balance in the brain [3]. Non-ergot DAs [4], including pramipexole, ropinirole, piribedil, rotigotine, and apomorphine, are widely used in PD treatment [5, 6]. They serve as the preferred initial therapy for early-onset PD and as adjuncts to levodopa in advanced PD [7]. Additionally, these agents are used in the management of restless legs syndrome [8, 9]. In contrast, ergot-derived dopamine agonists (ergot-DAs), such as bromocriptine, cabergoline, and pergolide are no longer recommended for PD treatment [7, 10] and are currently primarily indicated for hyperprolactinemia, galactorrhea, lactation, amenorrhea, and specific forms of acromegaly [8, 11].
Given the widespread distribution of dopamine receptors throughout the human body and the distinct physiological roles of various receptor subtypes, such as regulating prolactin secretion, motor function, and behavior, medications targeting multiple dopamine receptor sites may induce a range of adverse drug reactions (ADRs) in addition to their therapeutic benefits [12]. While several systematic reviews and meta-analyses [7, 13] have evaluated the safety profiles of non-ergot DAs, and our multicenter retrospective study [14] found that pramipexole is associated with a lower incidence of adverse events (AEs) compared to piribedil, no head-to-head clinical trials have been conducted to compare AE incidence rates between different DAs, especially for non-motor symptom-related AEs. Current evidence primarily relies on indirect comparisons from meta-analyses or observational studies, which may be confounded by heterogeneity in study designs and patient populations. This research gap limits the clinical use of these drugs, particularly for PD patients presenting with multiple non-motor symptoms and comorbid conditions [15].
Pharmaceutical post-marketing safety surveillance relies heavily on spontaneous reporting systems. The U.S. Food and Drug Administration’s Adverse Event Reporting System (FAERS) is an open and transparent system database containing diverse AE reports from healthcare professionals, drug manufacturers, and patients worldwide [16]. These reports provide valuable real-world data for post-marketing drug safety evaluations. Previous study has analyzed the frequency of ADRs related to DAs or levodopa using the French pharmacovigilance database [17], which included several DAs and levodopa marketed in France from 1984 to 2008. However, these studies were geographically limited and did not new DAs like rotigotine, which has been available since 2007. Additionally, the FAERS database offers a broader perspective by collecting data from diverse countries and ethnic backgrounds. Furthermore, the FAERS database allows for more robust analysis by restricting suspect strength, thereby reducing the risk of false positives in the analysis results.
Therefore, this study aimed to conduct a safety comparison of DAs using real-world AE data from the FAERS database, analyze the disproportionality measures of ADR signals, and identify unknown or potential signals to inform rational drug use in clinical practice.
Materials and methods
Data sources
This study utilized publicly available data from the FAERS database to collect AE reports. Given that rotigotine was the latest among the eight DAs to be marketed (9 May 2007), the data collection period was limited to the second quarter of 2007 to the fourth quarter of 2023. This timeframe was selected to minimize the bias arising from differences in marketing duration. However, this approach may introduce reporting bias due to regulatory changes over time. The study included only patients treated with DAs. The primary suspicion (PS) drugs were identified through fuzzy matching in the “drug name” field, using both trade and generic names, and referencing the FDA-approved drugs and ADRs. The DAs included were: pramipexole (“PRAMIPEXOLE”, “MIRAPEX” and “SIFROL”), piribedil (“PIRIBEDIL” and “TRIVASTAL”), ropinirole (“ROPINIROLE” and “REQUIP”), rotigotine (“ROTIGOTINE” and “NEUPRO”), apomorphine (“APOMORPHINE” and “APOKYN”), bromocriptine (“BROMOCRIPTINE”, “PARLODEL” and “CYCLOSET”), pergolide (“PERGOLIDE” and “PERMAX”), cabergoline (“CABERGOLINE” and “DOSTINEX”).
Data processing and analysis
Data extracted from the FAERS database were imported into SAS software (version 9.4) for collation and analysis. The Medical Dictionary for Regulatory Activities (MedDRA) (version 26.1) was used to systematically code AEs at system organ class (SOC) and preferred term levels. Since FAERS data are based on spontaneous reporting, the database contains duplicate or withdrawn/deleted reports. Deduplication was performed according to the FDA-recommended method: reports were selected based on PRIMARYID, CASEID, and FDA_DT from the DEMO table, sorted by CASEID, FDA_DT, and PRIMARYID. For reports with the same CASEID, the one with largest FDA_DT was retained; for those with the same CASEID and FDA_DT, the report with the largest PRIMARYID was kept. Four disproportionality analysis methods were employed to identify associations between drugs and ADRs: reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian confidence propagation neural network (BCPNN) and Muti-item Gamma Poisson Shrinker (MGPS). Each method has unique advantages and limitations. The ROR method is computationally simple and high sensitivity, but has lower specificity [18]; in contrast, BCPNN is computationally complex, has moderate sensitivity, but offers high specificity and stable signal [19]. Combining multiple disproportionality analysis methods reduces false-positive and false-negative signals, thereby minimizing bias from reliance on a single algorithm. A signal was generated when the following criteria were met: number of reports a ≥ 3 lower bounds of 95% confidence intervals for ROR and PRR (ROR, PRR) > 1, lower bounds of 95% confidence intervals for IC (IC025) > 0 and 95% confidence intervals for EBGM (EBGM05) > 2. Sensitivity analyses were limited to AEs in which the drug of interest was reported as “PS”.
Statistical description and analysis
Descriptive analyses were conducted on gender, age, report source, and country of the origin for DAs-related AEs. Counting data were presented as case numbers and composition ratio. All statistical analyses were performed using Microsoft Excel software, and a heatmap was generated using TBtools-II software [20].
Results
Descriptive analysis of basic information on AE reports
After deduplication, a total of 19,745,533 AE reports were obtained from the FAERS database, covering 67 quarters from the second quarter of 2007 to the fourth quarter of 2023 (Fig. 1). Among them, there were 8076 AE reports for pramipexole, 7287 reports for rotigotine, 5905 reports for ropinirole, 7313 AE reports for apomorphine, 870 reports for bromocriptine, and 2131 reports for cabergoline. Since piribedil in DAs has not been approved by the FDA in the United States, there is no statistical data about its AEs as the PS drug in the FAERS database. Pergolide was withdrawn from the market in 2007 due to its potential risk of causing damage to heart valves [21]. Following its market withdrawal, only seven cases were reported, which lacked significant statistical power and were excluded from further studies. Consequently, this study included six DAs. The age distribution of users for all six DAs was primarily between 60 and 80 years, with most AEs classified as serious. The median time from initiation of different DAs to AE onset, from shortest to longest, was as follows: apomorphine (0.00 days, interquartile range[IQR] 0.00–21.00 days), bromocriptine (4.50 days, IQR 0.00–244.00 days), cabergoline (21.00 days, IQR 0.00–534.00 days), ropinirole (28.00 days, IQR 0.00–244.00 days), pramipexole(49.00 days, IQR 0.00–425.00 days), and rotigotine (84.00 days, IQR 1.00–427.00 days). Detailed information on reported countries, patient gender and age is presented in Table 1.
Fig. 1.
The flow diagram of selecting DA-related AEs from the FAERS database. AEs, Adverse events; BCPNN, Bayesian confidence propagation neural network; DAs, Dopamine receptor agonists; MGPS, Muti-item Gamma Poisson Shrinker; PRR, Proportional reporting ratio; PS, primary suspicion; PTs, Preferred terms; ROR, Reporting odds ratio
Table 1.
Basic characteristics of AEs reports on the six dopamine receptor agonists (%)
| Characteristics | Apomorphine | Bromocriptine | Cabergoline | Pramipexole | Ropinirole | Rotigotine |
|---|---|---|---|---|---|---|
| Gender | ||||||
| Male | 4166(56.97) | 245(28.16) | 660(30.97) | 2993(37.06) | 2359(39.95) | 3072(42.16) |
| Female | 2725(37.26) | 488(56.09) | 1155(54.20) | 3748(46.41) | 2844(48.16) | 3208(44.02) |
| Not Specified | 422(5.77) | 137(15.75) | 316(14.83) | 1335(16.53) | 702(11.89) | 1007(13.82) |
| Age (years) | ||||||
| <60 | 907(12.40) | 417(47.93) | 1126(52.84) | 1484(18.38) | 1231(20.85) | 677(9.29) |
| 60 ~ 80 | 3803(52.00) | 104(11.95) | 236(11.07) | 2054(25.43) | 1892(32.04) | 2328(31.95) |
| >80 | 660(9.03) | 25(2.87) | 29(1.36) | 511(6.33) | 363(6.15) | 743(10.20) |
| Not Specified | 1943(26.57) | 324(37.24) | 740(34.73) | 4027(49.86) | 2419(40.97) | 3539(48.57) |
| Median | 70(63–76) | 72(63–79) | 40(30–55) | 65(54–74) | 65(55–73) | 37(29–58) |
| Reported countries (top five) | ||||||
| First |
US 6365(87.04) |
US 169(19.43) |
US 438(20.55) |
US 3271(40.50) |
US 3685(62.40) |
US 3421(46.95) |
| Second |
Germany 336(4.59) |
Brazil 168(19.31) |
Brazil 229(10.75) |
Not Specified 1786(22.11) |
UK 548(9.28) |
Germany 817(11.21) |
| Third |
UK 223(3.05) |
Japan 123(14.14) |
Japan 228(10.70) |
Japan 499(6.18) |
Japan 527(8.92) |
Colombia 610(8.37) |
| Fourth |
Not Specified 131(1.79) |
France 85(9.77) |
UK 221(10.37) |
UK 450(5.57) |
France 295(5) |
Japan 492(6.75) |
| Fifth |
Australia 43(0.59) |
China 46(5.29) |
France 198(9.29) |
Germany 351(4.35) |
Germany 216(3.66) |
Mexico 401(5.50) |
| Severity of AE | ||||||
| Serious | 5334(72.94) | 825(94.83) | 1918(90.00) | 4585(56.77) | 3285(55.63) | 4024(55.22) |
| Non-Serious | 1979(27.06) | 45(5.17) | 213(10.00) | 3491(43.23) | 2620(44.37) | 3263(44.78) |
| Outcome | ||||||
| Life-Threatening | 53(0.72) | 37(4.25) | 78(3.66) | 287(3.55) | 163(2.76) | 79(1.08) |
| Hospitalization - Initial or Prolonged | 990(13.54) | 256(29.43) | 500(23.46) | 1820(22.54) | 1189(20.14) | 1469(20.16) |
| Disability | 29(0.40) | 21(2.41) | 58(2.72) | 421(5.21) | 207(3.51) | 117(1.61) |
| Death | 401(5.48) | 51(5.86) | 62(2.91) | 289(3.58) | 171(2.90) | 1143(15.69) |
| Congenital Anomaly | 1(0.01) | 13(1.49) | 66(3.10) | 19(0.24) | 13(0.22) | 1(0.01) |
| Required Intervention | 13(0.18) | 4(0.46) | 4(0.19) | 23(0.28) | 33(0.56) | 3(0.04) |
| Other | 811(11.09) | 617(70.92) | 1511(70.91) | 2760(34.18) | 2233(37.82) | 2189(30.04) |
| Onset time (days) | ||||||
| N(Missing) | 3724(3589) | 178(692) | 341(1790) | 2047(6029) | 1514(4391) | 1809(5478) |
| Mean(SD) |
101.08 (401.51) |
449.09 (1078.68) |
497.78 (916.88) |
419.70 (848.95) |
315.70 (743.15) |
354.33 (601.21) |
| Median(Q1,Q3) |
0.00 (0.00,21.00) |
4.50 (0.00,244.00) |
21.00 (0.00,534.00) |
49.00 (0.00,425.00) |
28.00 (0.00,244.00) |
84.00 (1.00,427.00) |
| Min, Max | 0.00,7313.00 | 0.00,6634.00 | 0.00,4683.00 | 0.00,7716.00 | 0.00,6796.00 | 0.00,5384.00 |
Analysis of AE reports by SOC
AE signals for the six DAs were categorized according to their corresponding SOC levels using MedDRA. The results (Supplementary Table S1) showed that pramipexole, ropinirole, cabergoline, rotigotine, bromocriptine and apomorphine involved 23, 22, 21, 19, 19, and 17 SOC systems, respectively, and generated 269, 246, 202, 163, 146 and 135 signals, respectively. All six DAs were predominantly associated with “nervous system disorders” (apomorphine: 4837 cases; pramipexole: 4352 cases; ropinirole: 3876 cases; rotigotine: 3304 cases; cabergoline: 950 cases; bromocriptine: 739 cases). For non-ergot DAs, “psychiatric disorders” were the most frequently reported (pramipexole: 9481 cases; ropinirole: 4448 cases; rotigotine: 2255 cases; apomorphine: 1364 cases). In contrast, cabergoline and bromocriptine exhibited higher proportion of “cardiac disorders” compared to other non-ergot DAs (12.87% and 12.33%, respectively) (Fig. 2).
Fig. 2.
The proportion of DAs-related AEs at the System Organ Class (SOC) level. AEs, Adverse events; DAs, Dopamine receptor agonists
Signal differences in AEs among the six DAs
After excluding signals unrelated to drug use (e.g., “technical errors in product usage”, “device adhesion issues”), the top 20 reported signal strengths for each DA were analyzed. Results indicated that among the top 20 signals, pramipexole (Table 2) and ropinirole (Table 3) were predominantly associated with “psychiatric disorders” (pramipexole: 13/20; ropinirole: 11/20). Common psychiatric AEs included impulse-control disorders (ICDs) such as gambling disorder, compulsive shopping, and sexual function abnormalities (e.g., hypersexuality). Supplementary Table S2 shows that all DAs are associated with sleep-related risks, non-ergot DAs exhibited high signal intensities for “sleep attacks” (apomorphine ROR = 32.35; pramipexole ROR = 142.10; ropinirole ROR = 156.87; rotigotine ROR = 172.72) and “sudden onset of sleep” (apomorphine ROR = 7.85; pramipexole ROR = 169.30; ropinirole ROR = 354.96; rotigotine ROR = 38.42). In addition, bromocriptine has also been reported to cause a more typical “hypersomnia-bulimia syndrome” (ROR = 484.24). For “general disorders and administration site conditions”, rotigotine and apomorphine reported a higher frequency of AEs (rotigotine: 9/20; apomorphine: 7/20) (Fig. 3D and A, respectively). Common administration site AEs for rotigotine included eczema, dermatitis, itching, hypersensitivity, erythema, rash, and vesicles, while apomorphine-related AEs included necrosis, nodules, and injury at the injection site. In contrast, bromocriptine (Fig. 3B) and cabergoline (Fig. 3C) exhibited higher signal intensities for “cardiac disorders” (cabergoline ROR = 422.70 for cardiac valvular sclerosis; bromocriptine ROR = 1661.06 for tricuspid valvular sclerosis).
Table 2.
Top 20 AEs of Pramipexole with signal strength in target drug signal detection
| System Organ Class(SOC) | Preferred terms (PT) |
Case (n) |
ROR (95% CI Lower) |
PRR (X2) |
IC-2SD | EBGM05 |
|---|---|---|---|---|---|---|
| Psychiatric disorders | Delusional disorder, jealous type | 6 | 1688.67(544.59) | 1688.32(5058.95) | 1.44 | 272.40 |
| Psychiatric disorders | Voyeurism | 8 | 1501.14(579.13) | 1500.73(6347.55) | 1.98 | 306.69 |
| Psychiatric disorders | Gambling disorder | 1188 | 762.04(711.17) | 730.68(604239.00) | 8.38 | 476.21 |
| Psychiatric disorders | Jealous delusion | 56 | 544.42(402.77) | 543.37(22935.30) | 5.22 | 304.30 |
| Psychiatric disorders | Hypersexuality | 346 | 494.76(438.72) | 488.83(130629.00) | 7.33 | 336.34 |
| Psychiatric disorders | Paraphilia | 23 | 485.78(305.48) | 485.39(8635.39) | 3.85 | 237.22 |
| Psychiatric disorders | Paedophilia | 4 | 482.44(158.79) | 482.38(1494.51) | 0.86 | 123.56 |
| Nervous system disorders | Parkinsonism hyperpyrexia syndrome | 24 | 445.64(284.18) | 445.27(8418.35) | 3.91 | 224.82 |
| Psychiatric disorders | Impulse-control disorder | 334 | 440.53(390.39) | 435.44(115092.00) | 7.24 | 306.94 |
| Nervous system disorders | IVth nerve paresis | 4 | 422.14(141.12) | 422.08(1344.26) | 0.87 | 112.95 |
| General disorders and administration site conditions | Vaccination site oedema | 3 | 422.12(119.11) | 422.08(1008.20) | 0.37 | 95.34 |
| Social circumstances | Gambling | 191 | 415.11(354.22) | 412.36(62995.20) | 6.70 | 282.97 |
| Psychiatric disorders | Compulsive shopping | 342 | 379.92(337.72) | 375.43(104484.00) | 7.17 | 273.17 |
| Psychiatric disorders | Dopamine dysregulation syndrome | 65 | 364.20(278.51) | 363.38(19329.70) | 5.37 | 228.80 |
| Investigations | Ultrasound uterus abnormal | 7 | 262.69(118.46) | 262.63(1578.82) | 1.86 | 102.55 |
| Nervous system disorders | Restless arm syndrome | 12 | 259.85(141.49) | 259.74(2680.55) | 2.77 | 122.65 |
| Psychiatric disorders | Excessive sexual fantasies | 5 | 255.85(99.87) | 255.81(1102.07) | 1.29 | 86.77 |
| Musculoskeletal and connective tissue disorders | Camptocormia | 22 | 202.02(129.81) | 201.86(3927.54) | 3.72 | 115.93 |
| Psychiatric disorders | Parkinson’s disease psychosis | 14 | 195.44(112.37) | 195.34(2426.16) | 3.01 | 100.73 |
| Psychiatric disorders | Obsessive-compulsive disorder | 672 | 193.94(178.92) | 189.44(113270) | 6.97 | 157.23 |
Table 3.
Top 20 AEs of ropinirole with signal strength in target drug signal detection
| System Organ Class(SOC) | Preferred terms (PT) |
Case (n) |
ROR (95% CI Lower) |
PRR (X2) |
IC-2SD | EBGM05 |
|---|---|---|---|---|---|---|
| Nervous system disorders | Dyskinesia hyperpyrexia syndrome | 3 | 1811.25(405.34) | 1810.98(3101.11) | 0.23 | 231.69 |
| Nervous system disorders | Alien limb syndrome | 5 | 1097.84(381.41) | 1097.57(3766.01) | 1.20 | 262.26 |
| Psychiatric disorders | Excessive sexual fantasies | 11 | 984.28(488.18) | 983.74(7673.12) | 2.61 | 346.82 |
| Psychiatric disorders | Voyeurism | 4 | 743.11(242.28) | 742.97(2266.49) | 0.85 | 185.31 |
| Psychiatric disorders | Dopamine dysregulation syndrome | 61 | 482.81(366.66) | 481.35(24380.60) | 5.35 | 304.92 |
| Nervous system disorders | Parkinsonism hyperpyrexia syndrome | 18 | 448.48(271.14) | 448.08(6772.68) | 3.47 | 228.6 |
| Social circumstances | Gambling | 141 | 412.09(344.54) | 409.21(49097.8) | 6.40 | 292.67 |
| Social circumstances | Promiscuity | 4 | 357.80(125.18) | 357.73(1239.31) | 0.92 | 109.05 |
| Nervous system disorders | Sudden onset of sleep | 150 | 354.96(298.91) | 352.33(45859.80) | 6.41 | 259.02 |
| Psychiatric disorders | Jealous delusion | 28 | 335.17(225.69) | 334.70(8181.72) | 4.16 | 198.02 |
| Nervous system disorders | Parkinsonian crisis | 9 | 306.22(153.03) | 306.08(2428.88) | 2.31 | 135.81 |
| Psychiatric disorders | Hypersexuality | 118 | 201.40(166.84) | 200.23(21601.30) | 5.91 | 153.23 |
| Psychiatric disorders | Impulse-control disorder | 107 | 170.66(140.21) | 169.75(16771.90) | 5.72 | 130.36 |
| Psychiatric disorders | Sleep attacks | 22 | 156.87(101.90) | 156.70(3196.20) | 3.70 | 95.63 |
| Psychiatric disorders | Gambling disorder | 223 | 146.75(128.11) | 145.14(30114.00) | 6.21 | 119.57 |
| Social circumstances | Sexual activity increased | 5 | 140.42(56.99) | 140.39(653.95) | 1.32 | 53.87 |
| Musculoskeletal and connective tissue disorders | Camptocormia | 11 | 136.28(74.23) | 136.21(1397.57) | 2.61 | 70.26 |
| Psychiatric disorders | Somatic delusion | 11 | 132.22(72.05) | 132.14(1357.40) | 2.61 | 68.30 |
| Psychiatric disorders | Obsessive-compulsive personality disorder | 11 | 130.91(71.35) | 130.84(1344.51) | 2.61 | 67.68 |
| Psychiatric disorders | Paraphilia | 5 | 123.23(50.16) | 123.20(576.60) | 1.31 | 47.74 |
Fig. 3.
The bar plot shows the statistics of SOC regarding PTs of AEs. The ROR values in the figure represent the first 20 signal strengths of the four DAs. AEs, Adverse events; DAs, Dopamine receptor agonists; PTs, Preferred Terms; ROR, Reporting odds ratio; SOC, System organ class
Psychiatric AEs of the six DAs
A comparative analysis of the psychiatric AEs associated with the six DAs revealed distinct patterns of risk (Fig. 4). Pramipexole exhibited the highest diversity and overall signal intensity of psychiatric AEs at the high-level term (HLT) level. Notably, pramipexole was associated with the most significant signals for paraphilias and paraphilic disorders (ROR = 366.77), obsessive-compulsive disorders and symptoms (ROR = 169.12), and ICDs (ROR = 141.25). Ropinirole demonstrated a similar profile to pramipexole in terms of psychiatric AE categories, but exhibited slightly lower overall signal intensity, with the exception of “narcolepsy and associated conditions” and “dyssomnias”, for which ropinirole and rotigotine showed higher signal intensities compared to pramipexole. Among non-ergot DAs, the incidence “hallucinations (excluding sleep-related)” was higher than that observed with ergot-DAs. Specifically, ropinirole exhibited a stronger signal intensity for hallucinations compared to pramipexole (ROR for ropinirole = 15.76 vs. ROR for pramipexole = 11.23) (Fig. 4). Additionally, cabergoline and bromocriptine showed distinct psychiatric AE profiles. Cabergoline was associated with a prominent signal for amnestic symptoms (ROR = 340.54), while bromocriptine exhibited a higher signal intensity for brief psychotic disorders (ROR = 130.16).
Fig. 4.
The heatmap shows the ROR value for the HLT level of psychiatric AEs in the FAERS database under different DAs treatment strategies. The numbers in the grid represent the ROR value, gray cells indicate null values and the colors blue to yellow and red represent the signal intensity of HLT events from low to high (Due to the limitation of the figure, PT signals are not specifically displayed, and only HLT level signals are listed). AEs, Adverse events; DAs, Dopamine receptor agonists; HLT, High-Level Term; PT, Preferred Term; ROR, Reporting odds ratio
Discussion
This study provides a comprehensive analysis of the AE profiles of six DAs using the FAERS database. However, several limitations and contextual factors must be considered when interpreting these findings.
Notably, AE reporting rates were higher among females than males for most DAs, except for apomorphine. This gender discrepancy may suggest a potential gender correlation between gender and the occurrence of AEs associated with DAs. However, the gender-related AEs profile of apomorphine may be influenced by its therapeutic indications. According to the Chinese drug insert and consensus, apomorphine sublingual tablets are indicated for the treatment of erectile dysfunction in men [22]. The demographics of users, particularly age, also influenced the reported characteristics of AEs. Most AEs were observed in individuals aged 60 to 80 years, although bromocriptine and carbamazepine exhibited a broader age distribution due to their use in treating hyperprolactinemia. Furthermore, the majority of AEs in the six DAs were classified as serious, highlighting the need for vigilance regarding the safety of these agents.
Psychiatric AEs were most prominently associated with non-ergot DAs, particularly pramipexole, which accounted for 112 signals. Compared to other DAs, pramipexole exhibited a higher impact on ICDs, psychotic behavioral abnormalities, sexual dysfunction, and disorders of thinking and perception. In contrast, rotergotine and apomorphine were less likely to induce psychotic AEs. This difference may be attributed to the specific dopamine receptors targeted by these agents. Studies have shown that pramipexole and ropinirole primarily act on D2 subfamily receptors (D2, D3, and D4), while rotigotine and apomorphine mainly target D1 subfamily receptors (D1 and D5) and D2 receptors. Pramipexole has high selectivity for the D3 receptor [23, 24], which is localized to limbic areas of the brain associated with cognitive, emotional and endocrine functions. Moreover, Heidbreder CA et al. [25] proposed that the D3 receptor is related to compulsive behaviors. This receptor is also linked to compulsive behaviors, explaining the higher incidence of behavior-related AEs, such as pathological gambling, compulsive shopping, and hypersexuality. A recent meta-analysis [26] concluded that rotigotine had a lower incidence of ICDs compared to pramipexole and ropinirole, suggesting that rotigotine is approximately three times less likely to cause ICDs than pramipexole and ropinirole. Moreover, two studies [27, 28] have shown that the incidence of ICDs with rotigotine is lower than that with pramipexole and ropinirole, and it has been suggested that rotigotine should be considered a special DA that can be used to treat PD patients at risk of ICDs. Whereas, ICDs are less common in hyperprolactinemic patients due to the lower dose of DAs used to treat hyperprolactinemia and their low affinity for D3 receptors [8, 29]. These AE signals may prompt that in PD patients with psychiatric disorders, particularly those with ICDs, it may be advisable to consider gradually discontinuing or reducing DAs and prioritizing alternative non-DA anti-PD treatment options. If the use of DAs is necessary, rotigotine may be a preferred option for PD patients at risk of ICDs.
“Sleep attacks” and “sudden onset of sleep” are considered as the same type of event, which is a common non-motor symptom manifestation in PD [30]. Considering that most of the AEs reporters in the real world can not distinguish between the two AEs, we unify the ROR values of them as “sleep attacks” for discussion. Our results showed that non-ergot DAs overall reported high sleep-related AEs, and ropinirole reported the highest total signal of “sleep attacks” and “sudden onset of sleep”. Ferraiolo M et al. [1] explored the properties of pramipexole, ropinirole and rotigotine from both the clinical and molecular perspectives, indicating thatigotine had better tolerability profile with a lower risk of causing sleep attacks. Several studies [31, 32] have shown that D1 receptor agonists promote wakefulness, while D2 receptor agonists increase sleep. And Möller, J. C. et al. [33] proposed that pramipexole and ropinirole may have triggered sleep attacks by downregulating dopaminergic input to the reticular activating system, possibly by acting on presynaptic receptors. Overall, there is a risk of sleep attacks with all DAs, and it is particularly important to inform patients taking these medications to be aware of risk.
Non-ergot DAs reported more hallucination AEs compared to ergot-DAs, with ropinirole showing a higher signal intensity than pramipexole. However, the incidence of hallucinations varies across studies, with some reporting higher rates with ropinirole [34, 35], while others found opposite results [36] or no significant difference [37]. The conclusions drawn from the research results are inconsistent and fail to provide a definitive consensus, whereas our findings objectively reflect the safety signals of hallucination risks associated with these two drugs in the real world. However, it is relatively clear that both have a higher incidence of hallucinations than rotigotine based on multiple studies [1, 38]. Meanwhile, the conclusion that is also confirmed by the results of this study. There is no authoritative conclusion on the mechanism of hallucinations. From neurobiological perspectives, the basis and the mechanism of hallucinations induced by DAs is not fully understood but may involve the combined action of multiple neurotransmitter systems, including dopamine, serotonin (5-HT), norepinephrine and acetylcholine. The differences in the incidence of hallucinations related to DAs may result from the combined action of multiple receptors rather than a single receptor-induced response. Further research is needed to elucidate the precise mechanisms underlying these differences.
Rotigotine and apomorphine exhibited higher signal intensities for skin AEs related to the administration site. This may be attributed to their drug formulations: rotigotine is available as a transdermal patch, while apomorphine is administered via injection or sublingual tablet. In contrast, ergot-DAs such as bromocriptine and cabergoline are more frequently associated with cardiovascular AEs [39, 40].
This study employed four disproportionality analysis methods to comprehensively evaluate AE risk signals associated with DAs. Combining these methods improved the sensitivity of the analysis and highlighted certain research advantages. However, the FAERS database has limitations, including underreporting, misreporting, and inconsistent data quality. Additionally, reporting bias may exist among different reporting groups, potentially introducing bias into the analysis results. Furthermore, the FAERS database alone can not establish a causal relationship between drug use and AEs. Therefore, our findings should be interpreted as warning signals for clinicians and pharmacists to remain vigilant for potential AEs. Currently, piribedil is widely used in clinical practice, and the inability to include data related to piribedil in this study represents a limitation.
Conclusion
This study’s AE signal mining results reveal that non-ergot DAs are predominantly associated with psychiatric and the nervous system disorders. Specifically, pramipexole exhibited the highest number of cases and the strongest signal intensity related to ICDs. Overall, non-ergot DAs are more likely to induce psychiatric AEs such as hallucinations and sleep-related compared to ergot-DAs. Among them, compared with pramipexole and ropinirole, rotigotine generally exhibits milder signals in terms of psychiatric disorders such as hallucinations, ICDs, and sleep-related AEs. Additionally, administration site-related ADRs were particularly notable for apomorphine and rotigotine. Meanwhile, ergot-DAs such as bromocriptine and cabergoline were associated with valvular heart disease AEs, suggesting they should be considered secondary options in clinical practice, with heightened caution during their use.
These findings underscore the need for vigilance in monitoring psychiatric AEs associated with DAs, especially in patients with PD who exhibit non-motor symptoms. Future research should focus on elucidating the specific mechanisms underlying the occurrence of psychiatric AEs related to DAs through experimental and clinical studies. This study provides valuable insights into the safety profiles of DAs and is expected to assist clinicians in making informed decisions regarding medication choices for various PD patients, particularly those with non-motor symptoms.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We extend special thanks to Dr. Na Liu, Xiangyi Liu and Na He from Peking University Third Hospital for expert consultation, and we acknowledge all the persons reporting AEs in FAERS in this study. We are also grateful to Ms. Yu-Meng Lv from the Peking University Health Science Center for her help with editing. This manuscript was invited to be presented as a poster at the National Academic Conference of Chinese Society of Clinical Pharmacy 2024.
Author contributions
Li Mu: Software; Data curation; Formal analysis; Investigation; Methodology; Project administration; Resources; Writing – original draft; Writing – review & editing. Jing Xu: Formal analysis; Investigation; Methodology; Writing – review & editing. Xiaomei Ye: Formal analysis; Investigation; Methodology; Writing – review & editing. Yongxian Jiang: Investigation; Methodology; Writing – review & editing. Zhanmiao Yi: Conceptualization; Supervision; Funding acquisition; Project administration; Resources; Writing – review & editing. All authors contributed to the article and approved the submitted version.
Funding
This study was funded by the National Natural Science Foundation of China (72104003).
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
This is not applicable since this study was based on publicly available anonymous data from the FDA Adverse Event Reporting System database (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html).
Consent for publication
Not applicable since this was a brief report presenting the secondary analysis of a large dataset.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Ferraiolo M, Hermans E. The complex molecular pharmacology of the dopamine D(2) receptor: implications for pramipexole, ropinirole, and rotigotine. Pharmacol Ther. 2023;245:108392. 10.1016/j.pharmthera.2023.108392. [DOI] [PubMed] [Google Scholar]
- 2.Blandini F, Armentero MT. Dopamine receptor agonists for Parkinson’s disease. Expert Opin Investig Drugs. 2014;23(3):387–410. 10.1517/13543784.2014.869209. [DOI] [PubMed] [Google Scholar]
- 3.Jing XZ, Yang HJ, Taximaimaiti R, et al. Advances in the therapeutic use of non-ergot dopamine agonists in the treatment of motor and non-motor symptoms of Parkinson’s disease. Curr Neuropharmacol. 2023;21(5):1224–40. 10.2174/1570159x20666220915091022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Höglinger GA-O, Trenkwalder C. Diagnosis and treatment of Parkinson´s disease (guideline of the German society for Neurology). Neurol Res Pract. 2024;6(1):30. 10.1186/s42466-024-00325-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Torti M, Bravi D, Vacca L, et al. Are all dopamine agonists essentially the same?? Drugs. 2019;79(7):693–703. 10.1007/s40265-019-01103-2. [DOI] [PubMed] [Google Scholar]
- 6.Fox SH, Katzenschlager R, Lim SY, et al. International Parkinson and movement disorder society evidence-based medicine review: update on treatments for the motor symptoms of Parkinson’s disease. Mov Disord. 2018;33(8):1248–66. 10.1002/mds.27372. [DOI] [PubMed] [Google Scholar]
- 7.Chen XT, Zhang Q, Wen SY, et al. Efficacy and safety of non-ergot dopamine-receptor agonists as an adjunct to levodopa in advanced Parkinson’s disease: A network meta-analysis. Eur J Neurol. 2023;30(3):762–73. 10.1111/ene.15635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Almalki MH, Alsuraikh MA, Almalki E, et al. Impulse control disorders in patients with dopamine agonist-treated pituitary adenomas: a cross-sectional multicenter study. Pituitary. 2024;27(2):197–203. 10.1007/s11102-024-01383-2. [DOI] [PubMed] [Google Scholar]
- 9.Gossard TR, Trotti LM, Videnovic A, et al. Restless legs syndrome: contemporary diagnosis and treatment. Neurotherapeutics. 2021;18(1):140–55. 10.1007/s13311-021-01019-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zhang J, Liu B, Zheng Y, et al. Pramipexole for Chinese people with primary restless legs syndrome: a 12-week multicenter, randomized, double-blind study. Sleep Med. 2015;16(1):181–5. 10.1016/j.sleep.2014.09.015. [DOI] [PubMed] [Google Scholar]
- 11.Woitalla D, Buhmann C, Hilker-Roggendorf R, et al. Role of dopamine agonists in Parkinson’s disease therapy. J Neural Transm (Vienna). 2023;130(6):863–73. 10.1007/s00702-023-02647-0. [DOI] [PubMed] [Google Scholar]
- 12.Latif S, Jahangeer M, Maknoon Razia D, et al. Dopamine in Parkinson’s disease. Clin Chim Acta. 2021;522:114–26. 10.1016/j.cca.2021.08.009. [DOI] [PubMed] [Google Scholar]
- 13.Chen XT, Zhang Q, Chen FF, et al. Comparative efficacy and safety of six non-ergot dopamine-receptor agonists in early Parkinson’s disease: a systematic review and network meta-analysis. Front Neurol. 2023;14:1183823. 10.3389/fneur.2023.1183823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li W, Zhang H, Zhang Y, et al. Comparison of the effectiveness, safety, and costs of anti-Parkinson drugs: A multiple-center retrospective study. CNS Neurosci Ther. 2024;30(4):e14531. 10.1111/cns.14531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yi ZM, Willis S, Zhang Y, et al. Impact of a collaborative pharmaceutical care service for patients with Parkinson’s disease. Front Pharmacol. 2021;12:793361. 10.3389/fphar.2021.793361. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Zhu H, Qu Y, Du Z, et al. Mining and analysis of adverse event signals of Cariprazine based on the real-world data of FAERS database. J Affect Disord. 2024;347:45–50. 10.1016/j.jad.2023.11.076. [DOI] [PubMed] [Google Scholar]
- 17.Perez-Lloret S, Bondon-Guitton E, Rascol O, et al. Adverse drug reactions to dopamine agonists: a comparative study in the French pharmacovigilance database. Mov Disord. 2010;25(12):1876–80. 10.1002/mds.23204. [DOI] [PubMed] [Google Scholar]
- 18.Xiang Y, Xie Y, Fau - Yi D, Yi D. [Pharmacovigilance of major parmaceutical innovation]. China J Chin Materia Med. 2011;36(20):2831. 10.1111/j.1744-7909.2011.01037.x. [PubMed] [Google Scholar]
- 19.Norén GN, Bate A, Fau - Orre R, Orre R, Fau - Edwards IR, et al. Extending the methods used to screen the WHO drug safety database towards analysis of complex associations and improved accuracy for rare events. Stat Med. 2006;25(21):3740–57. 10.1002/sim.2473. [DOI] [PubMed] [Google Scholar]
- 20.Chen C, Wu Y, Li J, et al. TBtools-II: A one for all, all for one bioinformatics platform for biological big-data mining. Mol Plant. 2023;16(11):1733–42. 10.1016/j.molp.2023.09.010. [DOI] [PubMed] [Google Scholar]
- 21.Corvol JC, Anzouan-Kacou JB, Fauveau E, et al. Heart valve regurgitation, pergolide use, and Parkinson disease: an observational study and meta-analysis. Arch Neurol. 2007;64(12):1721–6. 10.1001/archneur.64.12.1721. [DOI] [PubMed] [Google Scholar]
- 22.Hongjun L. Advances and concensus on erectile dysfunction. Chin J Hum Sexuality. 2011.
- 23.Wilson SM, Wurst MG, Whatley MF, et al. Classics in chemical neuroscience: Pramipexole. ACS Chem Neurosci. 2020;11(17):2506–12. 10.1021/acschemneuro.0c00332. [DOI] [PubMed] [Google Scholar]
- 24.Tundo AA-OX, Betro SA-O, de Filippis RA-OX, et al. Pramipexole augmentation for Treatment-Resistant unipolar and bipolar depression in the real world: A systematic review and meta-analysis. Life (Basel). 2023;13(4):1043. 10.3390/life13041043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Heidbreder CA, Newman AH. Current perspectives on selective dopamine D(3) receptor antagonists as pharmacotherapeutics for addictions and related disorders. Ann N Y Acad Sci. 2010;1187:4–34. 10.1111/j.1749-6632.2009.05149.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Soileau LG, Talbot NC, Storey NR, et al. Impulse control disorders in Parkinson’s disease patients treated with pramipexole and ropinirole: a systematic review and meta-analysis. Neurol Sci. 2024;45(4):1399–408. 10.1007/s10072-023-07254-1. [DOI] [PubMed] [Google Scholar]
- 27.Frampton JE. Rotigotine transdermal patch: A review in Parkinson’s disease. CNS Drugs. 2019;33(7):707–18. 10.1007/s40263-019-00646-y. [DOI] [PubMed] [Google Scholar]
- 28.Raeder V, Boura I, Leta V, et al. Rotigotine transdermal patch for motor and non-motor Parkinson’s disease: A review of 12 years’ clinical experience. CNS Drugs. 2021;35(2):215–31. 10.1007/s40263-020-00788-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hamblin R, Karavitaki N. Impulse control disorders in patients with pituitary tumors treated with dopamine agonists: A systematic review. Arch Med Res. 2023;54(8):102910. 10.1016/j.arcmed.2023.102910. [DOI] [PubMed] [Google Scholar]
- 30.Távora DG, de Bruin VM, Lopes Gama R, et al. The nature of excessive sleepiness and sudden sleep onset in Parkinson׳s disease. Sleep Sci. 2014;7(1):13–8. 10.1016/j.slsci.2014.07.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.McCarthy AP, Svensson KA, Shanks E, et al. The dopamine D1 receptor positive allosteric modulator Mevidalen (LY3154207) enhances wakefulness in the humanized D1 mouse and in Sleep-Deprived healthy male volunteers. J Pharmacol Exp Ther. 2022;380(3):143–52. 10.1124/jpet.121.000719. [DOI] [PubMed] [Google Scholar]
- 32.Hyacinthe C, Barraud Q, Tison F, et al. D1 receptor agonist improves sleep-wake parameters in experimental parkinsonism. Neurobiol Dis. 2014;63:20–4. 10.1016/j.nbd.2013.10.029. [DOI] [PubMed] [Google Scholar]
- 33.Möller JC, Eggert Km Fau - Unger M, Unger M Fau - Odin P, et al. Clinical risk-benefit assessment of dopamine agonists. Eur J Neurol. 2008;15 Suppl 2:15–23. 10.1111/j.1468-1331.2008.02214.x [DOI] [PubMed]
- 34.Sako WA-O, Kogo YA-O, Koebis M, et al. Comparative efficacy and safety of adjunctive drugs to levodopa for fluctuating Parkinson’s disease - network meta-analysis. NPJ Parkinsons Dis. 2023;9(1):143. 10.1038/s41531-023-00589-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Singer C. Adverse effects in the treatment of Parkinson’s disease. Expert Rev Neurother. 2002;2(1):105–18. 10.1586/14737175.2.1.105. [DOI] [PubMed] [Google Scholar]
- 36.Etminan M, Gill S, Fau - Samii A, Samii A. Comparison of the risk of adverse events with pramipexole and ropinirole in patients with Parkinson’s disease: a meta-analysis. Drug Saf. 2003;26(6):439–44. 10.2165/00002018-200326060-00005. [DOI] [PubMed] [Google Scholar]
- 37.Zhao H, Ning Y, Cooper J, et al. Indirect comparison of ropinirole and pramipexole as levodopa adjunctive therapy in advanced Parkinson’s disease: A systematic review and network meta-analysis. Adv Ther. 2019;36(6):1252–65. 10.1007/s12325-019-00938-1. [DOI] [PubMed] [Google Scholar]
- 38.Poewe WH, Rascol O, Fau - Quinn N, Quinn N, Fau - Tolosa E, et al. Efficacy of pramipexole and transdermal rotigotine in advanced Parkinson’s disease: a double-blind, double-dummy, randomised controlled trial. Lancet Neurol. 2007;6(6):513–20. 10.1016/S1474-4422(07)70108-4. [DOI] [PubMed] [Google Scholar]
- 39.Csoti I, Jost WH, Reichmann H. Parkinson’s disease between internal medicine and neurology. J Neural Transm (Vienna). 2016;123(1):3–17. 10.1007/s00702-015-1443-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Elenkova A, Shabani R, Kalinov K, et al. Increased prevalence of subclinical cardiac valve fibrosis in patients with prolactinomas on long-term bromocriptine and cabergoline treatment. Eur J Endocrinol. 2012;167(1):17–25. 10.1530/eje-12-0121. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.