Introduction
Antipsychotic drugs effectively reduce psychotic symptoms and prevent relapse (Efthimiou et al., Reference Efthimiou, Taipale, Radua, Schneider-Thoma, Pinzón-Espinosa, Ortuño and Luykx2024; Højlund, Kemp, Haddad, Neill, & Correll, Reference Højlund, Kemp, Haddad, Neill and Correll2021; Leucht et al., Reference Leucht, Bauer, Siafis, Hamza, Wu, Schneider-Thoma and Davis2021, Reference Leucht, Tardy, Komossa, Heres, Kissling, Salanti and Davis2012). These effects are thought to result from reduced striatal dopamine signalling (Kaar, Natesan, McCutcheon, & Howes, Reference Kaar, Natesan, McCutcheon and Howes2020; Strange, Reference Strange2008). Antipsychotic drugs can be divided into D2 receptor (D2R) antagonists (e.g. olanzapine, haloperidol, risperidone), which decrease dopamine signalling by antagonizing dopamine D2R, and partial D2R agonists (e.g. aripiprazole, brexpiprazole), which stimulate D2R but to a lower extent than endogenous dopamine (Burris et al., Reference Burris, Molski, Xu, Ryan, Tottori, Kikuchi and Molinoff2002; Gründer, Kungel, Ebrecht, Göröcs, & Modell, Reference Gründer, Kungel, Ebrecht, Göröcs and Modell2006; Kinghorn & McEvoy, Reference Kinghorn and McEvoy2005).
Long-term exposure to antipsychotic treatment has been shown to alter pre- and postsynaptic dopamine neurotransmission and has been implied to upregulate D2R (Chouinard et al., Reference Chouinard, Samaha, Chouinard, Peretti, Kanahara, Takase and Iyo2017). This phenomenon has been reported for D2R antagonists, but little is known regarding the effects of partial agonists. The dopamine supersensitivity hypothesis states that long-term striatal D2R blockade by antipsychotics results in a compensatory upregulation of D2R expression and/or increase in the fraction of D2 receptors in the high-affinity state (Graff-Guerrero et al., Reference Graff-Guerrero, Mamo, Shammi, Mizrahi, Marcon, Barsoum and Kapur2009). Antipsychotic dose reduction or discontinuation after D2R upregulation is hypothesized to increase dopamine-mediated signalling, followed by a worsening of psychotic symptoms, called ‘supersensitivity psychosis’ (Servonnet & Samaha, Reference Servonnet and Samaha2020; Yin, Barr, Ramos-Miguel, & Procyshyn, Reference Yin, Barr, Ramos-Miguel and Procyshyn2017). Antipsychotic-induced D2R upregulation may therefore increase the risk for relapse and might contribute to the high relapse rates after antipsychotic discontinuation (Di Capite, Upthegrove, & Mallikarjun, Reference Di Capite, Upthegrove and Mallikarjun2018; Kishi et al., Reference Kishi, Ikuta, Matsui, Inada, Matsuda, Mishima and Iwata2019; Leucht et al., Reference Leucht, Tardy, Komossa, Heres, Kissling, Salanti and Davis2012; Thompson et al., Reference Thompson, Winsper, Marwaha, Haynes, Alvarez-Jimenez, Hetrick and Sullivan2018; Zipursky, Menezes, & Streiner, Reference Zipursky, Menezes and Streiner2014). Yet, it remains unknown whether such a D2R upregulation occurs directly after tapering and whether it is observed in both antagonist and partial agonist users.
Evidence for antipsychotic-induced D2R upregulation in humans is sparse and inconclusive. In a meta-analysis of human PET studies, a minor increase in D2/3R availability was observed in schizophrenia patients who had previously used antipsychotics, which was absent in drug-naïve patients (Howes et al., Reference Howes, Kambeitz, Kim, Stahl, Slifstein, Abi-Dargham and Kapur2012). A PET study including nine patients with schizophrenia showed a 30% increase in D2/3R availability 14 days after discontinuation of moderate to high doses of haloperidol, perphenazine, risperidone, or olanzapine compared to eight antipsychotic-naïve patients (Silvestri et al., Reference Silvestri, Seeman, Negrete, Houle, Shammi, Remington and Seeman2000). In contrast, no differences in striatal D2/3R availability were found between 14 controls and 25 FEP patients 7 weeks after discontinuation of aripiprazole, amisulpride, blonanserin, olanzapine, paliperidone, quetiapine, or risperidone (S. Kim et al., Reference Kim, Shin, Santangelo, Veronese, Kang, Lee and Kim2021).
For partial D2R agonists, this mechanism of D2R upregulation may not take place, as dopamine transmission is less affected compared to D2R antagonistic drugs. Animal studies indicate that partial D2R agonists do not cause dopamine supersensitivity (Chouinard et al., Reference Chouinard, Samaha, Chouinard, Peretti, Kanahara, Takase and Iyo2017). Rat studies showed that exposure to aripiprazole or brexpiprazole, in contrast to haloperidol, did not relate to increased D2R density (Amada et al., Reference Amada, Akazawa, Ohgi, Maeda, Sugino, Kurahashi and Futamura2019; Tadokoro et al., Reference Tadokoro, Okamura, Sekine, Kanahara, Hashimoto and Iyo2012), which led to the suggestion that partial D2R agonists may prevent the development of dopamine supersensitivity (Tadokoro et al., Reference Tadokoro, Okamura, Sekine, Kanahara, Hashimoto and Iyo2012). Recently, high D2R affinity antagonist antipsychotics have been associated with a greater risk of relapse than low-affinity antagonists and partial agonist antipsychotics (Gangadin et al., Reference Gangadin, de Beer, Wijnen, Begemann, van Beveren, Boonstra and Sommer2025). This effect could result from stronger upregulation with high-affinity D2R antagonists. To better predict relapse risk and enable safe tapering of antipsychotic medication, a better understanding of neurobiological aspects of relapse risk after discontinuation is needed.
The present study examines the longitudinal effects of discontinuing partial D2R agonists and D2R antagonists at two timepoints: 1 week and 6–8 weeks after discontinuation, as dopamine supersensitivity may be a time-dependent process. D2/3R availability in striatum was measured with [11C]raclopride PET, using the non-displaceable binding potential (BPND) as outcome parameter in 22 symptomatic remitted FEP patients and 15 matched controls for comparison. We hypothesized that discontinuation of D2R antagonists would result in higher D2/3R availability than partial D2R agonist discontinuation and controls after 1 week, with a normalization of D2R availability after 6–8 weeks. For partial D2R agonists, we did not expect any short or longer-term differences in D2/3R BPND compared to controls.
Methods and Materials
Participants
Patients were recruited from the HAMLETT trial (Handling Antipsychotic Medication: Long-term Evaluation of Targeted Treatment), a multicentre, single-blind randomized controlled trial in the Netherlands on antipsychotic medication reduction/discontinuation (Begemann et al., Reference Begemann, Thompson, Veling, Gangadin, Geraets, Van’T Hag and Sommer2020) Participation in the present PET study was optional for those who completely discontinued their antipsychotic treatment. For the HAMLETT trial, participants were included if they were aged 16 to 60 years; used antipsychotic medication; had achieved symptomatic remission according to the treating psychiatrist for 3–6 months; had a DSM-5 or ICD-10 diagnosis of first episode of schizophrenia, schizoaffective disorder, schizophreniform disorder, brief psychotic disorder, delusional disorder, or unspecified schizophrenia spectrum and other psychotic disorder. Patients were excluded if they had exhibited dangerous or harmful behaviour during FEP. For the PET study, patients were excluded in case of a neurological disorder, substance dependency, and pregnancy.
Male healthy controls were scanned in a prior study (NL48500.042.14), and female healthy controls were recruited via advertisement. Control subjects had to be aged 18–60 years and were excluded from the study in case of a neurological disorder, substance dependency, pregnancy, MRI incompatible implants, history of psychiatric disorders or psychotropic medication use. This group was matched for age and sex to the patient group. All participants provided written informed consent after oral and written explanation of the study. Ethical approval was obtained from the research and ethics committee of the University Medical Center Groningen (NL64040.042.17 and NL48500.042.14).
Study design
FEP patients who discontinued antipsychotic medication after a gradual tapering trajectory (tapering procedure and schedules are described in Supplementary Material 1) underwent two [11C]raclopride PET scans: 1 week and 6–8 weeks after discontinuation. Patients who relapsed or restarted antipsychotic medication after the first and before the second PET scan did not undergo the follow-up scan (n = 1 partial agonist user, n = 5 D2R antagonist users). Two antagonist discontinuers declined to participate in the follow-up scan. Blood samples were collected prior to the PET scan to determine antipsychotic concentrations. PET data of one follow-up scan of a participant with detectable antipsychotic serum concentrations (8
$ \mu $
g/L olanzapine) were excluded from the analyses because the participant had restarted antipsychotic medication without notification. Participants were followed up for 12 months to assess relapse. Relapse was defined as hospitalization due to exacerbation of psychotic symptoms, judgement by the treating clinician, or reinstatement of antipsychotic medication. Medication use before and after initial antipsychotic discontinuation was derived from clinician reports and dispensation data from the Foundation for Pharmaceutical Statistics, from which olanzapine equivalents were calculated (Leucht et al., Reference Leucht, Crippa, Siafis, Patel, Orsini and Davis2020; Leucht, Samara, Heres, & Davis, Reference Leucht, Samara, Heres and Davis2016). Antipsychotic drugs were classified into antagonists (i.e. olanzapine, risperidone, amisulpride) or partial D2R agonists (aripiprazole). Antipsychotic concentrations in plasma collected on the day of the first PET scan were not detectable, except in two patients who had discontinued aripiprazole and showed concentrations just above the lower limit of quantitation (12 and 13 μg/L; LLOQ 10 μg/L). Data from these patients were included in the analyses. Data from one person who used dexamphetamine was excluded from the analyses as dexamphetamine administration has been shown to lower D2/3R BPND (Weinstein et al., Reference Weinstein, van de Giessen, Rosengard, Xu, Ojeil, Brucato and Abi-Dargham2018). Supplementary Material 2 displays a flowchart of participants in the study. Prior to antipsychotic tapering, demographic characteristics and substance use were measured with the CASH (Andreassen, Flaum, & Arndt, Reference Andreassen, Flaum and Arndt1992), psychotic illness severity was measured with the Positive and Negative Syndrome Scale (Kay, Fiszbein, & Opler, Reference Kay, Fiszbein and Opler1987), and global functioning was measured with the GAF (Jones, Thornicroft, Coffey, & Dunn, Reference Jones, Thornicroft, Coffey and Dunn1995).
Imaging acquisition and kinetic analysis for [ 11 C]raclopride
[11C]raclopride PET scans were performed with a Biograph mCT40 PET/CT camera (Siemens Medical Solutions, Knoxville, Tennessee, USA). The head of participants was fixed in a headrest to prevent motion artifacts and placed in the centre of the field of view of the PET/CT camera. Participants were required to refrain from alcohol for 24 hours, smoking for 12 hours, and eating for 4 hours before the PET scan, since alcohol, smoking and eating can stimulate the release of dopamine. A low-dose computed tomography (CT) scan was performed to correct for attenuation. Participants received a bolus injection of [11C]raclopride over a period of 1 minute, starting 10 seconds prior to the 60 minutes dynamic acquisition procedure. The mean (SD) injected dose of [11C]raclopride was 210 (25) MBq in controls and 205 (45) MBq in FEP patients at the first scan and 206 (50) MBq at the second scan. The injected dose did not differ between controls and FEP patients at the first (t = 0.42, p = 0.68) or second scan (t = 0.27, p = 0.79). Images were corrected for attenuation, scatter, and radioactive decay and iteratively reconstructed (3 iterations, 21 subsets) into 25 timeframes: frame 1–7: 10s, frame 8–9: 30s, frame 10–12: 60s, frame 13–14: 120s, frame 15–16: 180s, frame 17–18: 180s, frame 19–23: 300s, and frame 24–25: 600s.
On the same day as the first PET scan, an anatomic 1mm isotropic 3D T1 MRI scan was conducted with a 3 Tesla Magnetom Prisma MRI system (Siemens Medical Solutions USA, Inc). The average of all frames of the PET scan was used to coregister the PET scan to the individual T1 MRI scan and spatially normalized to the Montreal Neurological Institute (MNI) template, using PMOD (version 4.1, PMOD Technologies Ltd, Zürich, Switzerland). In case of missing MRI scans (n = 4), the R1 map of the PET images, obtained with the simplified reference tissue model (SRTM) (Lammertsma & Hume, Reference Lammertsma and Hume1996), was used as anatomical input for normalization. The Hammers maximum probability atlas (Hammers N3083 1MM) (Hammers et al., Reference Hammers, Allom, Koepp, Free, Myers, Lemieux and Duncan2003) was used to define regions-of-interest for the cerebellum, putamen, caudate nucleus, nucleus accumbens and the striatum (i.e. caudate nucleus, nucleus accumbens, and putamen combined). Time activity curves were computed for all these brain areas. For each scan, the dopamine D2/3R BPND in striatal brain regions was calculated with SRTM2 (Wu & Carson, Reference Wu and Carson2002) using the cerebellum as a reference region. The median k2’ of the entire cortex was determined with SRTM and used as input for SRTM2.
Statistical analyses
A linear mixed effects model (LMEM) was applied to test whether the mean [11C]raclopride BPND differed between D2R antagonist users and partial D2R agonist users 1 week and 6–8 weeks after discontinuation, and controls (Supplementary Material 3). The BPND was used as dependent variable, with random intercepts for participants and fixed effects for time (first or second scan), the categorical variable group including D2R antagonist users, partial agonist users, and controls, and the interaction between time and group (Bates, D., Maechler, M., Bolker, B., & Walker, Reference Bates, Maechler, Bolker and Walker2015). Similarly, LMEM with fixed effect for group, including relapse, no relapse, and controls, was used to test whether BPND differed between participants with or without relapse and controls. Planned contrasts with LMEM estimated marginal means (Searle, Speed, & Milliken, Reference Searle, Speed and Milliken1980) tested BPND differences between the D2R antagonists, partial D2R agonists, and controls at the first and second scan, and the change over time for the D2R antagonist and partial D2R agonist group. Similarly, contrasts were used to compare BPND in participants with and without relapse, and controls at the first and second scan, and the change in BPND between the scans for the relapse and nonrelapse group. Sensitivity analyses include LMEM on only the putamen and the caudate nucleus (Supplementary Material 4 and 5). A χ2 test was performed to assess whether the incidence of relapse differed in D2R antagonist and partial agonist discontinuers. Statistical analyses were performed in R (version 4.3.2) via Rstudio (version 2023.12.1.402) (R Core Team, 2020).
Results
Demographic and clinical characteristics
The analyses include 22 FEP patients (n = 4, 18% females; n = 18, 82% males) who underwent a [11C]raclopride PET scan approximately one week after antipsychotic discontinuation. Sixteen participants (73%) discontinued D2R antagonist antipsychotics (e.g. olanzapine n = 8, amisulpride n = 2, quetiapine n = 2, haloperidol n = 2) and 6 participants (28%) discontinued the partial D2R agonist aripiprazole. The follow-up [11C]raclopride PET scan at 6–8 weeks was completed by eight former antagonist and five former partial agonist users. Fifteen controls (n = 4, 27% females; n = 11, 73% males) underwent a single [11C]raclopride PET scan. Table 1 shows the demographic and clinical characteristics of the participants.
Table 1. Sociodemographic and clinical characteristics of FEP patients and controls
a Positive and Negative Syndrome Scale (PANSS), measured at 3-6 months after remission from FEP.
b Global Assessment of Functioning (GAF), measured at 3-6 months after remission from FEP.
c In olanzapine equivalents mg/day.
Short-term effects of discontinuation of partial D 2 R agonist and D 2 R antagonist antipsychotics
The linear mixed effects model (LMEM) with the antagonist group as reference showed a significant effect for group (partial D2R agonist users: estimate = -1.21, CI = -1.68 to -0.75, p < 0.001; control group: estimate = -0.36, CI = -0.71 to -0.01, p < 0.001) and the interaction between time and group (estimate = 1.54, CI = 0.84 to 2.24, p < 0.0001 for partial agonist by 2-month follow-up interaction) (Supplementary Material 3).
One week after antipsychotic discontinuation, the BPND in striatum was significantly higher in FEP patients who discontinued D2R antagonists (mean 3.98, SD 0.48) compared to partial D2R agonist (mean 2.76, SD 0.66; t = 5.250, p < 0.001, CI = 0.749 to 1.681) and controls (mean 3.62, SD 0.43; t = 2.063, p = 0.045, CI = 0.008 to 0.708) (Figure 1). The striatal BPND was lower in participants who had discontinued partial D2R agonists in the past week than controls (t = -3.668, p = 0.001, CI = -1.326 to -0.386).
Figure 1. Averaged and normalized [11C]raclopride PET BPND maps showing D2/3R availability in the striatum for FEP patients 1 week after discontinuation of treatment with antagonist (left) or partial agonist (middle) antipsychotics and controls (right). Images are shown in axial (a) and sagittal (b) views.
Longer-term effects of discontinuation of partial D 2 R agonist and D2R antagonist antipsychotics
At 6-8 weeks follow-up, there was no difference in BPND between participants who had discontinued D2R antagonists and controls (t = 1.166, p = 0.250, CI = -0.184 to 0.691). Striatal BPND at 6-8 weeks follow-up was higher in former partial agonist users (mean 4.21, SD 0.45) than controls (mean 3.62, SD 0.43; t = 2.289, p = 0.027, CI = 0.069 to 1.085), but did not differ between former partial D2R agonist users and former D2R antagonist users (mean 3.88, SD 0.44; t = 1.147, p = 0.257, CI = -0.245 to 0.893) (Figure 2). In partial agonist discontinuers, there was a significant increase in striatal BPND over time (t = -5.063, p < 0.001, CI = -2.034 to -0.832), while the BPND did not differ significantly over time in former D2R antagonist users (t = 0.502, p = 0.621, CI = -0.329 to 0.539).
Figure 2. Averaged and normalized [11C]raclopride PET BPND maps showing D2/3R availability in striatum for FEP patients 6-8 weeks after discontinuation with antagonist antipsychotics (left) or partial agonist antipsychotics (middle) and controls (right). Images are shown in axial (a) and sagittal (b) views.
Psychotic relapse after antipsychotic discontinuation
Fourteen FEP participants (64%) experienced relapse within the 12 months after antipsychotic discontinuation, while eight participants (36%) did not experience relapse. Five former D2R antagonist users and one former partial D2R agonist user experienced a relapse or symptom return and restarted antipsychotic treatment before the second scan at two months follow-up. Relapses occurred more often in participants who discontinued D2R antagonists than partial D2R agonists (χ2 = 5.32, p = 0.021) (Figure 3).
Figure 3. Frequency of annual psychotic relapse following antipsychotic discontinuation in FEP patients who stopped D2 partial D2R agonist (n = 6) and D2R antagonist (n = 16) antipsychotics.
LMEM showed an interaction between time and group (estimate = 1.01, CI = 0.17 to 1.85, p = 0.020 for no relapse by 2-month follow-up interaction). The BPND in striatum one week after discontinuation showed a trend towards lower values in patients who remained stable (mean: 3.35, SD 0.89) than in those who relapsed (mean 3.82, SD 0.65) (t = -1.789, p = 0.080, CI = -0.983 to 0.058). There was no significant difference in BPND between controls and participants with (t = -0.901, p = 0.372, CI = -0.632 to 0.241) or without relapse (t = 1.046, p = 0.301, CI = -0.247 to 0.781) one week after discontinuation.
After being without medication for 2 months, the BPND in striatum was significantly higher in participants without relapse (mean 4.21, SD 0.38) than in controls (t = 2.324, p = 0.025, CI = 0.079 to 1.107). The BPND after 2 months did not differ between patients with later relapse (mean 3.67, SD 0.38) compared to those who remained stable (t = -1.583, p = 0.120, CI = -1.238 to 0.148), or controls (t = 0.154, p = 0.878, CI = -0.584 to 0.681). In FEP patients who remained stable, BPND showed a significant increase over time (+26%; t = -2.990, p = 0.009, CI = -1.475 to -0.245), while BPND did not change over time in those who relapsed within a year (-4%; t = 0.467, p = 0.644, CI = -0.497 to 0.791).
Discussion
The present study is the first to test short and longer-term molecular effects of discontinuing D2R antagonist and partial D2R agonists antipsychotics on D2/3R availability in 22 remitted FEP patients. One week after discontinuation, D2R antagonist discontinuers showed increased D2/3R availability compared to controls, while partial agonist users showed decreased D2/3R availability 1 week after discontinuation. After 2 months without medication, former antagonist users showed similar D2/3R availability to controls, while former partial agonist users showed increased D2/3R availability.
Short-term effects of discontinuation of D2R antagonist and partial D2R agonist antipsychotics
Our short-term findings support the dopamine supersensitivity hypothesis for antagonist antipsychotics and are in line with previous studies showing D2/3R upregulation after antipsychotic use in schizophrenia (Howes et al., Reference Howes, Kambeitz, Kim, Stahl, Slifstein, Abi-Dargham and Kapur2012). These results may also provide a neurobiological understanding of the particularly high relapse risk after tapering high D2R affinity antipsychotics in a previous study with FEP patients, from which the participants of the present study were a subset (Gangadin et al., Reference Gangadin, de Beer, Wijnen, Begemann, van Beveren, Boonstra and Sommer2025). Exposure to high D2R affinity antipsychotics could lead to a stronger D2R blockade and hence greater D2R upregulation, which could increase relapse risk after antipsychotic dose-reduction/discontinuation.
In partial D2R agonist discontinuers, D2/3R availability was lower compared to both antagonist discontinuers and healthy controls one week after discontinuation. As aripiprazole is a partial agonist, its use may have caused a downregulation of D2R as compared to controls. However, the lower availability could also reflect a transient increase in dopamine synthesis in the patients who discontinued aripiprazole. Given aripiprazole’s long half-life of 75 hours (Casey & Canal, Reference Casey and Canal2017), a minimal trace of the drug may still have been present 1 week after discontinuation during the first PET scan, potentially occupying some D2R. Although most partial agonist users had undetectable blood levels at scanning (mean 8 days post-discontinuation), two participants had low but detectable levels. Although residual drug traces may have had some effect, the robust differences we observed 1 week after antagonist and partial agonist cessation are too large to be solely the result of residual aripiprazole in the brain of former agonist users.
Longer-term effects of antipsychotic discontinuation and the risk of relapse
At 2 months follow-up, D2/3R availability in antagonist discontinuers was no longer significantly different from controls. Former users or the partial agonist aripiprazole showed a significant increase in D2/3R availability over the course of 2 months after discontinuation, to eventually higher D2/3R availability at 6-8 weeks follow-up than controls.
Consistent with findings from the larger group of which the participants in the present study were part of, aripiprazole users less frequently experienced relapse after discontinuation than high-affinity antagonist antipsychotics (Gangadin et al., Reference Gangadin, de Beer, Wijnen, Begemann, van Beveren, Boonstra and Sommer2025). Twelve months after discontinuation, 83% of former partial D2R agonist users remained relapse-free compared to only 17% of the antagonist group, despite similar baseline characteristics.
Participants without relapse showed a significant increase in D2/3R availability over time, resulting in higher D2/3R availability after two medication-free months in comparison to controls. As psychotic relapse has been associated with increased dopamine synthesis (Abi-Dargham et al., Reference Abi-Dargham, Kegeles, Zea-Ponce, Mawlawi, Martinez, Mitropoulou and Siever2004, Reference Abi-Dargham, Rodenhiser, Printz, Zea-Ponce, Gil, Kegeles and Laruelle2000; Cheng et al., Reference Cheng, Chang, Lo, Chan, Lee, Hui and Howes2020; Howes et al., Reference Howes, Bose, Turkheimer, Valli, Egerton, Valmaggia and McGuire2011, Reference Howes, Kambeitz, Kim, Stahl, Slifstein, Abi-Dargham and Kapur2012; S. Kim et al., Reference Kim, Shin, Santangelo, Veronese, Kang, Lee and Kim2021; Weidenauer et al., Reference Weidenauer, Bauer, Sauerzopf, Bartova, Nics, Pfaff and Willeit2020), a tentative explanation to reconcile the higher D2/3R availability with lower propensity for relapse is lower levels of endogenous dopamine in former partial agonist users and in those who remained relapse free. Alternatively, one may argue that both FEP patients who discontinued D2/3R antagonists as well as partial agonists eventually show greater D2/3R availability than controls. This effect, however, took longer to manifest in former partial agonist users, and the slower pace of increase in D2/3R availability may be protective against relapse.
Strengths and limitations
An important strength of this study is that we assessed both short and longer-term effects of antipsychotic discontinuation on D2/3R availability. Also, the differentiation between D2R antagonistic and partial agonistic antipsychotics is an important aspect, as these types of drugs may affect the dopamine system in different ways. Another strength is the focus on FEP patients as the study population, combined with the detailed clinical description of the participants prior to discontinuation.
A key limitation is the limited sample size, especially at the longer follow-up scan. From the 16 patients who discontinued, 6 restarted antipsychotics within 2 months, illustrating their vulnerability to relapse. Furthermore, a hurdle for the interpretation of results is the inability to separate the effects of different molecular processes affecting the [11C]raclopride BPND after antipsychotic discontinuation. The [11C]raclopride BPND is sensitive to the level of endogenous dopamine (Ginovart, Reference Ginovart2005) and the observed dorsal striatal binding potential results from both the number of D2R—and to a lesser extent D3R—, as well as from endogenous dopamine synthesis, release, and clearance (McCormick, Kapur, Seeman, & Wilson, Reference McCormick, Kapur, Seeman and Wilson2008). Dopamine neurotransmission is a complex process, and changes in dopamine production, release or transport also affect the BPND of the post-synaptic tracer. Additionally, we cannot rule out the minimal presence of antipsychotic drugs during the first scan, especially for those who discontinued aripiprazole. We could have standardized dopamine synthesis capacity after antipsychotic discontinuation by using the amphetamine challenge paradigm (Laruelle et al., Reference Laruelle, Abi-Dargham, van Dyck, Gil, D’Souza, Erdos and Innis1996, Reference Laruelle, Abi-Dargham, van Dyck, Rosenblatt, Zea-Ponce, Zoghbi and Kung1995; Volkow et al., Reference Volkow, Wang, Fowler, Logan, Schlyer, Hitzemann and Wolf1994). However, considering that participants were vulnerable for relapse, we did not deem this experimental paradigm ethically justified. Similarly, the dopamine depletion paradigm (Laruelle et al., Reference Laruelle, D’Souza, Baldwin, Abi-Dargham, Kanes, Fingado and Innis1997), which provides an estimate of D2R occupancy by endogenous dopamine and includes two PET scans within a 48-hour interval and alpha-methyl-para-tyrosine (AMPT) administration every 6 hours for 2-3 days, was considered too demanding for our delicate group. As participants were not randomized by the type of antipsychotic drug, there is a risk of bias by indication, even though baseline demographic and clinical characteristics were very similar between the D2R antagonist partial agonist groups. Another limitation is the relatively small sample size at follow-up, as patients who relapsed quickly after discontinuation had to be excluded, which may have limited our power in some of the analyses.
Conclusions
The present study is the first to assess the short and longer-term effects of the discontinuation of D2R antagonist and a partial D2R agonist antipsychotic on D2/3R availability in remitted FEP patients. D2R antagonist discontinuers showed increased D2/3R availability one week after cessation, and 81% experienced a relapse within 12 months. These findings are in line with the dopamine supersensitivity hypothesis. In contrast, partial agonist users had a much lower relapse incidence of 17% and showed decreased D2/3R availability shortly after discontinuation and an increased D2/3R availability after 2 months. These findings indicate that discontinuation of partial D2R agonists may affect D2/3R differently than D2R antagonist antipsychotics, which may be related to the lower relapse rates associated with partial D2R agonists. If replicated, current findings point to a preference for treatment with partial agonists.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S003329172510161X.
Acknowledgements
The authors wish to thank Gilles Stormezand for acquiring [11C]raclopride PET scans of controls, Ilse Thompson for the assistance in the acquisition of PET scans of FEP patients. The authors also thank all participants for their voluntary participation and effort in this study.
Author contributions
Franciska de Beer: conceptualization, methodology, formal analysis, investigation, data curation, writing – original draft, visualization, project administration. Erik F.J. de Vries: conceptualization, data curation, writing – review & editing, supervision. Ben Wijnen: methodology, writing – review & editing. Marieke J.H. Begemann: writing – review & editing, funding acquisition. Nico van Beveren: writing – review & editing. Nynke Boonstra: writing – review & editing. Shiral Gangadin: writing – review & editing. Lieuwe de Haan: writing – review & editing. Iris Hamers: investigation, writing – review & editing. Wim Veling: writing – review & editing. Sanne Koops: writing – review & editing. Iris E.C. Sommer: conceptualization, funding acquisition, supervision, writing – review & editing.
Funding statement
This study was supported by the Dutch Medical Science Foundation ZonMW (no. 636340001 to I.S.) and by the Brain and Behavior Research Foundation (BBRF) (no. 30974 to M.B.).
Declaration of interest
I.S. participates in a trial from Boehringer-Ingelheim, has received speakers’ fee from Otsuka, and a charity grant from Janssen. E.F.J.d.V. declares financial support from Hoffmann-La Roche, Eli Lilly, Bristol Myers Squibb, Ionis Pharmaceutical, Rodin Therapeutics, Lysosomal Therapeutics, Novartis, Janssen-Cilag BV, GE Healthcare and GlaxoSmithKline, for contracted research not related to this study, paid to the institution in the past 5 years. All other authors report no biomedical financial interests or potential conflicts of interest.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional guides on the care and use of laboratory animals.
Open access
