Introduction
Adequate resting cerebral blood flow (rCBF) is essential for healthy brain function, but rCBF is altered in medicated patients with psychotic disorders (Goozee, Handley, Kempton, & Dazzan, Reference Goozee, Handley, Kempton and Dazzan2014; Pollak et al., Reference Pollak, Drndarski, Stone, David, McGuire and Abbott2018). However, the cause and clinical impact of the rCBF alterations are sparsely investigated in patients with psychosis. RCBF alterations may reflect illness pathophysiology or an effect of antipsychotic treatment (Goozee et al., Reference Goozee, Handley, Kempton and Dazzan2014). Studies of twins and antipsychotic-naïve patients with psychotic disorder can provide insight into rCBF alterations in the illness pathophysiology, whereas studies of patients and healthy controls (HCs) before and after antipsychotic treatment or challenge studies may illuminate the effect of antipsychotics. Support for rCBF alterations in psychosis pathophysiology comes from studies of subjects at ultra-high risk for psychosis (UHR) and subclinical psychotic-like experiences (Allen et al., Reference Allen, Chaddock, Egerton, Howes, Bonoldi, Zelaya and McGuire2016, Reference Allen, Azis, Modinos, Bossong, Bonoldi, Samson and McGuire2018; Modinos et al., Reference Modinos, Egerton, McMullen, McLaughlin, Kumari, Barker and Zelaya2018), and a recent twin study that found a relation between increased striatal and thalamic rCBF and schizophrenia spectrum disorder (Legind et al., Reference Legind, Broberg, Brouwer, Mandl, Ebdrup, Anhoj and Rostrup2019). However, an effect of antipsychotic treatment cannot be ruled out in the twin study as patients were in the chronic phase of illness (Legind et al., Reference Legind, Broberg, Brouwer, Mandl, Ebdrup, Anhoj and Rostrup2019). Studies of antipsychotic-naïve patients can overcome the confounding effect of antipsychotic treatment. The majority of existing rCBF studies of antipsychotic-naïve patients do not find abnormal perfusion in striatal areas (Andreasen et al., Reference Andreasen, O'Leary, Flaum, Nopoulos, Watkins, Boles Ponto and Hichwa1997; Catafau et al., Reference Catafau, Parellada, Lomena, Bernardo, Pavia, Ros and Gonzalez-Monclus1994; Corson, O'Leary, Miller, & Andreasen, Reference Corson, O'Leary, Miller and Andreasen2002; Parellada et al., Reference Parellada, Catafau, Bernardo, Lomena, Gonzalez-Monclus and Setoain1994; Scottish Schizophrenia Research Group, 1998), although reduced (Vita et al., Reference Vita, Bressi, Perani, Invernizzi, Giobbio, Dieci and Fazio1995) and increased (Early, Reiman, Raichle, & Spitznagel, Reference Early, Reiman, Raichle and Spitznagel1987) striatal rCBF also have been observed. In thalamus, the majority reports no alterations (Catafau et al., Reference Catafau, Parellada, Lomena, Bernardo, Pavia, Ros and Gonzalez-Monclus1994; Corson et al., Reference Corson, O'Leary, Miller and Andreasen2002; Early et al., Reference Early, Reiman, Raichle and Spitznagel1987; Parellada et al., Reference Parellada, Catafau, Bernardo, Lomena, Gonzalez-Monclus and Setoain1994; Scottish Schizophrenia Research Group, 1998), although higher (Andreasen et al., Reference Andreasen, O'Leary, Flaum, Nopoulos, Watkins, Boles Ponto and Hichwa1997) and lower (Vita et al., Reference Vita, Bressi, Perani, Invernizzi, Giobbio, Dieci and Fazio1995) perfusion also have been found. However, sex differences and a low number of participants may have influenced findings. First, healthy female subjects have higher global cerebral perfusion (Alisch et al., Reference Alisch, Khattar, Kim, Cortina, Rejimon, Qian and Bouhrara2021; Gur et al., Reference Gur, Gur, Obrist, Hungerbuhler, Younkin, Rosen and Reivich1982; Liu et al., Reference Liu, Zhu, Feinberg, Guenther, Gregori, Weiner and Schuff2012; Mathew, Wilson, & Tant, Reference Mathew, Wilson and Tant1986) and higher subcortical rCBF in thalamus and striatum compared with males (Ghisleni et al., Reference Ghisleni, Bollmann, Biason-Lauber, Poil, Brandeis, Martin and O'Gorman2015). However, it has not been investigated if rCBF in antipsychotic-naïve female patients with first-episode psychosis differ from rCBF in male patients. Second, the number of included participants has, in general, been small, and significant findings are seen in studies with few participants (N = 6–17). Thus, larger studies of antipsychotic-naïve patients are needed to clarify the role of striatal and thalamic rCBF alterations and possible sex differences in the psychosis pathophysiology.
Antipsychotic treatment is known to alter rCBF among others in striatum and thalamus. Studies of initially antipsychotic-naïve patients and HCs receiving various first- and second generation antipsychotic compounds consistently report increased striatal perfusion (Corson et al., Reference Corson, O'Leary, Miller and Andreasen2002; Fernandez-Seara et al., Reference Fernandez-Seara, Aznarez-Sanado, Mengual, Irigoyen, Heukamp and Pastor2011; Hawkins et al., Reference Hawkins, Wood, Vernon, Bertolino, Sambataro, Dukart and Mehta2018; Mehta et al., Reference Mehta, McGowan, Lawrence, Aitken, Montgomery and Grasby2003; Michels, Scherpiet, Stampfli, Herwig, & Bruhl, Reference Michels, Scherpiet, Stampfli, Herwig and Bruhl2016; Scottish Schizophrenia Research Group, 1998; Viviani, Graf, Wiegers, & Abler, Reference Viviani, Graf, Wiegers and Abler2013) as well as increases in striatal volume (Andersen et al., Reference Andersen, Raghava, Svarer, Wulff, Johansen, Antonsen and Ebdrup2020). Cross-sectional studies of medicated patients also find increased perfusion in striatum and thalamus (Eisenberg et al., Reference Eisenberg, Yankowitz, Ianni, Rubinstein, Kohn, Hegarty and Berman2017; Kindler et al., Reference Kindler, Schultze-Lutter, Hauf, Dierks, Federspiel, Walther and Hubl2018; Lahti, Weiler, Holcomb, Tamminga, & Cropsey, Reference Lahti, Weiler, Holcomb, Tamminga and Cropsey2009; Miller et al., Reference Miller, Andreasen, O'Leary, Rezai, Watkins, Ponto and Hichwa1997a; Miller, Rezai, Alliger, & Andreasen, Reference Miller, Rezai, Alliger and Andreasen1997b; Oliveira et al., Reference Oliveira, Guimaraes, Souza, Dos Santos, Machado-de-Sousa, Hallak and Leoni2018; Ota et al., Reference Ota, Ishikawa, Sato, Okazaki, Maikusa, Hori and Kunugi2014; Rodriguez et al., Reference Rodriguez, Andree, Castejon, Zamora, Alvaro, Delgado and Vila1997; Xu et al., Reference Xu, Qin, Zhuo, Liu, Zhu and Yu2017; Zhu et al., Reference Zhu, Zhuo, Qin, Xu, Xu, Liu and Yu2015) as do studies of patients off-and-on antipsychotics (Eisenberg et al., Reference Eisenberg, Yankowitz, Ianni, Rubinstein, Kohn, Hegarty and Berman2017; Lahti et al., Reference Lahti, Weiler, Holcomb, Tamminga and Cropsey2009). However, the effect of monotherapy with a partial dopamine agonist on rCBF has not been investigated.
The clinical correlation between rCBF and psychotic symptoms is also sparsely investigated. Some studies of previously medicated patients have found an association between rCBF changes in striatum and thalamus and improvement in positive symptoms (Eisenberg et al., Reference Eisenberg, Yankowitz, Ianni, Rubinstein, Kohn, Hegarty and Berman2017; Lahti et al., Reference Lahti, Weiler, Holcomb, Tamminga and Cropsey2009), and striatal volume increases has also been related to treatment effect (Andersen et al., Reference Andersen, Raghava, Svarer, Wulff, Johansen, Antonsen and Ebdrup2020). We recently reported that baseline measures of striatal dopaminergic activity and thalamic levels of glutamate were related to subsequent reduction of positive symptoms in a partly overlapping sample of initially antipsychotic-naïve patients (Bojesen et al., Reference Bojesen, Ebdrup, Jessen, Sigvard, Tangmose, Edden and Glenthoj2020; Sigvard et al., Reference Sigvard, Nielsen, Gjedde, Bojesen, Fuglo, Tangmose and Glenthoj2022). Brain perfusion is believed to reflect neuronal activity (Attwell et al., Reference Attwell, Buchan, Charpak, Lauritzen, Macvicar and Newman2010), and baseline rCBF in striatum and thalamus may therefore also constitute markers of treatment outcome in antipsychotic-naïve patients.
To address these questions, we assessed rCBF before and after six weeks of monotherapy with a partial dopamine agonist in initially antipsychotic-naïve patients with first-episode psychosis and examined the associations with treatment outcome and sex. We tested the primary hypothesis that perfusion in striatum and thalamus would increase after treatment with a partial dopamine agonist and be higher in the antipsychotic-naïve state in female patients compared with male patients. Moreover, we explored if there was a different response to antipsychotic treatment in female patients compared with male patients. Our second hypothesis was that baseline striatal and thalamic perfusion would be related to reduction of positive symptoms after treatment.
Participants and methods
Participants
Antipsychotic-naïve patients with schizophrenia spectrum disorder and HCs were recruited from January 2014 to May 2019 as part of the larger, multimodal Pan European Collaboration on Antipsychotic Naïve Schizophrenia II (PECANSII) study previously described in detail (Bojesen et al., Reference Bojesen, Ebdrup, Jessen, Sigvard, Tangmose, Edden and Glenthoj2020). The study was approved by the Committee on Biomedical Research Ethics (H-3-2013-149), and all participants provided written informed consent after the study procedures were explained. Antipsychotic-naïve patients were recruited from mental health centers and outpatient services in the Capital Region of Denmark if the following inclusion criteria were fulfilled: a diagnosis of first-episode schizophrenia, schizoaffective disorder, or non-organic psychosis (ICD-10 criteria) as evaluated with the Schedules for Clinical Assessment in Neuropsychiatry (Wing et al., Reference Wing, Babor, Brugha, Burke, Cooper, Giel and Sartorius1990), never treated with antipsychotic compounds or central nervous system stimulants (as reported by patients and confirmed by medical record), age between 18–45 years, and being legally competent. Patients were excluded if they had been treated with an antidepressant within the last 30 days or was involuntarily admitted. HCs recruited through online advertisement (www.forsøgsperson.dk) were matched on age, sex, and parental socioeconomic status, but were excluded if they fulfilled the criteria for clinical-high-risk of psychosis according to the Comprehensive Assessment of At-Risk Mental States (Yung et al., Reference Yung, Phillips, McGorry, McFarlane, Francey, Harrigan and Jackson1998), had a lifetime psychiatric diagnosis, or a first-degree relative with a psychiatric diagnosis. The following exclusion criteria applied for all participants: severe medical condition, previous head injury with unconsciousness >5 min, contraindication to magnetic resonance (MR) scans, and substance abuse in the past three months (current occasional substance use was tolerated for patients). Participants reported on their use of alcohol, nicotine, cannabis, other recreational drugs, and benzodiazepines and had a urine drug-screen test done (Rapid Response, Jepsen HealthCare, Tune, DK). In patients only, hemoglobin was assessed as part of routine blood tests at both visits. Occasional use of benzodiazepines was accepted for patients although not later than 12 h before examinations. After baseline examination, patients were treated with aripiprazole in individual dosing of 5–30 mg/day. Aripiprazole was a tool-compound chosen due to the common use in first-episode patients and the partial dopamine agonist properties. Other psychoactive compounds (antipsychotics, antidepressants, or mood stabilizers) were not permitted. If patients had to change antipsychotic compound prior to six weeks follow-up examinations due to inadequate clinical effect or severe side-effects, they were excluded.
Participants have been included in other publications described in the online Supplemental Methods, but rCBF data have not been reported previously.
Clinical assessments
The Positive and Negative Syndrome Scale (PANSS) was used to assess psychopathology in patients by trained raters (Kay, Fiszbein, & Opler, Reference Kay, Fiszbein and Opler1987), and the minimum PANSS score was subtracted when calculating the percentage change in the PANSS positive score (Leucht, Davis, Engel, Kane, & Wagenpfeil, Reference Leucht, Davis, Engel, Kane and Wagenpfeil2007). Level of function in all participants was assessed with the Personal and Social Performance Scale (PSP) (Morosini, Magliano, Brambilla, Ugolini, & Pioli, Reference Morosini, Magliano, Brambilla, Ugolini and Pioli2000).
Magnetic resonance imaging acquisition
MR imaging was done as previously described (Bojesen et al., Reference Bojesen, Andersen, Rasmussen, Baandrup, Madsen, Glenthoj and Broberg2018) on a 3.0 Tesla Philips scanner with a 32-channel head coil. rCBF was acquired with the pseudo-Continuous Arterial Spin Labeling (PCASL) sequence (Dai, Garcia, de Bazelaire, & Alsop, Reference Dai, Garcia, de Bazelaire and Alsop2008) with 30 pairs of perfusion weighted and control scans (dual echo EPI; 16 slices of 5 mm with an in-plane resolution of 3.55 × 3.55 mm2; SENSE factor 2.3; TR = 4100 ms; TE = 12 ms/28.5 ms at a post labeling delay of 1600 ms; labeling duration 1650 ms; background inversion pulses at 1663 and 2850 ms after the start of labeling). A reference scan was acquired to estimate the fully magnetized signal (M0): TR/TE = 10 s/9 ms. The PCASL sequence was obtained after approximately 45 min MR scanning preceded by a T1-weighted structural scan, spectroscopy sequences, and a functional resting state sequence. Motion issues were controlled at subject level as described in the online Supplementary Methods.
Preprocessing of PCASL data
PCASL data were processed with the FSL software package (http://fsl.fmrib.ox.ac.uk/) using data acquired with the first echo. Extra-cerebral signal was removed from T1-weighted images with the ‘Brain extraction Tool’ (BET) and the raw PCASL scans were thereafter linearly co-registered with the skull-stripped T1-weighted image. The T1-weighted images were segmented with FAST from FSL software into gray- and white matter. Next, normalization of the T1-weighted image to Montreal Neurological Institute (MNI) standard space was performed. Maps of calibrated rCBF (ml/100 g/min) were calculated using tools provided by FSL (oxford_asl). Finally, rCBF maps were spatially transformed to MNI standard space, using a combination of the linear and non-linear transformation from anatomical data to MNI standard space.
Correction for gray and white matter partial volume effects
To account for possible inter-individual differences in gray and white matter, correction for gray and white matter partial volume effects was done using a linear regression model
Where S i is the scaled perfusion value of the i'th voxel within the region of interest (ROI) (such that 1 < i < N r, and N r is the number of voxels in the ROI), GMi is the partial volume fraction of gray matter, and WMi the partial volume fraction of white matter at the same voxel. We used β i as a measure of partial volume corrected regional gray matter perfusion. The value of β 0 corresponds to perfusion in CSF and was held constant at 0.
Region of interest analyses of rCBF
A ROI approach was chosen to extract gray matter rCBF values from the primary ROIs striatum and thalamus and the following exploratory ROIs: striatal subdivisions (putamen, nucleus accumbens, and caudate), hippocampus, and frontal lobe. Global CBF was estimated by averaging across the whole brain mask. ROIs from subcortical structures were defined by segmenting the structural T1 scan using FreeSurfer v. 6.0.0 (https://surfer.nmr.mgh.harvard.edu/) by applying individual anatomical regions to perfusion maps in structural space (subject space). The cortical MNI 152 atlas from FSL was used to estimate rCBF in frontal lobe.
Voxel-wise analysis of rCBF
Explorative whole brain voxel-wise analyses were performed to support the ROI analyses and explore rCBF differences at baseline and after treatment with age, sex, and global perfusion as co-variates and with correction for partial volume effects. Voxel-wise analyses were performed with and without masking deep white matter. The ‘randomise’ program from the FMRIB Software Library version (www.fmrib.ox.ac.uk/) using non-parametric, permutation-based statistical inference and threshold-free cluster enhancement was used. The threshold for statistical maps was set p < 0.05 (corrected for multiple comparisons).
Statistics
Demographic variables and clinical characteristics were compared with χ2, Fisher's exact test, or dependent and independent t tests as appropriate.
The following co-variates of no interest were included in all analyses: global rCBF (to eliminate global effects on regional rCBF), age and sex as well as smoking status that impact rCBF (Alisch et al., Reference Alisch, Khattar, Kim, Cortina, Rejimon, Qian and Bouhrara2021; Elbejjani et al., Reference Elbejjani, Auer, Dolui, Jacobs, Haight, Goff and Launer2019). Hemoglobin was adjusted for in post-hoc analyses of sex-differences in the patient group but was not available in HCs.
Baseline differences in rCBF were investigated with a general linear model (GLM) including a sex × group interaction. In cases with a significant interaction, post hoc analyses evaluated differences between female and male patients (main effect of sex). Differences between female and male HCs are reported as exploratory outcomes in the online Supplementary Results.
The primary hypothesis that treatment with a partial dopamine agonist would increase perfusion in striatum and thalamus was tested in a linear mixed model, where a group × time interaction evaluated the treatment effect. In cases with a significant group × time interaction, post hoc tests evaluated group differences after six weeks. Sex differences in the rCBF trajectory were tested by replacing the interaction term with a group × time × sex interaction. In cases with significant interactions, post hoc analyses of the patient group only evaluated a different trajectory between male and female patients by using the interaction term sex × time for the patient group in a GLM.
The statistical models are described in detail in the online Supplementary Methods.
The associations between reduction in PANSS positive after treatment and baseline rCBF or rCBF after treatment were evaluated in GLMs with the rCBF change being calculated as: (rCBF after treatment – rCBF before treatment)/ rCBF before treatment.
The significance level was adjusted to p = 0.025 due to two primary ROIs.
Statistical analyses were performed using SAS version 7.1 (SAS institute, Cary, NC).
Results
At baseline, 52 patients (58% females) with schizophrenia (N = 39), schizoaffective disorder (N = 1), or non-organic psychosis (N = 12) and 57 HCs (63% females) were included and after six weeks, 32 patients and 53 HCs were re-examined. Three patients and seven HCs did not have a usable pCASL scan at baseline and were excluded from the analyses, and 20 patients and one HCs were excluded after six weeks as specified in online Supplementary Fig. S1. Demographic and clinical characteristics of the remaining 49 patients and 50 HCs at baseline and 29 patients and 49 HCs after six weeks are provided in Table 1. There were no differences between groups with regards to age, parental socioeconomic status, and current cannabis use, but patients had significantly fewer years of education, smoked more, and had lower level of function than HCs as expected (Table 1). PANSS positive score significantly decreased after treatment (Table 1).
Abbreviations: FEP, First-episode psychosis patients; HCs, Healthy controls; F, Females; M, Males; s.d., Standard deviation; PANSS, Positive and negative syndrome scale; PSP, Personal and Social Performance scale.
a N states the number of participants with a usable pseudo-Continuous Arterial Spin Labeling scan.
b Minimum PANSS score was subtracted before calculating the percentage change.
c Statistics represent patient values at baseline compared with follow-up after six weeks.
The median dose of aripiprazole (25th–75th percentile) after six weeks was 10 mg (5.0–15.0 mg), and median serum aripiprazole level was 169.9 μg/l (111.2–246.6 μg/l) suggesting that patients were treated with sufficient doses and were compliant (Sparshatt, Taylor, Patel, & Kapur, Reference Sparshatt, Taylor, Patel and Kapur2010).
Resting CBF in antipsychotic-naïve patients at baseline
Region of interest analyses
Striatum: Striatal perfusion did not differ between patients and HCs at baseline (p = 0.99). There was a significant sex × group interaction (p = 0.014) and post hoc tests revealed that female patients had significantly lower striatal rCBF than male patients (p = 0.009).
Thalamus: Thalamic perfusion did not differ between patients and HCs at baseline (p = 0.79) and the sex × group interaction did not reach significance (p = 0.06).
Explorative regions: In the explorative regions of interest there were no significant group differences at baseline in nucleus accumbens (p = 0.41), caudate (p = 0.64), putamen (p = 0.55), hippocampus (p = 0.24), frontal lobe (p = 0.80), or in global rCBF (p = 0.58). The sex × group interaction was borderline significant for nucleus accumbens (p = 0.027), but the main effect of sex was not significant for the patients (p = 0.08). For the remaining regions, the sex × group interaction was insignificant (p = 0.11–0.82).
Voxel-wise analyses
Explorative voxel-wise analyses revealed no significant group differences at baseline.
rCBF after six weeks of treatment with partial dopamine agonism
Region of interest analyses
Striatum: The rCBF change from baseline to six weeks was significantly different in patients compared with HCs (significant group × time) due to increased rCBF in patients after treatment as summarized in Table 2 and Fig. 1a. There was a different trajectory in females compared to males (significant group × time × sex interaction: p = 0.020) and post hoc tests of the patient group only revealed a borderline significant different trajectory in female and male patients (month × sex: p = 0.025) due to lower rCBF at baseline in the female patients (p = 0.009) but no sex-difference after six weeks (p = 0.75; Fig. 1b).
Abbreviations: FEP, First-episode psychosis patients; HC, Healthy controls; rCBF, resting cerebral blood flow; s.d., standard deviation.
Table 2 shows resting cerebral blood flow (rCBF) in regions of interest in first-episode patients and healthy controls. Statistical analyses are adjusted for global rCBF, smoking status, age, and sex.
a The significance level was set to p < 0.025 due to two primary regions of interest.
b Striatal rCBF could not be calculated in three patients due to poor anatomical segmentation of striatum.
c Analyses of global rCBF are corrected for age and sex. The percent change of rCBF was calculated as: [(baseline-6 weeks)/baseline] × 100%.
Thalamus: There was a trend for a difference in rCBF change from baseline to six weeks in patients compared with HCs (group × time: p = 0.04, Table 2 and Fig. 1c). There was a significant different trajectory in females compared to males (group × time × sex interaction: p = 0.024) and post hoc test of the patient group only revealed a trend for a different trajectory in female and male patients (month × sex: p = 0.05) but not significant differences at baseline (p = 0.11) or after six weeks (p = 0.91).
Explorative regions: The rCBF change from baseline to six weeks was significantly different in patients compared with HCs for nucleus accumbens and putamen due to increased rCBF in patients after treatment, whereas the other explorative ROIs were unaffected by treatment (Table 2). For nucleus accumbens there was a significant group × time × sex interaction (p = 0.011) and post hoc tests of the patients group revealed a borderline significant month × sex interaction (p = 0.026) due to a trend for lower rCBF in female patients at baseline. For putamen, the group × time × sex interaction was borderline significant (p = 0.029), but post hoc tests were insignificant (month × sex: p = 0.27). The group × time × sex interaction was insignificant for the caudate (p = 0.56), hippocampus (p = 0.60), frontal lobe (p = 0.22), and global rCBF (p = 0.56).
The effect of treatment on rCBF without adjustment for global perfusion is provided in online Supplementary Table S1 and analyses without adjustment for any co-variates in online Supplementary Table S2.
Voxel-wise analyses
Explorative voxel-wise analyses without masking deep white matter revealed significantly increased rCBF in patients after six weeks of treatment in putamen, caudate, and white matter in frontal lobe when compared with HCs as shown in online Supplementary Fig. S3. When masking deep white matter, the CBF maps also revealed a striatal increase in perfusion after treatment although in a more restricted area as illustrated in Fig. 2.
Baseline rCBF and symptom improvement after treatment
Striatal and thalamic perfusion at baseline was not associated with improvement in PANSS positive symptoms (striatum: p = 0.32; thalamus: p = 0.33). For the exploratory ROIs, there was a significant negative association between nucleus accumbens baseline perfusion and improvement in PANSS positive (p = 0.046), whereas the associations for the other ROIs were insignificant (p = 0.14–0.68).
Analyses without the co-variates age, sex, and global rCBF revealed significant negative associations between improvement in PANSS positive symptoms and perfusion in whole striatum (p = 0.015), nucleus accumbens (p = 0.006), putamen (p = 0.02), and caudate (p = 0.005) but not for any other ROIs (p = 0.09–0.74).
There was no evidence for sex-specific associations between CBF and improvement in PANSS positive (CBF × sex interactions for all ROIs insignificant: p = 0.26–0.81).
Exploratory analyses revealed no association between baseline rCBF in any of the ROIs and improvement in PANSS total (p = 0.36–0.93), PANSS general (p = 0.46–0.98), or PANSS negative (p = 0.21–0.94).
rCBF changes after treatment and symptom improvement
Changes in striatal and thalamic perfusion after treatment were not associated with improvement in PANSS positive score (striatum: p = 0.97; thalamus: p = 0.72; without co-variates: striatum: p = 0.94; thalamus: p = 0.92). Also, there was no significant associations between changes in rCBF in explorative ROIs and improvement in PANSS positive score (p = 0.18–0.96; without co-variates: p = 0.35–0.94).
The CBF × sex interactions for all ROIs were insignificant (p = 0.07–0.43).
Exploratory analyses revealed no association between changes in rCBF in any of the ROIs and improvement in PANSS total (p = 0.41–0.98), PANSS general (p = 0.21–0.87), or PANSS negative (p = 0.60–0.95). There was a trend for a negative association between aripiprazole serum levels and rCBF change in nucleus accumbens (p = 0.06) but not in any other ROI (p = 0.26–0.99).
Discussion
This study is the first to investigate striatal and thalamic perfusion in a large group of initially antipsychotic-naïve psychosis patients before and after monotherapy with a partial dopamine receptor agonist. The main finding is that partial dopamine agonism increases striatal perfusion. Moreover, findings suggest that treatment affects thalamic perfusion to a lesser degree, and that female patients have lower striatal perfusion than males from illness onset. Last, lower baseline rCBF in nucleus accumbens was related to a greater improvement of positive symptoms after treatment.
Our finding of no striatal rCBF alterations in the entire group of antipsychotic-naïve patients is in line with existing studies (Andreasen et al., Reference Andreasen, O'Leary, Flaum, Nopoulos, Watkins, Boles Ponto and Hichwa1997; Catafau et al., Reference Catafau, Parellada, Lomena, Bernardo, Pavia, Ros and Gonzalez-Monclus1994; Corson et al., Reference Corson, O'Leary, Miller and Andreasen2002; Early et al., Reference Early, Reiman, Raichle and Spitznagel1987; Parellada et al., Reference Parellada, Catafau, Bernardo, Lomena, Gonzalez-Monclus and Setoain1994; Scottish Schizophrenia Research Group, 1998; Vita et al., Reference Vita, Bressi, Perani, Invernizzi, Giobbio, Dieci and Fazio1995) but at odds with the theory of striatal overactivity as a driver of psychosis (Glenthoj & Hemmingsen, Reference Glenthoj and Hemmingsen1997; Howes & Murray, Reference Howes and Murray2014; Kapur, Reference Kapur2003). This discrepancy might be explained by sex differences since we observed lower striatal perfusion in antipsychotic-naïve female patients. However, the lower rCBF in female patients contrasts with our hypothesis based on healthy females with higher striatal rCBF compared with males (Ghisleni et al., Reference Ghisleni, Bollmann, Biason-Lauber, Poil, Brandeis, Martin and O'Gorman2015). It is therefore likely that reduced rCBF is part of the psychosis pathophysiology in female patients. Moreover, it is possible that striatal hyperactivity primarily is part of psychosis pathophysiology in male patients. In support, estrogen is believed to increase striatal dopaminergic sensitivity in females (Yoest, Quigley, & Becker, Reference Yoest, Quigley and Becker2018) and this may cause a sensitized striatal dopaminergic system prone for development of psychotic symptoms without measurable striatal overactivity at illness onset in females. In line with this, we previously found a significant association between psychotic symptoms and measures of presynaptic dopamine activity in an overlapping sample of antipsychotic-naïve psychosis patients with a majority of females (66%) without detecting differences in dopamine synthesis between patients and HCs (Sigvard et al., Reference Sigvard, Nielsen, Gjedde, Bojesen, Fuglo, Tangmose and Glenthoj2022). These potential sex differences are clinically relevant to investigate further. For example, females require lower serum levels of antipsychotics to reach striatal dopamine receptor occupancy (Brand et al., Reference Brand, Haveman, de Beer, de Boer, Dazzan and Sommer2021; Hoekstra et al., Reference Hoekstra, Bartz-Johannessen, Sinkeviciute, Reitan, Kroken, Loberg and Sommer2021) and female-specific description guidelines may therefore be needed.
Thalamic perfusion was not increased at baseline in the antipsychotic-naïve patients corresponding to findings in previous rCBF studies (Catafau et al., Reference Catafau, Parellada, Lomena, Bernardo, Pavia, Ros and Gonzalez-Monclus1994; Corson et al., Reference Corson, O'Leary, Miller and Andreasen2002; Early et al., Reference Early, Reiman, Raichle and Spitznagel1987; Parellada et al., Reference Parellada, Catafau, Bernardo, Lomena, Gonzalez-Monclus and Setoain1994; Scottish Schizophrenia Research Group, 1998). However, it contrasts our previously findings of increased thalamic glutamate levels in an overlapping cohort of antipsychotic-naïve patients with psychotic disorder (Bojesen et al., Reference Bojesen, Ebdrup, Jessen, Sigvard, Tangmose, Edden and Glenthoj2020, Reference Bojesen, Broberg, Fagerlund, Jessen, Thomas, Sigvard and Glenthoj2021). Glutamatergic neurotransmission is believed to increase perfusion (Attwell et al., Reference Attwell, Buchan, Charpak, Lauritzen, Macvicar and Newman2010) as confirmed in a previous in vivo study that reported a positive associations between glutamate levels and rCBF in HCs (Bojesen et al., Reference Bojesen, Andersen, Rasmussen, Baandrup, Madsen, Glenthoj and Broberg2018). Based on these findings, one would expect thalamic perfusion in antipsychotic-naïve patients to be increased from illness onset as well. However, perfusion is regulated by many neurotransmitters and is among others reduced by activity of GABAergic interneurons (Franklin et al., Reference Franklin, Wang, Sciortino, Harper, Li, Hakun and Childress2011; Krause et al., Reference Krause, Wijtenburg, Holcomb, Kochunov, Wang, Hong and Rowland2014) that also are found in thalamus (de Biasi, Frassoni, & Spreafico, Reference de Biasi, Frassoni and Spreafico1986). It is possible that the net effect of increased glutamatergic activity and a possible compensatory increased GABAergic activity in thalamus may be no alterations in thalamic perfusion.
Treatment with a partial dopamine receptor agonist for six weeks significantly increased striatal rCBF and affected thalamic perfusion as also seen in patients treated with first- and second generation antipsychotics (Eisenberg et al., Reference Eisenberg, Yankowitz, Ianni, Rubinstein, Kohn, Hegarty and Berman2017; Kindler et al., Reference Kindler, Schultze-Lutter, Hauf, Dierks, Federspiel, Walther and Hubl2018; Lahti et al., Reference Lahti, Weiler, Holcomb, Tamminga and Cropsey2009; Miller et al., Reference Miller, Rezai, Alliger and Andreasen1997b; Scheef et al., Reference Scheef, Manka, Daamen, Kuhn, Maier, Schild and Jessen2010; Xu et al., Reference Xu, Qin, Zhuo, Liu, Zhu and Yu2017; Zhu et al., Reference Zhu, Zhuo, Qin, Xu, Xu, Liu and Yu2015). This suggests that not only dopamine D2 antagonism but also partial dopamine D2 agonism increases metabolism after short term treatment in striatum and thalamus that are regions believed implicated in psychotic disorder (Carlsson, Waters, & Carlsson, Reference Carlsson, Waters and Carlsson1999). The effect of treatment with a partial dopamine agonist on striatal perfusion was most robust, as it was found in both ROI and voxel-wise analyses. Global perfusion impacts regional perfusion in between-group studies (Selvaggi et al., Reference Selvaggi, Jauhar, Kotoula, Pepper, Veronese, Santangelo and Howes2022; Turkheimer et al., Reference Turkheimer, Selvaggi, Mehta, Veronese, Zelaya, Dazzan and Vernon2020; Viviani et al., Reference Viviani, Sim, Lo, Richter, Haffer, Osterfeld and Beschoner2009), but, despite this, we found significantly increased striatal perfusion both with and without adjustment for global perfusion supporting the robustness of this finding. However, the magnitude of the striatal rCBF increase after partial dopamine agonism might have been less pronounced than after dopamine antagonism, as preclinical studies have found that dopamine D2/D3 receptor antagonism increases while agonism decreases striatal cerebral blood volume (Sander et al., Reference Sander, Hooker, Catana, Normandin, Alpert, Knudsen and Mandeville2013; Sander, Hooker, Catana, Rosen, & Mandeville, Reference Sander, Hooker, Catana, Rosen and Mandeville2016).
Thalamic perfusion after treatment was only affected in the ROI analyses and only at trend level. This suggests that the effect of partial dopamine receptor agonism is most pronounced in striatum, which might be explained by a higher concentration of D2 receptors in this region (Selvaggi et al., Reference Selvaggi, Hawkins, Dipasquale, Rizzo, Bertolino, Dukart and Mehta2019). However, studies of the long-term effect of treatment on thalamic rCBF are needed, as thalamic rCBF is reduced in medicated patients (Walther et al., Reference Walther, Federspiel, Horn, Razavi, Wiest, Dierks and Muller2011).
We did not find an association between baseline rCBF in whole striatum and reduction of positive symptoms after treatment but found a significant negative association in nucleus accumbens. This is in line with our previous finding of an association between baseline striatal decarboxylation rate and improvement in positive symptoms in an overlapping cohort of patients (Sigvard et al., Reference Sigvard, Nielsen, Gjedde, Bojesen, Fuglo, Tangmose and Glenthoj2022). Interestingly, the change in perfusion was not related to treatment effect, suggesting baseline perfusion as a better predictor for subsequent outcome.
We had expected an association between baseline perfusion in thalamus and symptoms improvement given our previous finding of an association between baseline thalamic glutamate levels and treatment outcome in an overlapping cohort (Bojesen et al., Reference Bojesen, Ebdrup, Jessen, Sigvard, Tangmose, Edden and Glenthoj2020). This may suggest that more direct measures of neurochemical function are better predictors than rCBF of treatment effect.
Last, it should be noted that rCBF might be related to other symptom dimensions in psychosis, as associations with formal thought disorders, altered affectivity, and abnormal motor behavior, among others, previously have been reported (Horn et al., Reference Horn, Federspiel, Wirth, Muller, Wiest, Wang and Strik2009; Kindler et al., Reference Kindler, Michel, Schultze-Lutter, Felber, Hauf, Schimmelmann and Walther2019; Stegmayer et al., Reference Stegmayer, Strik, Federspiel, Wiest, Bohlhalter and Walther2017; Walther et al., Reference Walther, Stegmayer, Federspiel, Bohlhalter, Wiest and Viher2017).
Frontal rCBF both with and without adjustment for global rCBF was not reduced in the antipsychotic-naïve state as found in a previous study of first-episode patients free from antipsychotic medication (Selvaggi et al., Reference Selvaggi, Jauhar, Kotoula, Pepper, Veronese, Santangelo and Howes2022), and treatment did not affect rCBF in frontal lobe as suggested by previous studies (Goozee et al., Reference Goozee, Handley, Kempton and Dazzan2014). This might indicate that frontal changes in perfusion only occur in subgroups from illness onset and is affected by treatment later in the illness. It is also possible that previous studies have assessed a combination of gray- and white matter perfusion since we found increased white matter rCBF in frontal lobe in the voxel-wise analysis when not masking white matter. Specialized ASL sequences can provide reliable white matter rCBF measures (Giezendanner et al., Reference Giezendanner, Fisler, Soravia, Andreotti, Walther, Wiest and Federspiel2016), but the sensitivity is considered too small in sequences optimized for gray matter rCBF (Alsop et al., Reference Alsop, Detre, Golay, Gunther, Hendrikse, Hernandez-Garcia and Zaharchuk2015).
Last, perfusion in the explorative ROI hippocampus was also not increased from illness onset as found in studies of patients at UHR for psychosis and subjects scoring high on schizotypy (Allen et al., Reference Allen, Chaddock, Egerton, Howes, Bonoldi, Zelaya and McGuire2016, Reference Allen, Azis, Modinos, Bossong, Bonoldi, Samson and McGuire2018; Modinos et al., Reference Modinos, Egerton, McMullen, McLaughlin, Kumari, Barker and Zelaya2018), which might be due to diagnostic heterogeneity between these populations and patients with first-episode psychosis.
Strength of the current study is a large group of initially antipsychotic-naïve patients treated with a partial dopamine D2 receptor agonist as monotherapy, but limitations should also be addressed. Most importantly, post hoc analyses of sex differences in patients after treatment might have lacked power due to drop-out after baseline (N = 29).
In conclusion, increased striatal rCBF is induced by antipsychotic treatment with a partial dopamine agonist as also seen after dopamine antagonism, and female patients have lower striatal perfusion from illness onset. Moreover, rCBF in nucleus accumbens at illness onset is related to treatment effect. The findings stress the importance of sex differences and treatment effects when investigating the neurobiology of psychosis.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0033291723000144
Financial support
This study was funded by a PhD grant from the Faculty of Health and Medical Sciences, University of Copenhagen (KB Bojesen); PhD grants from the Mental Health Services in the Capital Region of Denmark (AM Sigvard and K Tangmose); an independent grant from the Lundbeck Foundation (R155-2013-16337) to the Lundbeck Foundation Centre of Excellence for Clinical Intervention and Neuropsychiatric Schizophrenia Research (CINS) (BY Glenthøj); and grants from the Wørzner and Gerhard Linds Foundations; support from the Mental Health Services, Capital Region of Denmark (BY Glenthøj). The funding sources had no role in the design or conduction of the study design, nor in the collection, analyses and interpretation of data, or in the writing, review approval and submission of the manuscript for publication.
Conflict of interest
Drs KB Bojesen and K Tangmose received lecture fees from Lundbeck Pharma A/S.
Dr Glenthøj has been the leader of a Lundbeck Foundation Centre of Excellence for Clinical Intervention and Neuropsychiatric Schizophrenia Research (CINS) (January 2009 – December 2021), which was partially financed by an independent grant from the Lundbeck Foundation based on international review and partially financed by the Mental Health Services in the Capital Region of Denmark, the University of Copenhagen, and other foundations. All grants are the property of the Mental Health Services in the Capital Region of Denmark and administrated by them. She has no other conflicts to disclose.
Dr BH. Ebdrup received lecture fees and/or is part of the advisory board at Bristol-Myers Squibb, Eli Lilly and Company, Janssen-Cilag, Otsuka Pharma Scandinavia AB, Takeda Pharmaceutical Company, Boehringer Ingelheim, and Lundbeck Pharma A/S.
The rest of the authors have no conflicts of interest to disclose.
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.