We found 105 hits in the databases, and after evaluating the abstracts, we could include 16 studies covering our topic in the review. The studies with their results are summarized in the Tables , and .

In a further study including 235 schizophrenia patients (14 % with comorbid cannabis abuse, 8 % with cannabis dependence), patients with marijuana abuse/dependence had smaller fronto-temporal white matter (WM) volumes than those without heavy marijuana use [ 25 ]. Genotyping for cannabinoid receptor 1 (CB1/CNR1) demonstrated significant CNR1 (rs12720071) genotype-by-marijuana-use interaction effects on WM volumes and neurocognitive impairment. Furthermore, CNR1 genetic polymorphisms were associated with WM brain volume variation among schizophrenia patients. These findings suggest that heavy cannabis use in the context of specific CNR1 genotypes may contribute to greater WM volume deficits and cognitive impairment.

A MRI study including 17 patients with schizophrenia (mean illness duration, above 2 years) and 31 healthy controls with 48 % of the healthy group and 47 % of the patients being long-term heavy cannabis users was performed focussing on the cerebellum [ 52 ]. In cannabis users with and without schizophrenia, cerebellar white matter volume was reduced compared to healthy non-users, by 29.7 and 23.9 %, respectively, and by 17.7 % in patients without cannabis use. Healthy cannabis users did not differ in white matter volume from either of the schizophrenia groups. No group differences in cerebellar grey matter or total volumes could be detected. Interestingly, total cerebellar volume decreased as a function of duration of cannabis use in the healthy users, and psychotic symptoms as well as illness duration correlated with cerebellar measures differentially between patients with and without cannabis use. The authors conclude that long-term heavy cannabis use in healthy individuals is associated with smaller cerebellar white matter volume similar to that observed in schizophrenia, but volume reduction is even more pronounced in patients with schizophrenia using cannabis.

Volumetric sMRI studies in recent-onset or first-episode schizophrenia with comorbid cannabis abuse

Our literature research revealed that most published studies investigated first-episode or recent-onset schizophrenia patients in comparison with healthy subjects. However, the results of these neuroimaging studies are again heterogeneous and remain inconclusive.

In one study, patients with recent-onset schizophrenia (schizoaffective disorder) with comorbid substance abuse (n = 20, all THC, 7 patients with additional amphetamines/ecstasy consumption, 2 with opiates, 8 with cocaine, 1 with hallucinogens and 2 with alcohol) did not differ significantly in terms of volumetric morphometry in temporolimbic regions (superior temporal gyrus, amygdala–hippocampus complex, cingulum) from schizophrenia patients without previous substance abuse (n = 21) [66]. However, this study did not include a healthy control group. In a further study of this research group, another sample suffering from first-episode schizophrenia (n = 47) and healthy controls (n = 55) underwent MRI scanning, and volumes of hippocampus, superior temporal gyrus, prefrontal cortex, internal capsule (ALIC) and the cross-sectional areas of the corpus callosum were compared between subgroups of patients with (n = 19) (all cannabis, 4 with additional amphetamines/cocaine) and without previous substance abuse (n = 28). A significant reduction in the cross-sectional area of the C 2-segment of the corpus callosum in the non-abusing patients in comparison with the patients with previous cannabis abuse could be observed, suggesting more subtle brain abnormalities in the group without previous substance abuse [32].

Another study investigating brain volumes in cannabis-exposed patients with first-episode schizophrenia compared to non-exposed patients revealed no differences between the subgroups for total brain volume, total grey and white matter, ventricles, cerebellum and caudate except a decreased asymmetry of the lateral ventricles in the cannabis-exposed patients [7]. In a recent longitudinal study, the same research group could demonstrate that first-episode schizophrenia patients with continued use of cannabis (n = 19) showed increased loss of cerebral grey matter volume and larger increases in lateral and third ventricle volumes than patients who did not use cannabis during the follow-up (n = 32) [43]. Nevertheless, brain volumes of schizophrenia patients with and without cannabis abuse did not differ significantly at the beginning of the longitudinal study, although it has to be assumed that the patients abused cannabis several years before entering the study. The same group of researchers also performed an analysis of cortical thickness in these patient groups [44]. At the beginning of the investigation, cortical thickness did not differ between patients and controls and between cannabis-using and non-using patients. During the follow-up period, excessive thinning of the right supplementary motor cortex, inferior frontal cortex, superior temporal gyrus, angular gyrus, occipital and parietal lobe in patients relative to controls after controlling for the factor “cannabis use” was observed. Patients who used cannabis showed additional thinning in the left dorsolateral prefrontal cortex (DLPFC), left anterior cingulate cortex (ACC) and left occipital lobe as compared to those not using cannabis during the observation interval.

A further study investigating the volumes of the superior frontal gyrus, anterior cingulated gyrus and orbital frontal lobe found less anterior cingulate grey matter in cannabis-abusing first-episode schizophrenia patients compared with patients with no history of cannabis abuse and healthy volunteers [58]. The association between reduced volume in anterior cingulate gyrus (ACC) and the history of cannabis abuse in schizophrenia has been discussed in the context of a disturbed function of ACC with the consequence of poor decision-making and more compulsive drive towards drug use as a predisposition to substance abuse and not as a toxic effect of cannabis itself. This is in contrast to one study revealing no volumetric differences in the ACC, the amygdala–hippocampus complex, the STG and the lateral ventricles between recent-onset schizophrenia patients with and without previous cannabis abuse [66]. In this study, the patients without previous cannabis abuse tended to demonstrate more reversed temporal asymmetry as the non-users. This may be a subtle hint that the subgroup of cannabis abusers lower their threshold of psychosis through drug abuse itself and the other subgroup increases the vulnerability by presenting slightly more brain abnormalities.

In a more recent voxel-based morphometry (VBM) study, a decrease in grey matter density in the right posterior cingulate cortex (PCC) in the subgroup of first-episode schizophrenia with cannabis abuse was observed compared to those patients without cannabis abuse [2]. The other included brain regions with high density of CB1 receptors, for instance hippocampus, did not differ between these groups. The authors concluded that cannabis use may be associated with altered brain structure in particular regions (in their study PCC) with a high density of CB1 receptors.

A magnetic resonance imaging study focussing on cerebellar grey and white matter in first-episode patients with and without a history of cannabis use and non-psychiatric cannabis users revealed a lifetime dose-dependent regional reduction in grey matter in the right cerebellar lobules and a tendency for more profound grey matter reduction in lobule III with younger age at the onset of cannabis use [11]. First-episode patients had lower total cerebellar grey matter/total cerebellar volume ratio and marked grey matter loss in the vermis, pedunculi, flocculi and lobules compared to pairwise matched healthy controls. This pattern and degree of grey matter loss did not differ from age-matched first-episode patients with comorbid cannabis use, indicating only small dose-dependent effects of juvenile cannabis use on cerebellar neuropathology but no evidence of an additional effect of cannabis use on cerebellar grey matter pathology in psychosis.

In an adolescent MRI study, subjects with early-onset schizophrenia (EOS, n = 35), cannabis use disorder (CUD, n = 16), comorbid EOS with CUD (n = 13) and healthy controls (HC) (n = 51) were compared with regard to brain volume within frontal, temporal, parietal and subcortical ROIs [29]. If volumetric differences in one ROI were identified, cortical thickness and surface area of this ROI were additionally measured and compared between subgroups. The authors found a significant EOS-by-CUD interaction, in the left superior parietal region; both “pure” EOS and “pure” CUD had smaller grey matter volumes that were associated with lower surface area compared with HC. A similar alteration was observed in the comorbid group compared with HC, but there was no additive volumetric deficit in the comorbid group compared with the separate groups (EOS and CUD).

In a diffusion tensor imaging (DTI) study, 35 male recent-onset schizophrenia patients, with and without a history of cannabis use before the age of 17, and 21 matched healthy comparison men without illicit drug use were assessed [40]. In comparison with healthy controls, patients with cannabis use before the age of 17 showed increased directional coherence in the bilateral uncinate fasciculus, anterior internal capsule and frontal WM. These abnormalities were absent in patients without cannabis use before the age of 17, but not related to lifetime doses of cannabis or other illicit drugs. The authors concluded that recent-onset schizophrenia patients starting to use cannabis during early adolescence may represent a subgroup of schizophrenia patients with increased WM directional coherence reflecting structural hyperconnectivity. However, they contrast their results with the fact that in most DTI studies, schizophrenia patients were associated with hypoconnectivity [48].

Another study investigated male schizophrenia in early adolescence patients with regard to abnormalities in WM structure and integrity using high-resolution structural and diffusion tensor brain imaging and compared patients who started a regular use of cannabis (1) before the age of 15 years (early-onset cannabis users, n = 10) or (2) at the age of 17 years or later (late-onset cannabis users, n = 8), (3) cannabis-naïve patients (n = 8) and a healthy control group (n = 10) [12]. Cannabis-naïve patients showed reduced WM density and reduced fractional anisotropy in the splenium of the corpus callosum compared with patients with early-onset cannabis use. In the same brain area, cannabis-naïve patients showed reduced fractional anisotropy compared with healthy controls, indicating alterations in fibre density, degree of myelination and WM integrity. The authors suggest that the age of onset of cannabis use is not an identifying characteristic for WM abnormalities in schizophrenia patients and that cannabis-naïve schizophrenia patients may have a more vulnerable brain structure which may responsible for the onset of psychosis.

In a further VBM and DTI study comparing patients with adolescent-onset schizophrenia (AOS) with early cannabis use (more than 3 times/week for at least 6 months) (n = 16) to AOS without cannabis use (n = 16) and healthy adolescents (n = 28), those with cannabis abuse showed increased GM density loss in temporal fusiform gyrus, parahippocampal gyrus, ventral striatum, right middle temporal gyrus, insular cortex, precuneus, right paracingulate gyrus, dorsolateral prefrontal cortex, left postcentral gyrus, lateral occipital cortex and cerebellum [26]. In addition, cannabis-abusing AOS demonstrated decreased fractional anisotropy (FA) in particular in brain stem, internal capsule, corona radiata, superior and inferior longitudinal fasciculus compared to those without cannabis abuse. Interestingly, there were no cognitive differences between patients with and without cannabis abuse, but both subgroups impaired relative to controls.