Affected cortico-striatal-cerebellar network in schizophrenia with catatonia revealed by magnetic resonance imaging: indications for electroconvulsive therapy and repetitive transcranial magnetic stimulation

Abstract Catatonia is a psychomotor syndrome that can occur in a broad spectrum of brain disorders, including schizophrenia. Current findings suggest that the neurobiological process underlying catatonia symptoms in schizophrenia is poorly understood. However, emerging neuroimaging studies in catatonia patients have indicated that a disruption in anatomical connectivity of the cortico-striatal-cerebellar system is part of the neurobiology of catatonia, which could serve as a target of neurostimulation such as electroconvulsive therapy and repetitive transcranial magnetic stimulation.


Introduction
Catatonia is a striking psychomotor syndrome characterized by v arying motor, affectiv e , and beha vioral symptoms .T he clinical manifestations of the syndrome include r educed r esponsiv eness (catalepsy, stupor, m utism), r epetitiv e mov ements and behaviors (ster eotypy, ec holalia, ec hopr axia, mannerisms), bizarr e postur es (waxy flexibility, posturing), and intense anxiety or agitation (Fink and Taylor, 2009 ;Organization, 2013 ;Tandon et al., 2013 ).Catatonia has been found to occur in as many as 7-31% of psychiatric inpatients (Daniels, 2009 ;Lee et al., 2000 ;Rosebush et al., 1990 ;Rosebush and Mazur ek, 2010 ;Stuiv enga and Morr ens , 2014 ;Ta ylor and Fink, 2003 ) and is associated with high mortality if left untreated (Cornic et al., 2009 ;Tuerlings et al., 2010 ).Curr entl y, the tr eatment of catatonia includes medication and physical ther a py, suc h as electr oconvulsiv e ther a py (ECT) or r e petiti v e tr anscr anial ma gnetic stimulation (rTMS).ECT is currently the common treatment for catatonia.Even though catatonia has been known > 140 years a go, the exact neur al mec hanism behind the disease r emains unclear.Catatonia is most commonly associated with affective disorders and sc hizophr enia (Chalasani et al., 2005 ;Morrison, 1975 ;Tuerlings et al., 2010 ) but also occurs in neurologic and other medical conditions (Ahuja, 2000 ;Gelenberg, 1976 ;Sadda wi-K onefka et al., 2014 ).Irr espectiv e of the underlying illness, catatonia patients exhibit marked similarities in clinical symptoms and treatment response (Smith et al., 2012 ;Tandon et al., 2013 ), suggesting that there may be a common neurobiological mechanism.To date, a limited number of neur oima ging studies hav e attempted to unr av el the neur obiology of catatonia, mainl y using functional ma gnetic resonance imaging (fMRI).These studies show abnormalities of orbitofrontal, (medial) prefrontal (Northoff et al., 2004 ;Richter et al., 2010 ), and parietal regions (Northoff et al., 1999(Northoff et al., , 2000 ; ;Satoh et al., 1993 ), as well as the premotor cortex (Northoff et al., 2004(Northoff et al., , 1999 ; ;Richter et al., 2010 ), and supplementary motor area (SMA) (Sc heuer ec ker et al., 2009 ;Walther et al., 2017a ) in catatonia patients compared to healthy controls and schizophrenia without catatonia.If the results matter, how this collective of implicated br ain r egions contributes to the manifestation of catatonia on the behavior al le v el r emains to be determined.Her e, we intr oduce the r esearc h pr ogr ess of MRI c hanges of br ain structur e and function in sc hizophr enia with catatonia and their clinical implications for neur ostim ulation, i.e., ECT to rTMS, whic h could pr ovide a r eference for the clinical treatment of schizophrenia with catatonia.In the following sections, we summarize studies on sc hizophr enia with catatonia.This paper used PubMed to r etrie v e r ele v ant liter atur e .T he r etrie v ed documents included original r esearc h and r e vie w articles, and the search string was set as "Sc hizophr enia and Catalonia and Magnetic Resonance Imaging."Six reviews were excluded according to the title and abstract.After full-text r eading, 17 pa pers wer e excluded because they did not meet the inclusion criteria.Nine recent reports (listed in Table 1 ) were
Fritze et al. [Fritze et al ., 2020 ]: 111 SSD patients In catatonic patients, significant correlations were detected between NCRS motor scores and the whole brainstem.Hirjak et al. [Hirjak et al ., 2019 ]: 56 SSD patients Distinct dimensions of catatonia are associated with different patterns of abnormal brain structure.Hirjak et al. [Hirjak et al ., 2020 ]: 87 SSD patients NCRS behavioral scores were associated with a joint structural and functional system that pr edominantl y included cerebellar and prefrontal/cortical motor regions.Wasserthal et al. [Wasserthal et al ., 2020 ]: 111 SSD patients and 28 healthy controls Structur al r eor ganization of WM bundles connecting orbitofr ontal/parietal, thalamic, and striatal regions contributed to catatonia in SSD patients.
Psyc homotor cognitiv e functioning may differ entiate psyc hosis patients with catatonia from those without catatonia.

Ca ta tonia
One part of the network found to be affected in catatonia patients is the bilateral medial orbitofrontal cortex.This functionally complex region has been implicated in emotional a ppr aisal and reg-ulation via its link to limbic regions (Etkin et al., 2011 ).The medial orbitofrontal cortex mediates behavioral inhibition in negative emotions (Goldstein et al., 2007 ;Silbersweig et al., 2007 ).A previous study found decr eased structur al connectivity between the bilateral medial orbitofrontal cortex and striatum in catatonia patients, which could cause impaired emotion regulation of anxiety and contribute to reduced responsiveness of patients (Northoff et al. , 2004 ;Richter et al. , 2010 ).Future studies in larger samples of catatonia patients or using more sensitive measurements of anxiety may help elucidate the contribution of the orbitofrontal cortex to catatonia symptoms.
Pr e vious studies have suggested that increased radial diffusivity along fiber tracts is likely related to de-or d ysm yelination rather than axonal pathology, which is more likely to be reflected in axial diffusivity measurements (Budde et al., 2009 ;Harsan et al., 2006 ;Klawiter et al., 2011 ;Song et al., 2003Song et al., , 2002 ) ). Ho w e v er, factors such as axonal density and crossing fibers also influence radial diffusivity (Wheeler-Kingshott and Cercignani, 2009 ), and these factors are difficult to disentangle from myelination properties with current in vivo imaging techniques .T herefore , we tentatively suggest that catatonia ma y in volv e c hanges in the myelination of corticostriatal connections.Some of the pr e vious neur oima ging studies in catatonia demonstrated functional abnormalities in these region (Northoff, 2002 ), including alterations in the corticostriatal functional network connectivity and abnormalities in the SMA during rest (Walther et al., 2017a ).Other studies implicated the premotor cortex in catatonia (Northoff et al., 1999 ;Richter et al., 2010 ).Walther et al. reported that the catatonia factor correlated with functional connectivity between the left thalamus and the bilateral primary motor cortex (Walther et al., 2017b ).It is also said that higher connectivity is associated with se v er e catatonic symptoms.In addition, Walther et al. reported the severity of catatonia was correlated with the hyperperfusion of the SMA (Walther et al., 2017a ).
Some k e y cortical motor areas below the skull may be the entry node for non-inv asiv e br ain r egulation.Ther efor e, rTMS may help impr ov e the dysfunctional motor network of patients with sc hizophr enia.Curr entl y, most of the targets regarding rTMS for the treatment of catatonia are located in the dorsolater al pr efrontal cortex (DLPFC) (Sharma et al., 2018 ;Stip et al., 2017 ;Trojak et al., 2014 ).
The striatal regions implicated in previous studies, including the nucleus accumbens , caudate nucleus , and putamen, are involved in a wide range of volitional, motor, and emotional aspects of human behavior (Alexander and Crutcher, 1990 ;Cummings , 1993 ).T he striatum is an essential regulator of motor contr ol, particularl y in terms of initiation (Cunnington et al., 2002 ;Hauber, 1998 ) and inhibition (Chevrier et al., 2007 ;Frank et al., 2001 ;Verbruggen and Logan, 2008 ;Vink et al., 2005 ) of motor response.Reduced output from the striatum to the frontal cortex may lead to difficulties in initiating or terminating ongoing movements, which could underlie symptoms such as akinesia, posturing, ster eotypic mov ements, and ec holalia (Jahanshahi et al., 2015 ;Northoff, 2002 ).Indeed, patients with idiopathic basal ganglia calcification, an inherited neurological disorder, also exhibit catatonia-like symptoms (Brunoni et al., 2009 ;Ishitobi et al., 2014 ;Saito et al., 2010 ), and blockade of striatal dopamine receptors has been shown to cause catalepsy in animal studies (Hauber et al., 2001 ).In all, there is strong evidence to suggest a central role for the basal ganglia in the pathophysiology of catatonia.Extending these findings means that disruptions in the anatomical connections linking basal ganglia to the cortical regions contribute to the de v elopment of catatonia.
Sc hizophr enia spectrum disorders (SSD) include a variety of diseases, most of which are schizophrenia.A recent study examined brainstem volume in catatonic SSD patients, and a significant correlation between the NCRS (Northoff Catatonia Rating Scale) motor score and the whole brainstem was found.It suggests abnormal total brainstem volume is essential in catatonia (Fritze et al., 2020 ).Furthermore, Hirjak et al. ( 2019 ) argued that cortical changes in frontoparietal regions might drive the catatonic SSD.Studies have shown that orbital-frontal/parietal lobe, thalamus, and striatal structural abnormalities are mainly present in SSD patients with catatonia compared with SSD patients without catatonia (Hirjak et al., 2020 ;Wasserthal et al., 2020 ).Ho w e v er, Dean et al. ( 2020 ) sho w ed that there is no difference in brain structure in SSD patients with catatonia and those without catatonia.Ho w e v er, compar ed with healthy controls, the patient groups had r educed gr ay volume in se v er al ar eas, including insula, anterior cingulate, medial frontal, and temporal cortices.In addition, SSD patients with catatonia had more significant cognitive difficulties than SSD patients without catatonia (Dean et al., 2020 ).Compared to SSD patients without catatonia, patients with catatonia had increased static functional connectivity in the cerebellar networks and decreased lo w-frequenc y oscillations in the basal ganglia networks (Sambataro et al ., 2021 ).

Implications of MRI Findings in Schizophrenia with Catatonia for ECT and rTMS
Although some liter atur e describes the impr ov ement of catatonic symptoms after ECT, there is also negative findings .T he case report sho w ed a patient failed to impr ov e after pr olonged ECT and TMS treatment (Stip et al., 2017 ).In recent years, rTMS has been consider ed a ne w option for tr eating neur opsyc hiatric diseases (Wu et al., 2021 ).rTMS may be a safe alternative treatment strategy for patients who failed multiple drug interventions and had safety problems with ECT .T o date, only a few reports on TMS use in catatonia have been published.Although TMS appears promis-ing, conclusions can only be drawn with more studies or a more significant number of case reports.
In line with an individualized strategy for MRI-guided and navigated neur ostim ulation in sc hizophr enia (Wu et al., 2021 ), MRI findings might meet the request for proper treatment of catatonia in sc hizophr enia with ECT or rTMS.

Futur e Dir ections on Tr eatment
Lorazepam is the first choice for treating catatonia, and benzodiazepines are the first-line treatment option.Eighty percent of catatonia is r elie v ed by benzodiazepines or barbitur ates, while ECT will be adopted for those patients who fail (Appiani and Castr o, 2018 ).The r etr ospectiv e studies r e ported that it is effecti ve in 80 to 100% of all forms of catatonia, including a total of 171 patients (Dutt et al., 2011 ;Rav eendr anathan et al., 2012 ;Unal et al., 2013 ).It is considered the most effective intervention for catatonia (Ungvari et al., 2018 ).Unal et al. ( 2013 ) found that catatonia recov er ed thr ough combined tr eatment of benzodiazepine and ECT.When drug intervention fails and ECT has safety problems, rTMS is helpful for the acute and maintenance treatment of catatonia in sc hizophr enia.It is a non-inv asiv e stim ulation tec hnique with no adverse effects on cognition compared with ECT (Stip et al., 2018 ).Ho w e v er, due to small-sized studies, there is a lack of solid evidence to support clinically significant results and the application of rTMS in tr eating sc hizophr enia with catatonia is limited.In a case report (Stip et al., 2017 ), catatonic symptoms decreased significantly after receiving rTMS targeting bilateral DLPFC.rTMS may be an effective and safe treatment for catatonia.In conclusion, rTMS can be used in patients who are drug non-responsive or contraindicated to ECT.Ho w ever, more clinical studies are needed to e v aluate the suitability of rTMS for the treatment of catatonia.In the future, it may be possible to help the de v elopment of precision psyc hiatry thr ough the mec hanism of computational models .T he Bush-Francis Catatonia Rating Scale can be used as a measure of symptoms and then a model of the mechanism explaining how rTMS r elie v es the symptoms of catatonia can be de v eloped.
Our r e vie w shows that catatonia patients exhibit decreased connectivity in a network encompassing cortico-striatalcer ebellar pathway, suc h as decr eased structur al connectivity between the bilateral medial orbitofrontal cortex and striatum.Catatonia may involve changes in the myelination of corticostriatal connections.In addition, the increase in the se v erity of catatonia may be associated with high perfusion of SMA.Multiple br ain structur es differ in patients with catatonia compar ed to patients without catatonia.These findings suggest that the affected functional and anatomical circuitry of the corticostriatal system is part of the neurobiology of catatonia.Based on this e vidence, an alternativ e a ppr oac h is a stim ulation tar geting the affected corticostriatal network in sc hizophr enia with catatonia r e v ealed by MRI.Although most rTMS studies curr entl y tar get the DLPFC for the modulation of catatonia, future studies may target the cortico-striatal-cerebellar pathway to explore new targets for treating catatonia.Ho w ever, the specificity of this pathway needs to be further explored.

Conclusions
Neur obiological under pinnings contribute to the manifestation of catatonia remains to be determined.Nevertheless , rTMS ma y serve as an effective and safe treatment for catatonia.The cortico-striatal-cerebellar pathwa y ma y be served as a target of rTMS for catatonia.

Figure 1 :
Figure 1: Strategy of literature review in this study.Six reviews were excluded at this screening stage.After full-text reading, 17 papers were excluded because they did not meet the inclusion criteria.Nine recent r eports wer e included to r e vie w the MRI findings in sc hizophr enia with catatonia.

Table 1 :
Examples of studies on sc hizophr enia with catatonia.