Clozapine is the antipsychotic medication with the greatest efficacy in treatment-resistant schizophrenia (TRS). Unfortunately, clozapine is ceased in approximately 0.2% to 8.5% of people due to concerns about clozapine-associated myocarditis (CAM). The opportunity for clozapine rechallenge is important for people with TRS and CAM, due to limited alternative treatments. However, there is a lack of consensus regarding the optimal process, monitoring, and dose titration to achieve successful clozapine rechallenge. The study aimed to review the process, monitoring, and dose titration within cases of clozapine rechallenge after CAM, to identify features associated with successful rechallenge.

A systematic review of clozapine rechallenge cases following CAM was conducted. PubMed, EMBASE, Cinahl, and PsycINFO were searched for cases. Reference lists of retrieved articles and field experts were consulted to identify additional studies.

Forty-five cases were identified that described clozapine rechallenge, 31 of which were successful. Successful rechallenge cases generally used a slower dose titration regime with more frequent monitoring than standard clozapine initiation protocols; however, this data was not always completely recorded within cases. Six cases referred to published rechallenge protocols to guide their rechallenge.

The process, monitoring, and dose titration of clozapine rechallenge are inconsistently reported in the literature. Despite this, 69% of case reports detailed a successful rechallenge post CAM; noting limitations associated with reliance on case data. Ensuring published clozapine rechallenge cases report standardised data, including titration speed and monitoring frequencies, is required to guide the development and validation of guidelines for clozapine rechallenge.

To determine the association between blood markers of white matter injury (e.g., serum neurofilament light and phosphorylated neurofilament heavy) and a novel neuroimaging technique measuring microstructural white matter changes (e.g., diffusion kurtosis imaging) in regions (e.g., anterior thalamic radiation and uncinate fasciculus) known to be impacted in traumatic brain injury (TBI) and associated with symptoms common in those with chronic TBI (e.g., sleep disruption, cognitive and emotional disinhibition) in a heterogeneous sample of Veterans and non-Veterans with a history of remote TBI (i.e., >6 months).

Participants with complete imaging and blood data (N=24) were sampled from a larger multisite study of chronic mild-moderate TBI. Participants ranged in age from young to middle-aged (mean age = 34.17, SD age = 10.96, range = 19-58) and primarily male (66.7%). The number of distinct TBIs ranged from 1-5 and the time since most recent TBI ranged from 0-30 years. Scores on a cognitive screener (MoCA) ranged from 22-30 (mean = 26.75). We performed bivariate correlations with mean kurtosis (MK) in the anterior thalamic radiation (ATR; left, right) uncinate fasciculus (UF; left, right), and serum neurofilament light (NFL), and phosphorylated neurofilament heavy (pNFH). Both were log transformed for non-normality. Significance threshold was set at p<0.05.

pNFH was significantly and negatively correlated to MK in the right (r=-0.446) and left (r=-0.599) UF and right (r=-0.531) and left (r=-0.469) ATR. NFL showed moderate associations with MK in the right (r=-0.345) and left (r=-0.361) UF and little to small association in the right (r=-0.063) and left (r=-0.215) ATR. In post-hoc analyses, MK in both the left (r=0.434) and right (r=0.514) UF was positively associated with performance on a frontally-mediated list-learning task (California Verbal Learning Test, 2nd Edition; Trials 1-5 total).

Results suggest that serum pNFH may be a more sensitive blood marker of microstructural complexity in white matter regions frequently impacted by TBI in a chronic mild-moderate TBI sample. Further, it suggests that even years after a mild-moderate TBI, levels of pNFH may be informative regarding white matter integrity in regions related to executive functioning and emotional disinhibition, both of which are common presenting problems when these patients are seen in a clinical setting.

To determine the association between in-vivo spectroscopy metabolite data, the local connectome, and markers of initial injury severity (I.e., history of loss of consciousness; LoC) in traumatic brain injury (TBI), in a heterogenous sample of Veterans and non-Veterans with a history of remote mild-to-moderate TBI (I.e., >6 months).

Participants with complete PRESS magnetic resonance spectroscopy (MRS) and diffusion weighted imaging (DWI) data (N = 41) were sampled from a larger multisite study of chronic mild-to-moderate TBI (Nmiid = 38; Nmoderate = 3; 54% with LoC; 46% with multiple TBI). The sample was predominantly male (76%) with ages ranging from 23-59 (M = 36.9, SD = 10.1), with 98% holding at least a high school degree (M = 14.5 years of education, SD = 2.4). Fully tissue-and-relaxation-corrected metabolite concentration estimates in the dorsal anterior cingulate (30x30x30mm voxel) were modeled using Osprey 2.4.0. Total creatine (tCr), total choline (tCho), total N-acetylaspartate (tNAA), glutamate/glutamine (Glx), and myo-inositol (mI) were analyzed. Logistic regression was used to measure the association between metabolites and history of TBI with LoC. Correlational connectometry using the normalized spin distribution function was performed for metabolites associated with LoC, to characterize the local connectome associated with metabolites of interest, controlling for age and sex, and correcting for multiple comparisons (FDR < .050 with 4000 permutations). A profile approach was used to interpret diffusion metrics, contrasting quantitative anisotropy (QA) with fractional anisotropy (FA). Local connectome tracks were then clustered to identify the larger white matter tract.

Glx (p = .008) and tCr (p = .032) were significantly associated with history of TBI with LoC. Increased Glx was associated with increased QA in 11,001 tracks, accounting for 1.4% of the total white matter tracks in the brain. 90% of tracks were identified in bilateral cingulum (33%), bilateral thalamic (13%), bilateral corticospinal (13%), corpus callosum (12%), left arcuate fasciculus (9%), left frontoparietal aslant tracts (6%), and bilateral inferior fronto-occipital fasciculus (4%) tracts. In contrast, FA was not associated with Glx. The same pattern emerged for tCr, with 10,542 tracks identified predominantly in bilateral cingulum (29%), corpus callosum (21%), bilateral corticospinal (15%), bilateral corticostriatal (7%), bilateral medial lemniscus (7%), left cortico-pontine (3%), left thalamic (2%), and bilateral superior longitudinal fasciculus (2%) tracts. Post-hoc exploratory analyses of mean QA across regions of cingulum found that increased QA was associated with self-report measures of headache intensity, fatigue, and perceived change in executive functioning.

Results provided evidence that multimodal imaging can identify subtle markers of initial TBI severity years after injury. Neurometabolite concentrations were associated with diffuse changes in the local connectome; the pattern of discrepancy between FA and QA was suggestive of reduced potential for neuroplasticity. Exploratory analyses further indicated that variability in white matter density in the cingulum, an important connection for limbic regions, was associated with a range of problems commonly reported in clinical settings, which may be informative for diagnosis and treatment planning.

It has been previously identified that levels of peripheral inflammatory proteins, such as cytokines, are altered in people with schizophrenia spectrum disorders (SSD).

As there is considerable inconsistency in the literature with respect to how inflammatory profiles differ between acute and chronic stages of SSD, a systematic review and network meta-analysis was performed.

Records from CINAHL, the Cochrane Central Register of Controlled Trials, EMBASE, PubMed, and PsycINFO were systematically searched from inception until 31 March 2022 for published studies that had measured levels of inflammatory proteins in cases of SSD and healthy controls. Pairwise and network meta-analyses were performed to determine whether there were significant differences in mean peripheral protein concentrations between acute SSD, chronic SSD, and healthy controls.

After application of the screening process, 215 articles were included for data-analysis. One group of markers were consistently elevated (p<0·05) in both acute and chronic SSD, relative to healthy controls; this group comprised interleukin (IL)-1β, IL-1 receptor antagonist (IL-1RA), soluble interleukin-2 receptor (sIL-2R), IL-6, IL-8, IL-10, tumor necrosis factor (TNF)-α, and high sensitivity C-reactive protein (hsCRP). A second group of markers were inconsistently altered between illness stages: IL-2 and interferon (IFN)-γ were significantly elevated (p<0·05) in acute SSD, whilst IL-4, IL-12 and IFN-γ were significantly decreased (p<0·05) in chronic SSD.

These results indicate that a baseline level of inflammatory protein alteration occurs in SSD throughout the course of illness. This was evident from the group of markers that were consistently elevated in acute and chronic SSD (e.g., IL-6), representing possible trait markers. Moreover, superimposed immune activity may occur in acute SSD, given the group of possible state markers that were increased only in acute illness (e.g., IFN-γ). Further research is required to elucidate whether these peripheral changes are reflected within the central nervous system.

None Declared

Clozapine has been well established as the most efficacious medication for treatment refractory schizophrenia. Optimising the benefit during clozapine trial is an important clinical consideration. Therapeutic drug monitoring of clozapine plasma or serum levels has formed a critical part of this. Though there is no agreed standardised therapeutic range, advice traditionally recommends a clozapine level of >350ng/mL in order to effect best response. Most studies analysing the relationship between treatment response and clozapine level are older, have small sample sizes, and do not consider whether additional factors might assist in determining optimal clozapine level for response.

We conducted a systematic review of PubMed, PsycInfo and Embase for studies that provided individual participant level data on clozapine levels and response.

This data was analysed using Receiver Operating Characteristic (ROC) curves to determine the prediction performance of serum clozapine levels for treatment response.

We were able to include data on 294 individual participants. ROC analysis yielded an area under the curve (AUC) of 0.612. The clozapine level at the optimal Youden index was 372ng/mL, and at this level there was response sensitivity of 57.3%, and specificity of 65.7%. The interquartile range for treatment response was 223ng/mL – 558ng/mL. There was no improvement in ROC performance with mixed models including patient sex, age or length of trial.

Clozapine dose should be optimised based on clozapine therapeutic levels. We found that a range between 250 – 550ng/mL could be recommended, while noting that a level of >350ng/mL is most optimal for response.

None Declared

Although clozapine is the most efficacious medication for treatment-refractory schizophrenia, not all patients will have an adequate response. Optimising clozapine dose using therapeutic drug monitoring could therefore maximise response.

Using individual patient data, we undertook a receiver operating characteristic (ROC) curve analysis to determine an optimal therapeutic range for clozapine levels to guide clinical practice.

We conducted a systematic review of PubMed, PsycINFO and Embase for studies that provided individual participant level data on clozapine levels and response. These data were analysed using ROC curves to determine the prediction performance of plasma clozapine levels for treatment response.

We included data on 294 individual participants from nine studies. ROC analysis yielded an area under the curve of 0.612. The clozapine level at the point of optimal diagnostic benefit was 372 ng/mL; at this level, the response sensitivity was 57.3%, and specificity 65.7%. The interquartile range for treatment response was 223–558 ng/mL. There was no improvement in ROC performance with mixed models including patient gender, age or length of trial. Clozapine dose and clozapine concentration to dose ratio did not provide significantly meaningful prediction of response to clozapine.

Clozapine dose should be optimised based on clozapine therapeutic levels. We found that a range between 250 and 550 ng/mL could be recommended, while noting that a level of >350 ng/mL is the most optimal for response. Although some patients may not respond without clozapine levels >550 ng/mL, the benefits should be weighed against the increased risk of adverse drug reactions.

Despite a wide range of proposed risk factors and theoretical models, prediction of eating disorder (ED) onset remains poor. This study undertook the first comparison of two machine learning (ML) approaches [penalised logistic regression (LASSO), and prediction rule ensembles (PREs)] to conventional logistic regression (LR) models to enhance prediction of ED onset and differential ED diagnoses from a range of putative risk factors.

Data were part of a European Project and comprised 1402 participants, 642 ED patients [52% with anorexia nervosa (AN) and 40% with bulimia nervosa (BN)] and 760 controls. The Cross-Cultural Risk Factor Questionnaire, which assesses retrospectively a range of sociocultural and psychological ED risk factors occurring before the age of 12 years (46 predictors in total), was used.

All three statistical approaches had satisfactory model accuracy, with an average area under the curve (AUC) of 86% for predicting ED onset and 70% for predicting AN v. BN. Predictive performance was greatest for the two regression methods (LR and LASSO), although the PRE technique relied on fewer predictors with comparable accuracy. The individual risk factors differed depending on the outcome classification (EDs v. non-EDs and AN v. BN).

Even though the conventional LR performed comparably to the ML approaches in terms of predictive accuracy, the ML methods produced more parsimonious predictive models. ML approaches offer a viable way to modify screening practices for ED risk that balance accuracy against participant burden.