2-Hydroxyglutarate as an MR spectroscopic predictor of an IDH mutation in gliomas

The mutated enzyme isocitrate dehydrogenase (IDH) 1 and 2 has been detected in various tumor entities such as gliomas and can convert α-ketoglutarate into the oncometabolite 2-hydroxyglutarate (2-HG). This neuro-oncologically significant metabolic product can be detected by MR spectroscopy and is therefore suitable for noninvasive glioma classification and therapy monitoring.This paper provides an up-to-date overview of the methodology and relevance of 1H-MR spectroscopy (MRS) in the oncological primary and follow-up diagnosis of gliomas. The possibilities and limitations of this MR spectroscopic examination are evaluated on the basis of the available literature.By detecting 2-HG, MRS can in principle offer a noninvasive alternative to immunohistological analysis thus avoiding surgical intervention in some cases. However, in addition to an adapted and optimized examination protocol, the individual measurement conditions in the examination region are of decisive importance. Due to the inherently small signal of 2-HG, unfavorable measurement conditions can influence the reliability of detection. · MR spectroscopy enables the non-invasive detection of 2-hydroxyglutarate.. · The measurement of this metabolite allows the detection of an IDH mutation in gliomas.. · The choice of MR examination method is particularly important.. · Detection reliability is influenced by glioma size, necrotic tissue and the existing measurement conditions.. · Bauer J, Raum HN, Kugel H et al. 2-Hydroxyglutarate as an MR spectroscopic predictor of an IDH mutation in gliomas. Fortschr Röntgenstr 2024; DOI 10.1055/a-2285-4923.

Radiomics-Based Prediction Model for Outcome of Radioembolization in Metastatic Colorectal Cancer

PURPOSE

To evaluate the benefit of a contrast-enhanced computed tomography (CT) radiomics-based model for predicting response and survival in patients with colorectal liver metastases treated with transarterial Yttrium-90 radioembolization (TARE).

MATERIALS AND METHODS

Fifty-one patients who underwent TARE were included in this single-center retrospective study. Response to treatment was assessed using the Response Evaluation Criteria in Solid Tumors (RECIST 1.1) at 3-month follow-up. Patients were stratified as responders (complete/partial response and stable disease, n = 24) or non-responders (progressive disease, n = 27). Radiomic features (RF) were extracted from pre-TARE CT after segmentation of the liver tumor volume. A model was built based on a radiomic signature consisting of reliable RFs that allowed classification of response using multivariate logistic regression. Patients were assigned to high- or low-risk groups for disease progression after TARE according to a cutoff defined in the model. Kaplan-Meier analysis was performed to analyze survival between high- and low-risk groups.

RESULTS

Two independent RF [Energy, Maximal Correlation Coefficient (MCC)], reflecting tumor heterogeneity, discriminated well between responders and non-responders. In particular, patients with higher magnitude of voxel values in an image (Energy), and texture complexity (MCC), were more likely to fail TARE. For predicting treatment response, the area under the receiver operating characteristic curve of the radiomics-based model was 0.75 (95% CI 0.48-1). The high-risk group had a shorter overall survival than the low-risk group (3.4 vs. 6.4 months, p < 0.001).

CONCLUSION

Our CT radiomics model may predict the response and survival outcome by quantifying tumor heterogeneity in patients treated with TARE for colorectal liver metastases.

Long-term effects of electroconvulsive therapy on brain structure in major depression

BACKGROUND

Magnetic resonance imaging (MRI) studies on major depressive disorder (MDD) have predominantly found short-term electroconvulsive therapy (ECT)-related gray matter volume (GMV) increases, but research on the long-term stability of such changes is missing. Our aim was to investigate long-term GMV changes over a 2-year period after ECT administration and their associations with clinical outcome.

METHODS

In this nonrandomized longitudinal study, patients with MDD undergoing ECT (n = 17) are assessed three times by structural MRI: Before ECT (t0), after ECT (t1) and 2 years later (t2). A healthy (n = 21) and MDD non-ECT (n = 33) control group are also measured three times within an equivalent time interval. A 3(group) × 3(time) ANOVA on whole-brain level and correlation analyses with clinical outcome variables is performed.

RESULTS

Analyses yield a significant group × time interaction (pFWE < 0.001) resulting from significant volume increases from t0 to t1 and decreases from t1 to t2 in the ECT group, e.g., in limbic areas. There are no effects of time in both control groups. Volume increases from t0 to t1 correlate with immediate and delayed symptom increase, while volume decreases from t1 to t2 correlate with long-term depressive outcome (all p ⩽ 0.049).

CONCLUSIONS

Volume increases induced by ECT appear to be a transient phenomenon as volume strongly decreased 2 years after ECT. Short-term volume increases are associated with less symptom improvement suggesting that the antidepressant effect of ECT is not due to volume changes. Larger volume decreases are associated with poorer long-term outcome highlighting the interplay between disease progression and structural changes.

A Systematic Evaluation of Machine Learning-Based Biomarkers for Major Depressive Disorder

Importance

Biological psychiatry aims to understand mental disorders in terms of altered neurobiological pathways. However, for one of the most prevalent and disabling mental disorders, major depressive disorder (MDD), no informative biomarkers have been identified.

Objective

To evaluate whether machine learning (ML) can identify a multivariate biomarker for MDD.

Design, Setting, and Participants

This study used data from the Marburg-Münster Affective Disorders Cohort Study, a case-control clinical neuroimaging study. Patients with acute or lifetime MDD and healthy controls aged 18 to 65 years were recruited from primary care and the general population in Münster and Marburg, Germany, from September 11, 2014, to September 26, 2018. The Münster Neuroimaging Cohort (MNC) was used as an independent partial replication sample. Data were analyzed from April 2022 to June 2023.

Exposure

Patients with MDD and healthy controls.

Main Outcome and Measure

Diagnostic classification accuracy was quantified on an individual level using an extensive ML-based multivariate approach across a comprehensive range of neuroimaging modalities, including structural and functional magnetic resonance imaging and diffusion tensor imaging as well as a polygenic risk score for depression.

Results

Of 1801 included participants, 1162 (64.5%) were female, and the mean (SD) age was 36.1 (13.1) years. There were a total of 856 patients with MDD (47.5%) and 945 healthy controls (52.5%). The MNC replication sample included 1198 individuals (362 with MDD [30.1%] and 836 healthy controls [69.9%]). Training and testing a total of 4 million ML models, mean (SD) accuracies for diagnostic classification ranged between 48.1% (3.6%) and 62.0% (4.8%). Integrating neuroimaging modalities and stratifying individuals based on age, sex, treatment, or remission status does not enhance model performance. Findings were replicated within study sites and also observed in structural magnetic resonance imaging within MNC. Under simulated conditions of perfect reliability, performance did not significantly improve. Analyzing model errors suggests that symptom severity could be a potential focus for identifying MDD subgroups.

Conclusion and Relevance

Despite the improved predictive capability of multivariate compared with univariate neuroimaging markers, no informative individual-level MDD biomarker-even under extensive ML optimization in a large sample of diagnosed patients-could be identified.

Computed tomography hypoperfusion-hypodensity mismatch vs. automated perfusion mismatch to identify stroke patients eligible for thrombolysis

Background and purpose

Automated perfusion imaging can detect stroke patients with unknown time of symptom onset who are eligible for thrombolysis. However, the availability of this technique is limited. We, therefore, established the novel concept of computed tomography (CT) hypoperfusion-hypodensity mismatch, i.e., an ischemic core lesion visible on cerebral perfusion CT without visible hypodensity in the corresponding native cerebral CT. We compared both methods regarding their accuracy in identifying patients suitable for thrombolysis.

Methods

In a retrospective analysis of the MissPerfeCT observational cohort study, patients were classified as suitable or not for thrombolysis based on established time window and imaging criteria. We calculated predictive values for hypoperfusion-hypodensity mismatch and automated perfusion imaging to compare accuracy in the identification of patients suitable for thrombolysis.

Results

Of 247 patients, 219 (88.7%) were eligible for thrombolysis and 28 (11.3%) were not eligible for thrombolysis. Of 197 patients who were within 4.5 h of symptom onset, 190 (96.4%) were identified by hypoperfusion-hypodensity mismatch and 88 (44.7%) by automated perfusion mismatch (p < 0.001). Of 22 patients who were beyond 4.5 h of symptom onset but were eligible for thrombolysis, 5 patients (22.7%) were identified by hypoperfusion-hypodensity mismatch. Predictive values for the hypoperfusion-hypodensity mismatch vs. automated perfusion mismatch were as follows: sensitivity, 89.0% vs. 50.2%; specificity, 71.4% vs. 100.0%; positive predictive value, 96.1% vs. 100.0%; and negative predictive value, 45.5% vs. 20.4%.

Conclusion

The novel method of hypoperfusion-hypodensity mismatch can identify patients suitable for thrombolysis with higher sensitivity and lower specificity than established techniques. Using this simple method might therefore increase the proportion of patients treated with thrombolysis without the use of special automated software.The MissPerfeCT study is a retrospective observational multicenter cohort study and is registered with clinicaltrials.gov (NCT04277728).

Multi-organ Radiomics-Based Prediction of Future Remnant Liver Hypertrophy Following Portal Vein Embolization

BACKGROUND

Portal vein embolization (PVE) is used to induce remnant liver hypertrophy prior to major hepatectomy. The purpose of this study was to evaluate the predictive value of baseline computed tomography (CT) data for future remnant liver (FRL) hypertrophy after PVE.

METHODS

In this retrospective study, all consecutive patients undergoing right-sided PVE with or without hepatic vein embolization between 2018 and 2021 were included. CT volumetry was performed before and after PVE to assess standardized FRL volume (sFRLV). Radiomic features were extracted from baseline CT after segmenting liver (without tumor), spleen and bone marrow. For selecting features that allow classification of response (hypertrophy ≥ 1.33), a stepwise dimension reduction was performed. Logistic regression models were fitted and selected features were tested for their predictive value. Decision curve analysis was performed on the test dataset.

RESULTS

A total of 53 patients with liver tumor were included in this study. sFRLV increased significantly after PVE, with a mean hypertrophy of FRL of 1.5 ± 0.3-fold. sFRLV hypertrophy ≥ 1.33 was reached in 35 (66%) patients. Three independent radiomic features, i.e. liver-, spleen- and bone marrow-associated, differentiated well between responders and non-responders. A logistic regression model revealed the highest accuracy (area under the curve 0.875) for the prediction of response, with sensitivity of 1.0 and specificity of 0.5. Decision curve analysis revealed a positive net benefit when applying the model.

CONCLUSIONS

This proof-of-concept study provides first evidence of a potential predictive value of baseline multi-organ radiomics CT data for FRL hypertrophy after PVE.

Digital Breast Tomosynthesis versus Digital Mammography for Detection of Early-Stage Cancers Stratified by Grade: A TOSYMA Subanalysis

Background Breast cancer screening with digital breast tomosynthesis (DBT) plus synthesized mammography (SM) increases invasive tumor detection compared with digital mammography (DM). However, it is not known how the prognostic characteristics of the cancers detected with the two screening approaches differ. Purpose To compare invasive breast cancers detected with DBT plus SM (test arm) versus DM (control arm) screening with regard to tumor stage, histologic grade, patient age, and breast density. Materials and Methods This exploratory subanalysis of the Tomosynthesis plus Synthesized Mammography (TOSYMA) study, which is a multicenter randomized controlled trial embedded in the German mammography screening program, recruited women aged 50-70 years from July 2018 to December 2020. It compared invasive cancer detection rates (iCDRs), rate differences, and odds ratios (ORs) between the arms stratified by Union for International Cancer Control (UICC) stage (I vs II-IV), histologic grade (1 vs 2 or 3), age group (50-59 vs 60-70 years), and Breast Imaging Reporting and Data System categories of breast density (A or B vs C or D). Results In total, 49 462 (median age, 57 years [IQR, 53-62 years]) and 49 669 (median age, 57 years [IQR, 53-62 years]) participants were allocated to DBT plus SM and DM screening, respectively. The iCDR of stage I tumors with DBT plus SM was 51.6 per 10 000 women (255 of 49 462) and with DM it was 30.0 per 10 000 women (149 of 49 669). DBT plus SM depicted more stage I tumors with grade 2 or 3 (166 of 49 462, 33.7 per 10 000 women) than DM (106 of 49 669, 21.3 per 10 000 women; rate difference, +12.3 per 10 000 women [95% CI: 0.3, 24.9]; OR, 1.6 [95% CI: 0.9, 2.7]). DBT plus SM achieved the highest iCDR of stage I tumors with grade 2 or 3 among women aged 60-70 years with dense breasts (41 of 7364, 55.4 per 10 000 women; rate difference, +21.6 per 10 000 women [95% CI: -21.1, 64.3]; OR, 1.6 [95% CI: 0.6, 4.5]). Conclusion DBT plus SM screening appears to lead to higher detection of early-stage invasive breast cancers of grade 2 or 3 than DM screening, with the highest rate among women aged 60-70 years with dense breasts. Clinical trial registration no. NCT03377036 © RSNA, 2023 See also the editorial by Ha and Chang in this issue.

Intratumoral heterogeneity after targeted therapy in murine cancer models with differing degrees of malignancy

INTRODUCTION

Conventional morphologic and volumetric assessment of treatment response is not suitable for adequately assessing responses to targeted cancer therapy. The aim of this study was to evaluate changes in tumor composition after targeted therapy in murine models of breast cancer with differing degrees of malignancy via non-invasive magnetic resonance imaging (MRI).

MATERIALS AND METHODS

Mice bearing highly malignant 4T1 tumors or low malignant 67NR tumors were treated with either a combination of two immune checkpoint inhibitors (ICI, anti-PD1 and anti-CTLA-4) or the multi-tyrosine kinase inhibitor sorafenib, following experiments with macrophage-depleting clodronate-loaded liposomes and vessel-stabilizing angiopoietin-1. Mice were imaged on a 9.4 T small animal MRI system with a multiparametric (mp) protocol, comprising T1 and T2 mapping and diffusion-weighted imaging. Tumors were analyzed ex vivo with histology.

RESULTS AND DISCUSSIONS

All treatments led to an increase in non-viable areas, but therapy-induced intratumoral changes differed between the two tumor models and the different targeted treatments. While ICI treatment led to intratumoral hemorrhage, sorafenib treatment mainly induced intratumoral necrosis. Treated 4T1 tumors showed increasing and extensive areas of necrosis, in comparison to 67NR tumors with only small, but also increasing, necrotic areas. After either of the applied treatments, intratumoral heterogeneity, was increased in both tumor models, and confirmed ex vivo by histology. Apparent diffusion coefficient with subsequent histogram analysis proved to be the most sensitive MRI sequence. In conclusion, mp MRI enables to assess dedicated therapy-related intratumoral changes and may serve as a biomarker for treatment response assessment.

Multiparametric chemical exchange saturation transfer MRI detects metabolic changes in breast cancer following immunotherapy

BACKGROUND

With metabolic alterations of the tumor microenvironment (TME) contributing to cancer progression, metastatic spread and response to targeted therapies, non-invasive and repetitive imaging of tumor metabolism is of major importance. The purpose of this study was to investigate whether multiparametric chemical exchange saturation transfer magnetic resonance imaging (CEST-MRI) allows to detect differences in the metabolic profiles of the TME in murine breast cancer models with divergent degrees of malignancy and to assess their response to immunotherapy.

METHODS

Tumor characteristics of highly malignant 4T1 and low malignant 67NR murine breast cancer models were investigated, and their changes during tumor progression and immune checkpoint inhibitor (ICI) treatment were evaluated. For simultaneous analysis of different metabolites, multiparametric CEST-MRI with calculation of asymmetric magnetization transfer ratio (MTRasym) at 1.2 to 2.0 ppm for glucose-weighted, 2.0 ppm for creatine-weighted and 3.2 to 3.6 ppm for amide proton transfer- (APT-) weighted CEST contrast was conducted. Ex vivo validation of MRI results was achieved by 1H nuclear magnetic resonance spectroscopy, matrix-assisted laser desorption/ionization mass spectrometry imaging with laser postionization and immunohistochemistry.

RESULTS

During tumor progression, the two tumor models showed divergent trends for all examined CEST contrasts: While glucose- and APT-weighted CEST contrast decreased and creatine-weighted CEST contrast increased over time in the 4T1 model, 67NR tumors exhibited increased glucose- and APT-weighted CEST contrast during disease progression, accompanied by decreased creatine-weighted CEST contrast. Already three days after treatment initiation, CEST contrasts captured response to ICI therapy in both tumor models.

CONCLUSION

Multiparametric CEST-MRI enables non-invasive assessment of metabolic signatures of the TME, allowing both for estimation of the degree of tumor malignancy and for assessment of early response to immune checkpoint inhibition.

Use Test of Automated Machine Learning in Cancer Diagnostics

Our aim is to investigate the added value of automated machine learning (AutoML) for potential future applications in cancer diagnostics. Using two important diagnostic questions, the non-invasive determination of IDH mutation status and ATRX status, we analyze whether it is possible to use AutoML to develop models that are comparable in performance to conventional machine learning models (ML) developed by experts. For this purpose, we develop AutoML models using different feature preselection methods and compare the results with previously developed conventional ML models. The cohort used for our study comprises T2-weighted MRI images of 124 patients with histologically confirmed gliomas. Using AutoML, we were able to develop sophisticated models in a very short time with only a few lines of computer code. In predicting IDH mutation status, we obtained a mean AUC of 0.7400 and a mean AUPRC of 0.8582. ATRX mutation status was predicted with very similar discriminatory power, with a mean AUC of 0.7810 and a mean AUPRC of 0.8511. In both cases, AutoML was even able to achieve a discriminatory power slightly above that of the respective conventionally developed models in a very short computing time, thus making such methods accessible to non-experts in the near future.

Resting-state functional connectivity patterns associated with childhood maltreatment in a large bicentric cohort of adults with and without major depression

BACKGROUND

Childhood maltreatment (CM) represents a potent risk factor for major depressive disorder (MDD), including poorer treatment response. Altered resting-state connectivity in the fronto-limbic system has been reported in maltreated individuals. However, previous results in smaller samples differ largely regarding localization and direction of effects.

METHODS

We included healthy and depressed samples [n = 624 participants with MDD; n = 701 healthy control (HC) participants] that underwent resting-state functional MRI measurements and provided retrospective self-reports of maltreatment using the Childhood Trauma Questionnaire. A-priori defined regions of interest [ROI; amygdala, hippocampus, anterior cingulate cortex (ACC)] were used to calculate seed-to-voxel connectivities.

RESULTS

No significant associations between maltreatment and resting-state connectivity of any ROI were found across MDD and HC participants and no interaction effect with diagnosis became significant. Investigating MDD patients only yielded maltreatment-associated increased connectivity between the amygdala and dorsolateral frontal areas [pFDR < 0.001; η2partial = 0.050; 95%-CI (0.023-0.085)]. This effect was robust across various sensitivity analyses and was associated with concurrent and previous symptom severity. Particularly strong amygdala-frontal associations with maltreatment were observed in acutely depressed individuals [n = 264; pFDR < 0.001; η2partial = 0.091; 95%-CI (0.038-0.166)). Weaker evidence - not surviving correction for multiple ROI analyses - was found for altered supracallosal ACC connectivity in HC individuals associated with maltreatment.

CONCLUSIONS

The majority of previous resting-state connectivity correlates of CM could not be replicated in this large-scale study. The strongest evidence was found for clinically relevant maltreatment associations with altered adult amygdala-dorsolateral frontal connectivity in depression. Future studies should explore the relevance of this pathway for a maltreated subgroup of MDD patients.

Comparison of MRI Sequences to Predict ATRX Status Using Radiomics-Based Machine Learning

ATRX is an important molecular marker according to the 2021 WHO classification of adult-type diffuse glioma. We aim to predict the ATRX mutation status non-invasively using radiomics-based machine learning models on MRI and to determine which MRI sequence is best suited for this purpose. In this retrospective study, we used MRI images of patients with histologically confirmed glioma, including the sequences T1w without and with the administration of contrast agent, T2w, and the FLAIR. Radiomics features were extracted from the corresponding MRI images by hand-delineated regions of interest. Data partitioning into training data and independent test data was repeated 100 times to avoid random effects. Feature preselection and subsequent model development were performed using Lasso regression. The T2w sequence was found to be the most suitable and the FLAIR sequence the least suitable for predicting ATRX mutations using radiomics-based machine learning models. For the T2w sequence, our seven-feature model developed with Lasso regression achieved a mean AUC of 0.831, a mean accuracy of 0.746, a mean sensitivity of 0.772, and a mean specificity of 0.697. In conclusion, for the prediction of ATRX mutation using radiomics-based machine learning models, the T2w sequence is the most suitable among the commonly used MRI sequences.

Vascular response patterns to targeted therapies in murine breast cancer models with divergent degrees of malignancy

BACKGROUND

Response assessment of targeted cancer therapies is becoming increasingly challenging, as it is not adequately assessable with conventional morphological and volumetric analyses of tumor lesions. The tumor microenvironment is particularly constituted by tumor vasculature which is altered by various targeted therapies. The aim of this study was to noninvasively assess changes in tumor perfusion and vessel permeability after targeted therapy in murine models of breast cancer with divergent degrees of malignancy.

METHODS

Low malignant 67NR or highly malignant 4T1 tumor-bearing mice were treated with either the multi-kinase inhibitor sorafenib or immune checkpoint inhibitors (ICI, combination of anti-PD1 and anti-CTLA4). Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) with i.v. injection of albumin-binding gadofosveset was conducted on a 9.4 T small animal MRI. Ex vivo validation of MRI results was achieved by transmission electron microscopy, immunohistochemistry and laser ablation-inductively coupled plasma-mass spectrometry.

RESULTS

Therapy-induced changes in tumor vasculature differed between low and highly malignant tumors. Sorafenib treatment led to decreased tumor perfusion and endothelial permeability in low malignant 67NR tumors. In contrast, highly malignant 4T1 tumors demonstrated characteristics of a transient window of vascular normalization with an increase in tumor perfusion and permeability early after therapy initiation, followed by decreased perfusion and permeability parameters. In the low malignant 67NR model, ICI treatment also mediated vessel-stabilizing effects with decreased tumor perfusion and permeability, while ICI-treated 4T1 tumors exhibited increasing tumor perfusion with excessive vascular leakage.

CONCLUSION

DCE-MRI enables noninvasive assessment of early changes in tumor vasculature after targeted therapies, revealing different response patterns between tumors with divergent degrees of malignancy. DCE-derived tumor perfusion and permeability parameters may serve as vascular biomarkers that allow for repetitive examination of response to antiangiogenic treatment or immunotherapy.

Association of hospitalization with structural brain alterations in patients with affective disorders over nine years

Repeated hospitalizations are a characteristic of severe disease courses in patients with affective disorders (PAD). To elucidate how a hospitalization during a nine-year follow-up in PAD affects brain structure, a longitudinal case-control study (mean [SD] follow-up period 8.98 [2.20] years) was conducted using structural neuroimaging. We investigated PAD (N = 38) and healthy controls (N = 37) at two sites (University of Münster, Germany, Trinity College Dublin, Ireland). PAD were divided into two groups based on the experience of in-patient psychiatric treatment during follow-up. Since the Dublin-patients were outpatients at baseline, the re-hospitalization analysis was limited to the Münster site (N = 52). Voxel-based morphometry was employed to examine hippocampus, insula, dorsolateral prefrontal cortex and whole-brain gray matter in two models: (1) group (patients/controls)×time (baseline/follow-up) interaction; (2) group (hospitalized patients/not-hospitalized patients/controls)×time interaction. Patients lost significantly more whole-brain gray matter volume of superior temporal gyrus and temporal pole compared to HC (p_FWE = 0.008). Patients hospitalized during follow-up lost significantly more insular volume than healthy controls (p_FWE = 0.025) and more volume in their hippocampus compared to not-hospitalized patients (p_FWE = 0.023), while patients without re-hospitalization did not differ from controls. These effects of hospitalization remained stable in a smaller sample excluding patients with bipolar disorder. PAD show gray matter volume decline in temporo-limbic regions over nine years. A hospitalization during follow-up comes with intensified gray matter volume decline in the insula and hippocampus. Since hospitalizations are a correlate of severity, this finding corroborates and extends the hypothesis that a severe course of disease has detrimental long-term effects on temporo-limbic brain structure in PAD.

Initial Results of 68Ga-FAPI-46 PET/MRI to Assess Response to Neoadjuvant Chemotherapy in Breast Cancer

Improving imaging-based response after neoadjuvant chemotherapy (NAC) in breast cancer assessment could obviate histologic confirmation of pathologic complete response (pCR) and facilitate deescalation of chemotherapy or surgery. Fibroblast activation protein inhibitor (FAPI) PET/MRI is a promising novel molecular imaging agent for the tumor microenvironment with intense uptake in breast cancer. We assessed the diagnostic performance of follow-up breast 68Ga-FAPI-46 (68Ga-FAPI) PET/MRI in classifying the response status of local breast cancer and lymph node metastases after completion of NAC and validated this approach immunohistochemically. Methods: In women who completed NAC for invasive breast cancer, follow-up 68Ga-FAPI PET/MRI and corresponding fibroblast activation protein (FAP) immunostainings were retrospectively analyzed. Metrics of 68Ga-FAPI uptake and FAP immunoreactivity in women with or without pCR were compared using the Mann-Whitney U test. Diagnostic performance to detect remnant invasive cancer was calculated for tracer uptake metrics using receiver-operating-characteristic curves and for masked readers' visual assessment categories of PET/MRI and MRI alone. Results: Thirteen women (mean age ± SD, 47 ± 9 y) were evaluated. Seven of the 13 achieved pCR in the breast and 6 in the axilla. FAP immunoreactivity was significantly associated with response status. The 68Ga-FAPI PET/MRI mean breast tumor-to-background ratio was 0.9 (range, 0.6-1.2) for pCR and 2.1 (range, 1.4-3.1) for no pCR (P = 0.001). Integrated PET/MRI could classify breast response correctly in all 13 women based on readers' visual assessment or tumor-to-background ratio. Evaluation of MRI alone resulted in at least 2 false-positives. For lymph nodes, PET/MRI readers had at least 2 false-negative classifications, whereas MRI alone resulted in 2 false-negatives and 1 false-positive. Conclusion: To our knowledge, this was the first analysis of 68Ga-FAPI PET/MRI for response assessment after NAC for breast cancer. The diagnostic performance of PET/MRI in a small study sample trended toward a gain over MRI alone, clearly supporting future prospective studies.

Radiomics of Tumor Heterogeneity in 18F-FDG-PET-CT for Predicting Response to Immune Checkpoint Inhibition in Therapy-Naïve Patients with Advanced Non-Small-Cell Lung Cancer

We aimed to evaluate the predictive and prognostic value of baseline ¹⁸F-FDG-PET-CT (PET-CT) radiomic features (RFs) for immune checkpoint-inhibitor (CKI)-based first-line therapy in advanced non-small-cell lung cancer (NSCLC) patients. In this retrospective study 44 patients were included. Patients were treated with either CKI-monotherapy or combined CKI-based immunotherapy-chemotherapy as first-line treatment. Treatment response was assessed by the Response Evaluation Criteria in Solid Tumors (RECIST). After a median follow-up of 6.4 months patients were stratified into "responder" (n = 33) and "non-responder" (n = 11). RFs were extracted from baseline PET and CT data after segmenting PET-positive tumor volume of all lesions. A Radiomics-based model was developed based on a Radiomics signature consisting of reliable RFs that allow classification of response and overall progression using multivariate logistic regression. These RF were additionally tested for their prognostic value in all patients by applying a model-derived threshold. Two independent PET-based RFs differentiated well between responders and non-responders. For predicting response, the area under the curve (AUC) was 0.69 for "PET-Skewness" and 0.75 predicting overall progression for "PET-Median". In terms of progression-free survival analysis, patients with a lower value of PET-Skewness (threshold < 0.2014; hazard ratio (HR) 0.17, 95% CI 0.06-0.46; p < 0.001) and higher value of PET-Median (threshold > 0.5233; HR 0.23, 95% CI 0.11-0.49; p < 0.001) had a significantly lower probability of disease progression or death. Our Radiomics-based model might be able to predict response in advanced NSCLC patients treated with CKI-based first-line therapy.

Sex-specifics of ECT outcome

OBJECTIVE

Electroconvulsive therapy (ECT) is the most effective treatment for patients with severe major depressive disorder (MDD). Given the known sex differences in MDD, improved knowledge may provide more sex-specific recommendations in clinical guidelines and improve outcome. In the present study we examine sex differences in ECT outcome and its predictors.

METHODS

Clinical data from 20 independent sites participating in the Global ECT-MRI Research Collaboration (GEMRIC) were obtained for analysis, totaling 500 patients with MDD (58.6 % women) with a mean age of 54.8 years. Severity of depression before and after ECT was assessed with validated depression scales. Remission was defined as a HAM-D score of 7 points or below after ECT. Variables associated with remission were selected based on literature (i.e. depression severity at baseline, age, duration of index episode, and presence of psychotic symptoms).

RESULTS

Remission rates of ECT were independent of sex, 48.0 % in women and 45.7 % in men (X²(1) = 0.2, p = 0.70). In the logistic regression analyses, a shorter index duration was identified as a sex-specific predictor for ECT outcome in women (X²(1) = 7.05, p = 0.01). The corresponding predictive margins did show overlapping confidence intervals for men and women.

CONCLUSION

The evidence provided by our study suggests that ECT as a biological treatment for MDD is equally effective in women and men. A shorter duration of index episode was an additional sex- specific predictor for remission in women. Future research should establish whether the confidence intervals for the corresponding predictive margins are overlapping, as we find, or not.

Genetic, individual, and familial risk correlates of brain network controllability in major depressive disorder

Many therapeutic interventions in psychiatry can be viewed as attempts to influence the brain's large-scale, dynamic network state transitions. Building on connectome-based graph analysis and control theory, Network Control Theory is emerging as a powerful tool to quantify network controllability-i.e., the influence of one brain region over others regarding dynamic network state transitions. If and how network controllability is related to mental health remains elusive. Here, from Diffusion Tensor Imaging data, we inferred structural connectivity and inferred calculated network controllability parameters to investigate their association with genetic and familial risk in patients diagnosed with major depressive disorder (MDD, n = 692) and healthy controls (n = 820). First, we establish that controllability measures differ between healthy controls and MDD patients while not varying with current symptom severity or remission status. Second, we show that controllability in MDD patients is associated with polygenic scores for MDD and psychiatric cross-disorder risk. Finally, we provide evidence that controllability varies with familial risk of MDD and bipolar disorder as well as with body mass index. In summary, we show that network controllability is related to genetic, individual, and familial risk in MDD patients. We discuss how these insights into individual variation of network controllability may inform mechanistic models of treatment response prediction and personalized intervention-design in mental health.

Breast Density and Breast Cancer Screening with Digital Breast Tomosynthesis: A TOSYMA Trial Subanalysis

Background Digital breast tomosynthesis (DBT) plus synthesized mammography (SM) reduces the diagnostic pitfalls of tissue superimposition, which is a limitation of digital mammography (DM). Purpose To compare the invasive breast cancer detection rate (iCDR) of DBT plus SM versus DM screening for different breast density categories. Materials and Methods An exploratory subanalysis of the TOmosynthesis plus SYnthesized MAmmography (TOSYMA) study, a randomized, controlled, multicenter, parallel-group trial recruited within the German mammography screening program from July 2018 to December 2020. Women aged 50-69 years were randomly assigned (1:1) to DBT plus SM or DM screening examination. Breast density categories A-D were visually assessed according to the Breast Imaging Reporting and Data System Atlas. Exploratory analyses were performed of the iCDR in both study arms and stratified by breast density, and odds ratios and 95% CIs were determined. Results A total of 49 762 women allocated to DBT plus SM and 49 796 allocated to DM (median age, 57 years [IQR, 53-62 years]) were included. In the DM arm, the iCDR was 3.6 per 1000 screening examinations in category A (almost entirely fatty) (16 of 4475 screenings), 4.3 in category B (102 of 23 534 screenings), 6.1 in category C (116 of 19 051 screenings), and 2.3 in category D (extremely dense breasts) (six of 2629 screenings). The iCDR in the DBT plus SM arm was 2.7 per 1000 screening examinations in category A (12 of 4439 screenings), 6.9 in category B (154 of 22 328 screenings), 8.3 in category C (156 of 18 772 screenings), and 8.1 in category D (32 of 3940 screenings). The odds ratio for DM versus DBT plus SM in category D was 3.8 (95% CI: 1.5, 11.1). The invasive cancers detected with DBT plus SM were most often grade 2 tumors; in category C, it was 58% (91 of 156 invasive cancers), and in category D, it was 47% (15 of 32 invasive cancers). Conclusion The TOmosynthesis plus SYnthesized MAmmography trial revealed higher invasive cancer detection rates with digital breast tomosynthesis plus synthesized mammography than digital mammography in dense breasts, relatively and absolutely most marked among women with extremely dense breasts. ClinicalTrials.gov registration no.: NCT03377036 © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Lee and Moy in this issue.

Towards a network control theory of electroconvulsive therapy response

Electroconvulsive Therapy (ECT) is arguably the most effective intervention for treatment-resistant depression. While large interindividual variability exists, a theory capable of explaining individual response to ECT remains elusive. To address this, we posit a quantitative, mechanistic framework of ECT response based on Network Control Theory (NCT). Then, we empirically test our approach and employ it to predict ECT treatment response. To this end, we derive a formal association between Postictal Suppression Index (PSI)-an ECT seizure quality index-and whole-brain modal and average controllability, NCT metrics based on white-matter brain network architecture, respectively. Exploiting the known association of ECT response and PSI, we then hypothesized an association between our controllability metrics and ECT response mediated by PSI. We formally tested this conjecture in N = 50 depressive patients undergoing ECT. We show that whole-brain controllability metrics based on pre-ECT structural connectome data predict ECT response in accordance with our hypotheses. In addition, we show the expected mediation effects via PSI. Importantly, our theoretically motivated metrics are at least on par with extensive machine learning models based on pre-ECT connectome data. In summary, we derived and tested a control-theoretic framework capable of predicting ECT response based on individual brain network architecture. It makes testable, quantitative predictions regarding individual therapeutic response, which are corroborated by strong empirical evidence. Our work might constitute a starting point for a comprehensive, quantitative theory of personalized ECT interventions rooted in control theory.

A magnetic resonance imaging based radiomics model to predict mitosis cycles in intracranial meningioma

The aim of this study was to develop a magnetic resonance imaging (MRI) based radiomics model to predict mitosis cycles in intracranial meningioma grading prior to surgery. Preoperative contrast-enhanced T1-weighted (T1CE) cerebral MRI data of 167 meningioma patients between 2015 and 2020 were obtained, preprocessed and segmented using the 3D Slicer software and the PyRadiomics plugin. In total 145 radiomics features of the T1CE MRI images were computed. The criterion on the basis of which the feature selection was made is whether the number of mitoses per 10 high power field (HPF) is greater than or equal to zero. Our analyses show that machine learning algorithms can be used to make accurate predictions about whether the number of mitoses per 10 HPF is greater than or equal to zero. We obtained our best model using Ridge regression for feature pre-selection, followed by stepwise logistic regression for final model construction. Using independent test data, this model resulted in an AUC (Area under the Curve) of 0.8523, an accuracy of 0.7941, a sensitivity of 0.8182, a specificity of 0.7500 and a Cohen's Kappa of 0.5576. We analyzed the performance of this model as a function of the number of mitoses per 10 HPF. The model performs well for cases with zero mitoses as well as for cases with more than one mitosis per 10 HPF. The worst model performance (accuracy = 0.6250) is obtained for cases with one mitosis per 10 HPF. Our results show that MRI-based radiomics may be a promising approach to predict the mitosis cycles in intracranial meningioma prior to surgery. Specifically, our approach may offer a non-invasive means of detecting the early stages of a malignant process in meningiomas prior to the onset of clinical symptoms.

Artificial Intelligence for Indication of Invasive Assessment of Calcifications in Mammography Screening

PURPOSE

Lesion-related evaluation of the diagnostic performance of an individual artificial intelligence (AI) system to assess mamographically detected and histologically proven calcifications.

MATERIALS AND METHODS

This retrospective study included 634 women of one screening unit (July 2012 - June 2018) who completed the invasive assessment of calcifications. For each leasion, the AI-system calculated a score between 0 and 98. Lesions scored > 0 were classified as AI-positive. The performance of the system was evaluated based on its positive predictive value of invasive assessment (PPV3), the false-negative rate and the true-negative rate.

RESULTS

The PPV3 increased across the categories (readers: 4a: 21.2 %, 4b: 57.7 %, 5: 100 %, overall 30.3 %; AI: 4a: 20.8 %, 4b: 57.8 %, 5: 100 %, overall: 30.7 %). The AI system yielded a false-negative rate of 7.2 % (95 %-CI: 4.3 %: 11.4 %) and a true-negative rate of 9.1 % (95 %-CI: 6.6 %; 11.9 %). These rates were highest in category 4a, 12.5 % and 10.4 % retrospectively. The lowest median AI score was observed for benign lesions (61, interquartile range (IQR): 45-74). Invasive cancers yielded the highest median AI score (81, IQR: 64-86). Median AI scores for ductal carcinoma in situ were: 74 (IQR: 63-84) for low grade, 70 (IQR: 52-79) for intermediate grade and 74 (IQR: 66-83) for high grade.

CONCLUSION

At the lowest threshold, the AI system yielded calcification-related PPV3 values that increased across categories, similar as seen in human evaluation. The strongest loss in AI-based breast cancer detection was observed for invasively assessed calcifications with the lowest suspicion of malignancy, yet with a comparable decrease in the false-positive rate. An AI-score based stratification of malignant lesions could not be determined.

KEY POINTS

· The AI-based PPV3 for calcifications is comparable to human assessment.. · AI showed a lower detection performance of screen-positive and screen-negative lesions in category 4a.. · Histological subgroups could not be discriminated by AI scores..

CITATION FORMAT

· Weigel S, Brehl AK, Heindel W et al. Artificial Intelligence for Indication of Invasive Assessment of Calcifications in Mammography Screening. Fortschr Röntgenstr 2023; 195: 38 - 46.

Multiparametric MRI enables for differentiation of different degrees of malignancy in two murine models of breast cancer

Objective

The objective of this study was to non-invasively differentiate the degree of malignancy in two murine breast cancer models based on identification of distinct tissue characteristics in a metastatic and non-metastatic tumor model using a multiparametric Magnetic Resonance Imaging (MRI) approach.

Methods

The highly metastatic 4T1 breast cancer model was compared to the non-metastatic 67NR model. Imaging was conducted on a 9.4 T small animal MRI. The protocol was used to characterize tumors regarding their structural composition, including heterogeneity, intratumoral edema and hemorrhage, as well as endothelial permeability using apparent diffusion coefficient (ADC), T1/T2 mapping and dynamic contrast-enhanced (DCE) imaging. Mice were assessed on either day three, six or nine, with an i.v. injection of the albumin-binding contrast agent gadofosveset. Ex vivo validation of the results was performed with laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), histology, immunhistochemistry and electron microscopy.

Results

Significant differences in tumor composition were observed over time and between 4T1 and 67NR tumors. 4T1 tumors showed distorted blood vessels with a thin endothelial layer, resulting in a slower increase in signal intensity after injection of the contrast agent. Higher permeability was further reflected in higher K_trans values, with consecutive retention of gadolinium in the tumor interstitium visible in MRI. 67NR tumors exhibited blood vessels with a thicker and more intact endothelial layer, resulting in higher peak enhancement, as well as higher maximum slope and area under the curve, but also a visible wash-out of the contrast agent and thus lower K_trans values. A decreasing accumulation of gadolinium during tumor progression was also visible in both models in LA-ICP-MS. Tissue composition of 4T1 tumors was more heterogeneous, with intratumoral hemorrhage and necrosis and corresponding higher T1 and T2 relaxation times, while 67NR tumors mainly consisted of densely packed tumor cells. Histogram analysis of ADC showed higher values of mean ADC, histogram kurtosis, range and the 90^th percentile (p90), as markers for the heterogenous structural composition of 4T1 tumors. Principal component analysis (PCA) discriminated well between the two tumor models.

Conclusions

Multiparametric MRI as presented in this study enables for the estimation of malignant potential in the two studied tumor models via the assessment of certain tumor features over time.

Simplified Assessment of Lesion Water Uptake for Identification of Patients within 4.5 Hours of Stroke Onset: An Analysis of the MissPerfeCT Study

Background and Purpose Many patients with stroke cannot receive intravenous thrombolysis because the time of symptom onset is unknown. We tested whether a simple method of computed tomography (CT)-based quantification of water uptake in the ischemic tissue can identify patients with stroke onset within 4.5 hours. Methods This retrospective analysis of the MissPerfeCT study (August 2009 to November 2017) includes consecutive patients with known onset of symptoms from seven tertiary stroke centers. We developed a simplified algorithm based on region of interest (ROI) measurements to quantify water uptake of the ischemic lesion and thereby quantify time of symptom onset within and beyond 4.5 hours. Perfusion CT was used to identify ischemic brain tissue, and its density was measured in non-contrast CT and related to the density of the corresponding area of the contralateral hemisphere to quantify lesion water uptake. Results Of 263 patients, 204 (77.6%) had CT within 4.5 hours. Water uptake was significantly lower in patients with stroke onset within (6.7%; 95% confidence interval [CI], 6.0% to 7.4%) compared to beyond 4.5 hours (12.7%; 95% CI, 10.7% to 14.7%). The area under the curve for distinguishing these patient groups according to percentage water uptake was 0.744 with an optimal cut-off value of 9.5%. According to this cut-off the positive predictive value was 88.8%, sensitivity was 73.5%, specificity 67.8%, negative predictive value was 42.6%. Conclusions Ischemic stroke patients with unknown time of symptom onset can be identified as being within a timeframe of 4.5 hours using a ROI-based method to assess water uptake on admission non-contrast head CT.

Respiratory motion correction in F-18-FDG PET/CT impacts lymph node assessment in lung cancer patients

BACKGROUNDS

Elastic motion correction in PET has been shown to increase image quality and quantitative measurements of PET datasets affected by respiratory motion. However, little is known on the impact of respiratory motion correction on clinical image evaluation in oncologic PET. This study evaluated the impact of motion correction on expert readers' lymph node assessment of lung cancer patients.

METHODS

Forty-three patients undergoing F-18-FDG PET/CT for the staging of suspected lung cancer were included. Three different PET reconstructions were investigated: non-motion-corrected ("static"), belt gating-based motion-corrected ("BG-MC") and data-driven gating-based motion-corrected ("DDG-MC"). Assessment was conducted independently by two nuclear medicine specialists blinded to the reconstruction method on a six-point scale [Formula: see text] ranging from "certainly negative" (1) to "certainly positive" (6). Differences in [Formula: see text] between reconstruction methods, accounting for variation caused by readers, were assessed by nonparametric regression analysis of longitudinal data. From [Formula: see text], a dichotomous score for N1, N2, and N3 ("negative," "positive") and a subjective certainty score were derived. SUV and metabolic tumor volumes (MTV) were compared between reconstruction methods.

RESULTS

BG-MC resulted in higher scores for N1 compared to static (p = 0.001), whereas DDG-MC resulted in higher scores for N2 compared to static (p = 0.016). Motion correction resulted in the migration of N1 from tumor free to metastatic on the dichotomized score, consensually for both readers, in 3/43 cases and in 2 cases for N2. SUV was significantly higher for motion-corrected PET, while MTV was significantly lower (all p < 0.003). No significant differences in the certainty scores were noted.

CONCLUSIONS

PET motion correction resulted in significantly higher lymph node assessment scores of expert readers. Significant effects on quantitative PET parameters were seen; however, subjective reader certainty was not improved.

Quantifying Deviations of Brain Structure and Function in Major Depressive Disorder Across Neuroimaging Modalities

IMPORTANCE

Identifying neurobiological differences between patients with major depressive disorder (MDD) and healthy individuals has been a mainstay of clinical neuroscience for decades. However, recent meta-analyses have raised concerns regarding the replicability and clinical relevance of brain alterations in depression.

OBJECTIVE

To quantify the upper bounds of univariate effect sizes, estimated predictive utility, and distributional dissimilarity of healthy individuals and those with depression across structural magnetic resonance imaging (MRI), diffusion-tensor imaging, and functional task-based as well as resting-state MRI, and to compare results with an MDD polygenic risk score (PRS) and environmental variables.

DESIGN, SETTING, AND PARTICIPANTS

This was a cross-sectional, case-control clinical neuroimaging study. Data were part of the Marburg-Münster Affective Disorders Cohort Study. Patients with depression and healthy controls were recruited from primary care and the general population in Münster and Marburg, Germany. Study recruitment was performed from September 11, 2014, to September 26, 2018. The sample comprised patients with acute and chronic MDD as well as healthy controls in the age range of 18 to 65 years. Data were analyzed from October 29, 2020, to April 7, 2022.

MAIN OUTCOMES AND MEASURES

Primary analyses included univariate partial effect size (η2), classification accuracy, and distributional overlapping coefficient for healthy individuals and those with depression across neuroimaging modalities, controlling for age, sex, and additional modality-specific confounding variables. Secondary analyses included patient subgroups for acute or chronic depressive status.

RESULTS

A total of 1809 individuals (861 patients [47.6%] and 948 controls [52.4%]) were included in the analysis (mean [SD] age, 35.6 [13.2] years; 1165 female patients [64.4%]). The upper bound of the effect sizes of the single univariate measures displaying the largest group difference ranged from partial η2 of 0.004 to 0.017, and distributions overlapped between 87% and 95%, with classification accuracies ranging between 54% and 56% across neuroimaging modalities. This pattern remained virtually unchanged when considering either only patients with acute or chronic depression. Differences were comparable with those found for PRS but substantially smaller than for environmental variables.

CONCLUSIONS AND RELEVANCE

Results of this case-control study suggest that even for maximum univariate biological differences, deviations between patients with MDD and healthy controls were remarkably small, single-participant prediction was not possible, and similarity between study groups dominated. Biological psychiatry should facilitate meaningful outcome measures or predictive approaches to increase the potential for a personalization of the clinical practice.

Brain structural correlates of recurrence following the first episode in patients with major depressive disorder

Former prospective studies showed that the occurrence of relapse in Major Depressive Disorder (MDD) is associated with volume loss in the insula, hippocampus and dorsolateral prefrontal cortex (DLPFC). However, these studies were confounded by the patient's lifetime disease history, as the number of previous episodes predict future recurrence. In order to analyze neural correlates of recurrence irrespective of prior disease course, this study prospectively examined changes in brain structure in patients with first-episode depression (FED) over 2 years. N = 63 FED patients and n = 63 healthy controls (HC) underwent structural magnetic resonance imaging at baseline and after 2 years. According to their disease course during the follow-up interval, patients were grouped into n = 21 FED patients with recurrence (FEDrec) during follow-up and n = 42 FED patients with stable remission (FEDrem). Gray matter volume changes were analysed using group by time interaction analyses of covariance for the DLPFC, hippocampus and insula. Significant group by time interactions in the DLPFC and insula emerged. Pairwise comparisons showed that FEDrec had greater volume decline in the DLPFC and insula from baseline to follow-up compared with FEDrem and HC. No group by time interactions in the hippocampus were found. Cross-sectional analyses at baseline and follow-up revealed no differences between groups. This longitudinal study provides evidence for neural alterations in the DLPFC and insula related to a detrimental course in MDD. These effects of recurrence are already detectable at initial stages of MDD and seem to occur without any prior disease history, emphasizing the importance of early interventions preventing depressive recurrence.

Virtual Ontogeny of Cortical Growth Preceding Mental Illness

BACKGROUND

Morphology of the human cerebral cortex differs across psychiatric disorders, with neurobiology and developmental origins mostly undetermined. Deviations in the tangential growth of the cerebral cortex during pre/perinatal periods may be reflected in individual variations in cortical surface area later in life.

METHODS

Interregional profiles of group differences in surface area between cases and controls were generated using T1-weighted magnetic resonance imaging from 27,359 individuals including those with attention-deficit/hyperactivity disorder, autism spectrum disorder, bipolar disorder, major depressive disorder, schizophrenia, and high general psychopathology (through the Child Behavior Checklist). Similarity of interregional profiles of group differences in surface area and prenatal cell-specific gene expression was assessed.

RESULTS

Across the 11 cortical regions, group differences in cortical area for attention-deficit/hyperactivity disorder, schizophrenia, and Child Behavior Checklist were dominant in multimodal association cortices. The same interregional profiles were also associated with interregional profiles of (prenatal) gene expression specific to proliferative cells, namely radial glia and intermediate progenitor cells (greater expression, larger difference), as well as differentiated cells, namely excitatory neurons and endothelial and mural cells (greater expression, smaller difference). Finally, these cell types were implicated in known pre/perinatal risk factors for psychosis. Genes coexpressed with radial glia were enriched with genes implicated in congenital abnormalities, birth weight, hypoxia, and starvation. Genes coexpressed with endothelial and mural genes were enriched with genes associated with maternal hypertension and preterm birth.

CONCLUSIONS

Our findings support a neurodevelopmental model of vulnerability to mental illness whereby prenatal risk factors acting through cell-specific processes lead to deviations from typical brain development during pregnancy.

Testing the applicability and performance of Auto ML for potential applications in diagnostic neuroradiology

To investigate the applicability and performance of automated machine learning (AutoML) for potential applications in diagnostic neuroradiology. In the medical sector, there is a rapidly growing demand for machine learning methods, but only a limited number of corresponding experts. The comparatively simple handling of AutoML should enable even non-experts to develop adequate machine learning models with manageable effort. We aim to investigate the feasibility as well as the advantages and disadvantages of developing AutoML models compared to developing conventional machine learning models. We discuss the results in relation to a concrete example of a medical prediction application. In this retrospective IRB-approved study, a cohort of 107 patients who underwent gross total meningioma resection and a second cohort of 31 patients who underwent subtotal resection were included. Image segmentation of the contrast enhancing parts of the tumor was performed semi-automatically using the open-source software platform 3D Slicer. A total of 107 radiomic features were extracted by hand-delineated regions of interest from the pre-treatment MRI images of each patient. Within the AutoML approach, 20 different machine learning algorithms were trained and tested simultaneously. For comparison, a neural network and different conventional machine learning algorithms were trained and tested. With respect to the exemplary medical prediction application used in this study to evaluate the performance of Auto ML, namely the pre-treatment prediction of the achievable resection status of meningioma, AutoML achieved remarkable performance nearly equivalent to that of a feed-forward neural network with a single hidden layer. However, in the clinical case study considered here, logistic regression outperformed the AutoML algorithm. Using independent test data, we observed the following classification results (AutoML/neural network/logistic regression): mean area under the curve = 0.849/0.879/0.900, mean accuracy = 0.821/0.839/0.881, mean kappa = 0.465/0.491/0.644, mean sensitivity = 0.578/0.577/0.692 and mean specificity = 0.891/0.914/0.936. The results obtained with AutoML are therefore very promising. However, the AutoML models in our study did not yet show the corresponding performance of the best models obtained with conventional machine learning methods. While AutoML may facilitate and simplify the task of training and testing machine learning algorithms as applied in the field of neuroradiology and medical imaging, a considerable amount of expert knowledge may still be needed to develop models with the highest possible discriminatory power for diagnostic neuroradiology.

Forensische Altersdiagnostik mittels Niederfeld-Magnetresonanztomographie

Hintergrund

Ein rechtlich zulässiges medizinisches Altersfeststellungsverfahren mit möglichst schonenden und zuverlässigen Methoden bei unbegleiteten und fraglich minderjährigen Ausländern ist erforderlich, um den tatsächlich Minderjährigen den für sie gesetzlich vorgesehenen besonderen Schutz zukommen zu lassen.

Ziel der Arbeit

Studiendesign und erste Ergebnisse einer Studie zur forensischen Altersdiagnostik mittels Niederfeld-Magnetresonanztomographie werden vorgestellt.

Material und Methoden

Geplant ist die prospektive Untersuchung von 650 Studienteilnehmenden, gleichmäßig verteilt innerhalb der Altersgruppe 12 bis 24 Jahre zu je 25 Personen je Geschlecht. Es sollen Entwicklungsstadien von distalem Radius, distaler Ulna, distalem Femur und proximaler Tibia mithilfe eines 0,31-Tesla(T)-Niederfeld-MRT-Scanners erfasst und mittels Anwendung der Vieth-Klassifikation ausgewertet werden.

Ergebnisse

Nach Auswertung der Untersuchungen der bisher 487 untersuchten Studienteilnehmenden ist ersichtlich, dass die gewählten Sequenzen die für die Vieth-Klassifikation relevanten Charakteristika der Ossifikation der untersuchten Epiphysen erfassen. Die Untersuchungen erweisen sich überdies als einfach in der Durchführung.

Schlussfolgerung

Untersuchungen zur forensischen Altersschätzung am Lebenden unter Anwendung der Vieth-Klassifikation sind mit Niederfeld-MRT-Scannern bei einer Feldstärke von 0,31 T durchführbar. Es bleibt zu klären, ob die mit einem 0,31-T-Niederfeld-MRT-Scanner nachweisbaren Ossifikationsmerkmale am Handgelenk und am Knie einen zweifelsfreien Volljährigkeitsnachweis erlauben.

Radiomics for pseudoprogression prediction in high grade gliomas: added value of MR contrast agent

Objective

Material & methods

Results

Conclusion

•

Radiomics allows for prediction of pseudoprogression in high-grade gliomas.

•

Use of contrast media boosts the performance of the Radiomics prediction model.

Association of disease course and brain structural alterations in major depressive disorder

INTRODUCTION

The investigation of disease course-associated brain structural alterations in Major Depressive Disorder (MDD) have resulted in heterogeneous findings, possibly due to low reliability of single clinical variables used for defining disease course. The present study employed a principal component analysis (PCA) on multiple clinical variables to investigate effects of cumulative lifetime illness burden on brain structure in a large and heterogeneous sample of MDD patients.

METHODS

Gray matter volumes (GMV) was estimated in n = 681 MDD patients (mean age: 35.87 years; SD = 12.89; 66.6% female) using voxel-based-morphometry. Five clinical variables were included in a PCA to obtain components reflecting disease course to associate resulting components with GMVs.

RESULTS

The PCA yielded two main components: Hospitalization reflected by patients' frequency and duration of inpatient treatment and Duration of Illness reflected by the frequency and duration of depressive episodes. Hospitalization revealed negative associations with bilateral dorsolateral prefrontal cortex (DLPFC) and left insula volumes. Duration of Illness showed significant negative associations with left hippocampus and right DLPFC volumes. Results in the DLPFC and hippocampus remained significant after additional control for depressive symptom severity, psychopharmacotherapy, psychiatric comorbidities, and remission status.

CONCLUSION

This study shows that a more severe and chronic lifetime disease course in MDD is associated with reduced volume in brain regions relevant for executive and cognitive functions and emotion regulation in a large sample of patients representing the broad heterogeneity of MDD disease course. These findings were only partly influenced by other clinical characteristics (e.g., remission status, psychopharmacological treatment).

Genetic variants associated with longitudinal changes in brain structure across the lifespan

Human brain structure changes throughout the lifespan. Altered brain growth or rates of decline are implicated in a vast range of psychiatric, developmental and neurodegenerative diseases. In this study, we identified common genetic variants that affect rates of brain growth or atrophy in what is, to our knowledge, the first genome-wide association meta-analysis of changes in brain morphology across the lifespan. Longitudinal magnetic resonance imaging data from 15,640 individuals were used to compute rates of change for 15 brain structures. The most robustly identified genes GPR139, DACH1 and APOE are associated with metabolic processes. We demonstrate global genetic overlap with depression, schizophrenia, cognitive functioning, insomnia, height, body mass index and smoking. Gene set findings implicate both early brain development and neurodegenerative processes in the rates of brain changes. Identifying variants involved in structural brain changes may help to determine biological pathways underlying optimal and dysfunctional brain development and aging.

Longitudinal Structural Brain Changes in Bipolar Disorder: A Multicenter Neuroimaging Study of 1232 Individuals by the ENIGMA Bipolar Disorder Working Group

BACKGROUND

Bipolar disorder (BD) is associated with cortical and subcortical structural brain abnormalities. It is unclear whether such alterations progressively change over time, and how this is related to the number of mood episodes. To address this question, we analyzed a large and diverse international sample with longitudinal magnetic resonance imaging (MRI) and clinical data to examine structural brain changes over time in BD.

METHODS

Longitudinal structural MRI and clinical data from the ENIGMA (Enhancing Neuro Imaging Genetics through Meta Analysis) BD Working Group, including 307 patients with BD and 925 healthy control subjects, were collected from 14 sites worldwide. Male and female participants, aged 40 ± 17 years, underwent MRI at 2 time points. Cortical thickness, surface area, and subcortical volumes were estimated using FreeSurfer. Annualized change rates for each imaging phenotype were compared between patients with BD and healthy control subjects. Within patients, we related brain change rates to the number of mood episodes between time points and tested for effects of demographic and clinical variables.

RESULTS

Compared with healthy control subjects, patients with BD showed faster enlargement of ventricular volumes and slower thinning of the fusiform and parahippocampal cortex (0.18 <d < 0.22). More (hypo)manic episodes were associated with faster cortical thinning, primarily in the prefrontal cortex.

CONCLUSIONS

In the hitherto largest longitudinal MRI study on BD, we did not detect accelerated cortical thinning but noted faster ventricular enlargements in BD. However, abnormal frontocortical thinning was observed in association with frequent manic episodes. Our study yields insights into disease progression in BD and highlights the importance of mania prevention in BD treatment.

Dimensions of Formal Thought Disorder and Their Relation to Gray- and White Matter Brain Structure in Affective and Psychotic Disorders

Factorial dimensions and neurobiological underpinnings of formal thought disorders (FTD) have been extensively investigated in schizophrenia spectrum disorders (SSD). However, FTD are also highly prevalent in other disorders. Still, there is a lack of knowledge about transdiagnostic, structural brain correlates of FTD. In N = 1071 patients suffering from DSM-IV major depressive disorder, bipolar disorder, or SSD, we calculated a psychopathological factor model of FTD based on the SAPS and SANS scales. We tested the association of FTD dimensions with 3 T MRI measured gray matter volume (GMV) and white matter fractional anisotropy (FA) using regression and interaction models in SPM12. We performed post hoc confirmatory analyses in diagnostically equally distributed, age- and sex-matched sub-samples to test whether results were driven by diagnostic categories. Cross-validation (explorative and confirmatory) factor analyses revealed three psychopathological FTD factors: disorganization, emptiness, and incoherence. Disorganization was negatively correlated with a GMV cluster comprising parts of the middle occipital and angular gyri and positively with FA in the right posterior cingulum bundle and inferior longitudinal fascicle. Emptiness was negatively associated with left hippocampus and thalamus GMV. Incoherence was negatively associated with FA in bilateral anterior thalamic radiation, and positively with the hippocampal part of the right cingulum bundle. None of the gray or white matter associations interacted with diagnosis. Our results provide a refined mapping of cross-disorder FTD phenotype dimensions. For the first time, we demonstrated that their neuroanatomical signatures are associated with language-related gray and white matter structures independent of diagnosis.

CT Hypoperfusion-Hypodensity Mismatch to Identify Patients With Acute Ischemic Stroke Within 4.5 Hours of Symptom Onset

BACKGROUND AND OBJECTIVES

To test the hypothesis that CT hypoperfusion-hypodensity mismatch identifies patients with ischemic stroke within 4.5 hours of symptom onset.

METHODS

We therefore performed the Retrospective Multicenter Hypoperfusion-Hypodensity Mismatch for The identification of Patients With Stroke Within 4.5 Hours study of patients with acute ischemic stroke and known time of symptom onset. The predictive values of hypoperfusion-hypodensity mismatch for the identification of patients with symptom onset within 4.5 hours were the main outcome measure.

RESULTS

Of 666 patients, 548 (82.3%) had multimodal CT within 4.5 hours and 118 (17.7%) beyond 4.5 hours. Hypoperfusion-hypodensity mismatch was visible in 516 (94.2%) patients with symptom onset within and in 30 (25.4%) patients beyond 4.5 hours. CT hypoperfusion-hypodensity mismatch identified patients within 4.5 hours of stroke onset with 94.2% (95% confidence interval [CI] 91.9%-95.8%) sensitivity, 74.6% (95% CI 66.0%-81.6%) specificity, 94.5% (95% CI 92.3%-96.1%) positive predictive value, and 73.3% (95% CI 64.8%-80.4%) negative predictive value. Interobserver agreement for hypoperfusion-hypodensity mismatch was substantial (κ = 0.61, 95% CI 0.53-0.69).

DISCUSSION

Patients with acute ischemic stroke with absence of a hypodensity on native CT (NCCT) within the hypoperfused core lesion on perfusion CT (hypoperfusion-hypodensity mismatch) are likely to be within the time window of thrombolysis. Applying this method may guide the decision to use thrombolysis in patients with unknown time of stroke onset.

TRIAL REGISTRATION INFORMATION

ClinicalTrials.gov Identifier: NCT04277728.

CLASSIFICATION OF EVIDENCE

This study provides Class III evidence that CT hypoperfusion-hypodensity mismatch identifies patients with stroke within 4.5 hours of onset.

Resolving immune cells with patrolling behaviour by magnetic resonance time-lapse single cell tracking

BACKGROUND

Immune cells show distinct motion patterns that change upon inflammatory stimuli. Monocytes patrol the vasculature to screen for pathogens, thereby exerting an early task of innate immunity. Here, we aimed to non-invasively analyse single patrolling monocyte behaviour upon inflammatory stimuli.

METHODS

We used time-lapse Magnetic Resonance Imaging (MRI) of the murine brain to dynamically track single patrolling monocytes within the circulation distant to the actual site of inflammation in different inflammatory conditions, ranging from a subcutaneous pellet model to severe peritonitis and bacteraemia.

FINDINGS

Single patrolling immune cells with a velocity of <1 µm/s could be detected and followed dynamically using time-lapse MRI. We show, that due to local and systemic stimuli the slowly patrolling behaviour of monocytes is altered systemically and differs with type, duration and strength of the underlying stimulus.

INTERPRETATION

Using time-lapse MRI, it is now possible to investigate the behaviour of single circulating monocytes over the course of the systemic immune response. Monocyte patrolling behaviour is altered systemically even before the onset of clinical symptoms distant to and depending on the underlying inflammatory stimulus.

FUNDING

This study was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - CRC 1009 - 194468054 to AZ, CF and - CRC 1450 - 431460824 to MM, SN, HB, AZ, CF, the Joachim Herz Foundation (Add-on Fellowship for Interdisciplinary Life Sciences to MM), the Interdisciplinary Centre for Clinical Research (IZKF, core unit PIX) and the Medical Faculty of the University of Muenster (MEDK fellowship to FF and IF).

Simultaneous FAPI PET/MRI Targeting the Fibroblast-Activation Protein for Breast Cancer

Background Integrated PET/MRI is a promising modality for breast assessment. The most frequently used tracer, fluorine 18 (¹⁸F) fluorodeoxyglucose (FDG), is applied for whole-body staging in advanced breast cancer but has limited accuracy in evaluating primary breast lesions. The fibroblast-activation protein (FAP) is abundantly expressed in invasive breast cancer. FAP-directed PET tracers have recently become available, but results in primary breast tumors remain lacking. Purpose To evaluate the use of FAP inhibitor (FAPI) breast PET/MRI in assessing breast lesions and of FAPI whole-body scanning for lymph node (LN) and distant staging using the ligand gallium 68 (⁶⁸Ga)-FAPI-46. Materials and Methods In women with histologically confirmed invasive breast cancer, all primary ⁶⁸Ga-FAPI-46 breast and whole-body PET/MRI and PET/CT examinations conducted at the authors' center between October 2019 and December 2020 were retrospectively analyzed. MRI lesion characteristics and standardized uptake values (SUVs) were quantified with dedicated software. Mann-Whitney U tests were used to compare tumor SUVs across different tumor types. The Pearson correlation coefficient was calculated between SUV and measures of MRI morphologic characteristics. Results Nineteen women (mean age, 49 years ± 9 [standard deviation]) were evaluated-18 to complement initial staging and one for restaging after therapy for distant metastases. Strong tracer accumulation was observed in all 18 untreated primary breast malignancies (mean maximum SUV [SUV_max] = 13.9 [range, 7.9-29.9]; median lesion diameter = 26 mm [range, 9-155 mm]), resulting in clear tumor delineation across different gradings, receptors, and histologic types. All preoperatively verified LN metastases in 13 women showed strong tracer accumulation (mean SUV_max= 12.2 [range, 3.3-22.4]; mean diameter = 21 mm [range, 14-35 mm]). Tracer uptake established or supported extra-axillary LN involvement in seven women and affected therapy decisions in three women. Conclusion This retrospective analysis indicates use of ⁶⁸Ga fibroblast-activation protein inhibitor tracers for breast cancer diagnosis and staging. © RSNA, 2021 Online supplemental material is available for this article. See also the editorial by Mankoff and Sellmyer in this issue.

Using deep learning to assist readers during the arbitration process: a lesion-based retrospective evaluation of breast cancer screening performance

Objectives

To evaluate if artificial intelligence (AI) can discriminate recalled benign from recalled malignant mammographic screening abnormalities to improve screening performance.

Methods

A total of 2257 full-field digital mammography screening examinations, obtained 2011–2013, of women aged 50–69 years which were recalled for further assessment of 295 malignant out of 305 truly malignant lesions and 2289 benign lesions after independent double-reading with arbitration, were included in this retrospective study. A deep learning AI system was used to obtain a score (0–95) for each recalled lesion, representing the likelihood of breast cancer. The sensitivity on the lesion level and the proportion of women without false-positive ratings (non-FPR) resulting under AI were estimated as a function of the classification cutoff and compared to that of human readers.

Results

Using a cutoff of 1, AI decreased the proportion of women with false-positives from 89.9 to 62.0%, non-FPR 11.1% vs. 38.0% (difference 26.9%, 95% confidence interval 25.1–28.8%; p < .001), preventing 30.1% of reader-induced false-positive recalls, while reducing sensitivity from 96.7 to 91.1% (5.6%, 3.1–8.0%) as compared to human reading. The positive predictive value of recall (PPV-1) increased from 12.8 to 16.5% (3.7%, 3.5–4.0%). In women with mass-related lesions (n = 900), the non-FPR was 14.2% for humans vs. 36.7% for AI (22.4%, 19.8–25.3%) at a sensitivity of 98.5% vs. 97.1% (1.5%, 0–3.5%).

Conclusion

The application of AI during consensus conference might especially help readers to reduce false-positive recalls of masses at the expense of a small sensitivity reduction. Prospective studies are needed to further evaluate the screening benefit of AI in practice.

Key Points

• Integrating the use of artificial intelligence in the arbitration process reduces benign recalls and increases the positive predictive value of recall at the expense of some sensitivity loss.

• Application of the artificial intelligence system to aid the decision to recall a woman seems particularly beneficial for masses, where the system reaches comparable sensitivity to that of the readers, but with considerably reduced false-positives.

• About one-fourth of all recalled malignant lesions are not automatically marked by the system such that their evaluation (AI score) must be retrieved manually by the reader. A thorough reading of screening mammograms by readers to identify suspicious lesions therefore remains mandatory.

Social support and hippocampal volume are negatively associated in adults with previous experience of childhood maltreatment

BACKGROUND

Childhood maltreatment has been associated with reduced hippocampal volume in healthy individuals, whereas social support, a protective factor, has been positively associated with hippocampal volumes. In this study, we investigated how social support is associated with hippocampal volume in healthy people with previous experience of childhood maltreatment.

METHODS

We separated a sample of 446 healthy participants into 2 groups using the Childhood Trauma Questionnaire: 265 people without maltreatment and 181 people with maltreatment. We measured perceived social support using a short version of the Social Support Questionnaire. We examined hippocampal volume using automated segmentation (Freesurfer). We conducted a social support × group analysis of covariance on hippocampal volumes controlling for age, sex, total intracranial volume, site and verbal intelligence.

RESULTS

Our analysis revealed significantly lower left hippocampal volume in people with maltreatment (left F1,432 = 5.686, p = 0.018; right F1,433 = 3.371, p = 0.07), but no main effect of social support emerged. However, we did find a significant social support × group interaction for left hippocampal volume (left F1,432 = 5.712, p = 0.017; right F1,433 = 3.480, p = 0.06). In people without maltreatment, we observed a trend toward a positive association between social support and hippocampal volume. In contrast, social support was negatively associated with hippocampal volume in people with maltreatment.

LIMITATIONS

Because of the correlative nature of our study, we could not infer causal relationships between social support, maltreatment and hippocampal volume.

CONCLUSION

Our results point to a complex dynamic between environmental risk, protective factors and brain structure - in line with previous evidence - suggesting a detrimental effect of maltreatment on hippocampal development.

Dose escalation and expansion phase I studies with the tumour-targeting antibody-tumour necrosis factor fusion protein L19TNF plus doxorubicin in patients with advanced tumours, including sarcomas

BACKGROUND

L19TNF is a recombinant fusion protein composed of a human antibody fragment and human tumour necrosis factor. L19TNF targets the EDB domain of oncofetal fibronectin highly expressed in tumour vasculature and induces tumour remission in mouse tumours. We summarise two phase I trials testing a combination of L19TNF with doxorubicin in patients with solid tumours, particularly soft tissue sarcomas (STS).

PATIENTS AND METHODS

The first study, an open-label, dose-escalation and expansion phase I study of L19TNF plus doxorubicin, enrolled 27 patients. Three cohorts (10.4-17 μg/kg L19TNF) of patients received L19TNF intravenously at days 1, 3, and 5 and doxorubicin (75 mg/m², then 60 mg/m²) on day 1 every 3 weeks. The expansion cohort enrolled patients with STS. The second study tried to re-escalate the doxorubicin dose to 75 mg/m² with 13 μg/kg L19TNF. Among primary objectives was the establishment of a recommended dose (RD).

RESULTS

The combination was safely applicable. Dose-limiting toxicity occurred either at 17 μg/kg L19TNF or at 75 mg/m² doxorubicin. RD is 13 μg/kg L19TNF plus 60 mg/m² doxorubicin. In 15 STS patients of the extension cohort evaluable for efficacy, antitumour activity was observed with complete remission in 1, partial remission in 1 and minor tumour shrinkage in 7 patients. The median overall survival for this heavily pretreated cohort was 14.9 months.

CONCLUSION

L19TNF can be safely applied in combination with doxorubicin and induces encouraging tumour remissions in patients with soft tissue sarcomas.

[Systematic and quality-assured early diagnosis of sporadic breast cancer : Update on screening effects and scientific studies]

BACKGROUND

A quality-assured mammography screening programme has been available since 2009, nationwide, to all women in Germany between the ages of 50 and 69. The programme is based on the European Guidelines. In this review article the authors summarize the current status of scientific assessments of this national early detection programme for breast cancer and provide an outlook regarding ongoing studies on effectiveness tests and further development.

RESULTS

We expect a decline in mortality rates relating to breast cancer as a result of successfully bringing diagnoses forward and a decrease in advanced breast cancer after a repeated screening. The extent will be shown in the current ZEBra study on mortality evaluation.

CONCLUSION

Potential for a further increase in the effectiveness of the systematic early detection of breast cancer can be identified in four areas: (1) More women should take advantage of the early detection opportunities offered by the medical insurance funds; so far, on average, only about 50% of the women between 50 and 69 who are entitled to a screening examination actually take part in the programme. (2) Entitlement to take part in the programme should be extended to women over 70. (3) The further development of digital mammography towards digital breast tomosynthesis promises to reduce the number of false positive recalls while at the same time increasing sensitivity. (4) There should be scientific studies relating to an extension of screening strategies for the small number of women in the entitlement range who have extremly dense breasts.

Portal and hepatic vein embolization prior to major hepatectomy

PURPOSE

 To analyze safety and effectiveness of simultaneous portal and hepatic vein embolization (PHVE) or sequential hepatic vein embolization (HVE) compared to portal vein embolization (PVE) for future remnant liver (FRL) hypertrophy prior to major hepatic surgery.

METHODS

 Patients undergoing PVE, PHVE or HVE at our tertiary care center between 2018 and 2020 were retrospectively included. FRLV, standardized FRLV (sFRLV) and sFRLV growth rate per day were assessed via volumetry, as well as laboratory parameters.

RESULTS

 36 patients (f = 15, m = 21; median 64.5 y) were included, 16 patients received PHVE and 20 patients PVE, of which 4 received sequential HVE. Significant increase of FRLV was achieved with both PVE and PHVE compared to baseline (p < 0.0001). sFRLV growth rate did not significantly differ following PHVE (2.2 ± 1.2 %/d) or PVE (2.2 ± 1.7 %/d, p = 0.94). Left portal vein thrombosis (LPVT) was observed after PHVE in 6 patients and in 1 patient after PVE. Sequential HVE showed a considerably high growth rate of 1.42 ± 0.45 %/d after PVE.

CONCLUSION

 PHVE effectively induces FRL hypertrophy but yields comparable sFRLV to PVE. Sequential HVE further induces hypertrophy after insufficient growth due to PVE. Considering a potentially higher rate of LPVT after PHVE, PVE might be preferred in patients with moderate baseline sFRLV, with optional sequential HVE in non-sufficient responders.

Feasibility, Safety, and Outcome of Endovascular Recanalization in Childhood Stroke: The Save ChildS Study

Importance

Randomized clinical trials have shown the efficacy of thrombectomy of large intracranial vessel occlusions in adults; however, any association of therapy with clinical outcomes in children is unknown.

Objective

To evaluate the use of endovascular recanalization in pediatric patients with arterial ischemic stroke.

Design, Setting, and Participants

This retrospective, multicenter cohort study, conducted from January 1, 2000, to December 31, 2018, analyzed the databases from 27 stroke centers in Europe and the United States. Included were all pediatric patients (<18 years) with ischemic stroke who underwent endovascular recanalization. Median follow-up time was 16 months.

Exposures

Endovascular recanalization.

Main Outcomes and Measures

The decrease of the Pediatric National Institutes of Health Stroke Scale (PedNIHSS) score from admission to day 7 was the primary outcome (score range: 0 [no deficit] to 34 [maximum deficit]). Secondary clinical outcomes included the modified Rankin scale (mRS) (score range: 0 [no deficit] to 6 [death]) at 6 and 24 months and rate of complications.

Results

Seventy-three children from 27 participating stroke centers were included. Median age was 11.3 years (interquartile range [IQR], 7.0-15.0); 37 patients (51%) were boys, and 36 patients (49%) were girls. Sixty-three children (86%) received treatment for anterior circulation occlusion and 10 patients (14%) received treatment for posterior circulation occlusion; 16 patients (22%) received concomitant intravenous thrombolysis. Neurologic outcome improved from a median PedNIHSS score of 14.0 (IQR, 9.2-20.0) at admission to 4.0 (IQR, 2.0-7.3) at day 7. Median mRS score was 1.0 (IQR, 0-1.6) at 6 months and 1.0 (IQR, 0-1.0) at 24 months. One patient (1%) developed a postinterventional bleeding complication and 4 patients (5%) developed transient peri-interventional vasospasm. The proportion of symptomatic intracerebral hemorrhage events in the HERMES meta-analysis of trials with adults was 2.79 (95% CI, 0.42-6.66) and in Save ChildS was 1.37 (95% CI, 0.03-7.40).

Conclusions and Relevance

The results of this study suggest that the safety profile of thrombectomy in childhood stroke does not differ from the safety profile in randomized clinical trials for adults; most of the treated children had favorable neurologic outcomes. This study may support clinicians' practice of off-label thrombectomy in childhood stroke in the absence of high-level evidence.