Percutaneous Disc Biopsy versus Bone Biopsy for the Identification of Infectious Agents in Osteomyelitis/Discitis.

PURPOSE: To determine whether sampling of the disc or bone is more likely to yield positive tissue culture results in patients with vertebral discitis and osteomyelitis (VDO). MATERIALS AND METHODS: Retrospective review was performed of consecutive patients who underwent vertebral disc or vertebral body biopsy at a single institution between February 2019 and May 2023. Inclusion criteria were age ≥18 years, presumed VDO on spinal magnetic resonance (MR) imaging, absence of paraspinal abscess, and technically successful percutaneous biopsy with fluoroscopic guidance. The primary outcome was a positive biopsy culture result, and secondary outcomes included complications such as nerve injury and segmental artery injury. RESULTS: Sixty-six patients met the inclusion criteria; 36 patients (55%) underwent disc biopsy, and 30 patients (45%) underwent bone biopsy. Six patients required a repeat biopsy for an initially negative culture result. No significant demographic, laboratory, antibiotic administration, or pain medication use differences were observed between the 2 groups. Patients who underwent bone biopsy were more likely to have a history of intravenous drug use (26.7%) compared with patients who underwent disc biopsy (5.5%; P = .017). Positive tissue culture results were observed in 41% of patients who underwent disc biopsy and 15% of patients who underwent bone biopsy (P = .016). No vessel or nerve injuries were detected after procedure in either group. CONCLUSIONS: Percutaneous disc biopsy is more likely to yield a positive tissue culture result than vertebral body biopsy in patients with VDO.

Myeloperoxidase Molecular MRI Reveals Synergistic Combination Therapy in Murine Experimental Autoimmune Neuroinflammation.

Background Despite advances in immunomodulatory agents, most current therapies for multiple sclerosis target lymphocytes or lymphocytic function. However, therapy response may be less than optimal due to demyelination and axonal damage caused by myeloid cells. Purpose To determine if myeloperoxidase (MPO) molecular MRI can evaluate whether combination therapy targeting both lymphoid and myeloid inflammation can improve autoimmune neuroinflammation compared with either drug alone, even at suboptimal doses. Materials and Methods Four groups of 94 female mice (8-10 weeks old) were induced with experimental autoimmune encephalomyelitis (EAE) from August 2, 2016, to March 30, 2018, and divided into saline control (n = 22), 4-aminobenzoic acid hydrazide (ABAH) therapy group (n = 19), glatiramer acetate (GA) therapy group (n = 22), and combination therapy group (n = 31). Mice were administered suboptimal doses of ABAH, an irreversible inhibitor of MPO; GA, a first-line multiple sclerosis drug; both ABAH and GA; or saline (control). Mice were imaged with bis-5-hydroxytryptamide-diethylenetriaminepentaacetate gadolinium (hereafter, MPO-Gd) MRI. One-way analysis of variance, two-way analysis of variance, Kurskal-Wallis, and log-rank tests were used. P < .05 was considered to indicate statistical significance. Results The combination-treated group showed delayed disease onset (day 11.3 vs day 9.8 for ABAH, day 10.4 for GA, day 9.9 for control; P < .05) and reduced disease severity (clinical score during the acute exacerbation period of 1.8 vs 3.8 for ABAH, 3.1 for GA, 3.9 for control; P < .05). The combination-treated group demonstrated fewer MPO-positive lesions (30.2 vs 73.7 for ABAH, 64.8 for GA, 67.2 for control; P < .05), smaller MPO-positive lesion volume (16.7 mm3 vs 65.2 mm3 for ABAH, 69.9 mm3 for GA, 66.0 mm3 for control; P < .05), and lower intensity of MPO-Gd lesion activation ratio (0.7 vs 1.9 for ABAH, 3.2 for GA, 2.3 for control; P < .05). Reduced disease severity in the combination group was confirmed at histopathologic analysis, where MPO expression (1779 vs 2673 for ABAH, 2898 for GA; P < .05) and demyelination (5.3% vs 9.0% for ABAH, 10.6% for GA; P < .05) were ameliorated. Conclusion Myeloperoxidase molecular MRI can track the treatment response from immunomodulatory drugs even if the drug does not directly target myeloperoxidase, and establishes that combination therapy targeting both myeloid and lymphocytic inflammation is effective for murine autoimmune neuroinflammation, even at suboptimal doses. © RSNA, 2019 Online supplemental material is available for this article. See also the editorial by Walczak in this issue.

Spinal Cord Inflammation: Molecular Imaging after Thoracic Aortic Ischemia Reperfusion Injury.

Purpose To evaluate whether noninvasive molecular imaging technologies targeting myeloperoxidase (MPO) can reveal early inflammation associated with spinal cord injury after thoracic aortic ischemia-reperfusion (TAR) in mice. Materials and Methods The study was approved by the institutional animal care and use committee. C57BL6 mice that were 8-10 weeks old underwent TAR (n = 55) or sham (n = 26) surgery. Magnetic resonance (MR) imaging (n = 6) or single photon emission computed tomography (SPECT)/computed tomography (CT) (n = 15) studies targeting MPO activity were performed after intravenous injection of MPO sensors (bis-5-hydroxytryptamide-tetraazacyclododecane [HT]-diethyneletriaminepentaacetic acid [DTPA]-gadolinium or indium 111-bis-5-HT-DTPA, respectively). Immunohistochemistry and flow cytometry were used to identify myeloid cells and neuronal loss. Proinflammatory cytokines, keratinocyte chemoattractant (KC), and interleukin 6 (IL-6) were measured with enzyme-linked immunosorbent assay. Statistical analyses were performed by using nonparametric tests and the Pearson correlation coefficient. P < .05 was considered to indicate a significant difference. Results Myeloid cells infiltrated into the injured cord at 6 and 24 hours after TAR. MR imaging confirmed the presence of ischemic lesions associated with mild MPO-mediated enhancement in the thoracolumbar spine at 24 hours compared with the sham procedure. SPECT/CT imaging of MPO activity showed marked MPO-sensor retention at 6 hours (P = .003) that continued to increase at 24 hours after TAR (P = .0001). The number of motor neurons decreased substantially at 24 hours after TAR (P < .01), which correlated inversely with in vivo inflammatory changes detected at molecular imaging (r = 0.64, P = .0099). MPO was primarily secreted by neutrophils, followed by lymphocyte antigen 6 complexhigh monocytes and/or macrophages. There were corresponding increased levels of proinflammatory cytokines KC (P = .0001) and IL-6 (P = .0001) that mirrored changes in MPO activity. Conclusion MPO is a suitable imaging biomarker for identifying and tracking inflammatory damage in the spinal cord after TAR in a mouse model. © RSNA, 2016 Online supplemental material is available for this article.

Molecular MR Imaging of Myeloperoxidase Distinguishes Steatosis from Steatohepatitis in Nonalcoholic Fatty Liver Disease.

Purpose To test whether MPO-Gd, an activatable molecular magnetic resonance (MR) imaging agent specific for myeloperoxidase (MPO) activity, could detect MPO activity in nonalcoholic steatohepatitis (NASH) mouse models and human liver biopsy samples. Materials and Methods In this study, 20 leptin receptor-deficient and three MPO knockout mice were injected with endotoxin (lipopolysaccharide) or fed a methionine and choline-deficient (MCD) diet to induce experimental NASH and underwent MR imaging with MPO-Gd. Saline-injected and control diet-fed leptin receptor-deficient mice were used as respective controls. MPO protein and activity measurements and histologic analyses were performed. Eleven human liver biopsy samples underwent MPO-Gd-enhanced MR imaging ex vivo and subsequent histologic evaluation. Results were compared with Student t test or Mann-Whitney U test. Results With endotoxin, a significantly increased contrast-to-noise ratio (CNR) was found compared with sham (mean CNR, 1.81 [95% confidence interval {CI}: 1.53, 2.10] vs 1.02 [95% CI: 0.89, 1.14]; P = .03) at MPO-Gd MR imaging. In the diet-induced NASH model, an increased CNR was also found compared with sham mice (mean CNR, 1.33 [95% CI: 1.27, 1.40] vs 0.98 [95% CI: 0.83, 1.12]; P = .008). Conversely, CNR remained at baseline in NASH mice imaged with gadopentetate dimeglumine and in MPO knockout NASH mice with MPO-Gd, which proves specificity of MPO-Gd. Ex vivo molecular MR imaging of liver biopsy samples from NASH and control patients confirmed results from animal studies (mean CNR for NASH vs control patients, 2.61 [95% CI: 1.48, 3.74] vs 1.29 [95% CI: 1.06, 1.52]; P = .004). Conclusion MPO-Gd showed elevated MPO activity in NAFLD mouse models and human liver biopsy samples. © RSNA, 2017 Online supplemental material is available for this article. An earlier incorrect version of this article appeared online. This article was corrected on April 6, 2017.

Surface biotinylation of cytotoxic T lymphocytes for in vivo tracking of tumor immunotherapy in murine models.

Currently, there is no stable and flexible method to label and track cytotoxic T lymphocytes (CTLs) in vivo in CTL immunotherapy. We aimed to evaluate whether the sulfo-hydroxysuccinimide (NHS)-biotin-streptavidin (SA) platform could chemically modify the cell surface of CTLs for in vivo tracking. CD8+ T lymphocytes were labeled with sulfo-NHS-biotin under different conditions and then incubated with SA-Alexa647. Labeling efficiency was proportional to sulfo-NHS-biotin concentration. CD8+ T lymphocytes could be labeled with higher efficiency with sulfo-NHS-biotin in DPBS than in RPMI (P < 0.05). Incubation temperature was not a key factor. CTLs maintained sufficient labeling for at least 72 h (P < 0.05), without altering cell viability. After co-culturing labeled CTLs with mouse glioma stem cells (GSCs) engineered to present biotin on their surface, targeting CTLs could specifically target biotin-presenting GSCs and inhibited cell proliferation (P < 0.01) and tumor spheres formation. In a biotin-presenting GSC brain tumor model, targeting CTLs could be detected in biotin-presenting gliomas in mouse brains but not in the non-tumor-bearing contralateral hemispheres (P < 0.05). In vivo fluorescent molecular tomography imaging in a subcutaneous U87 mouse model confirmed that targeting CTLs homed in on the biotin-presenting U87 tumors but not the control U87 tumors. PET imaging with 89Zr-deferoxamine-biotin and SA showed a rapid clearance of the PET signal over 24 h in the control tumor, while only minimally decreased in the targeted tumor. Thus, sulfo-NHS-biotin-SA labeling is an efficient method to noninvasively track the migration of adoptive transferred CTLs and does not alter CTL viability or interfere with CTL-mediated cytotoxic activity.

Myeloperoxidase Nuclear Imaging for Epileptogenesis.

PURPOSE: To determine if myeloperoxidase (MPO) is involved in epileptogenesis and if molecular nuclear imaging can be used to noninvasively map inflammatory changes in epileptogenesis. MATERIALS AND METHODS: The animal and human studies were approved by the institutional review boards. Pilocarpine-induced epileptic mice were treated with 4-aminobenzoic acid hydrazide (n = 46), a specific irreversible MPO inhibitor, or saline (n = 42). Indium-111-bis-5-hydroxytryptamide-diethylenetriaminepentaacetate was used to image brain MPO activity (n = 6 in the 4-aminobenzoic acid hydrazide and saline groups; n = 5 in the sham group) by using single photon emission computed tomography/computed tomography. The role of MPO in the development of spontaneous recurrent seizures was assessed by means of clinical symptoms and biochemical and histopathologic data. Human brain specimens from a patient with epilepsy and a patient without epilepsy were stained for MPO. The Student t test, one-way analysis of variance, and Mann-Whitney and Kruskal-Wallis tests were used. Differences were regarded as significant if P was less than .05. RESULTS: MPO and leukocytes increased in the brain during epileptogenesis (P < .05). Blocking MPO delayed spontaneous recurrent seizures (99.6 vs 142 hours, P = .016), ameliorated the severity of spontaneous recurrent seizures (P < .05), and inhibited mossy fiber sprouting (Timm index, 0.31 vs 0.03; P = .003). Matrix metalloproteinase activity was upregulated during epileptogenesis in an MPO-dependent manner (1.44 vs 0.94 U/mg, P = .049), suggesting that MPO acts upstream of matrix metalloproteinases. MPO activity was mapped during epileptogenesis in vivo in the hippocampal regions. Resected temporal lobe tissue from a human patient with refractory epilepsy but not the temporal lobe tissue from a patient without seizures demonstrated positive MPO immunostaining, suggesting high translational potential for this imaging technology. CONCLUSION: The findings of this study highlight an important role for MPO in epileptogenesis and show MPO to be a potential therapeutic target and imaging biomarker for epilepsy.

Multiple sclerosis: myeloperoxidase immunoradiology improves detection of acute and chronic disease in experimental model.

PURPOSE: To test if MPO-Gd, a gadolinium-based magnetic resonance (MR) imaging probe that is sensitive and specific for the proinflammatory and oxidative enzyme myeloperoxidase (MPO), which is secreted by certain inflammatory cells, is more sensitive than diethylenetriaminepentaacetic acid (DTPA)-Gd in revealing early subclinical and chronic disease activity in the brain in experimental autoimmune encephalomyelitis (EAE), a mouse model of multiple sclerosis. MATERIALS AND METHODS: The protocol for animal experiments was approved by the institutional animal care committee. A total of 61 female SJL mice were induced with EAE. Mice underwent MPO-Gd- or DTPA-Gd-enhanced MR imaging on days 6, 8, and 10 after induction, before clinical disease develops, and during chronic disease at remission and the first relapse. Brains were harvested at these time points for flow cytometric evaluation of immune cell subtypes and immunohistochemistry. Statistical analysis was performed, and P < .05 was considered to indicate a significant difference. RESULTS: MPO-Gd helps detect earlier (5.2 vs 2.3 days before symptom onset, P = .004) and more (3.1 vs 0.3, P = .008) subclinical inflammatory lesions compared with DTPA-Gd, including in cases in which there was no evidence of overt blood-brain barrier (BBB) breakdown detected with DTPA-Gd enhancement. The number of MPO-Gd-enhancing lesions correlated with early infiltration of MPO-secreting monocytes and neutrophils into the brain (r = 0.91). MPO-Gd also helped detect more lesions during subclinical disease at remission (5.5 vs 1.3, P = .006) and at the first relapse (9.0 vs 2.7, P = .03) than DTPA-Gd, which also correlated well with the presence and accumulation of MPO-secreting inflammatory cells in the brain (r = 0.93). CONCLUSION: MPO-Gd specifically reveals lesions with inflammatory monocytes and neutrophils, which actively secrete MPO. These results demonstrate the feasibility of detection of subclinical inflammatory disease activity in vivo, which is different from overt BBB breakdown.

Dual-Energy CTA Iodine Map Reconstructions Improve Visualization of Residual Cerebral Aneurysms following Endovascular Coiling.

BACKGROUND AND PURPOSE: Material-specific reconstructions of dual-energy CTA (DECTA) can highlight iodinated contrast, subtract predefined materials, and reduce metal artifact. We present a technique to improve detection of residual aneurysms after endovascular coiling by which iodine-map DECTA (IM-DECTA) reconstructions subtract platinum coil artifacts in MIP images (MIP IM-DECTA) and assess if IM-DECTA offers improved detection over conventional CTA (CCTA) or monoenergetic DECTA. MATERIALS AND METHODS: We included consecutive patients who underwent endovascular aneurysm coiling with follow-up DECTA and DSA within 24 months. DECTA was performed at 80- and 150-kVp tube voltages on a rapid kV-switching single-source Revolution scanner. CCTA and IM-DECTA series were reconstructed. Reference-standard DSA was compared with CCTA, 50- and 70-keV virtual monochromatic DECTA, IM-DECTA, and MIP IM-DECTA. Blinded to DSA data, cross-section images were reviewed in consensus by 3 neurointerventionalists for residual aneurysms and assigned modified Raymond-Roy classifications (mRRC). Sensitivity, specificity, and accuracy of each series is reported relative to DSA, and single-factor ANOVA and pair-wise Spearman correlation coefficients compared the accuracy of each series. Readers provided ROI measurements of HU deviation adjacent to the aneurysm neck for quantitative noise assessment and qualitatively scored each series on a 3-point Likert-style scale ranging from uninterpretable to excellent image quality. RESULTS: Twenty-one patients with 25 coiled aneurysms were included. Mean time from DECTA to DSA was 286 ± 212 days. IM-DECTA and MIP IM-DECTA most sensitively (89% and 90%) and specifically (93% and 93%) detected residual aneurysms relative to CCTA (6% and 86%). Relative to DSA, IM-DECTA and MIP IM-DECTA most accurately detected (92% versus 28% for CCTA) and classified residual aneurysms by mRRC (ρC-CTA = -0.08; ρIM = 0.50; ρIM-MIP = 0.55; P < .001). Reader consensus reported the best image quality at the aneurysm neck with IM-DECTA and MIP IM-DECTA, with 56% of CCTAs considered uninterpretable versus 0% of IM-DECTAs, and image noise was significantly lower for IM-DECTA (27.9 ± 3.6 HU) or MIP IM-DECTA (26.8 ± 3.5 HU) than CCTA (103.2 ± 13.3 HU; P < .001). CONCLUSIONS: MIP IM-DECTA can subtract coil mass artifact and is more sensitive and specific than CCTA for the detection of residual aneurysms after endovascular coiling.

Advancing diagnostic precision of delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage: The potential for a vasospasm index score on perfusion imaging to detect vasospasm.

BACKGROUND: The occurrence of delayed cerebral ischemia and vasospasm following aneurysmal subarachnoid hemorrhage (aSAH) results in high morbidity and mortality, but the diagnosis remains challenging. This study aimed to identify neuroimaging perfusion parameters indicative of delayed cerebral ischemia in patients with suspected vasospasm. METHODS: This is a case-control study. Cases were adult aSAH patients who underwent magnetic resonance perfusion or computed tomography perfusion (CTP) imaging ≤ 24 h before digital subtraction angiography performed for vasospasm diagnosis and treatment. Controls were patients without aSAH who underwent CTP. Quantitative perfusion parameters at different thresholds, including Tmax 4-6-8-10 s delay, cerebral blood flow and cerebral blood volume were measured and compared between cases and controls. The Vasospasm Index Score was calculated as the ratio of brain volume with time-to-max (Tmax) delay > 6 s over volume with Tmax > 4 s. RESULTS: 54 patients with aSAH and 119 controls without aSAH were included. Perfusion parameters with the strongest prediction of vasospasm on cerebral angiography were the combination of the Vasospasm Index Score (Tmax6/Tmax4) + CBV ≤ 48 % (area under the curve value of 0.85 [95 % CI 0.78-0.91]) with a sensitivity of 63 % and specificity of 95 %. CONCLUSION: The Vasospasm Index Score in combination with CBV ≤ 48 % on cerebral perfusion imaging reliably identified vasospasm as the cause of DCI on perfusion imaging.

Perfusion imaging for delayed cerebral ischemia detection in patients following ruptured aneurysmal subarachnoid hemorrhage: Interrater reliability assessment.

BACKGROUND: Delayed cerebral ischemia (DCI) following aneurysmal subarachnoid hemorrhage (aSAH) is associated with adverse neurological outcomes. Early and accurate diagnosis of DCI is crucial to prevent cerebral infarction. This study aimed to assess the diagnostic accuracy and interrater agreement of the visual assessment of neuroimaging perfusion maps to detect DCI in patients suspected of vasospasm after aSAH. METHODS: In this case-control study, cases were adult aSAH patients with DCI who underwent magnetic resonance perfusion or computed tomography perfusion (CTP) imaging in the 24 h prior to digital subtraction angiography for vasospasm diagnosis. Controls were patients with dizziness and no aSAH on CTP imaging. Three independent raters, blinded to patients' clinical information, other neuroimaging studies, and angiographic results, visually assessed anonymized perfusion color maps to classify patients as either having DCI or not. Tmax delay was classified by symmetry into no delay, unilateral, or bilateral. RESULTS: Perfusion imaging of 54 patients with aSAH and 119 control patients without aSAH was assessed. Sensitivities for DCI diagnosis ranged from 0.65 to 0.78, and specificities ranged from 0.70 to 0.87, with interrater agreement ranging from 0.60 (moderate) to 0.68 (substantial). CONCLUSION: Visual assessment of perfusion color maps demonstrated moderate to substantial accuracy in diagnosing DCI in aSAH patients.

Random expert sampling for deep learning segmentation of acute ischemic stroke on non-contrast CT.

BACKGROUND: Outlining acutely infarcted tissue on non-contrast CT is a challenging task for which human inter-reader agreement is limited. We explored two different methods for training a supervised deep learning algorithm: one that used a segmentation defined by majority vote among experts and another that trained randomly on separate individual expert segmentations. METHODS: The data set consisted of 260 non-contrast CT studies in 233 patients with acute ischemic stroke recruited from the multicenter DEFUSE 3 (Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3) trial. Additional external validation was performed using 33 patients with matched stroke onset times from the University Hospital Lausanne. A benchmark U-Net was trained on the reference annotations of three experienced neuroradiologists to segment ischemic brain tissue using majority vote and random expert sampling training schemes. The median of volume, overlap, and distance segmentation metrics were determined for agreement in lesion segmentations between (1) three experts, (2) the majority model and each expert, and (3) the random model and each expert. The two sided Wilcoxon signed rank test was used to compare performances (1) to 2) and (1) to (3). We further compared volumes with the 24 hour follow-up diffusion weighted imaging (DWI, final infarct core) and correlations with clinical outcome (modified Rankin Scale (mRS) at 90 days) with the Spearman method. RESULTS: The random model outperformed the inter-expert agreement ((1) to (2)) and the majority model ((1) to (3)) (dice 0.51±0.04 vs 0.36±0.05 (P<0.0001) vs 0.45±0.05 (P<0.0001)). The random model predicted volume correlated with clinical outcome (0.19, P<0.05), whereas the median expert volume and majority model volume did not. There was no significant difference when comparing the volume correlations between random model, median expert volume, and majority model to 24 hour follow-up DWI volume (P>0.05, n=51). CONCLUSION: The random model for ischemic injury delineation on non-contrast CT surpassed the inter-expert agreement ((1) to (2)) and the performance of the majority model ((1) to (3)). We showed that the random model volumetric measures of the model were consistent with 24 hour follow-up DWI.

Mechanical Thrombectomy Versus Intravenous Thrombolysis in Distal Medium Vessel Acute Ischemic Stroke: A Multinational Multicenter Propensity Score-Matched Study.

Background and Purpose: The management of acute ischemic stroke (AIS) due to distal medium vessel occlusion (DMVO) remains uncertain, particularly in comparing the effectiveness of intravenous thrombolysis (IVT) plus mechanical thrombectomy (MT) versus IVT alone. This study aimed to evaluate the safety and efficacy in DMVO patients treated with either MT-IVT or IVT alone. Methods: This multinational study analyzed data from 37 centers across North America, Asia, and Europe. Patients with AIS due to DMVO were included, with data collected from September 2017 to July 2023. The primary outcome was functional independence, with secondary outcomes including mortality and safety measures such as types of intracerebral hemorrhage. Results: The study involved 1,057 patients before matching, and 640 patients post-matching. Functional outcomes at 90 days showed no significant difference between groups in achieving good functional recovery (modified Rankin Scale 0-1 and 0-2), with adjusted odds ratios (OR) of 1.21 (95% confidence interval [CI] 0.81 to 1.79; P=0.35) and 1.00 (95% CI 0.66 to 1.51; P>0.99), respectively. Mortality rates at 90 days were similar between the two groups (OR 0.75, 95% CI 0.44 to 1.29; P=0.30). The incidence of symptomatic intracerebral hemorrhage was comparable, but any type of intracranial hemorrhage was significantly higher in the MT-IVT group (OR 0.43, 95% CI 0.29 to 0.63; P<0.001). Conclusion: The results of this study indicate that while MT-IVT and IVT alone show similar functional and mortality outcomes in DMVO patients, MT-IVT presents a higher risk of hemorrhagic complications, thus MT-IVT may not routinely offer additional benefits over IVT alone for all DMVO stroke patients. Further prospective randomized trials are needed to identify patient subgroups most likely to benefit from MT-IVT treatment in DMVO.

Non-inferiority of deep learning ischemic stroke segmentation on non-contrast CT within 16-hours compared to expert neuroradiologists.

We determined if a convolutional neural network (CNN) deep learning model can accurately segment acute ischemic changes on non-contrast CT compared to neuroradiologists. Non-contrast CT (NCCT) examinations from 232 acute ischemic stroke patients who were enrolled in the DEFUSE 3 trial were included in this study. Three experienced neuroradiologists independently segmented hypodensity that reflected the ischemic core on each scan. The neuroradiologist with the most experience (expert A) served as the ground truth for deep learning model training. Two additional neuroradiologists' (experts B and C) segmentations were used for data testing. The 232 studies were randomly split into training and test sets. The training set was further randomly divided into 5 folds with training and validation sets. A 3-dimensional CNN architecture was trained and optimized to predict the segmentations of expert A from NCCT. The performance of the model was assessed using a set of volume, overlap, and distance metrics using non-inferiority thresholds of 20%, 3 ml, and 3 mm, respectively. The optimized model trained on expert A was compared to test experts B and C. We used a one-sided Wilcoxon signed-rank test to test for the non-inferiority of the model-expert compared to the inter-expert agreement. The final model performance for the ischemic core segmentation task reached a performance of 0.46 ± 0.09 Surface Dice at Tolerance 5mm and 0.47 ± 0.13 Dice when trained on expert A. Compared to the two test neuroradiologists the model-expert agreement was non-inferior to the inter-expert agreement, [Formula: see text]. The before, CNN accurately delineates the hypodense ischemic core on NCCT in acute ischemic stroke patients with an accuracy comparable to neuroradiologists.

Matched-pair analysis of patients with ischemic stroke undergoing thrombectomy using next-generation balloon guide catheters.

BACKGROUND: Balloon guide catheters (BGCs) have not been widely adopted, possibly due to the incompatibility of past-generation BGCs with large-bore intermediate catheters. The next-generation BGC is compatible with large-bore catheters. We compared outcomes of thrombectomy cases using BGCs versus conventional guide catheters. METHODS: We conducted a retrospective study of 110 thrombectomy cases using BGCs (n=55) and non-BGCs (n=55). Sixty consecutive thrombectomy cases in whom the BOBBY BGC was used at a single institution between February 2021 and March 2022 were identified. Of these, 55 BGC cases were 1:1 matched with non-BGC cases by proceduralists, age, gender, stent retriever + aspiration device versus aspiration-only, and site of occlusion. First-pass effect was defined as Thrombolysis In Cerebral Infarction 2b or higher with a single pass. RESULTS: The BGC and non-BGC cohorts had similar mean age (67.2 vs 68.9 years), gender distribution (43.6% vs 47.3% women), median initial National Institutes of Health Stroke Scale score (14 vs 15), and median pretreatment ischemic core volumes (12 mL vs 11.5 mL). BGC and non-BGC cases had similar rates of single pass (60.0% vs 54.6%), first-pass effect (58.2% vs 49.1%), and complications (1.8% vs 9.1%). In aspiration-only cases, the BGC cohort had a significantly higher rate of first-pass effect (100% vs 50.0%, p=0.01). BGC was associated with a higher likelihood of achieving a modified Rankin Scale score of 2 at discharge (OR 7.76, p=0.02). No additional procedural time was required for BGC cases (46.7 vs 48.2 min). CONCLUSION: BGCs may be safely adopted with comparable procedural efficacy, benefits to aspiration-only techniques, and earlier functional improvement compared with conventional guide catheters.

Prediction of delayed cerebral ischemia after cerebral aneurysm rupture using explainable machine learning approach.

BACKGROUND: Aneurysmal subarachnoid hemorrhage results in significant mortality and disability, which is worsened by the development of delayed cerebral ischemia. Tests to identify patients with delayed cerebral ischemia prospectively are of high interest. OBJECTIVE: We created a machine learning system based on clinical variables to predict delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage patients. We also determined which variables have the most impact on delayed cerebral ischemia prediction using SHapley Additive exPlanations method. METHODS: 500 aneurysmal subarachnoid hemorrhage patients were identified and 369 met inclusion criteria: 70 patients developed delayed cerebral ischemia (delayed cerebral ischemia+) and 299 did not (delayed cerebral ischemia-). The algorithm was trained based upon age, sex, hypertension (HTN), diabetes, hyperlipidemia, congestive heart failure, coronary artery disease, smoking history, family history of aneurysm, Fisher Grade, Hunt and Hess score, and external ventricular drain placement. Random Forest was selected for this project, and prediction outcome of the algorithm was delayed cerebral ischemia+. SHapley Additive exPlanations was used to visualize each feature's contribution to the model prediction. RESULTS: The Random Forest machine learning algorithm predicted delayed cerebral ischemia: accuracy 80.65% (95% CI: 72.62-88.68), area under the curve 0.780 (95% CI: 0.696-0.864), sensitivity 12.5% (95% CI: -3.7 to 28.7), specificity 94.81% (95% CI: 89.85-99.77), PPV 33.3% (95% CI: -4.39 to 71.05), and NPV 84.1% (95% CI: 76.38-91.82). SHapley Additive exPlanations value demonstrated Age, external ventricular drain placement, Fisher Grade, and Hunt and Hess score, and HTN had the highest predictive values for delayed cerebral ischemia. Lower age, absence of hypertension, higher Hunt and Hess score, higher Fisher Grade, and external ventricular drain placement increased risk of delayed cerebral ischemia. CONCLUSION: Machine learning models based upon clinical variables predict delayed cerebral ischemia with high specificity and good accuracy.

Short-hairpin RNA-mediated Heat shock protein 90 gene silencing inhibits human breast cancer cell growth in vitro and in vivo.

Hsp90 interacts with proteins that mediate signaling pathways involved in the regulation of essential processes such as proliferation, cell cycle control, angiogenesis and apoptosis. Hsp90 inhibition is therefore an attractive strategy for blocking abnormal pathways that are crucial for cancer cell growth. In the present study, the role of Hsp90 in human breast cancer MCF-7 cells was examined by stably silencing Hsp90 gene expression with an Hsp90-silencing vector (Hsp90-shRNA). RT-PCR and Western blot analyses showed that Hsp90-shRNA specifically and markedly down-regulated Hsp90 mRNA and protein expression. NF-kB and Akt protein levels were down-regulated in Hsp90-shRNA transfected cells, indicating that Hsp90 knockout caused a reduction of survival factors and induced apoptosis. Treatment with Hsp90-shRNA significantly increased apoptotic cell death and caused cell cycle arrest in the G1/S phase in MCF-7 cells, as shown by flow cytometry. Silencing of Hsp90 also reduced cell viability, as determined by MTT assay. In vivo experiments showed that MCF-7 cells stably transfected with Hsp90-shRNA grew slowly in nude mice as compared with control groups. In summary, the Hsp90-shRNA specifically silenced the Hsp90 gene, and inhibited MCF-7 cell growth in vitro and in vivo. Possible molecular mechanisms underlying the effects of Hsp90-shRNA include the degradation of Hsp90 breast cancer-related client proteins, the inhibition of survival signals and the upregulation of apoptotic pathways. shRNA-mediated interference may have potential therapeutic utility in human breast cancer.

Acute ischemic stroke: infarct core estimation on CT angiography source images depends on CT angiography protocol.

PURPOSE: To test whether the relationship between acute ischemic infarct size on concurrent computed tomographic (CT) angiography source images and diffusion-weighted (DW) magnetic resonance images is dependent on the parameters of CT angiography acquisition protocols. MATERIALS AND METHODS: This retrospective study had institutional review board approval, and all records were HIPAA compliant. Data in 100 patients with anterior-circulation acute ischemic stroke and large vessel occlusion who underwent concurrent CT angiography and DW imaging within 9 hours of symptom onset were analyzed. Measured areas of hyperintensity at acute DW imaging were used as the standard of reference for infarct size. Information regarding lesion volumes and CT angiography protocol parameters was collected for each patient. For analysis, patients were divided into two groups on the basis of CT angiography protocol differences (patients in group 1 were imaged with the older, slower protocol). Intermethod agreement for infarct size was evaluated by using the Wilcoxon signed rank test, as well as by using Spearman correlation and Bland-Altman analysis. Multivariate analysis was performed to identify predictors of marked (≥20%) overestimation of infarct size on CT angiography source images. RESULTS: In group 1 (n=35), median hypoattenuation volumes on CT angiography source images were slightly underestimated compared with DW imaging hyperintensity volumes (33.0 vs 41.6 mL, P=.01; ratio=0.83), with high correlation (ρ=0.91). In group 2 (n=65), median volume on CT angiography source images was much larger than that on DW images (94.8 vs 17.8 mL, P<.0001; ratio=3.5), with poor correlation (ρ=0.49). This overestimation on CT angiography source images would have inappropriately excluded from reperfusion therapy 44.4% or 90.3% of patients eligible according to DW imaging criteria on the basis of a 100-mL absolute threshold or a 20% or greater mismatch threshold, respectively. Atrial fibrillation and shorter time from contrast material injection to image acquisition were independent predictors of marked (≥20%) infarct size overestimation on CT angiography source images. CONCLUSION: CT angiography protocol changes designed to speed imaging and optimize arterial opacification are associated with significant overestimation of infarct size on CT angiography source images.

Demyelinating diseases: myeloperoxidase as an imaging biomarker and therapeutic target.

PURPOSE: To evaluate myeloperoxidase (MPO) as a newer therapeutic target and bis-5-hydroxytryptamide-diethylenetriaminepentaacetate-gadolinium (Gd) (MPO-Gd) as an imaging biomarker for demyelinating diseases such as multiple sclerosis (MS) by using experimental autoimmune encephalomyelitis (EAE), a murine model of MS. MATERIALS AND METHODS: Animal experiments were approved by the institutional animal care committee. EAE was induced in SJL mice by using proteolipid protein (PLP), and mice were treated with either 4-aminobenzoic acid hydrazide (ABAH), 40 mg/kg injected intraperitoneally, an irreversible inhibitor of MPO, or saline as control, and followed up to day 40 after induction. In another group of SJL mice, induction was performed without PLP as shams. The mice were imaged by using MPO-Gd to track changes in MPO activity noninvasively. Imaging results were corroborated by enzymatic assays, flow cytometry, and histopathologic analyses. Significance was computed by using the t test or Mann-Whitney U test. RESULTS: There was a 2.5-fold increase in myeloid cell infiltration in the brain (P = .026), with a concomitant increase in brain MPO level (P = .0087). Inhibiting MPO activity with ABAH resulted in decrease in MPO-Gd-positive lesion volume (P = .012), number (P = .009), and enhancement intensity (P = .03) at MR imaging, reflecting lower local MPO activity (P = .03), compared with controls. MPO inhibition was accompanied by decreased demyelination (P = .01) and lower inflammatory cell recruitment in the brain (P < .0001), suggesting a central MPO role in inflammatory demyelination. Clinically, MPO inhibition significantly reduced the severity of clinical symptoms (P = .0001) and improved survival (P = .0051) in mice with EAE. CONCLUSION: MPO may be a key mediator of myeloid inflammation and tissue damage in EAE. Therefore, MPO could represent a promising therapeutic target, as well as an imaging biomarker, for demyelinating diseases and potentially for other diseases in which MPO is implicated.