Cystic fibrosis and noninvasive liver fibrosis assessment methods in children.

BACKGROUND: Noninvasive assessments of liver fibrosis are currently used to evaluate cystic fibrosis (CF)-related liver disease. However, there is scarce data regarding their repeatability and reproducibility, especially in children with CF. The present study aimed to evaluate the repeatability and reproducibility of transient elastography (TE) (FibroScan®) and point shear-wave elastography using virtual touch quantification (pSWE VTQ) in children with CF. METHODS: TE and pSWE VTQ were performed in 56 children with CF by two different operators. Analysis of repeatability and reproducibility was available in 33 patients for TE and 46 patients for pSWE VTQ. Intra- and interobserver agreement were assessed using the intraclass correlation coefficient (ICC) and their 95% confidence interval (CI), and Bland and Altman graphs. RESULTS: For TE, ICC was 0.91 (0.83-0.95) for intraobserver agreement and 0.92 (95% CI: 0.86-0.96) for interobserver agreement. For pSWE VTQ, ICC was 0.83 (0.72-0.90) for intraobserver agreement and 0.67 (0.48-0.80) for interobserver agreement. CONCLUSIONS: Both technics can be proposed in the follow-up of patients, according to their availability in CF centers. IMPACT: This study shows that TE and pSWE VTQ are reliable methods to evaluate liver fibrosis in children with CF. This study shows for the first time that TE and pSWE VTQ are both repeatable and reproducible in children with CF. These data indicate that both TE and pSWE VTQ can be proposed for the follow-up of patients with CF, according to their availability in each CF center.

Childhood cancer risks estimates following CT scans: an update of the French CT cohort study.

OBJECTIVES: Increased risks of central nervous system (CNS) tumors and leukemia associated with computed tomography (CT) exposure during childhood have been reported in recent epidemiological studies. However, no evidence of increased risks was suggested in a previous analysis of the French CT cohort. This study benefits from an updated cohort with a longer follow-up and a larger sample size of patients. METHODS: The patients were followed from the date of their first CT (between 2000 and 2011) until their date of cohort exit defined as the earliest among the following: 31 December 2016, date of death, date of first cancer diagnosis or date of their 18th birthday. Cancer incidence, vital status, cancer predisposing factors (PFs), and additional CT scans were collected via external national databases. Hazard ratios (HRs) associated to cumulative organ doses and sex were estimated from Cox models. RESULTS: At the end of follow-up, mean cumulative doses were 27.7 and 10.3 mGy for the brain and the red bone marrow (RBM), respectively. In patients without PFs, an HR per 10 mGy of 1.05 (95% CI: 1.01-1.09) for CNS tumors, 1.17 (95% CI: 1.09-1.26) for leukemia, and 0.96 (95% CI: 0.63-1.45) for lymphoma was estimated. These estimates were not modified by the inclusion of CT scans performed outside the participating hospitals or after the inclusion period. CONCLUSIONS: This study shows statistically significant dose-response relationships for CNS tumors and leukemia for patients without PFs. KEY POINTS: • Computed tomography is the most important contributor to the collective dose for diagnostic imaging to the French population. • Concerns have been raised about possible cancer risks, particularly after exposure to CT in childhood, due to the greater radiation sensitivity of children and to their longer life expectancy. • Analysis of the updated French CT cohort shows statistically significant dose-response relationships for CNS tumors and leukemia.

A practical guide to optimize arterial spin labeling in neonates at 1.5 Tesla: what the radiologist needs to know.

Arterial spin labeling magnetic resonance imaging is highly suited to the exploration of brain perfusion in neonates and has the potential to provide relevant complementary information to neuroimaging studies, with insights into neurodevelopmental outcomes. Applying this technique within the first days of life is challenging and requires specific technical adaptations. The literature on this topic is scarce and heterogeneous, especially on 1.5-T scanners, limiting widespread clinical adoption. This paper aims to describe a simple approach for arterial spin labeling in neonates, with key considerations for radiologists.

Improved stenosis outcome in stroke-free sickle cell anemia children after transplantation compared to chronic transfusion.

We report here the 3-year stenosis outcome in 60 stroke-free children with sickle cell anaemia (SCA) and an abnormal transcranial Doppler history, enrolled in the DREPAGREFFE trial, which compared stem cell transplantation (SCT) with standard-care (chronic transfusion for 1-year minimum). Twenty-eight patients with matched sibling donors were transplanted, while 32 remained on standard-care. Stenosis scores were calculated after performing cerebral/cervical 3D time-of-flight magnetic resonance angiography. Fourteen patients had stenosis at enrollment, but only five SCT versus 10 standard-care patients still had stenosis at 3 years. Stenosis scores remained stable on standard-care, but significantly improved after SCT (P = 0·006). No patient developed stenosis after SCT, while two on standard-care did, indicating better stenosis prevention and improved outcome after SCT.

Brain fetal magnetic resonance imaging to evaluate maturation of normal white matter during the third trimester of pregnancy.

BACKGROUND: Quantitative magnetic resonance imaging (MRI) could improve the estimation of fetal brain maturation and the interpretation of white matter signal intensity in pathological conditions. OBJECTIVE: To investigate T2-based and diffusion-weighted imaging (DWI) measurements for the evaluation of fetal brain maturation during the last trimester of pregnancy. MATERIALS AND METHODS: One hundred sixty-eight fetal brain MRIs were retrospectively analyzed (age range: 28-37 weeks of gestation) after ensuring that none of the children developed psychomotor or cognitive impairment (median follow-up: 4.7 years). Bilateral regions of interest were drawn on the frontal, occipital, parietal and temporal lobes from T2-W imaging and DWI, when available, to evaluate signal intensity and apparent diffusion coefficient (ADC) values. Ratios were calculated with two references (pons or thalamus and cerebrospinal fluid) to standardize signal intensities. Reproducibility was evaluated with intraclass correlation coefficients (ICCs) and Bland-Altman plots. Correlations with gestational age were evaluated with univariate and multivariate linear regressions. RESULTS: T2 measurements were achieved in all cases, and DWI was available in 37 cases. Measurements and ratios were reproducible in eight localizations (i.e. intra- and interobserver ICCs >0.5): frontal T2/thalamus, parietal T2/thalamus, occipital T2/pons, parietal ADC/thalamus, occipital ADC/pons, temporal ADC/pons, occipital ADC and temporal ADC. The frontal T2/thalamus and parietal T2/thalamus correlated with gestational age (P<0.0001 and P=0.014, respectively). In the multivariate modeling, frontal T2/thalamus remained an independent predictor of the gestational age (P<0.0001). CONCLUSION: The frontal T2/thalamus ratio emerged as a potential additional biomarker of fetal brain maturation during the last trimester of pregnancy.

Chest computed tomography findings for a cohort of children with pulmonary Langerhans cell histiocytosis.

OBJECTIVE: This study was undertaken to describe the spectrum of lung computed-tomography (CT) findings in children with pulmonary Langerhans cell histiocytosis (PLCH) and to evaluate for this population the CT-scan nodule and cyst scores proposed by adult pulmonologists at diagnosis and during follow-up. METHODS: Among 175 children with PLCH identified in the French national population-based Langerhans cell histiocytosis cohort, 60 were retrospectively selected by the availability of CT for a central review by three pediatric radiologists. These 60 patients are representative of childhood PLCH for almost all clinical aspects, except a lower percentage of risk organ involvement (38% vs 54%; P = 0.05). RESULTS: The 60 children's chest CT scans (n = 218) were reviewed. At diagnosis, 63% of them had nodules, 53% had cysts, and 29% had both. The percentages of patients with nodules or cysts increased from diagnosis to peak disease activity, respectively, from 63% to 73% and from 53% to 66%. The costophrenic angle was involved in 71%. Patients with pneumothorax (25%) had a higher median cyst score. Alveolar consolidation was observed in 34%. Patients with low CT-scan nodule and cyst scores had no long-term pulmonary sequelae. CONCLUSIONS: Well-known characteristics of adult PLCH (nodules and cysts) were observed in children. The chest CT scores proposed by adult pulmonologists could easily be applied to childhood PLCH. Lesions in children, unlike those in adults, are frequently located near the costophrenic angles. Alveolar consolidation might be considered an atypical feature of childhood PLCH.

Pediatric minor head injury imaging practices: results from an ESPR survey.

PURPOSE: Computed tomography (CT) for minor head injury exposes a large number of children to ionizing radiation, with an associated increased lifetime risk of malignancy. To study imaging practices for children with minor head injury and the level of awareness of radiologists of the current clinical decision rules for minor traumatic brain injury (TBI). METHODS: A questionnaire consisting of 17 questions was distributed electronically to 472 ESPR members. The questionnaire covered demographic information, employment status, years of experience and the current practice setting of the participants, the number of CTs performed for pediatric head trauma, awareness of clinical decision rules and use of shielding, pediatric CT protocol and MRI. RESULTS: The response rate was 18.4%. The majority of participants was aged over 50 years and was full-time consultants. Regarding decision rules, 73.8% of respondents cited the NICE head injury guidelines, and 79% reported that the decision to perform CT was agreed between specialists. Shielding was used by 58.3% and 67.4% applied a specific pediatric protocol. MRI was not used for pediatric head trauma by 70.6% of respondents, although always available in 68.6% of cases. The reported obstacles to MRI use were machine availability (42.7%), the long acquisition time (39%) and patients' intolerance (18.3%), and less frequently the cost and the need for sedation. CONCLUSION: There is room for decreasing the use of CT for pediatric minor TBI. The use of shielding and application of pediatric CT protocols constitute areas for improvement.

Magnetic resonance imaging diagnosis of subependymal giant cell astrocytomas in follow-up of children with tuberous sclerosis complex: should we always use contrast enhancement?

BACKGROUND: Subependymal giant cell astrocytomas (SEGAs) arise in 10-26% of tuberous sclerosis complex (TSC) patients. SEGAs cause obstructive hydrocephalus and increase morbi-mortality. It is recommended that TSC patients be followed with contrast enhanced magnetic resonance imaging (CE-MRI), but repetitive use of gadolinium-based contrast-agents (GBCAs) may cause organ deposits. OBJECTIVE: To compare the diagnostic performances of non-CE- and CE-MRI to differentiate SEGAs from subependymal nodules in TSC patients during follow-up. MATERIALS AND METHODS: Thirty-five TSC patients (median age: 2.4 years) were enrolled in this retrospective single-center study from September 2007 to January 2019. Inclusion criteria were a certain diagnosis of TSC and at least three follow-up brain MRIs with GBCA injection. Two consecutive MRI scans per patient were selected and anonymized. Three radiologists performed a blinded review of non-enhanced and enhanced MRI sequences during different sessions. The diagnostic performances were compared (sensitivity, specificity, positive/negative predictive values, accuracy, inter/intra-observer agreements). RESULTS: The accuracies for detecting SEGAs were good and similar between the non-enhanced and enhanced MRI sequences. The sensitivity and specificity of non-CE-MRI to diagnose SEGA ranged from 75% to 100% and from 94% to 100%, respectively. The differences in numbers of false-positive and false-negative patients between non-CE- and CE-MRI never exceeded one case. Nodules size >10 mm, location near the Monro foramen, hydrocephalus and modifications between two consecutive MRI scans were significantly associated with the diagnosis of SEGA for the three readers (all P-values <0.05). Inter- and intra-observer agreements were also excellent for non-enhanced and enhanced MRI sequences (kappa=0.85-1 and 0.81-0.93, respectively). CONCLUSION: The performances of non-enhanced and enhanced MRI sequences are comparable for detecting SEGAs, questioning the need for systematic GBCA injections for TSC patients.

Correction to: Diagnostic performance of an unenhanced MRI exam for tumor follow-up of the optic pathway gliomas in children.

In the article "Diagnostic performance of an unenhanced MRI exam for tumor follow-up of the optic pathway gliomas in children", Table 2 data were not presented correctly, with results placed beneath an incorrect heading. Confidence interval also added. The original article has been corrected.

Diagnostic performance of an unenhanced MRI exam for tumor follow-up of the optic pathway gliomas in children.

PURPOSE: Contrast-enhanced MRI (MRI + C) is considered as mandatory for brain tumors follow-up, but gadolinium brain depositions in relation with repeated injections have been reported. The aim of our work was to evaluate the diagnostic performance of an unenhanced MRI examination for the follow-up of optic pathway gliomas (OPG) in children. METHODS: Seventeen patients (with/without NF1) were selected from 2001 to 2017, with at least 5 MRI + C brain follow-up examinations. Privacy and data protection rights were addressed by the data protection officer (DPO) and the study was in accordance with the local ethical rules. Twenty-five cases of tumor progression and 25 cases of tumor stability mentioned in the conclusion of radiological reports (defined as gold standard) were isolated. Those exams were anonymized and independently reviewed by two radiologists, who analyzed both quantitative (such as tumor volume variation) and qualitative criteria (such as ventricular dilatation) on unenhanced images. Sensitivity, specificity, positive/negative predictive values (PPV, NPV), and inter/intra-observer agreement were calculated. RESULTS: The mean age of patients was 5.4 ± 3.4 years and mean follow-up length 6.7 years. The mean number of MRI + C was 13.5 (SD 7.2). The sensitivity of unenhanced MRI for tumor follow-up was 84-88% (95% CI 63.9-97.5). The specificity was 91.3-100% (95% CI 72-100). The PPV was 91.7% for reader 1 and 100% for reader 2. The NVP was 87.5% for reader 1 and 85.2% for reader 2. There was an excellent inter-observer agreement regarding tumor progression: kappa coefficient of 0.87 (p < 0.001). Inter/intra-variability for percentage of tumor volume variation between two exams were good (correlation coefficients of 0.97 and 0.94). CONCLUSION: Tumor volume variation is in most cases sufficient to assess OPG progression. Systematic MRI + C could be questionable.

Individual radiation exposure from computed tomography: a survey of paediatric practice in French university hospitals, 2010-2013.

OBJECTIVES: To describe computed tomography (CT) scanning parameters, volume CT dose index (CTDIvol) and dose-length product (DLP) in paediatric practice and compare them to current diagnostic reference levels (DRLs). METHODS: The survey was conducted in radiology departments of six major university hospitals in France in 2010-2013. Data collection was automatised to extract and standardise information on scanning parameters from DICOM-header files. CTDIvol and DLP were estimated based on Monte Carlo transport simulation and computational reference phantoms. RESULTS: CTDIvol and DLP were derived for 4,300 studies, four age groups and 18 protocols. CTDIvol was lower in younger patients for non-head scans, but did not vary with age for routine head scans. Ratios of 95th to 5th percentile CTDIvol values were 2-4 for most body parts, but 5-7 for abdominal examinations and 4-14 for mediastinum CT with contrast, depending on age. The 75th percentile CTDIvol values were below the national DRLs for chest (all ages) and head and abdominal scans (≥10 years). CONCLUSION: The results suggest the need for a better optimisation of scanning parameters for routine head scans and infrequent protocols with patient age, enhanced standardisation of practices across departments and revision of current DRLs for children. KEY POINTS: • CTDIvol varied little with age for routine head scans. • CTDIvol was lowest in youngest children for chest or abdominal scans. • Individual and inter-department variability warrant enhanced standardisation of practices. • Recent surveys support the need for revised diagnostic reference levels. • More attention should be given to specific protocols (sinuses, neck, spine, mediastinum).

Allergic Bronchopulmonary Aspergillosis in Cystic Fibrosis: MR Imaging of Airway Mucus Contrasts as a Tool for Diagnosis.

Purpose To assess the diagnostic accuracy of mucus contrast characterization by using magnetic resonance (MR) imaging to discriminate allergic bronchopulmonary aspergillosis (ABPA) in cystic fibrosis (CF). Materials and Methods The study was approved by the local Ethics Committee, and all patients or their parents gave written informed consent. One hundred ten consecutive patients with CF were screened between January 2014 and July 2015. All patients underwent a non-contrast material-enhanced MR protocol that included routine T1-weighted and T2-weighted sequences. The presence of mucus with both high T1 and low T2 signal intensities and the so-called inverted mucoid impaction signal (IMIS) sign was qualitatively and quantitatively assessed by two physicians who were blinded to all other data. The reference standard for a diagnosis of ABPA was the criteria of the Cystic Fibrosis Foundation Consensus Conference. ABPA status was followed up for 1 year. Reproducibility was assessed by using the κ test, correlation was assessed by using the Spearman coefficient, and diagnostic accuracy was assessed by calculating the sensitivity and specificity of IMIS. Results One hundred eight patients with CF were included (mean age, 20 years ± 11 [standard deviation]; range, 6-53 years): 18 patients with ABPA and 90 patients without ABPA. At the lobar level, inter- and intrareader reproducibility were very good (κ > 0.90). IMIS had 94% sensitivity (95% confidence interval [CI]: 73%, 99%) and 100% specificity (95% CI: 96%, 100%) for the diagnosis of ABPA. A complete resolution of IMIS was observed in patients with ABPA after 3 months of specific treatment that was significantly correlated with decrease in total immunoglobulin E level (ρ = 0.47; P = .04). Conclusion The IMIS sign was both specific and sensitive for the diagnosis of ABPA in CF. Allergic fungal inflammation appears to induce characteristic modifications of mucus contrasts that are assessable by using a noninvasive, contrast material-free, and radiation-free method. © RSNA, 2017 Online supplemental material is available for this article.

Lung morphology assessment of cystic fibrosis using MRI with ultra-short echo time at submillimeter spatial resolution.

OBJECTIVES: We hypothesized that non-contrast-enhanced PETRA (pointwise encoding time reduction with radial acquisition) MR (magnetic resonance) sequencing could be an alternative to unenhanced computed tomography (CT) in assessing cystic fibrosis (CF) lung structural alterations, as well as compared agreements and concordances with those of conventional T1-weighted and T2-weighted sequences. MATERIAL AND METHODS: Thirty consecutive CF patients completed both CT and MRI the same day. No contrast injection was used. Agreement in identifying structural alterations was evaluated at the segmental level using a kappa test. Intraclass correlation coefficients (ICC) and Bland-Altman analysis were used to assess concordances and reproducibility in Helbich-Bhalla disease severity scoring. RESULTS: Agreement between PETRA and CT was higher than that of T1- or T2-weighted sequences, notably in assessing the segmental presence of bronchiectasis (Kappa = 0.83; 0.51; 0.49, respectively). The concordance in Helbich-Bhalla scores was very good using PETRA (ICC = 0.97), independently from its magnitude (mean difference (MD) = -0.3 [-2.8; 2.2]), whereas scoring was underestimated using both conventional T1 and T2 sequences (MD = -3.6 [-7.4; 0.1]) and MD = -4.6 [-8.2; -1.0], respectively). Intra- and interobserver reproducibility were very good for all imaging modalities (ICC = 0.86-0.98). CONCLUSION: PETRA showed higher agreement in describing CF lung morphological changes than that of conventional sequences, whereas the Helbich-Bhalla scoring matched closely with that of CT. KEY POINTS: • Spatial resolution of lung MRI is limited using non-ultra-short TE MRI technique • Ultra-short echo time (UTE) technique enables submillimeter 3D-MRI of airways • 3D-UTE MRI shows very good concordance with CT in assessing cystic fibrosis • Radiation-free 3D-UTE MRI enables the Helbich-Bhalla scoring without a need for contrast injection.

Quiet Submillimeter MR Imaging of the Lung Is Feasible with a PETRA Sequence at 1.5 T.

PURPOSE: To assess lung magnetic resonance (MR) imaging with a respiratory-gated pointwise encoding time reduction with radial acquisition (PETRA) sequence at 1.5 T and compare it with imaging with a standard volumetric interpolated breath-hold examination (VIBE) sequence, with extra focus on the visibility of bronchi and the signal intensity of lung parenchyma. MATERIALS AND METHODS: The study was approved by the local ethics committee, and all subjects gave written informed consent. Twelve healthy volunteers were imaged with PETRA and VIBE sequences. Image quality was evaluated by using visual scoring, numbering of visible bronchi, and quantitative measurement of the apparent contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR). For preliminary clinical assessment, three young patients with cystic fibrosis underwent both MR imaging and computed tomography (CT). Comparisons were made by using the Wilcoxon signed-rank test for means and the McNemar test for ratios. Agreement between CT and MR imaging disease scores was assessed by using the κ test. RESULTS: PETRA imaging was performed with a voxel size of 0.86 mm(3). Overall image quality was good, with little motion artifact. Bronchi were visible consistently up to the fourth generation and in some cases up to the sixth generation. Mean CNR and SNR with PETRA were 32.4% ± 7.6 (standard deviation) and 322.2% ± 37.9, respectively, higher than those with VIBE (P < .001). Good agreement was found between CT and PETRA cystic fibrosis scores (κ = 1.0). CONCLUSION: PETRA enables silent, free-breathing, isotropic, and submillimeter imaging of the bronchi and lung parenchyma with high CNR and SNR and may be an alternative to CT for patients with cystic fibrosis.

A prospective study of risk for Sturge-Weber syndrome in children with upper facial port-wine stain.

BACKGROUND: Upper facial port-wine stain (PWS) is a feature of Sturge-Weber syndrome (SWS). Recent studies suggest that the distribution of the PWS corresponds to genetic mosaicism rather than to trigeminal nerve impairment. OBJECTIVES: We sought to refine the cutaneous distribution of upper facial PWS at risk for SWS. METHODS: This was a prospective multicenter study of consecutive cases of upper facial PWS larger than 1 cm² located in the ophthalmic division of trigeminal nerve distribution in infants aged less than 1 year, seen in 8 French pediatric dermatology departments between 2006 and 2012. Clinical data, magnetic resonance imaging, and photographs were systematically collected and studied. PWS were classified into 6 distinct patterns. RESULTS: In all, 66 patients were included. Eleven presented with SWS (magnetic resonance imaging signs and seizure). Four additional infants had suspected SWS without neurologic manifestations. Hemifacial (odds ratio 7.7, P = .003) and median (odds ratio 17.08, P = .008) PWS patterns were found to be at high risk for SWS. A nonmedian linear pattern was not associated with SWS. LIMITATIONS: Small number of patients translated to limited power of the study. CONCLUSIONS: Specific PWS distribution patterns are associated with an increased risk of SWS. These PWS patterns conform to areas of somatic mosaicism. Terminology stipulating ophthalmic division of trigeminal nerve territory involvement in SWS should be abandoned.

Thymus and mediastinal node involvement in childhood Langerhans cell histiocytosis: long-term follow-up from the French national cohort.

BACKGROUND: Mediastinal involvement (MI) in Langerhans cell histiocytosis (LCH) has been rarely reported. Here, we describe the clinical, radiological, and biological presentation, and the outcome of childhood LCH with MI. METHOD: From the French LCH register, which includes 1,423 patients aged less than 18 years, we retrieved the medical charts of patients with mediastinal enlargement detected on chest X-rays. RESULTS: Thirty-seven patients were retrieved, including 18 males; median age of diagnosis was 0.7 years, and median follow-up time was 6.2 years. The prevalence of MI varied with the age at diagnosis, ranging from 7% below 1 year old to less than 1% at >5 years. Thirteen cases (35%) were diagnosed because of MI-related symptoms, including respiratory distress (N = 4), superior venous cava syndrome (N = 2), and/or cough and polypnea (N = 10). CT scans performed in 32 cases at diagnosis showed tracheal compression (N = 5), cava thrombosis (N = 2), and/or calcification (N = 16). All patients presented multi-system disease at LCH diagnosis, and 35/37 were initially treated with vinblastine and corticosteroids. Death occurred in five cases, due to MI (N = 1) or hematological refractory involvement (N = 4). The overall 5-year survival was 87.1%, and immunodeficiency was not detected as a sequel. CONCLUSIONS: MI in LCH mainly occurs in young children, and diagnosis was based on CT showing thymus enlargement and calcifications.

Hemidystonia with polymicrogyria is part of ATP1A3-related disorders.

INTRODUCTION: Pathogenic variants in ATP1A3 cause various phenotypes of neurological disorders, including alternating hemiplegia of childhood 2, CAPOS syndrome (cerebellar ataxia, areflexia, pes cavus, optic atrophy, and sensorineural hearing loss) and rapid-onset dystonia-parkinsonism (RDP). Early developmental and epileptic encephalopathy has also been reported. Polymicrogyria has recently been added to the phenotypic spectrum of ATP1A3-related disorders. CASE REPORT: We report here a male patient with early developmental delay who at 12 months presented dystonia of the right arm which evolved into hemidystonia at the age of 2. A cerebral MRI showed bilateral perisylvian polymicrogyria with intact basal ganglia. Whole-exome and whole-genome sequencing analyses identified a de novo new ATP1A3 missense variant (p.Arg914Lys) predicted pathogenic. Hemidystonia was thought not to be due to polymicrogyria, but rather a consequence of this variant. CONCLUSION: This case expands the phenotypic spectrum of ATP1A3-related disorders with a new variant associated with hemidystonia and polymicrogyria and thereby, suggests a clinical continuum between the different phenotypes of this condition.

Intracranial empyema complicating sinusitis in childhood: Epidemiology, imaging findings and outcome.

BACKGROUND: To describe clinical presentations of intracranial sinusitis complications in childhood, their pitfalls and imaging findings. MATERIEL AND METHODS: This retrospective IRB-approved single-center study included infants diagnosed with sinusitis and empyema and/or other intracranial complications who underwent imaging between September 2008 and September 2019. Three radiologists individually reviewed clinical charts and imaging findings, including sinusitis complications and at-risk anatomical variations. RESULTS: 21 children (76% males and 24% females, mean age 13±3.1 years) with imaging pansinusitis were included. Headache (95%) and fever (90%) were the main clinical nonspecific signs. Ten (48%) children presented an extradural empyema, nine (43%) children had a subdural empyema and two (10%) children had both. Frontal location sinusitis was the most common (76%). In MRI, all empyema presented as a hypo intensity on pre-contrast T1-WI, a hyperintensity on T2-WI, a reduced apparent diffusion coefficient (ADC) on diffusion weighted imaging (DWI) and a peripheral contrast enhancement on post-contrast T1-WI. CT or MRI revealed intracranial complications such as a collection size increase (52%), a midline shift (62%), intraparenchymal abscesses (24%), a cerebral venous thrombosis (29%), an intracranial pressure increase (29%), cerebral ischemia (43%) and Pott's Puffy Tumor (10%). Imaging highlighted sinus anatomical abnormalities in 52% of cases. All children were treated with sinus drainage and/or neurosurgery. Long-term follow-up was favorable in 14 cases (67%). CONCLUSION: Complications of sinusitis are life threatening in the studied population. Empyema and cerebral complications may be misleading. Brain contrast-enhanced CT covering sinuses and orbits, is mainly the first examination done but MRI is mandatory.