In this blog, we provide a synopsis of the latest in Pediatric Imaging published during the month of June 2020.
ABDOMINAL IMAGING
Abdominal applications of pediatric body MR angiography: tailored optimization for a successful outcome
Edwards EA, Lee MM, Acord MR et al. Abdominal applications of pediatric body MR angiography: tailored optimization for a successful outcome. AJR 2020; 215: 1-9.
https://doi.org/10.2214/AJR.19.22289
Question(s) What are current common techniques and indications for pediatric abdominopelvic MR angiography (MRA) and strategies for optimizing them?
Design Review article
Commentary The authors summarize the techniques of abdominopelvic MRA for common indications in children. There is a wide range of MRA applications, including vascular anomalies, vasculitis, renal artery stenosis, tumor evaluation, portal hypertension, and transplant evaluation. Current common techniques discussed include contrast-enhanced MRA and unenhanced MRA. The article additionally includes new/up-and-coming techniques, such as four-dimensional MRA, whole-bodying imaging, and blood pool contrast agents (like ferumoxytol) as alternatives to gadolinium-based contrast agents.
Comparison of liver T1 relaxation times without and with iron correction in pediatric autoimmune liver disease
Dillman JR, Serai SD, Miethke AG et al. Comparison of liver T1 relaxation times without and with iron correction in pediatric autoimmune liver disease. Pediatr Radiol 2020; 50: 935-942.
https://doi.org/10.1007/s00247-020-04663-8
Question(s) What is the relationship between liver MRI T1 relaxometry measurements using a modified Look-Locker inversion recovery (MOLLI) pulse sequence without and with iron (T2*) correction (cT1) in pediatric autoimmune liver disease?
Design Cross-sectional study using a prospective institutional pediatric autoimmune liver disease registry (allowing retrospective and prospective review of data)
Setting Patient data from a registry
Participants 48 registry participants with known autoimmune liver disease
Intervention MOLLI pulse sequence without and with iron correction
Outcomes T1 relaxometry measurements (liver native T1, liver T2*, proton density fat fraction, ROI-based cT1, whole-liver median cT1, whole-liver mean cT1) and correlation coefficient (r) between native T1 and the cT1 via statistical analyses
Main Results Excellent positive correlation between native T1 and ROI-based cT1 (r=0.91, P<0.0001), native T1 and whole-liver median cT1 (r=0.9, P<0.0001), and native liver T1 and whole-liver mean cT1 (r=0.81, P<0.0001). No significance between the liver T2* and the difference between cT1 and native T1 values (r=0.15, P=0.32).
Commentary MRI can be used for detection and surveillance of chronic liver disease and can reduce or even preclude the need for liver biopsy. Certain quantitative MRI techniques, such as MR elastography or cT1 relaxometry/mapping correlate with hepatic fibrosis and inflammation. The authors of this article demonstrated excellent positive correlation between liver native T1 relaxometry measurements with and without iron correction and cT1. They conclude that T1 iron correction may not be needed in pediatric patients without excessive iron overload. However, the authors acknowledge that due to systematic bias related to mathematical formulas used, non-iron-corrected native T1 and cT1 values are not interchangeable.
Radiographic findings of irreducibility and surgical resection in ileocolic intussusception
Patel DM, Loewen JM, Braithwaite KA et al. Radiographic findings of irreducibility and surgical resection in ileocolic intussusception. Pediatr Radiol (2020).
https://doi.org/10.1007/s00247-020-04695-0
Question(s) Do abdominal radiographs offer prognostic information about the results of therapeutic air enema and surgical outcomes, including surgical complications?
Design Retrospective study of a 5-year period, 2 pediatric radiologists reviewed the radiographs and were blinded to enema and surgical outcomes but not to a diagnosis of intussusception, 14 radiologists performed intussusception reduction exams
Setting Single center: Emory University/Children’s Healthcare of Atlanta
Participants 182 cases
Intervention Evaluate abdominal radiographs for normal bowel gas pattern, soft tissue mass, paucity of bowel gas, meniscus sign, and bowel obstruction. Review medical records for enema and surgical outcomes.
Outcomes Success rates of air enema, complications with surgical reduction, bowel resection
Main Results When compared to patient with normal bowel gas patterns on radiographs, patients with bowel obstruction were less successfully treated with air enema (21% vs. 83%, P=0.0001), had more complications with surgical reductions (47% vs. 4%m, P=0.0012), and increased need for bowel resection (42% vs. 4%, P=0.003).
Commentary The reference standard for diagnosing intussusception is currently abdominal ultrasound (US). However, radiographs of the abdomen are often obtained as the initial work up of belly pain or for evaluation of complications, such as bowel obstruction and/or perforation. These abdominal radiographs may provide prognostic value with respect to success of air enema reduction and surgical outcomes.
FETAL IMAGING
Magnetic resonance imaging of the fetal brain in a monochorionic diamniotic twin gestation: correlation of cerebral injury with ultrasound staging and survival outcomes
Kocaoglu M, Kline-Fath BM, Calvo-Garcia MA et al. Magnetic resonance imaging of the fetal brain in a monochorionic diamniotic twin gestation: correlation of cerebral injury with ultrasound staging and survival outcomes. Pediatr Radiol(2020).
https://doi.org/10.1007/s00247-020-04661-w
Question(s) What is the incidence of cerebral damage on fetal MRI? What is the correlation among abnormal cerebral diffusion-weighted imaging (DWI) findings, survival outcome, and ultrasound (US) staging of twin-twin transfusion syndrome (TTTS)?
Design Retrospective study
Setting Single center: University of Cincinnati/Cincinnati Children’s Hospital Medical Center
Participants 34 pregnancies (68 fetuses)
Intervention MRI/DWI examination of fetal brains
Outcomes MRI/DWI findings of cerebral infarction or hemorrhage, survival outcome, and correlation with the Quintero US stage
Main Results The incidence of fetal cerebral abnormalities was 42 cases (18.5%); 33 of these were seen only on DWI and not by US. No significant correlation between cerebral infarction or hemorrhage on MRI/DWI with US staging (P=0.138). Significant decreased survival was observed in the presence of cerebral infarction or hemorrhage on MRI/DWI (P=0.009).
Commentary The study suggests that fetal cerebral abnormalities could be incorporated into the TTTS staging system as an independent risk factor. Higher sensitivity and accuracy seen with MRI/DWI compared to US and conventional MRI could be helpful toward early intervention and parental counseling.
INTERVENTIONAL RADIOLOGY
Bronchial artery embolization in pediatric pulmonary hemorrhage: a single-center study experience
Zhao QM, Zhao L, He L et al. Bronchial artery embolization in pediatric pulmonary hemorrhage: a single-center study experience. J Vasc Interv Radiol 2020; 31: 1103-1109.
https://doi.org/10.1016/j.jvir.2019.11.007
Question(s) Is bronchial artery embolization (BAE) for pulmonary hemorrhage in children safe and effective?
Design Retrospective study of clinical outcomes of BAE from Feb 2016 – Feb 2019
Setting Single center: Children’s Hospital of Fudan University
Participants 41 patients (44 embolization sessions, 137 embolized vessels) with indication of BAE (massive hemoptysis, recurrent hemoptysis, refractory anemia) due to pulmonary hemosiderosis, congenital heart disease, and infection
Intervention Bronchial artery embolization
Outcomes Technical success (ability to the embolize abnormal vessel) and clinical success as established by the SIR Standards of Practice Committee: the complete or partial resolution of hemoptysis within 30 days of embolization), procedure-related complications, bleeding-free survival rates (at 6, 12, 24, 36 months).
Main Results Technical success achieved in 97.6% of patients, clinical success achieved in 90.2%, 1 procedure-related complication (cerebral infarction) in 2.4%, and 1 death from multiorgan dysfunction in 2.4%. Survival rates at 6, 12, 24, and 36 months were 92.5%, 83.9%, 83.9%, and 70.8%, respectively.
Commentary This single-center study shows BAE for pulmonary hemorrhage to be a safe and effective procedure in children. Given the single but serious complication of cerebral infarction in this study, it is of utmost importance to recognize the connections between embolized vessels and systemic circulation and be vigilant to prospectively identify the retrograde flow of embolization agents into systemic vessels. The recommendation for particle size for BAE in children is >325 µm. The complication in this article occurred due to an error in the selection of the appropriate particle size (100-300 µm). The article reminds readers that impaired neurologic symptoms are frequently masked by general anesthesia used during the procedure.
MUSCULOSKELETAL RADIOLOGY
The development of the intercondylar notch in the pediatric population
Lima FM, Debieux P, Aihara AY et al. The development of the intercondylar notch in the pediatric population. The Knee2020; 27: 747-754.
https://doi.org/10.1016/j.knee.2020.04.020
Question(s) What are the normative values of the intercondylar notch during development with emphasis between males and females?
Design Retrospective study
Setting Single center: Universidade Federal de São Paulo (UNIFESP-EPM)
Participants 253 MRI examinations (130 males, 123 females) ages 6-18 years
Intervention Image review by 2 musculoskeletal radiologists (4 and 7 years of experience)
Outcomes Measurements (intercondylar notch width, bicondylar distance, notch width index (NWI). Statistical analyses (linear and fractional polynomial regression models) were used to evaluate the relationships between measurements.
Main Results Intercondylar notch width increased up to 10 years of age in females and 11 years of age in males. Stabilization occurred up to 13 years in females and 14 years in males. Bicondylar distance showed progressive growth with age for both girls and boys. NWI discretely and homogenously reduced with age in both sexes.
Commentary Anterior cruciate ligament (ACL) repair is now being performed on younger children, so it is important to recognize the normal developmental changes of the intercondylar notch for successful graft placement. The authors offer normative data – albeit from a single center perspective – regarding intercondylar notch development in males and females.
Deep learning measurement of leg length discrepancy in children based on radiographs
Zheng, Q, Shellikeri S, Huang H et al. Deep learning measurement of leg length discrepancy in children based on radiographs. Radiology 2020; 296: 152-158.
https://doi.org/10.1148/radiol.2020192003
Question(s) How does deep learning (DL) measurements of leg length discrepancy (LLD) compare to measurements obtained by radiologists?
Design Retrospective study
Setting Single center: University of Pennsylvania/Children’s Hospital of Philadelphia
Participants 179 patients total identified between January-August 2018. 26 random cases were used for assessment of measurement times.
Intervention Computer-aided automatic measurement of LLD based on DL model compared to radiologists of 1-30 years of post-fellowship experience
Outcomes Spatial overlap and dice score (for training and validation of the DL model), calculation time between the DL model vs. 1 radiologist, correlation between radiology reports and the DL-calculated lengths (for testing of the DL model)
Main Results High spatial overlap between automatic and manual segmentation of pediatric legs (Dice value 0.97 for coarse segmentation and 0.94 for fine segmentation). There was significant correlation (P<0.001) between both in separated femurs and tibias (r=0.99), full pediatric leg lengths (r=0.99), and full LLD (r=0.84). There was no difference in calculated leg lengths between radiology reports and DL calculation (P>0.05). Measurements from radiology reports correlated with DL calculations (P<0.01) in separate femurs and tibias (r=0.99), full pediatric leg lengths (r=0.99), and full LLD (r=0.92). The DL algorithm took approximately 1 second to measure LLD for each pediatric patient compared to the radiologist, who averaged 96 seconds (P<0.01).
Commentary LLD studies are commonly obtained by orthopedic surgeons to guide treatment and surgical planning. Their evaluation is time-consuming and labor-intensive. A highly trained and validated DL model can significantly decrease the time needed to determine the LLD for each study while maintaining accuracy comparable to that of a human reader. Other potential additional benefits include measurement of leg angles and evaluation of hardware.
NEURORADIOLOGY
Imaging utilization in children with headaches: current status and opportunities for improvement
Trofimova AV, Kishore D, Urquia L et al. Imaging utilization in children with headaches: current status and opportunities for improvement. J Am Coll Radiol 2020; 17: 574-583.
https://doi.org/10.1016/j.jacr.2020.01.008
Question(s) What is the use and estimated costs of imaging in children with headaches?
Design Retrospective, cross-sectional study
Setting Single center: Emory University/Children’s Healthcare of Atlanta (free-standing tertiary care 255-bed pediatric hospital with regional level I trauma center)
Participants 4,257 imaging studies in 2015; 3,098 studies (73%) were excluded because they met appropriate indications and 1,159 studies (27%) presumably did not meet imaging guideline recommendations, based on evidence from the ACR Appropriateness Criteria, American Academy of Neurology practice parameters and the Pediatric Emergency Care Applied Research Network
Intervention Chart review of relevant imaging examinations, containing the terms “headache” or “migraine”
Outcomes Rates of diagnoses that were likely/potentially causative of headaches, total estimated societal cost of imaging studies for presumed primary headache, loss of imaging time
Main Results 19.8% of studies had diagnoses that were likely/potentially causative of headaches and 71.2% were normal. The total estimated cost in 2015 was $322,422. Loss of imaging time was 845.3 hours.
Commentary Based on the data, the study offers several areas of improvement including provider education on imaging appropriateness criteria for various clinical signs. For example, there seems to be a disconnect between initial clinical management of sinus disease (i.e., watchful waiting or antibiotics) versus imaging guidelines, which indicate cross-sectional imaging for complicated sinusitis (i.e., orbital or intracranial involvement). Therefore, provider education along with guidelines adherence, establishing care pathways, providing feedback, and using electronic systems to guide provider decisions may help reduce cost and loss of imaging time.
Role of susceptibility-weighted imaging and intratumoral susceptibility signals in grading and differentiating pediatric brain tumors at 1.5T: a preliminary study
Gaudino S, Marziali G, Pezzullo G et al. Role of susceptibility-weighted imaging and intratumoral susceptibility signals in grading and differentiating pediatric brain tumors at 1.5T: a preliminary study. Neuroradiol 2020; 62: 705-713.
https://doi.org/10.1007/s00234-020-02386-z
Question(s) What is the role of susceptibility-weighted imaging (SWI) for pediatric tumor grading and differentiation by assessment of intratumoral susceptibility signal intensity (ITSS)?
Design Retrospective study
Setting Single center: Fondazione Policlinico Universitario A. Gemelli IRCCS
Participants 96 children with histopathologically diagnosed brain tumors from Feb 2013 to April 2019
Intervention Image analysis by 2 readers (neuroradiologist with 10+ years of experience and resident radiologist with 4 years of experience) for assignment of ITSS scores (0 = absence of ITSS, 1 = presence of 1-10 ITSS, and 2 = presence of >11 ITSS).
Outcomes ITSS score to distinguish between low-grade (LG) and high-grade (HG) tumors, histological differentiation, tumor location, and the inter-reader agreement (k)
Main Results For all tumors, there was a statistically significant difference between LG and HG tumors with ITSS scores of 0 and 2, respectively (P=0.002). ITSS scores of 0 and 2 were strong discriminating factors for astrocytoma (score 0), embryonal, choroid plexus, germ-cell, pineal and ependymomal tumors (score 2) (P=0.001). The inter-observer agreement was almost perfect (k = 0.844).
Commentary The results support possible diagnostic value of SWI for pediatric brain tumor grading and histologic differentiation. However, with regard to glioma grading, the results seen in the pediatric population do not necessarily reflect those observed in the adult population. Further studies are needed to better explore the potential of SWI in combination with perfusion and DWI data.
NUCLEAR MEDICINE
Pediatric applications of Dotatate: early diagnostic and therapeutic experience
McElroy KM, Binkovitz LA, Trout AT et al. Pediatric applications of Dotatate: early diagnostic and therapeutic experience. Pediatr Radiol 2020; 50: 882-897.
https://doi.org/10.1007/s00247-020-04688-z
Question(s) What is the experience of using Dotatate for the diagnosis and treatment of neuroendocrine tumors?
Design Case series
Setting 5 centers in the USA (Mayo Clinic, UPMC Children’s Hospital of Pittsburgh, Cincinnati Children’s Hospital Medical Center, Texas Children’s Hospital, Children’s Hospital of Philadelphia)
Participants Patients <19 years old with 68Ga-Dotatate PET/CTs
Intervention 68Ga-Dotatate for diagnostic imaging, 177Lu-, 90Y-, or 111In-Dotatate for treatment
Outcomes Localization of primary, metastatic and recurrent tumors. Imaging and clinical response to treatment with Dotatate.
Main Results There are multiple advantages of 68Ga-Dotatate in tumor localization over the standard agents 123I-MIBG, Octreoscan, and 18F-FDG PET/CT. Dotatate therapy, when paired with 177Lu, 90Y, or 111In, may offer prospective treatment options which result in both imaging and clinical response.
Commentary This multicenter collaboration describes the experience of using Dotatate, a novel somatostatin receptor (SSTR) ligand, for the diagnosis and potential treatment of common pediatric tumors that express SSTRs. These include sporadic neuroendocrine tumors, such as neuroblastoma, carcinoid, pheochromocytoma/paraganglioma, and also hereditary neuroendocrine tumors, like multiple endocrine neoplasia (MEN) syndromes, pheochromocytoma-paraganglioma syndromes, Von Hippel-Lindau (VHL) syndrome, and neurofibromatosis type 1 (NF-1). The article showcases the spectrum of potential diagnostic and therapeutic uses of Dotatate with representative cases from multiple institutions.
ReferencesEdwards EA, Lee MM, Acord MR et al. Abdominal applications of pediatric body MR angiography: tailored optimization for successful outcome. AJR 2020; 215: 1-9.
https://doi.org/10.2214/AJR.19.22289
Dillman JR, Serai SD, Miethke AG et al. Comparison of liver T1 relaxation times without and with iron correction in pediatric autoimmune liver disease. Pediatr Radiol 2020; 50: 935-942.
https://doi.org/10.1007/s00247-020-04663-8
Patel DM, Loewen JM, Braithwaite KA et al. Radiographic findings of irreducibility and surgical resection in ileocolic intussusception. Pediatr Radiol (2020).
https://doi.org/10.1007/s00247-020-04695-0
Kocaoglu M, Kline-Fath BM, Calvo-Garcia MA et al. Magnetic resonance imaging of the fetal brain in monochorionic diamniotic twin gestation: correlation of cerebral injury with ultrasound staging and survival outcomes. Pediatr Radiol(2020).
https://doi.org/10.1007/s00247-020-04661-w
Zhao QM, Zhao L, He L et al. Bronchial artery embolization in pediatric pulmonary hemorrhage: a single-center study experience. J Vasc Interv Radiol 2020; 31: 1103-1109.
https://doi.org/10.1016/j.jvir.2019.11.007
Lima FM, Debieux P, Aihara AY et al. The development of the intercondylar notch in the pediatric population. The Knee2020; 27: 747-754.
https://doi.org/10.1016/j.knee.2020.04.020
Zheng, Q, Shellikeri S, Huang H et al. Deep learning measurement of leg length discrepancy in children based on radiographs. Radiology 2020; 296: 152-158.
https://doi.org/10.1148/radiol.2020192003
Trofimova AV, Kishore D, Urquia L et al. Imaging utilization in children with headaches: current status and opportunities for improvement. J Am Coll Radiol 2020; 17: 574-583.
https://doi.org/10.1016/j.jacr.2020.01.008
Gaudino S, Marziali G, Pezzullo G et al. Role of susceptibility-weighted imaging and intratumoral susceptibility signals in grading and differentiating pediatric brain tumors at 1.5T: a preliminary study. Neuroradiol 2020; 62: 705-713.
https://doi.org/10.1007/s00234-020-02386-z
McElroy KM, Binkovitz LA, Trout AT et al. Pediatric applications of Dotatate: early diagnostic and therapeutic experience. Pediatr Radiol 2020; 50: 882-897.
https://doi.org/10.1007/s00247-020-04688-z