What’s new in Pediatric Imaging – February 2021

12 months ago

 

BODY IMAGING

 

Pneumatosis intestinalis after hematopoietic stem cell transplantation: when not doing anything is good enough

Wallace G, Rosen N, Towbin AJ et al.

J Ped Surg 2021. Journal Pre-Proof.

https://doi.org/10.1016/j.jpedsurg.2020.12.020

Question(s) What is the significance and management of pneumatosis intestinalis (PI) in hematopoietic stem cell transplant (HSCT) recipients?

Design Retrospective study from January 2008 through November 2019

Setting Single center

Participants 53 HSCT recipients with radiologic evidence of PI

Outcomes Clinical PI presentation (disease characteristics: time of diagnosis after transplantation, symptoms, time of PI resolution), management, and outcomes

Results The median time of diagnosis was 113 days (IQR 34-136 days) after transplantation. Six patients (11%) were asymptomatic when PI was incidentally identified on radiologic studies performed for other indications, such as feeding tube placement or post-procedurally (GI endoscopy, oscillator placement, intubation). Forty-seven patients (89%) presented with clinical symptoms of abdominal pain, distension or diarrhea. Forty-three patients (81%) did not have a targeted clinical intervention for PI; they were closely monitored. The remaining 10 patients received conservative management, which consisted of daily monitoring for inpatients and at least twice a week clinic visits for outpatients; these patients were also provided additional intravenous hydration as needed. No patients received surgical management. The median time to resolve PI was 15 days (IQR 3-61 days). Eight patients died within 100 days of PI diagnosis, but bowel ischemia was not the cause of death in any patients.

Commentary Pneumatosis intestinalis can be seen in HSCT recipients receiving systemic steroid therapy. This study describes its clinical characteristics and demonstrates that conservative management has been largely successful in these patients, so decisions should be guided by relevant clinical symptoms rather than the presence or absence of PI. It is important to recognize that although PI in HSCT recipients may rarely result in surgical complications, surgical evaluation is necessary in cases of acute clinical change.

 

Use of small bowel ultrasound to predict response to infliximab induction in pediatric Crohn’s disease

Dolinger MT, Choi JJ, Phan BL et al.

J Clin Gastroenterol 2020;00:1-4.

https://doi.org/10.1097/MCG.0000000000001367

Question(s) What is the utility of small bowel ultrasound (SBUS) as a noninvasive tool to assess induction response to infliximab (IFX) in pediatric Crohn’s disease (CD)?

Design Prospective study

Setting Single center

Participants 15 patients enrolled in the study, 13 patients included in the analysis (2 patients were excluded due to denial of maintenance insurance coverage and inability to complete induction therapy secondary to anaphylaxis)

Intervention Small bowel ultrasound performed at baseline (T0) and at week 14 (T14)

Outcomes The primary outcome is to describe changes in SBUS parameters pre- and post-induction and how they are associated with clinical and biomarker response: maximum length of continuous involved segment (bowel segment length, BSL in cm), bowel wall hyperemia (BWH), bowel wall thickness >3 cm (BWT), strictures, creeping fat, free fluid, abscess, and/or obstruction and qualitative degree of bowel dilatation (none, mild, moderate, or severe), degree of peristalsis (diminished, normal, mildly increased, moderately increased, or severely increased), and stool burden (normal, mildly increased, moderately increased or excessive). The clinical response measured was achievement of steroid-free remission. Biomarkers analyzed were erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), fecal calprotectin (FC), and weighted Pediatric Crohn’s Disease Activity Index (wPCDAI).

Results There was decrease in BWH and BSL involved (p=0.01 and p=0.07, respectively) after induction. Decrease in FC at 1 week (T1) moderately correlated with a decrease in BSL (r=0.57, p=0.04). No correlation was seen with BWT, BWH, ESR, CRP or patient-reported clinical disease activity after induction. Clinically, all 13 patients who underwent induction therapy achieved steroid-free remission and had significant decrease in wPCDAI (p<0.001), ESR (p=0.001), and CRP (p=0.01). Of the 13 patients, 9 showed decreased FC post-induction (p=0.06).

Commentary The authors conclude that BWH is the first parameter on SBUS to change in response to IFX induction, which echoes the current literature. Major limitations include small sample size and abbreviated length of longitudinal follow up.

 

BREAST IMAGING

 

Use of Breast Imaging-Reporting and Data System (BI-RADS) ultrasound classification in pediatric and adolescent patients overestimates likelihood of malignancy

Davis J, Liang J, Roh A et al.

J Ped Surg 2021. Journal Pre-proof.

https://doi.org/10.1016/j.jpedsurg.2020.12.025

Question(s) What is the validity of the BI-RADS ultrasound classification for malignant breast lesions?

Design Retrospective study

Setting Multicenter (3 centers); exams were interpreted independently by board-certified breast imaging radiologists.

Participants 227 patients with available ultrasound exam, under age 21 with breast pathology (other than gynecomastia, abscess, or acute mastitis) were included from January 2003 through November 2016. Phyllodes tumors were included as ‘benign’ or ‘malignant’ based on the clinical documentation.

Intervention BI-RADS ultrasound classification compared to histologic diagnosis

Outcomes Proportion of BI-RADS assessment applied in ultrasound reports, proportion of each BI-RADS category (1, 2, 3, 4, or 5), false-positive rate

Results BI-RADS assessment was applied in 84.1%. 4 patients (1.8%) were assigned BI-RADS 1 (negative or no findings), 38 patients (16.7%) were assigned BI-RADS 2 (benign, 0% malignancy), 24 patients (10.6%) received a BI-RADS 3 category (probably benign, <2% malignancy), 124 patients (54.6%) received a BI-RADS 4 category (suspicious for malignancy, 2-94% malignancy), and 1 patient (0.4%) was assigned BI-RADS 5 (highly suspicious for malignancy, >95% malignancy). Both BI-RADS 4 and 5 categories lead to tissue biopsy and the rate of malignancy in these categories was 0%; thus, there was a 100% false-positive rate for BI-RADS 4 and 5. Note, the single BI-RADS 5 assignment proved to be fat necrosis.

Commentary The authors conclude that use of BI-RADS ultrasound classification in the pediatric and adolescent population severely overestimates the risk of malignancy; therefore, the BI-RADS ultrasound classification is not valid for this population. This is not unexpected as the BI-RADS classification was developed for adult patients. It might be important to evaluate why the majority of these breast ultrasound reports even applied BI-RADS for pediatric and adolescent patients. It is unclear why only 84.1% of included patients had a BI-RADS category assignment; the authors list the benign pathologies of the remaining 15.9% who did not receive at BI-RADS assignment, and it would be helpful to clarify the reasons, for example, if this was due to inconsistencies in clinical practice. In addition to the retrospective study design, another limitation includes the extremely low incidence of breast malignancy in the pediatric population, which often precludes the need for tissue biopsy; the authors suggest that if BI-RADS is to be used in our pediatric/adolescent population, then the associated biopsy recommendations should be disregarded.

 

CARDIAC IMAGING

 

Echocardiographic findings in pediatric multisystem inflammatory syndrome associated with COVID-19 in the United States

Matsubara D, Kauffman HL, Wang Y et al.

JACC 2020;76(17):1947-61.

https://doi.org/10.1016/j.jacc.2020.08.056

Question(s) What are the anatomic and functional echocardiographic manifestations in multisystem inflammatory syndrome in children (MIS-C)? Is MIS-C associated with reduced systolic and diastolic function, similar to other viral myocarditis?

Design Retrospective study from April to June 2020

Setting Single center (The Children’s Hospital of Philadelphia, Philadelphia, PA)

Participants 28 MIS-C patients, 20 classic Kawasaki disease (KD) patients, and 20 healthy control subjects

Intervention Echocardiographic parameters in the acute phase of disease (MIS-C and KD groups) and during the subacute period in the MIS-C group (5.2+3 days)

Outcomes Coronary artery dilatation, left ventricular (LV) systolic and diastolic function by deformation parameters (including global longitudinal strain, global longitudinal strain rate, longitudinal early diastolic strain rate, and peak left atrial strain), and myocardial injury

Results 1 of 28 patients with MIS-C manifested coronary artery dilatation (ectasia). LV systolic function was worse in patients with MIS-C compared to both healthy controls and patients with classic KD (LVEF 57% vs. 64% vs. 66%, p<0.001 and p=0.003). LV diastolic function by all deformation parameters were worse in patients with MIS-C. In MIS-C patients, the strongest predictors associated with myocardial injury, defined by elevated biomarkers brain natriuretic peptide BNP) and/or troponin-I, were global longitudinal strain, global circumferential strain, left atrial strain, and peak longitudinal strain of the right ventricular free wall.

Commentary The study concludes that coronary arteries may be spared in patients with MIS-C, but cardiac dysfunction may be more common than in KD patients. However, the small sample size and unavailability of prolonged longitudinal follow-up are major limitations of this study.

 

CHEST IMAGING

 

Atypical pulmonary metastases in children: pictorial review of imaging patterns

Gagnon MH, Wallace AB, Yedururi S et al.

Pediatric Radiology 2021;51:131-139.

https://doi.org/10.1007/s00247-020-04821-y

Question(s) What are the imaging findings of atypical pulmonary metastatic disease in children?

Design Pictorial review

Commentary Typical metastatic pulmonary disease in children present as circumscribed, solid nodules, commonly in the lung bases. This article demonstrates the atypical presentations – cavitary or calcified nodules, peritumoral halo, tumor embolism, miliary nodules, lymphangitic carcinomatosis, pleural metastasis, spontaneous pneumothorax, and atelectasis/lobar collapse – of metastatic pulmonary disease from common primary pediatric cancers. It is important for radiologists to recognize these atypical presentations to avoid underdiagnosis and understaging.

 

CTA utilization for evaluation of suspected pulmonary embolism in a tertiary pediatric emergency department

Barrera CA, Otero HJ, Fenlon E et al.

Clinical Imaging 2020;0(00). Journal Pre-Proof.

https://doi.org/10.1016/j.clinimag.2020.12.024

Question(s) What are the changes in the use of computed tomography angiography (CTA) for evaluating suspected pulmonary embolism (PE)? What is the positive rate of ancillary findings for studies negative for PE?

Design Retrospective study from 2005 to 2017

Setting Single center

Participants 307 chest CTAs for suspected PE were included

Outcomes Rate of positive chest CTA for PE, rate of negative chest CTA, rate of negative chest CTA but positive ancillary findings.

Results 50 cases (16%) were reported positive, 91 cases (30%) were negative but positive for ancillary findings, which were most commonly pneumonia and pleural effusions. The number of chest CTAs for PE increased by 3.2 studies per year, but the rate of chest CTAs positive for ancillary findings and positive for PE was stable throughout this time period (1.5 and 0.3, respectively).

Commentary The study states that the rate of CTAs ordered for evaluation of PE has increased over time but the overall rate of diagnosis of PE remained stable, though a rationale was not offered for the stability of the incidence. This reflects similar trends in the adult literature. The study describes common ancillary findings in the diagnosis of PE, but it does not address their clinical relevance, which the authors state could be examined in the future. Limitations include the retrospective nature of the study design and limited data (such as lack of imaging review and exclusion of outside CTA exams).

 

INTERVENTIONAL RADIOLOGY

 

Mechanical endovenous ablation of varicose veins in pediatric patients with Klippel-Trenaunay syndrome: feasibility, safety, and initial results

Lambert G, Teplisky D, Cabezas M et al.

J Vasc Interv Radiol 2021;32:80-86.

https://doi.org/10.1016/j.jvir.2020.08.019

Question(s) What are the feasibility, safety and results of endovenous mechanochemical ablation (MOCA) for treatment of embryonic and dysplastic veins in patients with Klippel-Trenaunay syndrome (KTS)?

Design Retrospective study

Setting Single center

Participants 11 pediatric patients (13 procedures) were included (mean age 11.9 years, range 4-16 years)

Intervention MOCA via ClariVein device (Merit Medical Systems, South Jordan, UT) with mean follow-up of 16 months (range 6-25 months)

Outcomes Technical success rate, primary occlusion rate, adverse effects, and recanalization rates

Results All 13 patients achieved technical success and primary occlusion, which lasted for 1 year. At follow up, 2 of 13 patients (18%) showed partial recanalization with symptom recurrence. A repeat MOCA procedure was successful in both patients. No adverse events occurred.

Commentary Surgical treatment of KTS has largely been replaced by minimally invasive radiologic techniques, including sclerotherapy, embolization, and more recently, endovenous thermal ablation, such as radiofrequency and laser ablation. MOCA is a more recent technique that combines mechanical damage via a rotating wire tip and chemical damage via a sclerosing agent to the venous endothelium. Unlike the previously described ablation techniques, MOCA does not require tumescent anesthesia and it is not associated with thermal complications. This study shows that MOCA is a safe and effective treatment option for their population. The main limitations of the study include its small sample size and that it did not precisely define “feasibility.”

 

NEURORADIOLOGY

 

Management of hydrocephalus in children: anatomic imaging appearances of CSF shunts and their complications

Khalatbari H and Parisi MT.

AJR 2021;216:187-199.

https://doi.org/10.2214/AJR.20.22888

Question(s) What is the management of hydrocephalus and CSF shunts?

Design Review

Commentary This is a detailed, educational review of pediatric hydrocephalus with a primary focus on its management. The authors describe the anatomy and mechanisms of various CSF shunts, including shunt valves and diversionary shunts. Sample CT and MRI protocols for are provided during discussion of different imaging modalities used to evaluate hydrocephalus and CSF shunts. The article concludes with mechanisms of shunt complications and a brief discussion of CSF shunt revision.

 

AI & MACHINE LEARNING

 

Improving image quality and reducing radiation dose for pediatric CT by using deep learning reconstruction

Brady SL, Trout AT, Somasundaram E et al.

Radiology 2021;298:180-188.

https://doi.org/10.1148/radiol.2020202317

Question(s) How does the deep learning reconstruction (DLR) algorithm affect image quality and radiation dose for pediatric CT?

Design Retrospective study using data from randomly selected CT exams performed February to December 2018

Setting Single center (Cincinnati Children’s Hospital Medical Center)

Participants 4 anatomic structures were selected for review: azygous vein, right hepatic vein, common bile duct, and superior mesenteric artery from 19 patients (total of 152 series)

Intervention Objective comparison of the Advanced Intelligent Clear-IQ Engine, the DLR algorithm (developed by Canon Medical Systems) versus existing filtered back projection (FBP), statistical-based iterative reconstruction (SBIR), and model-based iterative reconstruction (MBIR). A subjective comparison by 3 fellowship-trained pediatric radiologists was also performed for validation.

Outcomes Objective outcomes included image quality (as object detectability and contrast difference) and dose reduction estimation (as noise magnitude and volume CT dose index value). Observer outcomes included ratings (scale of 1-10) of edge definition, quantum noise level, and object conspicuity.

Results DLR demonstrated improved object detectability compared to FBR (by 51%), SBIR (by 18%) and MBIR (by 11%). DLR also demonstrated 52% greater dose reduction than SBIR. DLR was preferred over FBR, SBIR, or MBIR by radiologists.

Commentary The study demonstrates that this DLR algorithm was able to improve image quality and has greater radiation dose reduction potential than existing alternative algorithms. The authors acknowledge a few major limitations, including but not limited to: a small sample size, retrospective study design, and lack of pathology evaluation.

References
  • Share